oceansweep commited on
Commit
fa9a583
1 Parent(s): 086d7f1

Syncing latest changes, lets see what breaks

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. App_Function_Libraries/Article_Summarization_Lib.py +220 -291
  2. App_Function_Libraries/Audio_Files.py +691 -691
  3. App_Function_Libraries/Audio_Transcription_Lib.py +191 -191
  4. App_Function_Libraries/Chat.py +273 -0
  5. App_Function_Libraries/Chat_related_functions.py +41 -0
  6. App_Function_Libraries/ChromaDB_Library.py +225 -0
  7. App_Function_Libraries/Chunk_Lib.py +586 -582
  8. App_Function_Libraries/DB_Manager.py +472 -0
  9. App_Function_Libraries/Diarization_Lib.py +1 -6
  10. App_Function_Libraries/Gradio_Related.py +0 -0
  11. App_Function_Libraries/Gradio_UI/Audio_ingestion_tab.py +152 -0
  12. App_Function_Libraries/Gradio_UI/Chat_ui.py +1017 -0
  13. App_Function_Libraries/Gradio_UI/Explain_summarize_tab.py +193 -0
  14. App_Function_Libraries/Gradio_UI/Export_Functionality.py +314 -0
  15. App_Function_Libraries/Gradio_UI/Gradio_Shared.py +284 -0
  16. App_Function_Libraries/Gradio_UI/Import_Functionality.py +473 -0
  17. App_Function_Libraries/Gradio_UI/Introduction_tab.py +161 -0
  18. App_Function_Libraries/Gradio_UI/Keywords.py +65 -0
  19. App_Function_Libraries/Gradio_UI/Llamafile_tab.py +122 -0
  20. App_Function_Libraries/Gradio_UI/Media_edit.py +273 -0
  21. App_Function_Libraries/Gradio_UI/PDF_ingestion_tab.py +152 -0
  22. App_Function_Libraries/Gradio_UI/Podcast_tab.py +164 -0
  23. App_Function_Libraries/Gradio_UI/Re_summarize_tab.py +268 -0
  24. App_Function_Libraries/Gradio_UI/Search_Tab.py +487 -0
  25. App_Function_Libraries/Gradio_UI/Transcript_comparison.py +94 -0
  26. App_Function_Libraries/Gradio_UI/Trash.py +134 -0
  27. App_Function_Libraries/Gradio_UI/Utilities.py +118 -0
  28. App_Function_Libraries/Gradio_UI/Video_transcription_tab.py +691 -0
  29. App_Function_Libraries/Gradio_UI/Website_scraping_tab.py +113 -0
  30. App_Function_Libraries/Gradio_UI/Writing.py +700 -0
  31. App_Function_Libraries/Gradio_UI/__init__.py +0 -0
  32. App_Function_Libraries/Gradio_UI/__pycache__/Audio_ingestion_tab.cpython-312.pyc +0 -0
  33. App_Function_Libraries/Gradio_UI/__pycache__/Chat_ui.cpython-312.pyc +0 -0
  34. App_Function_Libraries/Gradio_UI/__pycache__/Explain_summarize_tab.cpython-312.pyc +0 -0
  35. App_Function_Libraries/Gradio_UI/__pycache__/Export_Functionality.cpython-312.pyc +0 -0
  36. App_Function_Libraries/Gradio_UI/__pycache__/Gradio_Shared.cpython-312.pyc +0 -0
  37. App_Function_Libraries/Gradio_UI/__pycache__/Import_Functionality.cpython-312.pyc +0 -0
  38. App_Function_Libraries/Gradio_UI/__pycache__/Introduction_tab.cpython-312.pyc +0 -0
  39. App_Function_Libraries/Gradio_UI/__pycache__/Keywords.cpython-312.pyc +0 -0
  40. App_Function_Libraries/Gradio_UI/__pycache__/Llamafile_tab.cpython-312.pyc +0 -0
  41. App_Function_Libraries/Gradio_UI/__pycache__/Media_edit.cpython-312.pyc +0 -0
  42. App_Function_Libraries/Gradio_UI/__pycache__/PDF_ingestion_tab.cpython-312.pyc +0 -0
  43. App_Function_Libraries/Gradio_UI/__pycache__/Podcast_tab.cpython-312.pyc +0 -0
  44. App_Function_Libraries/Gradio_UI/__pycache__/Re_summarize_tab.cpython-312.pyc +0 -0
  45. App_Function_Libraries/Gradio_UI/__pycache__/Search_Tab.cpython-312.pyc +0 -0
  46. App_Function_Libraries/Gradio_UI/__pycache__/Trash.cpython-312.pyc +0 -0
  47. App_Function_Libraries/Gradio_UI/__pycache__/Utilities.cpython-312.pyc +0 -0
  48. App_Function_Libraries/Gradio_UI/__pycache__/Video_transcription_tab.cpython-312.pyc +0 -0
  49. App_Function_Libraries/Gradio_UI/__pycache__/Website_scraping_tab.cpython-312.pyc +0 -0
  50. App_Function_Libraries/Gradio_UI/__pycache__/Writing.cpython-312.pyc +0 -0
App_Function_Libraries/Article_Summarization_Lib.py CHANGED
@@ -1,292 +1,221 @@
1
- # Article_Summarization_Lib.py
2
- #########################################
3
- # Article Summarization Library
4
- # This library is used to handle summarization of articles.
5
-
6
- #
7
- ####
8
- #
9
- ####################
10
- # Function List
11
- #
12
- # 1.
13
- #
14
- ####################
15
- #
16
- # Import necessary libraries
17
- import datetime
18
- from datetime import datetime
19
- import gradio as gr
20
- import json
21
- import os
22
- import logging
23
- import requests
24
- # 3rd-Party Imports
25
- from tqdm import tqdm
26
-
27
- from App_Function_Libraries.Utils import sanitize_filename
28
- # Local Imports
29
- from Article_Extractor_Lib import scrape_article
30
- from Local_Summarization_Lib import summarize_with_llama, summarize_with_oobabooga, summarize_with_tabbyapi, \
31
- summarize_with_vllm, summarize_with_kobold, save_summary_to_file, summarize_with_local_llm
32
- from Summarization_General_Lib import summarize_with_openai, summarize_with_anthropic, summarize_with_cohere, \
33
- summarize_with_groq, summarize_with_openrouter, summarize_with_deepseek, summarize_with_huggingface, \
34
- summarize_with_mistral
35
- from SQLite_DB import Database, create_tables, add_media_with_keywords
36
- #
37
- #######################################################################################################################
38
- # Function Definitions
39
- #
40
-
41
- def ingest_article_to_db(url, title, author, content, keywords, summary, ingestion_date, custom_prompt):
42
- try:
43
- # Check if content is not empty or whitespace
44
- if not content.strip():
45
- raise ValueError("Content is empty.")
46
-
47
- db = Database()
48
- create_tables()
49
- keyword_list = keywords.split(',') if keywords else ["default"]
50
- keyword_str = ', '.join(keyword_list)
51
-
52
- # Set default values for missing fields
53
- url = url or 'Unknown'
54
- title = title or 'Unknown'
55
- author = author or 'Unknown'
56
- keywords = keywords or 'default'
57
- summary = summary or 'No summary available'
58
- ingestion_date = ingestion_date or datetime.datetime.now().strftime('%Y-%m-%d')
59
-
60
- # Log the values of all fields before calling add_media_with_keywords
61
- logging.debug(f"URL: {url}")
62
- logging.debug(f"Title: {title}")
63
- logging.debug(f"Author: {author}")
64
- logging.debug(f"Content: {content[:50]}... (length: {len(content)})") # Log first 50 characters of content
65
- logging.debug(f"Keywords: {keywords}")
66
- logging.debug(f"Summary: {summary}")
67
- logging.debug(f"Ingestion Date: {ingestion_date}")
68
- logging.debug(f"Custom Prompt: {custom_prompt}")
69
-
70
- # Check if any required field is empty and log the specific missing field
71
- if not url:
72
- logging.error("URL is missing.")
73
- raise ValueError("URL is missing.")
74
- if not title:
75
- logging.error("Title is missing.")
76
- raise ValueError("Title is missing.")
77
- if not content:
78
- logging.error("Content is missing.")
79
- raise ValueError("Content is missing.")
80
- if not keywords:
81
- logging.error("Keywords are missing.")
82
- raise ValueError("Keywords are missing.")
83
- if not summary:
84
- logging.error("Summary is missing.")
85
- raise ValueError("Summary is missing.")
86
- if not ingestion_date:
87
- logging.error("Ingestion date is missing.")
88
- raise ValueError("Ingestion date is missing.")
89
- if not custom_prompt:
90
- logging.error("Custom prompt is missing.")
91
- raise ValueError("Custom prompt is missing.")
92
-
93
- # Add media with keywords to the database
94
- result = add_media_with_keywords(
95
- url=url,
96
- title=title,
97
- media_type='article',
98
- content=content,
99
- keywords=keyword_str or "article_default",
100
- prompt=custom_prompt or None,
101
- summary=summary or "No summary generated",
102
- transcription_model=None, # or some default value if applicable
103
- author=author or 'Unknown',
104
- ingestion_date=ingestion_date
105
- )
106
- return result
107
- except Exception as e:
108
- logging.error(f"Failed to ingest article to the database: {e}")
109
- return str(e)
110
-
111
-
112
- def scrape_and_summarize_multiple(urls, custom_prompt_arg, api_name, api_key, keywords, custom_article_titles, system_message=None):
113
- urls = [url.strip() for url in urls.split('\n') if url.strip()]
114
- custom_titles = custom_article_titles.split('\n') if custom_article_titles else []
115
-
116
- results = []
117
- errors = []
118
-
119
- # Create a progress bar
120
- progress = gr.Progress()
121
-
122
- for i, url in tqdm(enumerate(urls), total=len(urls), desc="Processing URLs"):
123
- custom_title = custom_titles[i] if i < len(custom_titles) else None
124
- try:
125
- result = scrape_and_summarize(url, custom_prompt_arg, api_name, api_key, keywords, custom_title, system_message)
126
- results.append(f"Results for URL {i + 1}:\n{result}")
127
- except Exception as e:
128
- error_message = f"Error processing URL {i + 1} ({url}): {str(e)}"
129
- errors.append(error_message)
130
- results.append(f"Failed to process URL {i + 1}: {url}")
131
-
132
- # Update progress
133
- progress((i + 1) / len(urls), desc=f"Processed {i + 1}/{len(urls)} URLs")
134
-
135
- # Combine results and errors
136
- combined_output = "\n".join(results)
137
- if errors:
138
- combined_output += "\n\nErrors encountered:\n" + "\n".join(errors)
139
-
140
- return combined_output
141
-
142
-
143
- def scrape_and_summarize(url, custom_prompt_arg, api_name, api_key, keywords, custom_article_title, system_message=None):
144
- try:
145
- # Step 1: Scrape the article
146
- article_data = scrape_article(url)
147
- print(f"Scraped Article Data: {article_data}") # Debugging statement
148
- if not article_data:
149
- return "Failed to scrape the article."
150
-
151
- # Use the custom title if provided, otherwise use the scraped title
152
- title = custom_article_title.strip() if custom_article_title else article_data.get('title', 'Untitled')
153
- author = article_data.get('author', 'Unknown')
154
- content = article_data.get('content', '')
155
- ingestion_date = datetime.now().strftime('%Y-%m-%d')
156
-
157
- print(f"Title: {title}, Author: {author}, Content Length: {len(content)}") # Debugging statement
158
-
159
- # Custom system prompt for the article
160
- system_message = system_message or "Act as a professional summarizer and summarize this article."
161
- # Custom prompt for the article
162
- article_custom_prompt = custom_prompt_arg or "Act as a professional summarizer and summarize this article."
163
-
164
- # Step 2: Summarize the article
165
- summary = None
166
- if api_name:
167
- logging.debug(f"Article_Summarizer: Summarization being performed by {api_name}")
168
-
169
- # Sanitize filename for saving the JSON file
170
- sanitized_title = sanitize_filename(title)
171
- json_file_path = os.path.join("Results", f"{sanitized_title}_segments.json")
172
-
173
- with open(json_file_path, 'w') as json_file:
174
- json.dump([{'text': content}], json_file, indent=2)
175
-
176
- # FIXME - Swap out this if/else to use the dedicated function....
177
- try:
178
- if api_name.lower() == 'openai':
179
- # def summarize_with_openai(api_key, input_data, custom_prompt_arg)
180
- summary = summarize_with_openai(api_key, json_file_path, article_custom_prompt, system_message)
181
-
182
- elif api_name.lower() == "anthropic":
183
- # def summarize_with_anthropic(api_key, input_data, model, custom_prompt_arg, max_retries=3, retry_delay=5):
184
- summary = summarize_with_anthropic(api_key, json_file_path, article_custom_prompt, system_message)
185
- elif api_name.lower() == "cohere":
186
- # def summarize_with_cohere(api_key, input_data, model, custom_prompt_arg)
187
- summary = summarize_with_cohere(api_key, json_file_path, article_custom_prompt, system_message)
188
-
189
- elif api_name.lower() == "groq":
190
- logging.debug(f"MAIN: Trying to summarize with groq")
191
- # def summarize_with_groq(api_key, input_data, model, custom_prompt_arg):
192
- summary = summarize_with_groq(api_key, json_file_path, article_custom_prompt, system_message)
193
-
194
- elif api_name.lower() == "openrouter":
195
- logging.debug(f"MAIN: Trying to summarize with OpenRouter")
196
- # def summarize_with_openrouter(api_key, input_data, custom_prompt_arg):
197
- summary = summarize_with_openrouter(api_key, json_file_path, article_custom_prompt, system_message)
198
-
199
- elif api_name.lower() == "deepseek":
200
- logging.debug(f"MAIN: Trying to summarize with DeepSeek")
201
- # def summarize_with_deepseek(api_key, input_data, custom_prompt_arg):
202
- summary = summarize_with_deepseek(api_key, json_file_path, article_custom_prompt, system_message)
203
-
204
- elif api_name.lower() == "mistral":
205
- summary = summarize_with_mistral(api_key, json_file_path, article_custom_prompt, system_message)
206
-
207
- elif api_name.lower() == "llama.cpp":
208
- logging.debug(f"MAIN: Trying to summarize with Llama.cpp")
209
- # def summarize_with_llama(api_url, file_path, token, custom_prompt)
210
- summary = summarize_with_llama(json_file_path, article_custom_prompt, system_message)
211
-
212
- elif api_name.lower() == "kobold":
213
- logging.debug(f"MAIN: Trying to summarize with Kobold.cpp")
214
- # def summarize_with_kobold(input_data, kobold_api_token, custom_prompt_input, api_url):
215
- summary = summarize_with_kobold(json_file_path, api_key, article_custom_prompt, system_message)
216
-
217
- elif api_name.lower() == "ooba":
218
- # def summarize_with_oobabooga(input_data, api_key, custom_prompt, api_url):
219
- summary = summarize_with_oobabooga(json_file_path, api_key, article_custom_prompt, system_message)
220
-
221
- elif api_name.lower() == "tabbyapi":
222
- # def summarize_with_tabbyapi(input_data, tabby_model, custom_prompt_input, api_key=None, api_IP):
223
- summary = summarize_with_tabbyapi(json_file_path, article_custom_prompt, system_message)
224
-
225
- elif api_name.lower() == "vllm":
226
- logging.debug(f"MAIN: Trying to summarize with VLLM")
227
- # def summarize_with_vllm(api_key, input_data, custom_prompt_input):
228
- summary = summarize_with_vllm(json_file_path, article_custom_prompt, system_message)
229
-
230
- elif api_name.lower() == "local-llm":
231
- logging.debug(f"MAIN: Trying to summarize with Local LLM")
232
- summary = summarize_with_local_llm(json_file_path, article_custom_prompt, system_message)
233
-
234
- elif api_name.lower() == "huggingface":
235
- logging.debug(f"MAIN: Trying to summarize with huggingface")
236
- # def summarize_with_huggingface(api_key, input_data, custom_prompt_arg):
237
- summarize_with_huggingface(api_key, json_file_path, article_custom_prompt, system_message)
238
- # Add additional API handlers here...
239
- except requests.exceptions.ConnectionError as e:
240
- logging.error(f"Connection error while trying to summarize with {api_name}: {str(e)}")
241
-
242
- if summary:
243
- logging.info(f"Article_Summarizer: Summary generated using {api_name} API")
244
- save_summary_to_file(summary, json_file_path)
245
- else:
246
- summary = "Summary not available"
247
- logging.warning(f"Failed to generate summary using {api_name} API")
248
-
249
- else:
250
- summary = "Article Summarization: No API provided for summarization."
251
-
252
- print(f"Summary: {summary}") # Debugging statement
253
-
254
- # Step 3: Ingest the article into the database
255
- ingestion_result = ingest_article_to_db(url, title, author, content, keywords, summary, ingestion_date,
256
- article_custom_prompt)
257
-
258
- return f"Title: {title}\nAuthor: {author}\nIngestion Result: {ingestion_result}\n\nSummary: {summary}\n\nArticle Contents: {content}"
259
- except Exception as e:
260
- logging.error(f"Error processing URL {url}: {str(e)}")
261
- return f"Failed to process URL {url}: {str(e)}"
262
-
263
-
264
- def ingest_unstructured_text(text, custom_prompt, api_name, api_key, keywords, custom_article_title, system_message=None):
265
- title = custom_article_title.strip() if custom_article_title else "Unstructured Text"
266
- author = "Unknown"
267
- ingestion_date = datetime.now().strftime('%Y-%m-%d')
268
-
269
- # Summarize the unstructured text
270
- if api_name:
271
- json_file_path = f"Results/{title.replace(' ', '_')}_segments.json"
272
- with open(json_file_path, 'w') as json_file:
273
- json.dump([{'text': text}], json_file, indent=2)
274
-
275
- if api_name.lower() == 'openai':
276
- summary = summarize_with_openai(api_key, json_file_path, custom_prompt, system_message)
277
- # Add other APIs as needed
278
- else:
279
- summary = "Unsupported API."
280
- else:
281
- summary = "No API provided for summarization."
282
-
283
- # Ingest the unstructured text into the database
284
- ingestion_result = ingest_article_to_db('Unstructured Text', title, author, text, keywords, summary, ingestion_date,
285
- custom_prompt)
286
- return f"Title: {title}\nSummary: {summary}\nIngestion Result: {ingestion_result}"
287
-
288
-
289
-
290
- #
291
- #
292
  #######################################################################################################################
 
1
+ # Article_Summarization_Lib.py
2
+ #########################################
3
+ # Article Summarization Library
4
+ # This library is used to handle summarization of articles.
5
+
6
+ #
7
+ ####
8
+ #
9
+ ####################
10
+ # Function List
11
+ #
12
+ # 1.
13
+ #
14
+ ####################
15
+ #
16
+ # Import necessary libraries
17
+ import datetime
18
+ from datetime import datetime
19
+ import gradio as gr
20
+ import json
21
+ import os
22
+ import logging
23
+ import requests
24
+ # 3rd-Party Imports
25
+ from tqdm import tqdm
26
+
27
+ from App_Function_Libraries.Utils import sanitize_filename
28
+ # Local Imports
29
+ from Article_Extractor_Lib import scrape_article
30
+ from Local_Summarization_Lib import summarize_with_llama, summarize_with_oobabooga, summarize_with_tabbyapi, \
31
+ summarize_with_vllm, summarize_with_kobold, save_summary_to_file, summarize_with_local_llm
32
+ from Summarization_General_Lib import summarize_with_openai, summarize_with_anthropic, summarize_with_cohere, \
33
+ summarize_with_groq, summarize_with_openrouter, summarize_with_deepseek, summarize_with_huggingface, \
34
+ summarize_with_mistral
35
+ from App_Function_Libraries.DB_Manager import ingest_article_to_db
36
+ #
37
+ #######################################################################################################################
38
+ # Function Definitions
39
+ #
40
+
41
+ def scrape_and_summarize_multiple(urls, custom_prompt_arg, api_name, api_key, keywords, custom_article_titles, system_message=None):
42
+ urls = [url.strip() for url in urls.split('\n') if url.strip()]
43
+ custom_titles = custom_article_titles.split('\n') if custom_article_titles else []
44
+
45
+ results = []
46
+ errors = []
47
+
48
+ # Create a progress bar
49
+ progress = gr.Progress()
50
+
51
+ for i, url in tqdm(enumerate(urls), total=len(urls), desc="Processing URLs"):
52
+ custom_title = custom_titles[i] if i < len(custom_titles) else None
53
+ try:
54
+ result = scrape_and_summarize(url, custom_prompt_arg, api_name, api_key, keywords, custom_title, system_message)
55
+ results.append(f"Results for URL {i + 1}:\n{result}")
56
+ except Exception as e:
57
+ error_message = f"Error processing URL {i + 1} ({url}): {str(e)}"
58
+ errors.append(error_message)
59
+ results.append(f"Failed to process URL {i + 1}: {url}")
60
+
61
+ # Update progress
62
+ progress((i + 1) / len(urls), desc=f"Processed {i + 1}/{len(urls)} URLs")
63
+
64
+ # Combine results and errors
65
+ combined_output = "\n".join(results)
66
+ if errors:
67
+ combined_output += "\n\nErrors encountered:\n" + "\n".join(errors)
68
+
69
+ return combined_output
70
+
71
+
72
+ def scrape_and_summarize(url, custom_prompt_arg, api_name, api_key, keywords, custom_article_title, system_message=None):
73
+ try:
74
+ # Step 1: Scrape the article
75
+ article_data = scrape_article(url)
76
+ print(f"Scraped Article Data: {article_data}") # Debugging statement
77
+ if not article_data:
78
+ return "Failed to scrape the article."
79
+
80
+ # Use the custom title if provided, otherwise use the scraped title
81
+ title = custom_article_title.strip() if custom_article_title else article_data.get('title', 'Untitled')
82
+ author = article_data.get('author', 'Unknown')
83
+ content = article_data.get('content', '')
84
+ ingestion_date = datetime.now().strftime('%Y-%m-%d')
85
+
86
+ print(f"Title: {title}, Author: {author}, Content Length: {len(content)}") # Debugging statement
87
+
88
+ # Custom system prompt for the article
89
+ system_message = system_message or "Act as a professional summarizer and summarize this article."
90
+ # Custom prompt for the article
91
+ article_custom_prompt = custom_prompt_arg or "Act as a professional summarizer and summarize this article."
92
+
93
+ # Step 2: Summarize the article
94
+ summary = None
95
+ if api_name:
96
+ logging.debug(f"Article_Summarizer: Summarization being performed by {api_name}")
97
+
98
+ # Sanitize filename for saving the JSON file
99
+ sanitized_title = sanitize_filename(title)
100
+ json_file_path = os.path.join("Results", f"{sanitized_title}_segments.json")
101
+
102
+ with open(json_file_path, 'w') as json_file:
103
+ json.dump([{'text': content}], json_file, indent=2)
104
+
105
+ # FIXME - Swap out this if/else to use the dedicated function....
106
+ try:
107
+ if api_name.lower() == 'openai':
108
+ # def summarize_with_openai(api_key, input_data, custom_prompt_arg)
109
+ summary = summarize_with_openai(api_key, json_file_path, article_custom_prompt, system_message)
110
+
111
+ elif api_name.lower() == "anthropic":
112
+ # def summarize_with_anthropic(api_key, input_data, model, custom_prompt_arg, max_retries=3, retry_delay=5):
113
+ summary = summarize_with_anthropic(api_key, json_file_path, article_custom_prompt, system_message)
114
+ elif api_name.lower() == "cohere":
115
+ # def summarize_with_cohere(api_key, input_data, model, custom_prompt_arg)
116
+ summary = summarize_with_cohere(api_key, json_file_path, article_custom_prompt, system_message)
117
+
118
+ elif api_name.lower() == "groq":
119
+ logging.debug(f"MAIN: Trying to summarize with groq")
120
+ # def summarize_with_groq(api_key, input_data, model, custom_prompt_arg):
121
+ summary = summarize_with_groq(api_key, json_file_path, article_custom_prompt, system_message)
122
+
123
+ elif api_name.lower() == "openrouter":
124
+ logging.debug(f"MAIN: Trying to summarize with OpenRouter")
125
+ # def summarize_with_openrouter(api_key, input_data, custom_prompt_arg):
126
+ summary = summarize_with_openrouter(api_key, json_file_path, article_custom_prompt, system_message)
127
+
128
+ elif api_name.lower() == "deepseek":
129
+ logging.debug(f"MAIN: Trying to summarize with DeepSeek")
130
+ # def summarize_with_deepseek(api_key, input_data, custom_prompt_arg):
131
+ summary = summarize_with_deepseek(api_key, json_file_path, article_custom_prompt, system_message)
132
+
133
+ elif api_name.lower() == "mistral":
134
+ summary = summarize_with_mistral(api_key, json_file_path, article_custom_prompt, system_message)
135
+
136
+ elif api_name.lower() == "llama.cpp":
137
+ logging.debug(f"MAIN: Trying to summarize with Llama.cpp")
138
+ # def summarize_with_llama(api_url, file_path, token, custom_prompt)
139
+ summary = summarize_with_llama(json_file_path, article_custom_prompt, system_message)
140
+
141
+ elif api_name.lower() == "kobold":
142
+ logging.debug(f"MAIN: Trying to summarize with Kobold.cpp")
143
+ # def summarize_with_kobold(input_data, kobold_api_token, custom_prompt_input, api_url):
144
+ summary = summarize_with_kobold(json_file_path, api_key, article_custom_prompt, system_message)
145
+
146
+ elif api_name.lower() == "ooba":
147
+ # def summarize_with_oobabooga(input_data, api_key, custom_prompt, api_url):
148
+ summary = summarize_with_oobabooga(json_file_path, api_key, article_custom_prompt, system_message)
149
+
150
+ elif api_name.lower() == "tabbyapi":
151
+ # def summarize_with_tabbyapi(input_data, tabby_model, custom_prompt_input, api_key=None, api_IP):
152
+ summary = summarize_with_tabbyapi(json_file_path, article_custom_prompt, system_message)
153
+
154
+ elif api_name.lower() == "vllm":
155
+ logging.debug(f"MAIN: Trying to summarize with VLLM")
156
+ # def summarize_with_vllm(api_key, input_data, custom_prompt_input):
157
+ summary = summarize_with_vllm(json_file_path, article_custom_prompt, system_message)
158
+
159
+ elif api_name.lower() == "local-llm":
160
+ logging.debug(f"MAIN: Trying to summarize with Local LLM")
161
+ summary = summarize_with_local_llm(json_file_path, article_custom_prompt, system_message)
162
+
163
+ elif api_name.lower() == "huggingface":
164
+ logging.debug(f"MAIN: Trying to summarize with huggingface")
165
+ # def summarize_with_huggingface(api_key, input_data, custom_prompt_arg):
166
+ summarize_with_huggingface(api_key, json_file_path, article_custom_prompt, system_message)
167
+ # Add additional API handlers here...
168
+ except requests.exceptions.ConnectionError as e:
169
+ logging.error(f"Connection error while trying to summarize with {api_name}: {str(e)}")
170
+
171
+ if summary:
172
+ logging.info(f"Article_Summarizer: Summary generated using {api_name} API")
173
+ save_summary_to_file(summary, json_file_path)
174
+ else:
175
+ summary = "Summary not available"
176
+ logging.warning(f"Failed to generate summary using {api_name} API")
177
+
178
+ else:
179
+ summary = "Article Summarization: No API provided for summarization."
180
+
181
+ print(f"Summary: {summary}") # Debugging statement
182
+
183
+ # Step 3: Ingest the article into the database
184
+ ingestion_result = ingest_article_to_db(url, title, author, content, keywords, summary, ingestion_date,
185
+ article_custom_prompt)
186
+
187
+ return f"Title: {title}\nAuthor: {author}\nIngestion Result: {ingestion_result}\n\nSummary: {summary}\n\nArticle Contents: {content}"
188
+ except Exception as e:
189
+ logging.error(f"Error processing URL {url}: {str(e)}")
190
+ return f"Failed to process URL {url}: {str(e)}"
191
+
192
+
193
+ def ingest_unstructured_text(text, custom_prompt, api_name, api_key, keywords, custom_article_title, system_message=None):
194
+ title = custom_article_title.strip() if custom_article_title else "Unstructured Text"
195
+ author = "Unknown"
196
+ ingestion_date = datetime.now().strftime('%Y-%m-%d')
197
+
198
+ # Summarize the unstructured text
199
+ if api_name:
200
+ json_file_path = f"Results/{title.replace(' ', '_')}_segments.json"
201
+ with open(json_file_path, 'w') as json_file:
202
+ json.dump([{'text': text}], json_file, indent=2)
203
+
204
+ if api_name.lower() == 'openai':
205
+ summary = summarize_with_openai(api_key, json_file_path, custom_prompt, system_message)
206
+ # Add other APIs as needed
207
+ else:
208
+ summary = "Unsupported API."
209
+ else:
210
+ summary = "No API provided for summarization."
211
+
212
+ # Ingest the unstructured text into the database
213
+ ingestion_result = ingest_article_to_db('Unstructured Text', title, author, text, keywords, summary, ingestion_date,
214
+ custom_prompt)
215
+ return f"Title: {title}\nSummary: {summary}\nIngestion Result: {ingestion_result}"
216
+
217
+
218
+
219
+ #
220
+ #
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
221
  #######################################################################################################################
App_Function_Libraries/Audio_Files.py CHANGED
@@ -1,692 +1,692 @@
1
- # Audio_Files.py
2
- #########################################
3
- # Audio Processing Library
4
- # This library is used to download or load audio files from a local directory.
5
- #
6
- ####
7
- #
8
- # Functions:
9
- #
10
- # download_audio_file(url, save_path)
11
- # process_audio(
12
- # process_audio_file(audio_url, audio_file, whisper_model="small.en", api_name=None, api_key=None)
13
- #
14
- #
15
- #########################################
16
- # Imports
17
- import json
18
- import logging
19
- import os
20
- import subprocess
21
- import tempfile
22
- import uuid
23
- from datetime import datetime
24
- from pathlib import Path
25
-
26
- import requests
27
- import yt_dlp
28
-
29
- from App_Function_Libraries.Audio_Transcription_Lib import speech_to_text
30
- from App_Function_Libraries.Chunk_Lib import improved_chunking_process
31
- #
32
- # Local Imports
33
- from App_Function_Libraries.SQLite_DB import add_media_to_database, add_media_with_keywords, \
34
- check_media_and_whisper_model
35
- from App_Function_Libraries.Summarization_General_Lib import save_transcription_and_summary, perform_transcription, \
36
- perform_summarization
37
- from App_Function_Libraries.Utils import create_download_directory, save_segments_to_json, downloaded_files, \
38
- sanitize_filename
39
- from App_Function_Libraries.Video_DL_Ingestion_Lib import extract_metadata
40
-
41
- #
42
- #######################################################################################################################
43
- # Function Definitions
44
- #
45
-
46
- MAX_FILE_SIZE = 500 * 1024 * 1024
47
-
48
-
49
- def download_audio_file(url, current_whisper_model="", use_cookies=False, cookies=None):
50
- try:
51
- # Check if media already exists in the database and compare whisper models
52
- should_download, reason = check_media_and_whisper_model(
53
- url=url,
54
- current_whisper_model=current_whisper_model
55
- )
56
-
57
- if not should_download:
58
- logging.info(f"Skipping audio download: {reason}")
59
- return None
60
-
61
- logging.info(f"Proceeding with audio download: {reason}")
62
-
63
- # Set up the request headers
64
- headers = {}
65
- if use_cookies and cookies:
66
- try:
67
- cookie_dict = json.loads(cookies)
68
- headers['Cookie'] = '; '.join([f'{k}={v}' for k, v in cookie_dict.items()])
69
- except json.JSONDecodeError:
70
- logging.warning("Invalid cookie format. Proceeding without cookies.")
71
-
72
- # Make the request
73
- response = requests.get(url, headers=headers, stream=True)
74
- # Raise an exception for bad status codes
75
- response.raise_for_status()
76
-
77
- # Get the file size
78
- file_size = int(response.headers.get('content-length', 0))
79
- if file_size > 500 * 1024 * 1024: # 500 MB limit
80
- raise ValueError("File size exceeds the 500MB limit.")
81
-
82
- # Generate a unique filename
83
- file_name = f"audio_{uuid.uuid4().hex[:8]}.mp3"
84
- save_path = os.path.join('downloads', file_name)
85
-
86
- # Ensure the downloads directory exists
87
- os.makedirs('downloads', exist_ok=True)
88
-
89
-
90
- # Download the file
91
- with open(save_path, 'wb') as f:
92
- for chunk in response.iter_content(chunk_size=8192):
93
- if chunk:
94
- f.write(chunk)
95
-
96
- logging.info(f"Audio file downloaded successfully: {save_path}")
97
- return save_path
98
-
99
- except requests.RequestException as e:
100
- logging.error(f"Error downloading audio file: {str(e)}")
101
- raise
102
- except ValueError as e:
103
- logging.error(str(e))
104
- raise
105
- except Exception as e:
106
- logging.error(f"Unexpected error downloading audio file: {str(e)}")
107
- raise
108
-
109
-
110
- def process_audio(
111
- audio_file_path,
112
- num_speakers=2,
113
- whisper_model="small.en",
114
- custom_prompt_input=None,
115
- offset=0,
116
- api_name=None,
117
- api_key=None,
118
- vad_filter=False,
119
- rolling_summarization=False,
120
- detail_level=0.01,
121
- keywords="default,no_keyword_set",
122
- chunk_text_by_words=False,
123
- max_words=0,
124
- chunk_text_by_sentences=False,
125
- max_sentences=0,
126
- chunk_text_by_paragraphs=False,
127
- max_paragraphs=0,
128
- chunk_text_by_tokens=False,
129
- max_tokens=0
130
- ):
131
- try:
132
-
133
- # Perform transcription
134
- audio_file_path, segments = perform_transcription(audio_file_path, offset, whisper_model, vad_filter)
135
-
136
- if audio_file_path is None or segments is None:
137
- logging.error("Process_Audio: Transcription failed or segments not available.")
138
- return "Process_Audio: Transcription failed.", None, None, None, None, None
139
-
140
- logging.debug(f"Process_Audio: Transcription audio_file: {audio_file_path}")
141
- logging.debug(f"Process_Audio: Transcription segments: {segments}")
142
-
143
- transcription_text = {'audio_file': audio_file_path, 'transcription': segments}
144
- logging.debug(f"Process_Audio: Transcription text: {transcription_text}")
145
-
146
- # Save segments to JSON
147
- segments_json_path = save_segments_to_json(segments)
148
-
149
- # Perform summarization
150
- summary_text = None
151
- if api_name:
152
- if rolling_summarization is not None:
153
- pass
154
- # FIXME rolling summarization
155
- # summary_text = rolling_summarize_function(
156
- # transcription_text,
157
- # detail=detail_level,
158
- # api_name=api_name,
159
- # api_key=api_key,
160
- # custom_prompt=custom_prompt_input,
161
- # chunk_by_words=chunk_text_by_words,
162
- # max_words=max_words,
163
- # chunk_by_sentences=chunk_text_by_sentences,
164
- # max_sentences=max_sentences,
165
- # chunk_by_paragraphs=chunk_text_by_paragraphs,
166
- # max_paragraphs=max_paragraphs,
167
- # chunk_by_tokens=chunk_text_by_tokens,
168
- # max_tokens=max_tokens
169
- # )
170
- else:
171
- summary_text = perform_summarization(api_name, segments_json_path, custom_prompt_input, api_key)
172
-
173
- if summary_text is None:
174
- logging.error("Summary text is None. Check summarization function.")
175
- summary_file_path = None
176
- else:
177
- summary_text = 'Summary not available'
178
- summary_file_path = None
179
-
180
- # Save transcription and summary
181
- download_path = create_download_directory("Audio_Processing")
182
- json_file_path, summary_file_path = save_transcription_and_summary(transcription_text, summary_text,
183
- download_path)
184
-
185
- # Update function call to add_media_to_database so that it properly applies the title, author and file type
186
- # Add to database
187
- add_media_to_database(None, {'title': 'Audio File', 'author': 'Unknown'}, segments, summary_text, keywords,
188
- custom_prompt_input, whisper_model)
189
-
190
- return transcription_text, summary_text, json_file_path, summary_file_path, None, None
191
-
192
- except Exception as e:
193
- logging.error(f"Error in process_audio: {str(e)}")
194
- return str(e), None, None, None, None, None
195
-
196
-
197
- def process_single_audio(audio_file_path, whisper_model, api_name, api_key, keep_original,custom_keywords, source,
198
- custom_prompt_input, chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
199
- use_multi_level_chunking, chunk_language):
200
- progress = []
201
- transcription = ""
202
- summary = ""
203
-
204
- def update_progress(message):
205
- progress.append(message)
206
- return "\n".join(progress)
207
-
208
- try:
209
- # Check file size before processing
210
- file_size = os.path.getsize(audio_file_path)
211
- if file_size > MAX_FILE_SIZE:
212
- update_progress(f"File size ({file_size / (1024 * 1024):.2f} MB) exceeds the maximum limit of {MAX_FILE_SIZE / (1024 * 1024):.2f} MB. Skipping this file.")
213
- return "\n".join(progress), "", ""
214
-
215
- # Perform transcription
216
- update_progress("Starting transcription...")
217
- segments = speech_to_text(audio_file_path, whisper_model=whisper_model)
218
- transcription = " ".join([segment['Text'] for segment in segments])
219
- update_progress("Audio transcribed successfully.")
220
-
221
- # Perform summarization if API is provided
222
- if api_name and api_key:
223
- update_progress("Starting summarization...")
224
- summary = perform_summarization(api_name, transcription, "Summarize the following audio transcript",
225
- api_key)
226
- update_progress("Audio summarized successfully.")
227
- else:
228
- summary = "No summary available"
229
-
230
- # Prepare keywords
231
- keywords = "audio,transcription"
232
- if custom_keywords:
233
- keywords += f",{custom_keywords}"
234
-
235
- # Add to database
236
- add_media_with_keywords(
237
- url=source,
238
- title=os.path.basename(audio_file_path),
239
- media_type='audio',
240
- content=transcription,
241
- keywords=keywords,
242
- prompt="Summarize the following audio transcript",
243
- summary=summary,
244
- transcription_model=whisper_model,
245
- author="Unknown",
246
- ingestion_date=None # This will use the current date
247
- )
248
- update_progress("Audio file added to database successfully.")
249
-
250
- if not keep_original and source != "Uploaded File":
251
- os.remove(audio_file_path)
252
- update_progress(f"Temporary file {audio_file_path} removed.")
253
- elif keep_original and source != "Uploaded File":
254
- update_progress(f"Original audio file kept at: {audio_file_path}")
255
-
256
- except Exception as e:
257
- update_progress(f"Error processing {source}: {str(e)}")
258
- transcription = f"Error: {str(e)}"
259
- summary = "No summary due to error"
260
-
261
- return "\n".join(progress), transcription, summary
262
-
263
-
264
- def process_audio_files(audio_urls, audio_file, whisper_model, api_name, api_key, use_cookies, cookies, keep_original,
265
- custom_keywords, custom_prompt_input, chunk_method, max_chunk_size, chunk_overlap,
266
- use_adaptive_chunking, use_multi_level_chunking, chunk_language, diarize):
267
- progress = []
268
- temp_files = []
269
- all_transcriptions = []
270
- all_summaries = []
271
-
272
- def update_progress(message):
273
- progress.append(message)
274
- return "\n".join(progress)
275
-
276
- def cleanup_files():
277
- for file in temp_files:
278
- try:
279
- if os.path.exists(file):
280
- os.remove(file)
281
- update_progress(f"Temporary file {file} removed.")
282
- except Exception as e:
283
- update_progress(f"Failed to remove temporary file {file}: {str(e)}")
284
-
285
- def reencode_mp3(mp3_file_path):
286
- try:
287
- reencoded_mp3_path = mp3_file_path.replace(".mp3", "_reencoded.mp3")
288
- subprocess.run([ffmpeg_cmd, '-i', mp3_file_path, '-codec:a', 'libmp3lame', reencoded_mp3_path], check=True)
289
- update_progress(f"Re-encoded {mp3_file_path} to {reencoded_mp3_path}.")
290
- return reencoded_mp3_path
291
- except subprocess.CalledProcessError as e:
292
- update_progress(f"Error re-encoding {mp3_file_path}: {str(e)}")
293
- raise
294
-
295
- def convert_mp3_to_wav(mp3_file_path):
296
- try:
297
- wav_file_path = mp3_file_path.replace(".mp3", ".wav")
298
- subprocess.run([ffmpeg_cmd, '-i', mp3_file_path, wav_file_path], check=True)
299
- update_progress(f"Converted {mp3_file_path} to {wav_file_path}.")
300
- return wav_file_path
301
- except subprocess.CalledProcessError as e:
302
- update_progress(f"Error converting {mp3_file_path} to WAV: {str(e)}")
303
- raise
304
-
305
- try:
306
- # Check and set the ffmpeg command
307
- global ffmpeg_cmd
308
- if os.name == "nt":
309
- logging.debug("Running on Windows")
310
- ffmpeg_cmd = os.path.join(os.getcwd(), "Bin", "ffmpeg.exe")
311
- else:
312
- ffmpeg_cmd = 'ffmpeg' # Assume 'ffmpeg' is in PATH for non-Windows systems
313
-
314
- # Ensure ffmpeg is accessible
315
- if not os.path.exists(ffmpeg_cmd) and os.name == "nt":
316
- raise FileNotFoundError(f"ffmpeg executable not found at path: {ffmpeg_cmd}")
317
-
318
- # Define chunk options early to avoid undefined errors
319
- chunk_options = {
320
- 'method': chunk_method,
321
- 'max_size': max_chunk_size,
322
- 'overlap': chunk_overlap,
323
- 'adaptive': use_adaptive_chunking,
324
- 'multi_level': use_multi_level_chunking,
325
- 'language': chunk_language
326
- }
327
-
328
- # Process multiple URLs
329
- urls = [url.strip() for url in audio_urls.split('\n') if url.strip()]
330
-
331
- for i, url in enumerate(urls):
332
- update_progress(f"Processing URL {i + 1}/{len(urls)}: {url}")
333
-
334
- # Download and process audio file
335
- audio_file_path = download_audio_file(url, use_cookies, cookies)
336
- if not os.path.exists(audio_file_path):
337
- update_progress(f"Downloaded file not found: {audio_file_path}")
338
- continue
339
-
340
- temp_files.append(audio_file_path)
341
- update_progress("Audio file downloaded successfully.")
342
-
343
- # Re-encode MP3 to fix potential issues
344
- reencoded_mp3_path = reencode_mp3(audio_file_path)
345
- if not os.path.exists(reencoded_mp3_path):
346
- update_progress(f"Re-encoded file not found: {reencoded_mp3_path}")
347
- continue
348
-
349
- temp_files.append(reencoded_mp3_path)
350
-
351
- # Convert re-encoded MP3 to WAV
352
- wav_file_path = convert_mp3_to_wav(reencoded_mp3_path)
353
- if not os.path.exists(wav_file_path):
354
- update_progress(f"Converted WAV file not found: {wav_file_path}")
355
- continue
356
-
357
- temp_files.append(wav_file_path)
358
-
359
- # Initialize transcription
360
- transcription = ""
361
-
362
- # Transcribe audio
363
- if diarize:
364
- segments = speech_to_text(wav_file_path, whisper_model=whisper_model, diarize=True)
365
- else:
366
- segments = speech_to_text(wav_file_path, whisper_model=whisper_model)
367
-
368
- # Handle segments nested under 'segments' key
369
- if isinstance(segments, dict) and 'segments' in segments:
370
- segments = segments['segments']
371
-
372
- if isinstance(segments, list):
373
- transcription = " ".join([segment.get('Text', '') for segment in segments])
374
- update_progress("Audio transcribed successfully.")
375
- else:
376
- update_progress("Unexpected segments format received from speech_to_text.")
377
- logging.error(f"Unexpected segments format: {segments}")
378
- continue
379
-
380
- if not transcription.strip():
381
- update_progress("Transcription is empty.")
382
- else:
383
- # Apply chunking
384
- chunked_text = improved_chunking_process(transcription, chunk_options)
385
-
386
- # Summarize
387
- if api_name:
388
- try:
389
- summary = perform_summarization(api_name, chunked_text, custom_prompt_input, api_key)
390
- update_progress("Audio summarized successfully.")
391
- except Exception as e:
392
- logging.error(f"Error during summarization: {str(e)}")
393
- summary = "Summary generation failed"
394
- else:
395
- summary = "No summary available (API not provided)"
396
-
397
- all_transcriptions.append(transcription)
398
- all_summaries.append(summary)
399
-
400
- # Add to database
401
- add_media_with_keywords(
402
- url=url,
403
- title=os.path.basename(wav_file_path),
404
- media_type='audio',
405
- content=transcription,
406
- keywords=custom_keywords,
407
- prompt=custom_prompt_input,
408
- summary=summary,
409
- transcription_model=whisper_model,
410
- author="Unknown",
411
- ingestion_date=datetime.now().strftime('%Y-%m-%d')
412
- )
413
- update_progress("Audio file processed and added to database.")
414
-
415
- # Process uploaded file if provided
416
- if audio_file:
417
- if os.path.getsize(audio_file.name) > MAX_FILE_SIZE:
418
- update_progress(
419
- f"Uploaded file size exceeds the maximum limit of {MAX_FILE_SIZE / (1024 * 1024):.2f}MB. Skipping this file.")
420
- else:
421
- # Re-encode MP3 to fix potential issues
422
- reencoded_mp3_path = reencode_mp3(audio_file.name)
423
- if not os.path.exists(reencoded_mp3_path):
424
- update_progress(f"Re-encoded file not found: {reencoded_mp3_path}")
425
- return update_progress("Processing failed: Re-encoded file not found"), "", ""
426
-
427
- temp_files.append(reencoded_mp3_path)
428
-
429
- # Convert re-encoded MP3 to WAV
430
- wav_file_path = convert_mp3_to_wav(reencoded_mp3_path)
431
- if not os.path.exists(wav_file_path):
432
- update_progress(f"Converted WAV file not found: {wav_file_path}")
433
- return update_progress("Processing failed: Converted WAV file not found"), "", ""
434
-
435
- temp_files.append(wav_file_path)
436
-
437
- # Initialize transcription
438
- transcription = ""
439
-
440
- if diarize:
441
- segments = speech_to_text(wav_file_path, whisper_model=whisper_model, diarize=True)
442
- else:
443
- segments = speech_to_text(wav_file_path, whisper_model=whisper_model)
444
-
445
- # Handle segments nested under 'segments' key
446
- if isinstance(segments, dict) and 'segments' in segments:
447
- segments = segments['segments']
448
-
449
- if isinstance(segments, list):
450
- transcription = " ".join([segment.get('Text', '') for segment in segments])
451
- else:
452
- update_progress("Unexpected segments format received from speech_to_text.")
453
- logging.error(f"Unexpected segments format: {segments}")
454
-
455
- chunked_text = improved_chunking_process(transcription, chunk_options)
456
-
457
- if api_name and api_key:
458
- try:
459
- summary = perform_summarization(api_name, chunked_text, custom_prompt_input, api_key)
460
- update_progress("Audio summarized successfully.")
461
- except Exception as e:
462
- logging.error(f"Error during summarization: {str(e)}")
463
- summary = "Summary generation failed"
464
- else:
465
- summary = "No summary available (API not provided)"
466
-
467
- all_transcriptions.append(transcription)
468
- all_summaries.append(summary)
469
-
470
- add_media_with_keywords(
471
- url="Uploaded File",
472
- title=os.path.basename(wav_file_path),
473
- media_type='audio',
474
- content=transcription,
475
- keywords=custom_keywords,
476
- prompt=custom_prompt_input,
477
- summary=summary,
478
- transcription_model=whisper_model,
479
- author="Unknown",
480
- ingestion_date=datetime.now().strftime('%Y-%m-%d')
481
- )
482
- update_progress("Uploaded file processed and added to database.")
483
-
484
- # Final cleanup
485
- if not keep_original:
486
- cleanup_files()
487
-
488
- final_progress = update_progress("All processing complete.")
489
- final_transcriptions = "\n\n".join(all_transcriptions)
490
- final_summaries = "\n\n".join(all_summaries)
491
-
492
- return final_progress, final_transcriptions, final_summaries
493
-
494
- except Exception as e:
495
- logging.error(f"Error processing audio files: {str(e)}")
496
- cleanup_files()
497
- return update_progress(f"Processing failed: {str(e)}"), "", ""
498
-
499
-
500
- def download_youtube_audio(url):
501
- try:
502
- # Determine ffmpeg path based on the operating system.
503
- ffmpeg_path = './Bin/ffmpeg.exe' if os.name == 'nt' else 'ffmpeg'
504
-
505
- # Create a temporary directory
506
- with tempfile.TemporaryDirectory() as temp_dir:
507
- # Extract information about the video
508
- with yt_dlp.YoutubeDL({'quiet': True}) as ydl:
509
- info_dict = ydl.extract_info(url, download=False)
510
- sanitized_title = sanitize_filename(info_dict['title'])
511
-
512
- # Setup the temporary filenames
513
- temp_video_path = Path(temp_dir) / f"{sanitized_title}_temp.mp4"
514
- temp_audio_path = Path(temp_dir) / f"{sanitized_title}.mp3"
515
-
516
- # Initialize yt-dlp with options for downloading
517
- ydl_opts = {
518
- 'format': 'bestaudio[ext=m4a]/best[height<=480]', # Prefer best audio, or video up to 480p
519
- 'ffmpeg_location': ffmpeg_path,
520
- 'outtmpl': str(temp_video_path),
521
- 'noplaylist': True,
522
- 'quiet': True
523
- }
524
-
525
- # Execute yt-dlp to download the video/audio
526
- with yt_dlp.YoutubeDL(ydl_opts) as ydl:
527
- ydl.download([url])
528
-
529
- # Check if the file exists
530
- if not temp_video_path.exists():
531
- raise FileNotFoundError(f"Expected file was not found: {temp_video_path}")
532
-
533
- # Use ffmpeg to extract audio
534
- ffmpeg_command = [
535
- ffmpeg_path,
536
- '-i', str(temp_video_path),
537
- '-vn', # No video
538
- '-acodec', 'libmp3lame',
539
- '-b:a', '192k',
540
- str(temp_audio_path)
541
- ]
542
- subprocess.run(ffmpeg_command, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
543
-
544
- # Check if the audio file was created
545
- if not temp_audio_path.exists():
546
- raise FileNotFoundError(f"Expected audio file was not found: {temp_audio_path}")
547
-
548
- # Create a persistent directory for the download if it doesn't exist
549
- persistent_dir = Path("downloads")
550
- persistent_dir.mkdir(exist_ok=True)
551
-
552
- # Move the file from the temporary directory to the persistent directory
553
- persistent_file_path = persistent_dir / f"{sanitized_title}.mp3"
554
- os.replace(str(temp_audio_path), str(persistent_file_path))
555
-
556
- # Add the file to the list of downloaded files
557
- downloaded_files.append(str(persistent_file_path))
558
-
559
- return str(persistent_file_path), f"Audio downloaded successfully: {sanitized_title}.mp3"
560
- except Exception as e:
561
- return None, f"Error downloading audio: {str(e)}"
562
-
563
-
564
- def process_podcast(url, title, author, keywords, custom_prompt, api_name, api_key, whisper_model,
565
- keep_original=False, enable_diarization=False, use_cookies=False, cookies=None,
566
- chunk_method=None, max_chunk_size=300, chunk_overlap=0, use_adaptive_chunking=False,
567
- use_multi_level_chunking=False, chunk_language='english'):
568
- progress = []
569
- error_message = ""
570
- temp_files = []
571
-
572
- def update_progress(message):
573
- progress.append(message)
574
- return "\n".join(progress)
575
-
576
- def cleanup_files():
577
- if not keep_original:
578
- for file in temp_files:
579
- try:
580
- if os.path.exists(file):
581
- os.remove(file)
582
- update_progress(f"Temporary file {file} removed.")
583
- except Exception as e:
584
- update_progress(f"Failed to remove temporary file {file}: {str(e)}")
585
-
586
- try:
587
- # Download podcast
588
- audio_file = download_audio_file(url, use_cookies, cookies)
589
- temp_files.append(audio_file)
590
- update_progress("Podcast downloaded successfully.")
591
-
592
- # Extract metadata
593
- metadata = extract_metadata(url)
594
- title = title or metadata.get('title', 'Unknown Podcast')
595
- author = author or metadata.get('uploader', 'Unknown Author')
596
-
597
- # Format metadata for storage
598
- metadata_text = f"""
599
- Metadata:
600
- Title: {title}
601
- Author: {author}
602
- Series: {metadata.get('series', 'N/A')}
603
- Episode: {metadata.get('episode', 'N/A')}
604
- Season: {metadata.get('season', 'N/A')}
605
- Upload Date: {metadata.get('upload_date', 'N/A')}
606
- Duration: {metadata.get('duration', 'N/A')} seconds
607
- Description: {metadata.get('description', 'N/A')}
608
- """
609
-
610
- # Update keywords
611
- new_keywords = []
612
- if metadata.get('series'):
613
- new_keywords.append(f"series:{metadata['series']}")
614
- if metadata.get('episode'):
615
- new_keywords.append(f"episode:{metadata['episode']}")
616
- if metadata.get('season'):
617
- new_keywords.append(f"season:{metadata['season']}")
618
-
619
- keywords = f"{keywords},{','.join(new_keywords)}" if keywords else ','.join(new_keywords)
620
-
621
- update_progress(f"Metadata extracted - Title: {title}, Author: {author}, Keywords: {keywords}")
622
-
623
- # Transcribe the podcast
624
- try:
625
- if enable_diarization:
626
- segments = speech_to_text(audio_file, whisper_model=whisper_model, diarize=True)
627
- else:
628
- segments = speech_to_text(audio_file, whisper_model=whisper_model)
629
- transcription = " ".join([segment['Text'] for segment in segments])
630
- update_progress("Podcast transcribed successfully.")
631
- except Exception as e:
632
- error_message = f"Transcription failed: {str(e)}"
633
- raise
634
-
635
- # Apply chunking
636
- chunk_options = {
637
- 'method': chunk_method,
638
- 'max_size': max_chunk_size,
639
- 'overlap': chunk_overlap,
640
- 'adaptive': use_adaptive_chunking,
641
- 'multi_level': use_multi_level_chunking,
642
- 'language': chunk_language
643
- }
644
- chunked_text = improved_chunking_process(transcription, chunk_options)
645
-
646
- # Combine metadata and transcription
647
- full_content = metadata_text + "\n\nTranscription:\n" + transcription
648
-
649
- # Summarize if API is provided
650
- summary = None
651
- if api_name and api_key:
652
- try:
653
- summary = perform_summarization(api_name, chunked_text, custom_prompt, api_key)
654
- update_progress("Podcast summarized successfully.")
655
- except Exception as e:
656
- error_message = f"Summarization failed: {str(e)}"
657
- raise
658
-
659
- # Add to database
660
- try:
661
- add_media_with_keywords(
662
- url=url,
663
- title=title,
664
- media_type='podcast',
665
- content=full_content,
666
- keywords=keywords,
667
- prompt=custom_prompt,
668
- summary=summary or "No summary available",
669
- transcription_model=whisper_model,
670
- author=author,
671
- ingestion_date=datetime.now().strftime('%Y-%m-%d')
672
- )
673
- update_progress("Podcast added to database successfully.")
674
- except Exception as e:
675
- error_message = f"Error adding podcast to database: {str(e)}"
676
- raise
677
-
678
- # Cleanup
679
- cleanup_files()
680
-
681
- return (update_progress("Processing complete."), full_content, summary or "No summary generated.",
682
- title, author, keywords, error_message)
683
-
684
- except Exception as e:
685
- logging.error(f"Error processing podcast: {str(e)}")
686
- cleanup_files()
687
- return update_progress(f"Processing failed: {str(e)}"), "", "", "", "", "", str(e)
688
-
689
-
690
- #
691
- #
692
  #######################################################################################################################
 
1
+ # Audio_Files.py
2
+ #########################################
3
+ # Audio Processing Library
4
+ # This library is used to download or load audio files from a local directory.
5
+ #
6
+ ####
7
+ #
8
+ # Functions:
9
+ #
10
+ # download_audio_file(url, save_path)
11
+ # process_audio(
12
+ # process_audio_file(audio_url, audio_file, whisper_model="small.en", api_name=None, api_key=None)
13
+ #
14
+ #
15
+ #########################################
16
+ # Imports
17
+ import json
18
+ import logging
19
+ import os
20
+ import subprocess
21
+ import tempfile
22
+ import uuid
23
+ from datetime import datetime
24
+ from pathlib import Path
25
+
26
+ import requests
27
+ import yt_dlp
28
+
29
+ from App_Function_Libraries.Audio_Transcription_Lib import speech_to_text
30
+ from App_Function_Libraries.Chunk_Lib import improved_chunking_process
31
+ #
32
+ # Local Imports
33
+ from App_Function_Libraries.DB_Manager import add_media_to_database, add_media_with_keywords, \
34
+ check_media_and_whisper_model
35
+ from App_Function_Libraries.Summarization_General_Lib import save_transcription_and_summary, perform_transcription, \
36
+ perform_summarization
37
+ from App_Function_Libraries.Utils import create_download_directory, save_segments_to_json, downloaded_files, \
38
+ sanitize_filename
39
+ from App_Function_Libraries.Video_DL_Ingestion_Lib import extract_metadata
40
+
41
+ #
42
+ #######################################################################################################################
43
+ # Function Definitions
44
+ #
45
+
46
+ MAX_FILE_SIZE = 500 * 1024 * 1024
47
+
48
+
49
+ def download_audio_file(url, current_whisper_model="", use_cookies=False, cookies=None):
50
+ try:
51
+ # Check if media already exists in the database and compare whisper models
52
+ should_download, reason = check_media_and_whisper_model(
53
+ url=url,
54
+ current_whisper_model=current_whisper_model
55
+ )
56
+
57
+ if not should_download:
58
+ logging.info(f"Skipping audio download: {reason}")
59
+ return None
60
+
61
+ logging.info(f"Proceeding with audio download: {reason}")
62
+
63
+ # Set up the request headers
64
+ headers = {}
65
+ if use_cookies and cookies:
66
+ try:
67
+ cookie_dict = json.loads(cookies)
68
+ headers['Cookie'] = '; '.join([f'{k}={v}' for k, v in cookie_dict.items()])
69
+ except json.JSONDecodeError:
70
+ logging.warning("Invalid cookie format. Proceeding without cookies.")
71
+
72
+ # Make the request
73
+ response = requests.get(url, headers=headers, stream=True)
74
+ # Raise an exception for bad status codes
75
+ response.raise_for_status()
76
+
77
+ # Get the file size
78
+ file_size = int(response.headers.get('content-length', 0))
79
+ if file_size > 500 * 1024 * 1024: # 500 MB limit
80
+ raise ValueError("File size exceeds the 500MB limit.")
81
+
82
+ # Generate a unique filename
83
+ file_name = f"audio_{uuid.uuid4().hex[:8]}.mp3"
84
+ save_path = os.path.join('downloads', file_name)
85
+
86
+ # Ensure the downloads directory exists
87
+ os.makedirs('downloads', exist_ok=True)
88
+
89
+
90
+ # Download the file
91
+ with open(save_path, 'wb') as f:
92
+ for chunk in response.iter_content(chunk_size=8192):
93
+ if chunk:
94
+ f.write(chunk)
95
+
96
+ logging.info(f"Audio file downloaded successfully: {save_path}")
97
+ return save_path
98
+
99
+ except requests.RequestException as e:
100
+ logging.error(f"Error downloading audio file: {str(e)}")
101
+ raise
102
+ except ValueError as e:
103
+ logging.error(str(e))
104
+ raise
105
+ except Exception as e:
106
+ logging.error(f"Unexpected error downloading audio file: {str(e)}")
107
+ raise
108
+
109
+
110
+ def process_audio(
111
+ audio_file_path,
112
+ num_speakers=2,
113
+ whisper_model="small.en",
114
+ custom_prompt_input=None,
115
+ offset=0,
116
+ api_name=None,
117
+ api_key=None,
118
+ vad_filter=False,
119
+ rolling_summarization=False,
120
+ detail_level=0.01,
121
+ keywords="default,no_keyword_set",
122
+ chunk_text_by_words=False,
123
+ max_words=0,
124
+ chunk_text_by_sentences=False,
125
+ max_sentences=0,
126
+ chunk_text_by_paragraphs=False,
127
+ max_paragraphs=0,
128
+ chunk_text_by_tokens=False,
129
+ max_tokens=0
130
+ ):
131
+ try:
132
+
133
+ # Perform transcription
134
+ audio_file_path, segments = perform_transcription(audio_file_path, offset, whisper_model, vad_filter)
135
+
136
+ if audio_file_path is None or segments is None:
137
+ logging.error("Process_Audio: Transcription failed or segments not available.")
138
+ return "Process_Audio: Transcription failed.", None, None, None, None, None
139
+
140
+ logging.debug(f"Process_Audio: Transcription audio_file: {audio_file_path}")
141
+ logging.debug(f"Process_Audio: Transcription segments: {segments}")
142
+
143
+ transcription_text = {'audio_file': audio_file_path, 'transcription': segments}
144
+ logging.debug(f"Process_Audio: Transcription text: {transcription_text}")
145
+
146
+ # Save segments to JSON
147
+ segments_json_path = save_segments_to_json(segments)
148
+
149
+ # Perform summarization
150
+ summary_text = None
151
+ if api_name:
152
+ if rolling_summarization is not None:
153
+ pass
154
+ # FIXME rolling summarization
155
+ # summary_text = rolling_summarize_function(
156
+ # transcription_text,
157
+ # detail=detail_level,
158
+ # api_name=api_name,
159
+ # api_key=api_key,
160
+ # custom_prompt=custom_prompt_input,
161
+ # chunk_by_words=chunk_text_by_words,
162
+ # max_words=max_words,
163
+ # chunk_by_sentences=chunk_text_by_sentences,
164
+ # max_sentences=max_sentences,
165
+ # chunk_by_paragraphs=chunk_text_by_paragraphs,
166
+ # max_paragraphs=max_paragraphs,
167
+ # chunk_by_tokens=chunk_text_by_tokens,
168
+ # max_tokens=max_tokens
169
+ # )
170
+ else:
171
+ summary_text = perform_summarization(api_name, segments_json_path, custom_prompt_input, api_key)
172
+
173
+ if summary_text is None:
174
+ logging.error("Summary text is None. Check summarization function.")
175
+ summary_file_path = None
176
+ else:
177
+ summary_text = 'Summary not available'
178
+ summary_file_path = None
179
+
180
+ # Save transcription and summary
181
+ download_path = create_download_directory("Audio_Processing")
182
+ json_file_path, summary_file_path = save_transcription_and_summary(transcription_text, summary_text,
183
+ download_path)
184
+
185
+ # Update function call to add_media_to_database so that it properly applies the title, author and file type
186
+ # Add to database
187
+ add_media_to_database(None, {'title': 'Audio File', 'author': 'Unknown'}, segments, summary_text, keywords,
188
+ custom_prompt_input, whisper_model)
189
+
190
+ return transcription_text, summary_text, json_file_path, summary_file_path, None, None
191
+
192
+ except Exception as e:
193
+ logging.error(f"Error in process_audio: {str(e)}")
194
+ return str(e), None, None, None, None, None
195
+
196
+
197
+ def process_single_audio(audio_file_path, whisper_model, api_name, api_key, keep_original,custom_keywords, source,
198
+ custom_prompt_input, chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
199
+ use_multi_level_chunking, chunk_language):
200
+ progress = []
201
+ transcription = ""
202
+ summary = ""
203
+
204
+ def update_progress(message):
205
+ progress.append(message)
206
+ return "\n".join(progress)
207
+
208
+ try:
209
+ # Check file size before processing
210
+ file_size = os.path.getsize(audio_file_path)
211
+ if file_size > MAX_FILE_SIZE:
212
+ update_progress(f"File size ({file_size / (1024 * 1024):.2f} MB) exceeds the maximum limit of {MAX_FILE_SIZE / (1024 * 1024):.2f} MB. Skipping this file.")
213
+ return "\n".join(progress), "", ""
214
+
215
+ # Perform transcription
216
+ update_progress("Starting transcription...")
217
+ segments = speech_to_text(audio_file_path, whisper_model=whisper_model)
218
+ transcription = " ".join([segment['Text'] for segment in segments])
219
+ update_progress("Audio transcribed successfully.")
220
+
221
+ # Perform summarization if API is provided
222
+ if api_name and api_key:
223
+ update_progress("Starting summarization...")
224
+ summary = perform_summarization(api_name, transcription, "Summarize the following audio transcript",
225
+ api_key)
226
+ update_progress("Audio summarized successfully.")
227
+ else:
228
+ summary = "No summary available"
229
+
230
+ # Prepare keywords
231
+ keywords = "audio,transcription"
232
+ if custom_keywords:
233
+ keywords += f",{custom_keywords}"
234
+
235
+ # Add to database
236
+ add_media_with_keywords(
237
+ url=source,
238
+ title=os.path.basename(audio_file_path),
239
+ media_type='audio',
240
+ content=transcription,
241
+ keywords=keywords,
242
+ prompt="Summarize the following audio transcript",
243
+ summary=summary,
244
+ transcription_model=whisper_model,
245
+ author="Unknown",
246
+ ingestion_date=None # This will use the current date
247
+ )
248
+ update_progress("Audio file added to database successfully.")
249
+
250
+ if not keep_original and source != "Uploaded File":
251
+ os.remove(audio_file_path)
252
+ update_progress(f"Temporary file {audio_file_path} removed.")
253
+ elif keep_original and source != "Uploaded File":
254
+ update_progress(f"Original audio file kept at: {audio_file_path}")
255
+
256
+ except Exception as e:
257
+ update_progress(f"Error processing {source}: {str(e)}")
258
+ transcription = f"Error: {str(e)}"
259
+ summary = "No summary due to error"
260
+
261
+ return "\n".join(progress), transcription, summary
262
+
263
+
264
+ def process_audio_files(audio_urls, audio_file, whisper_model, api_name, api_key, use_cookies, cookies, keep_original,
265
+ custom_keywords, custom_prompt_input, chunk_method, max_chunk_size, chunk_overlap,
266
+ use_adaptive_chunking, use_multi_level_chunking, chunk_language, diarize):
267
+ progress = []
268
+ temp_files = []
269
+ all_transcriptions = []
270
+ all_summaries = []
271
+
272
+ def update_progress(message):
273
+ progress.append(message)
274
+ return "\n".join(progress)
275
+
276
+ def cleanup_files():
277
+ for file in temp_files:
278
+ try:
279
+ if os.path.exists(file):
280
+ os.remove(file)
281
+ update_progress(f"Temporary file {file} removed.")
282
+ except Exception as e:
283
+ update_progress(f"Failed to remove temporary file {file}: {str(e)}")
284
+
285
+ def reencode_mp3(mp3_file_path):
286
+ try:
287
+ reencoded_mp3_path = mp3_file_path.replace(".mp3", "_reencoded.mp3")
288
+ subprocess.run([ffmpeg_cmd, '-i', mp3_file_path, '-codec:a', 'libmp3lame', reencoded_mp3_path], check=True)
289
+ update_progress(f"Re-encoded {mp3_file_path} to {reencoded_mp3_path}.")
290
+ return reencoded_mp3_path
291
+ except subprocess.CalledProcessError as e:
292
+ update_progress(f"Error re-encoding {mp3_file_path}: {str(e)}")
293
+ raise
294
+
295
+ def convert_mp3_to_wav(mp3_file_path):
296
+ try:
297
+ wav_file_path = mp3_file_path.replace(".mp3", ".wav")
298
+ subprocess.run([ffmpeg_cmd, '-i', mp3_file_path, wav_file_path], check=True)
299
+ update_progress(f"Converted {mp3_file_path} to {wav_file_path}.")
300
+ return wav_file_path
301
+ except subprocess.CalledProcessError as e:
302
+ update_progress(f"Error converting {mp3_file_path} to WAV: {str(e)}")
303
+ raise
304
+
305
+ try:
306
+ # Check and set the ffmpeg command
307
+ global ffmpeg_cmd
308
+ if os.name == "nt":
309
+ logging.debug("Running on Windows")
310
+ ffmpeg_cmd = os.path.join(os.getcwd(), "Bin", "ffmpeg.exe")
311
+ else:
312
+ ffmpeg_cmd = 'ffmpeg' # Assume 'ffmpeg' is in PATH for non-Windows systems
313
+
314
+ # Ensure ffmpeg is accessible
315
+ if not os.path.exists(ffmpeg_cmd) and os.name == "nt":
316
+ raise FileNotFoundError(f"ffmpeg executable not found at path: {ffmpeg_cmd}")
317
+
318
+ # Define chunk options early to avoid undefined errors
319
+ chunk_options = {
320
+ 'method': chunk_method,
321
+ 'max_size': max_chunk_size,
322
+ 'overlap': chunk_overlap,
323
+ 'adaptive': use_adaptive_chunking,
324
+ 'multi_level': use_multi_level_chunking,
325
+ 'language': chunk_language
326
+ }
327
+
328
+ # Process multiple URLs
329
+ urls = [url.strip() for url in audio_urls.split('\n') if url.strip()]
330
+
331
+ for i, url in enumerate(urls):
332
+ update_progress(f"Processing URL {i + 1}/{len(urls)}: {url}")
333
+
334
+ # Download and process audio file
335
+ audio_file_path = download_audio_file(url, use_cookies, cookies)
336
+ if not os.path.exists(audio_file_path):
337
+ update_progress(f"Downloaded file not found: {audio_file_path}")
338
+ continue
339
+
340
+ temp_files.append(audio_file_path)
341
+ update_progress("Audio file downloaded successfully.")
342
+
343
+ # Re-encode MP3 to fix potential issues
344
+ reencoded_mp3_path = reencode_mp3(audio_file_path)
345
+ if not os.path.exists(reencoded_mp3_path):
346
+ update_progress(f"Re-encoded file not found: {reencoded_mp3_path}")
347
+ continue
348
+
349
+ temp_files.append(reencoded_mp3_path)
350
+
351
+ # Convert re-encoded MP3 to WAV
352
+ wav_file_path = convert_mp3_to_wav(reencoded_mp3_path)
353
+ if not os.path.exists(wav_file_path):
354
+ update_progress(f"Converted WAV file not found: {wav_file_path}")
355
+ continue
356
+
357
+ temp_files.append(wav_file_path)
358
+
359
+ # Initialize transcription
360
+ transcription = ""
361
+
362
+ # Transcribe audio
363
+ if diarize:
364
+ segments = speech_to_text(wav_file_path, whisper_model=whisper_model, diarize=True)
365
+ else:
366
+ segments = speech_to_text(wav_file_path, whisper_model=whisper_model)
367
+
368
+ # Handle segments nested under 'segments' key
369
+ if isinstance(segments, dict) and 'segments' in segments:
370
+ segments = segments['segments']
371
+
372
+ if isinstance(segments, list):
373
+ transcription = " ".join([segment.get('Text', '') for segment in segments])
374
+ update_progress("Audio transcribed successfully.")
375
+ else:
376
+ update_progress("Unexpected segments format received from speech_to_text.")
377
+ logging.error(f"Unexpected segments format: {segments}")
378
+ continue
379
+
380
+ if not transcription.strip():
381
+ update_progress("Transcription is empty.")
382
+ else:
383
+ # Apply chunking
384
+ chunked_text = improved_chunking_process(transcription, chunk_options)
385
+
386
+ # Summarize
387
+ if api_name:
388
+ try:
389
+ summary = perform_summarization(api_name, chunked_text, custom_prompt_input, api_key)
390
+ update_progress("Audio summarized successfully.")
391
+ except Exception as e:
392
+ logging.error(f"Error during summarization: {str(e)}")
393
+ summary = "Summary generation failed"
394
+ else:
395
+ summary = "No summary available (API not provided)"
396
+
397
+ all_transcriptions.append(transcription)
398
+ all_summaries.append(summary)
399
+
400
+ # Add to database
401
+ add_media_with_keywords(
402
+ url=url,
403
+ title=os.path.basename(wav_file_path),
404
+ media_type='audio',
405
+ content=transcription,
406
+ keywords=custom_keywords,
407
+ prompt=custom_prompt_input,
408
+ summary=summary,
409
+ transcription_model=whisper_model,
410
+ author="Unknown",
411
+ ingestion_date=datetime.now().strftime('%Y-%m-%d')
412
+ )
413
+ update_progress("Audio file processed and added to database.")
414
+
415
+ # Process uploaded file if provided
416
+ if audio_file:
417
+ if os.path.getsize(audio_file.name) > MAX_FILE_SIZE:
418
+ update_progress(
419
+ f"Uploaded file size exceeds the maximum limit of {MAX_FILE_SIZE / (1024 * 1024):.2f}MB. Skipping this file.")
420
+ else:
421
+ # Re-encode MP3 to fix potential issues
422
+ reencoded_mp3_path = reencode_mp3(audio_file.name)
423
+ if not os.path.exists(reencoded_mp3_path):
424
+ update_progress(f"Re-encoded file not found: {reencoded_mp3_path}")
425
+ return update_progress("Processing failed: Re-encoded file not found"), "", ""
426
+
427
+ temp_files.append(reencoded_mp3_path)
428
+
429
+ # Convert re-encoded MP3 to WAV
430
+ wav_file_path = convert_mp3_to_wav(reencoded_mp3_path)
431
+ if not os.path.exists(wav_file_path):
432
+ update_progress(f"Converted WAV file not found: {wav_file_path}")
433
+ return update_progress("Processing failed: Converted WAV file not found"), "", ""
434
+
435
+ temp_files.append(wav_file_path)
436
+
437
+ # Initialize transcription
438
+ transcription = ""
439
+
440
+ if diarize:
441
+ segments = speech_to_text(wav_file_path, whisper_model=whisper_model, diarize=True)
442
+ else:
443
+ segments = speech_to_text(wav_file_path, whisper_model=whisper_model)
444
+
445
+ # Handle segments nested under 'segments' key
446
+ if isinstance(segments, dict) and 'segments' in segments:
447
+ segments = segments['segments']
448
+
449
+ if isinstance(segments, list):
450
+ transcription = " ".join([segment.get('Text', '') for segment in segments])
451
+ else:
452
+ update_progress("Unexpected segments format received from speech_to_text.")
453
+ logging.error(f"Unexpected segments format: {segments}")
454
+
455
+ chunked_text = improved_chunking_process(transcription, chunk_options)
456
+
457
+ if api_name and api_key:
458
+ try:
459
+ summary = perform_summarization(api_name, chunked_text, custom_prompt_input, api_key)
460
+ update_progress("Audio summarized successfully.")
461
+ except Exception as e:
462
+ logging.error(f"Error during summarization: {str(e)}")
463
+ summary = "Summary generation failed"
464
+ else:
465
+ summary = "No summary available (API not provided)"
466
+
467
+ all_transcriptions.append(transcription)
468
+ all_summaries.append(summary)
469
+
470
+ add_media_with_keywords(
471
+ url="Uploaded File",
472
+ title=os.path.basename(wav_file_path),
473
+ media_type='audio',
474
+ content=transcription,
475
+ keywords=custom_keywords,
476
+ prompt=custom_prompt_input,
477
+ summary=summary,
478
+ transcription_model=whisper_model,
479
+ author="Unknown",
480
+ ingestion_date=datetime.now().strftime('%Y-%m-%d')
481
+ )
482
+ update_progress("Uploaded file processed and added to database.")
483
+
484
+ # Final cleanup
485
+ if not keep_original:
486
+ cleanup_files()
487
+
488
+ final_progress = update_progress("All processing complete.")
489
+ final_transcriptions = "\n\n".join(all_transcriptions)
490
+ final_summaries = "\n\n".join(all_summaries)
491
+
492
+ return final_progress, final_transcriptions, final_summaries
493
+
494
+ except Exception as e:
495
+ logging.error(f"Error processing audio files: {str(e)}")
496
+ cleanup_files()
497
+ return update_progress(f"Processing failed: {str(e)}"), "", ""
498
+
499
+
500
+ def download_youtube_audio(url):
501
+ try:
502
+ # Determine ffmpeg path based on the operating system.
503
+ ffmpeg_path = './Bin/ffmpeg.exe' if os.name == 'nt' else 'ffmpeg'
504
+
505
+ # Create a temporary directory
506
+ with tempfile.TemporaryDirectory() as temp_dir:
507
+ # Extract information about the video
508
+ with yt_dlp.YoutubeDL({'quiet': True}) as ydl:
509
+ info_dict = ydl.extract_info(url, download=False)
510
+ sanitized_title = sanitize_filename(info_dict['title'])
511
+
512
+ # Setup the temporary filenames
513
+ temp_video_path = Path(temp_dir) / f"{sanitized_title}_temp.mp4"
514
+ temp_audio_path = Path(temp_dir) / f"{sanitized_title}.mp3"
515
+
516
+ # Initialize yt-dlp with options for downloading
517
+ ydl_opts = {
518
+ 'format': 'bestaudio[ext=m4a]/best[height<=480]', # Prefer best audio, or video up to 480p
519
+ 'ffmpeg_location': ffmpeg_path,
520
+ 'outtmpl': str(temp_video_path),
521
+ 'noplaylist': True,
522
+ 'quiet': True
523
+ }
524
+
525
+ # Execute yt-dlp to download the video/audio
526
+ with yt_dlp.YoutubeDL(ydl_opts) as ydl:
527
+ ydl.download([url])
528
+
529
+ # Check if the file exists
530
+ if not temp_video_path.exists():
531
+ raise FileNotFoundError(f"Expected file was not found: {temp_video_path}")
532
+
533
+ # Use ffmpeg to extract audio
534
+ ffmpeg_command = [
535
+ ffmpeg_path,
536
+ '-i', str(temp_video_path),
537
+ '-vn', # No video
538
+ '-acodec', 'libmp3lame',
539
+ '-b:a', '192k',
540
+ str(temp_audio_path)
541
+ ]
542
+ subprocess.run(ffmpeg_command, check=True, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
543
+
544
+ # Check if the audio file was created
545
+ if not temp_audio_path.exists():
546
+ raise FileNotFoundError(f"Expected audio file was not found: {temp_audio_path}")
547
+
548
+ # Create a persistent directory for the download if it doesn't exist
549
+ persistent_dir = Path("downloads")
550
+ persistent_dir.mkdir(exist_ok=True)
551
+
552
+ # Move the file from the temporary directory to the persistent directory
553
+ persistent_file_path = persistent_dir / f"{sanitized_title}.mp3"
554
+ os.replace(str(temp_audio_path), str(persistent_file_path))
555
+
556
+ # Add the file to the list of downloaded files
557
+ downloaded_files.append(str(persistent_file_path))
558
+
559
+ return str(persistent_file_path), f"Audio downloaded successfully: {sanitized_title}.mp3"
560
+ except Exception as e:
561
+ return None, f"Error downloading audio: {str(e)}"
562
+
563
+
564
+ def process_podcast(url, title, author, keywords, custom_prompt, api_name, api_key, whisper_model,
565
+ keep_original=False, enable_diarization=False, use_cookies=False, cookies=None,
566
+ chunk_method=None, max_chunk_size=300, chunk_overlap=0, use_adaptive_chunking=False,
567
+ use_multi_level_chunking=False, chunk_language='english'):
568
+ progress = []
569
+ error_message = ""
570
+ temp_files = []
571
+
572
+ def update_progress(message):
573
+ progress.append(message)
574
+ return "\n".join(progress)
575
+
576
+ def cleanup_files():
577
+ if not keep_original:
578
+ for file in temp_files:
579
+ try:
580
+ if os.path.exists(file):
581
+ os.remove(file)
582
+ update_progress(f"Temporary file {file} removed.")
583
+ except Exception as e:
584
+ update_progress(f"Failed to remove temporary file {file}: {str(e)}")
585
+
586
+ try:
587
+ # Download podcast
588
+ audio_file = download_audio_file(url, use_cookies, cookies)
589
+ temp_files.append(audio_file)
590
+ update_progress("Podcast downloaded successfully.")
591
+
592
+ # Extract metadata
593
+ metadata = extract_metadata(url)
594
+ title = title or metadata.get('title', 'Unknown Podcast')
595
+ author = author or metadata.get('uploader', 'Unknown Author')
596
+
597
+ # Format metadata for storage
598
+ metadata_text = f"""
599
+ Metadata:
600
+ Title: {title}
601
+ Author: {author}
602
+ Series: {metadata.get('series', 'N/A')}
603
+ Episode: {metadata.get('episode', 'N/A')}
604
+ Season: {metadata.get('season', 'N/A')}
605
+ Upload Date: {metadata.get('upload_date', 'N/A')}
606
+ Duration: {metadata.get('duration', 'N/A')} seconds
607
+ Description: {metadata.get('description', 'N/A')}
608
+ """
609
+
610
+ # Update keywords
611
+ new_keywords = []
612
+ if metadata.get('series'):
613
+ new_keywords.append(f"series:{metadata['series']}")
614
+ if metadata.get('episode'):
615
+ new_keywords.append(f"episode:{metadata['episode']}")
616
+ if metadata.get('season'):
617
+ new_keywords.append(f"season:{metadata['season']}")
618
+
619
+ keywords = f"{keywords},{','.join(new_keywords)}" if keywords else ','.join(new_keywords)
620
+
621
+ update_progress(f"Metadata extracted - Title: {title}, Author: {author}, Keywords: {keywords}")
622
+
623
+ # Transcribe the podcast
624
+ try:
625
+ if enable_diarization:
626
+ segments = speech_to_text(audio_file, whisper_model=whisper_model, diarize=True)
627
+ else:
628
+ segments = speech_to_text(audio_file, whisper_model=whisper_model)
629
+ transcription = " ".join([segment['Text'] for segment in segments])
630
+ update_progress("Podcast transcribed successfully.")
631
+ except Exception as e:
632
+ error_message = f"Transcription failed: {str(e)}"
633
+ raise
634
+
635
+ # Apply chunking
636
+ chunk_options = {
637
+ 'method': chunk_method,
638
+ 'max_size': max_chunk_size,
639
+ 'overlap': chunk_overlap,
640
+ 'adaptive': use_adaptive_chunking,
641
+ 'multi_level': use_multi_level_chunking,
642
+ 'language': chunk_language
643
+ }
644
+ chunked_text = improved_chunking_process(transcription, chunk_options)
645
+
646
+ # Combine metadata and transcription
647
+ full_content = metadata_text + "\n\nTranscription:\n" + transcription
648
+
649
+ # Summarize if API is provided
650
+ summary = None
651
+ if api_name and api_key:
652
+ try:
653
+ summary = perform_summarization(api_name, chunked_text, custom_prompt, api_key)
654
+ update_progress("Podcast summarized successfully.")
655
+ except Exception as e:
656
+ error_message = f"Summarization failed: {str(e)}"
657
+ raise
658
+
659
+ # Add to database
660
+ try:
661
+ add_media_with_keywords(
662
+ url=url,
663
+ title=title,
664
+ media_type='podcast',
665
+ content=full_content,
666
+ keywords=keywords,
667
+ prompt=custom_prompt,
668
+ summary=summary or "No summary available",
669
+ transcription_model=whisper_model,
670
+ author=author,
671
+ ingestion_date=datetime.now().strftime('%Y-%m-%d')
672
+ )
673
+ update_progress("Podcast added to database successfully.")
674
+ except Exception as e:
675
+ error_message = f"Error adding podcast to database: {str(e)}"
676
+ raise
677
+
678
+ # Cleanup
679
+ cleanup_files()
680
+
681
+ return (update_progress("Processing complete."), full_content, summary or "No summary generated.",
682
+ title, author, keywords, error_message)
683
+
684
+ except Exception as e:
685
+ logging.error(f"Error processing podcast: {str(e)}")
686
+ cleanup_files()
687
+ return update_progress(f"Processing failed: {str(e)}"), "", "", "", "", "", str(e)
688
+
689
+
690
+ #
691
+ #
692
  #######################################################################################################################
App_Function_Libraries/Audio_Transcription_Lib.py CHANGED
@@ -1,192 +1,192 @@
1
- # Audio_Transcription_Lib.py
2
- #########################################
3
- # Transcription Library
4
- # This library is used to perform transcription of audio files.
5
- # Currently, uses faster_whisper for transcription.
6
- #
7
- ####
8
- import configparser
9
- ####################
10
- # Function List
11
- #
12
- # 1. convert_to_wav(video_file_path, offset=0, overwrite=False)
13
- # 2. speech_to_text(audio_file_path, selected_source_lang='en', whisper_model='small.en', vad_filter=False)
14
- #
15
- ####################
16
- #
17
- # Import necessary libraries to run solo for testing
18
- import gc
19
- import json
20
- import logging
21
- import os
22
- import sys
23
- import subprocess
24
- import time
25
-
26
- # DEBUG Imports
27
- #from memory_profiler import profile
28
-
29
- # Import Local
30
- #
31
- #######################################################################################################################
32
- # Function Definitions
33
- #
34
-
35
- # Convert video .m4a into .wav using ffmpeg
36
- # ffmpeg -i "example.mp4" -ar 16000 -ac 1 -c:a pcm_s16le "output.wav"
37
- # https://www.gyan.dev/ffmpeg/builds/
38
- #
39
-
40
-
41
- whisper_model_instance = None
42
- # Retrieve processing choice from the configuration file
43
- config = configparser.ConfigParser()
44
- config.read('config.txt')
45
- processing_choice = config.get('Processing', 'processing_choice', fallback='cpu')
46
-
47
-
48
- # FIXME: This is a temporary solution.
49
- # This doesn't clear older models, which means potentially a lot of memory is being used...
50
- def get_whisper_model(model_name, device):
51
- global whisper_model_instance
52
- if whisper_model_instance is None:
53
- from faster_whisper import WhisperModel
54
- logging.info(f"Initializing new WhisperModel with size {model_name} on device {device}")
55
- whisper_model_instance = WhisperModel(model_name, device=device)
56
- return whisper_model_instance
57
-
58
-
59
- # os.system(r'.\Bin\ffmpeg.exe -ss 00:00:00 -i "{video_file_path}" -ar 16000 -ac 1 -c:a pcm_s16le "{out_path}"')
60
- #DEBUG
61
- #@profile
62
- def convert_to_wav(video_file_path, offset=0, overwrite=False):
63
- out_path = os.path.splitext(video_file_path)[0] + ".wav"
64
-
65
- if os.path.exists(out_path) and not overwrite:
66
- print(f"File '{out_path}' already exists. Skipping conversion.")
67
- logging.info(f"Skipping conversion as file already exists: {out_path}")
68
- return out_path
69
- print("Starting conversion process of .m4a to .WAV")
70
- out_path = os.path.splitext(video_file_path)[0] + ".wav"
71
-
72
- try:
73
- if os.name == "nt":
74
- logging.debug("ffmpeg being ran on windows")
75
-
76
- if sys.platform.startswith('win'):
77
- ffmpeg_cmd = ".\\Bin\\ffmpeg.exe"
78
- logging.debug(f"ffmpeg_cmd: {ffmpeg_cmd}")
79
- else:
80
- ffmpeg_cmd = 'ffmpeg' # Assume 'ffmpeg' is in PATH for non-Windows systems
81
-
82
- command = [
83
- ffmpeg_cmd, # Assuming the working directory is correctly set where .\Bin exists
84
- "-ss", "00:00:00", # Start at the beginning of the video
85
- "-i", video_file_path,
86
- "-ar", "16000", # Audio sample rate
87
- "-ac", "1", # Number of audio channels
88
- "-c:a", "pcm_s16le", # Audio codec
89
- out_path
90
- ]
91
- try:
92
- # Redirect stdin from null device to prevent ffmpeg from waiting for input
93
- with open(os.devnull, 'rb') as null_file:
94
- result = subprocess.run(command, stdin=null_file, text=True, capture_output=True)
95
- if result.returncode == 0:
96
- logging.info("FFmpeg executed successfully")
97
- logging.debug("FFmpeg output: %s", result.stdout)
98
- else:
99
- logging.error("Error in running FFmpeg")
100
- logging.error("FFmpeg stderr: %s", result.stderr)
101
- raise RuntimeError(f"FFmpeg error: {result.stderr}")
102
- except Exception as e:
103
- logging.error("Error occurred - ffmpeg doesn't like windows")
104
- raise RuntimeError("ffmpeg failed")
105
- elif os.name == "posix":
106
- os.system(f'ffmpeg -ss 00:00:00 -i "{video_file_path}" -ar 16000 -ac 1 -c:a pcm_s16le "{out_path}"')
107
- else:
108
- raise RuntimeError("Unsupported operating system")
109
- logging.info("Conversion to WAV completed: %s", out_path)
110
- except subprocess.CalledProcessError as e:
111
- logging.error("Error executing FFmpeg command: %s", str(e))
112
- raise RuntimeError("Error converting video file to WAV")
113
- except Exception as e:
114
- logging.error("speech-to-text: Error transcribing audio: %s", str(e))
115
- return {"error": str(e)}
116
- gc.collect()
117
- return out_path
118
-
119
-
120
- # Transcribe .wav into .segments.json
121
- #DEBUG
122
- #@profile
123
- def speech_to_text(audio_file_path, selected_source_lang='en', whisper_model='medium.en', vad_filter=False, diarize=False):
124
- global whisper_model_instance, processing_choice
125
- logging.info('speech-to-text: Loading faster_whisper model: %s', whisper_model)
126
-
127
- time_start = time.time()
128
- if audio_file_path is None:
129
- raise ValueError("speech-to-text: No audio file provided")
130
- logging.info("speech-to-text: Audio file path: %s", audio_file_path)
131
-
132
- try:
133
- _, file_ending = os.path.splitext(audio_file_path)
134
- out_file = audio_file_path.replace(file_ending, ".segments.json")
135
- prettified_out_file = audio_file_path.replace(file_ending, ".segments_pretty.json")
136
- if os.path.exists(out_file):
137
- logging.info("speech-to-text: Segments file already exists: %s", out_file)
138
- with open(out_file) as f:
139
- global segments
140
- segments = json.load(f)
141
- return segments
142
-
143
- logging.info('speech-to-text: Starting transcription...')
144
- options = dict(language=selected_source_lang, beam_size=5, best_of=5, vad_filter=vad_filter)
145
- transcribe_options = dict(task="transcribe", **options)
146
- # use function and config at top of file
147
- whisper_model_instance = get_whisper_model(whisper_model, processing_choice)
148
- segments_raw, info = whisper_model_instance.transcribe(audio_file_path, **transcribe_options)
149
-
150
- segments = []
151
- for segment_chunk in segments_raw:
152
- chunk = {
153
- "Time_Start": segment_chunk.start,
154
- "Time_End": segment_chunk.end,
155
- "Text": segment_chunk.text
156
- }
157
- logging.debug("Segment: %s", chunk)
158
- segments.append(chunk)
159
-
160
- if segments:
161
- segments[0]["Text"] = f"This text was transcribed using whisper model: {whisper_model}\n\n" + segments[0]["Text"]
162
-
163
- if not segments:
164
- raise RuntimeError("No transcription produced. The audio file may be invalid or empty.")
165
- logging.info("speech-to-text: Transcription completed in %.2f seconds", time.time() - time_start)
166
-
167
- # Save the segments to a JSON file - prettified and non-prettified
168
- # FIXME so this is an optional flag to save either the prettified json file or the normal one
169
- save_json = True
170
- if save_json:
171
- logging.info("speech-to-text: Saving segments to JSON file")
172
- output_data = {'segments': segments}
173
-
174
- logging.info("speech-to-text: Saving prettified JSON to %s", prettified_out_file)
175
- with open(prettified_out_file, 'w') as f:
176
- json.dump(output_data, f, indent=2)
177
-
178
- logging.info("speech-to-text: Saving JSON to %s", out_file)
179
- with open(out_file, 'w') as f:
180
- json.dump(output_data, f)
181
-
182
- logging.debug(f"speech-to-text: returning {segments[:500]}")
183
- gc.collect()
184
- return segments
185
-
186
- except Exception as e:
187
- logging.error("speech-to-text: Error transcribing audio: %s", str(e))
188
- raise RuntimeError("speech-to-text: Error transcribing audio")
189
-
190
- #
191
- #
192
  #######################################################################################################################
 
1
+ # Audio_Transcription_Lib.py
2
+ #########################################
3
+ # Transcription Library
4
+ # This library is used to perform transcription of audio files.
5
+ # Currently, uses faster_whisper for transcription.
6
+ #
7
+ ####
8
+ import configparser
9
+ ####################
10
+ # Function List
11
+ #
12
+ # 1. convert_to_wav(video_file_path, offset=0, overwrite=False)
13
+ # 2. speech_to_text(audio_file_path, selected_source_lang='en', whisper_model='small.en', vad_filter=False)
14
+ #
15
+ ####################
16
+ #
17
+ # Import necessary libraries to run solo for testing
18
+ import gc
19
+ import json
20
+ import logging
21
+ import os
22
+ import sys
23
+ import subprocess
24
+ import time
25
+
26
+ # DEBUG Imports
27
+ #from memory_profiler import profile
28
+
29
+ # Import Local
30
+ #
31
+ #######################################################################################################################
32
+ # Function Definitions
33
+ #
34
+
35
+ # Convert video .m4a into .wav using ffmpeg
36
+ # ffmpeg -i "example.mp4" -ar 16000 -ac 1 -c:a pcm_s16le "output.wav"
37
+ # https://www.gyan.dev/ffmpeg/builds/
38
+ #
39
+
40
+
41
+ whisper_model_instance = None
42
+ # Retrieve processing choice from the configuration file
43
+ config = configparser.ConfigParser()
44
+ config.read('config.txt')
45
+ processing_choice = config.get('Processing', 'processing_choice', fallback='cpu')
46
+
47
+
48
+ # FIXME: This is a temporary solution.
49
+ # This doesn't clear older models, which means potentially a lot of memory is being used...
50
+ def get_whisper_model(model_name, device):
51
+ global whisper_model_instance
52
+ if whisper_model_instance is None:
53
+ from faster_whisper import WhisperModel
54
+ logging.info(f"Initializing new WhisperModel with size {model_name} on device {device}")
55
+ whisper_model_instance = WhisperModel(model_name, device=device)
56
+ return whisper_model_instance
57
+
58
+
59
+ # os.system(r'.\Bin\ffmpeg.exe -ss 00:00:00 -i "{video_file_path}" -ar 16000 -ac 1 -c:a pcm_s16le "{out_path}"')
60
+ #DEBUG
61
+ #@profile
62
+ def convert_to_wav(video_file_path, offset=0, overwrite=False):
63
+ out_path = os.path.splitext(video_file_path)[0] + ".wav"
64
+
65
+ if os.path.exists(out_path) and not overwrite:
66
+ print(f"File '{out_path}' already exists. Skipping conversion.")
67
+ logging.info(f"Skipping conversion as file already exists: {out_path}")
68
+ return out_path
69
+ print("Starting conversion process of .m4a to .WAV")
70
+ out_path = os.path.splitext(video_file_path)[0] + ".wav"
71
+
72
+ try:
73
+ if os.name == "nt":
74
+ logging.debug("ffmpeg being ran on windows")
75
+
76
+ if sys.platform.startswith('win'):
77
+ ffmpeg_cmd = ".\\Bin\\ffmpeg.exe"
78
+ logging.debug(f"ffmpeg_cmd: {ffmpeg_cmd}")
79
+ else:
80
+ ffmpeg_cmd = 'ffmpeg' # Assume 'ffmpeg' is in PATH for non-Windows systems
81
+
82
+ command = [
83
+ ffmpeg_cmd, # Assuming the working directory is correctly set where .\Bin exists
84
+ "-ss", "00:00:00", # Start at the beginning of the video
85
+ "-i", video_file_path,
86
+ "-ar", "16000", # Audio sample rate
87
+ "-ac", "1", # Number of audio channels
88
+ "-c:a", "pcm_s16le", # Audio codec
89
+ out_path
90
+ ]
91
+ try:
92
+ # Redirect stdin from null device to prevent ffmpeg from waiting for input
93
+ with open(os.devnull, 'rb') as null_file:
94
+ result = subprocess.run(command, stdin=null_file, text=True, capture_output=True)
95
+ if result.returncode == 0:
96
+ logging.info("FFmpeg executed successfully")
97
+ logging.debug("FFmpeg output: %s", result.stdout)
98
+ else:
99
+ logging.error("Error in running FFmpeg")
100
+ logging.error("FFmpeg stderr: %s", result.stderr)
101
+ raise RuntimeError(f"FFmpeg error: {result.stderr}")
102
+ except Exception as e:
103
+ logging.error("Error occurred - ffmpeg doesn't like windows")
104
+ raise RuntimeError("ffmpeg failed")
105
+ elif os.name == "posix":
106
+ os.system(f'ffmpeg -ss 00:00:00 -i "{video_file_path}" -ar 16000 -ac 1 -c:a pcm_s16le "{out_path}"')
107
+ else:
108
+ raise RuntimeError("Unsupported operating system")
109
+ logging.info("Conversion to WAV completed: %s", out_path)
110
+ except subprocess.CalledProcessError as e:
111
+ logging.error("Error executing FFmpeg command: %s", str(e))
112
+ raise RuntimeError("Error converting video file to WAV")
113
+ except Exception as e:
114
+ logging.error("speech-to-text: Error transcribing audio: %s", str(e))
115
+ return {"error": str(e)}
116
+ gc.collect()
117
+ return out_path
118
+
119
+
120
+ # Transcribe .wav into .segments.json
121
+ #DEBUG
122
+ #@profile
123
+ def speech_to_text(audio_file_path, selected_source_lang='en', whisper_model='medium.en', vad_filter=False, diarize=False):
124
+ global whisper_model_instance, processing_choice
125
+ logging.info('speech-to-text: Loading faster_whisper model: %s', whisper_model)
126
+
127
+ time_start = time.time()
128
+ if audio_file_path is None:
129
+ raise ValueError("speech-to-text: No audio file provided")
130
+ logging.info("speech-to-text: Audio file path: %s", audio_file_path)
131
+
132
+ try:
133
+ _, file_ending = os.path.splitext(audio_file_path)
134
+ out_file = audio_file_path.replace(file_ending, ".segments.json")
135
+ prettified_out_file = audio_file_path.replace(file_ending, ".segments_pretty.json")
136
+ if os.path.exists(out_file):
137
+ logging.info("speech-to-text: Segments file already exists: %s", out_file)
138
+ with open(out_file) as f:
139
+ global segments
140
+ segments = json.load(f)
141
+ return segments
142
+
143
+ logging.info('speech-to-text: Starting transcription...')
144
+ options = dict(language=selected_source_lang, beam_size=5, best_of=5, vad_filter=vad_filter)
145
+ transcribe_options = dict(task="transcribe", **options)
146
+ # use function and config at top of file
147
+ whisper_model_instance = get_whisper_model(whisper_model, processing_choice)
148
+ segments_raw, info = whisper_model_instance.transcribe(audio_file_path, **transcribe_options)
149
+
150
+ segments = []
151
+ for segment_chunk in segments_raw:
152
+ chunk = {
153
+ "Time_Start": segment_chunk.start,
154
+ "Time_End": segment_chunk.end,
155
+ "Text": segment_chunk.text
156
+ }
157
+ logging.debug("Segment: %s", chunk)
158
+ segments.append(chunk)
159
+
160
+ if segments:
161
+ segments[0]["Text"] = f"This text was transcribed using whisper model: {whisper_model}\n\n" + segments[0]["Text"]
162
+
163
+ if not segments:
164
+ raise RuntimeError("No transcription produced. The audio file may be invalid or empty.")
165
+ logging.info("speech-to-text: Transcription completed in %.2f seconds", time.time() - time_start)
166
+
167
+ # Save the segments to a JSON file - prettified and non-prettified
168
+ # FIXME so this is an optional flag to save either the prettified json file or the normal one
169
+ save_json = True
170
+ if save_json:
171
+ logging.info("speech-to-text: Saving segments to JSON file")
172
+ output_data = {'segments': segments}
173
+
174
+ logging.info("speech-to-text: Saving prettified JSON to %s", prettified_out_file)
175
+ with open(prettified_out_file, 'w') as f:
176
+ json.dump(output_data, f, indent=2)
177
+
178
+ logging.info("speech-to-text: Saving JSON to %s", out_file)
179
+ with open(out_file, 'w') as f:
180
+ json.dump(output_data, f)
181
+
182
+ logging.debug(f"speech-to-text: returning {segments[:500]}")
183
+ gc.collect()
184
+ return segments
185
+
186
+ except Exception as e:
187
+ logging.error("speech-to-text: Error transcribing audio: %s", str(e))
188
+ raise RuntimeError("speech-to-text: Error transcribing audio")
189
+
190
+ #
191
+ #
192
  #######################################################################################################################
App_Function_Libraries/Chat.py ADDED
@@ -0,0 +1,273 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Chat.py
2
+ # Chat functions for interacting with the LLMs as chatbots
3
+
4
+ # Imports
5
+ import json
6
+ import logging
7
+ import os
8
+ import re
9
+ import tempfile
10
+ from datetime import datetime
11
+
12
+ from App_Function_Libraries.DB_Manager import get_conversation_name, save_chat_history_to_database
13
+ from App_Function_Libraries.LLM_API_Calls import chat_with_openai, chat_with_anthropic, chat_with_cohere, \
14
+ chat_with_groq, chat_with_openrouter, chat_with_deepseek, chat_with_mistral, chat_with_huggingface, chat_with_vllm
15
+ from App_Function_Libraries.LLM_API_Calls_Local import chat_with_aphrodite, chat_with_local_llm, chat_with_ollama, \
16
+ chat_with_kobold, chat_with_llama, chat_with_oobabooga, chat_with_tabbyapi
17
+ from App_Function_Libraries.SQLite_DB import load_media_content
18
+ from App_Function_Libraries.Utils import generate_unique_filename
19
+
20
+
21
+ #
22
+ # External Imports
23
+ #
24
+ # Local Imports
25
+ #
26
+
27
+ ####################################################################################################
28
+ def chat(message, history, media_content, selected_parts, api_endpoint, api_key, prompt, temperature,
29
+ system_message=None):
30
+ try:
31
+ logging.info(f"Debug - Chat Function - Message: {message}")
32
+ logging.info(f"Debug - Chat Function - Media Content: {media_content}")
33
+ logging.info(f"Debug - Chat Function - Selected Parts: {selected_parts}")
34
+ logging.info(f"Debug - Chat Function - API Endpoint: {api_endpoint}")
35
+ # logging.info(f"Debug - Chat Function - Prompt: {prompt}")
36
+
37
+ # Ensure selected_parts is a list
38
+ if not isinstance(selected_parts, (list, tuple)):
39
+ selected_parts = [selected_parts] if selected_parts else []
40
+
41
+ # logging.debug(f"Debug - Chat Function - Selected Parts (after check): {selected_parts}")
42
+
43
+ # Combine the selected parts of the media content
44
+ combined_content = "\n\n".join(
45
+ [f"{part.capitalize()}: {media_content.get(part, '')}" for part in selected_parts if part in media_content])
46
+ # Print first 500 chars
47
+ # logging.debug(f"Debug - Chat Function - Combined Content: {combined_content[:500]}...")
48
+
49
+ # Prepare the input for the API
50
+ if not history:
51
+ input_data = f"{combined_content}\n\nUser: {message}\n"
52
+ else:
53
+ input_data = f"User: {message}\n"
54
+ # Print first 500 chars
55
+ # logging.info(f"Debug - Chat Function - Input Data: {input_data[:500]}...")
56
+
57
+ if system_message:
58
+ print(f"System message: {system_message}")
59
+ logging.debug(f"Debug - Chat Function - System Message: {system_message}")
60
+ temperature = float(temperature) if temperature else 0.7
61
+ temp = temperature
62
+
63
+ logging.debug("Debug - Chat Function - Temperature: {temperature}")
64
+ logging.debug(f"Debug - Chat Function - API Key: {api_key[:10]}")
65
+ logging.debug(f"Debug - Chat Function - Prompt: {prompt}")
66
+
67
+ # Use the existing API request code based on the selected endpoint
68
+ logging.info(f"Debug - Chat Function - API Endpoint: {api_endpoint}")
69
+ if api_endpoint.lower() == 'openai':
70
+ response = chat_with_openai(api_key, input_data, prompt, temp, system_message)
71
+ elif api_endpoint.lower() == "anthropic":
72
+ response = chat_with_anthropic(api_key, input_data, prompt, temp, system_message)
73
+ elif api_endpoint.lower() == "cohere":
74
+ response = chat_with_cohere(api_key, input_data, prompt, temp, system_message)
75
+ elif api_endpoint.lower() == "groq":
76
+ response = chat_with_groq(api_key, input_data, prompt, temp, system_message)
77
+ elif api_endpoint.lower() == "openrouter":
78
+ response = chat_with_openrouter(api_key, input_data, prompt, temp, system_message)
79
+ elif api_endpoint.lower() == "deepseek":
80
+ response = chat_with_deepseek(api_key, input_data, prompt, temp, system_message)
81
+ elif api_endpoint.lower() == "mistral":
82
+ response = chat_with_mistral(api_key, input_data, prompt, temp, system_message)
83
+ elif api_endpoint.lower() == "llama.cpp":
84
+ response = chat_with_llama(input_data, prompt, temp, system_message)
85
+ elif api_endpoint.lower() == "kobold":
86
+ response = chat_with_kobold(input_data, api_key, prompt, temp, system_message)
87
+ elif api_endpoint.lower() == "ooba":
88
+ response = chat_with_oobabooga(input_data, api_key, prompt, temp, system_message)
89
+ elif api_endpoint.lower() == "tabbyapi":
90
+ response = chat_with_tabbyapi(input_data, prompt, temp, system_message)
91
+ elif api_endpoint.lower() == "vllm":
92
+ response = chat_with_vllm(input_data, prompt, system_message)
93
+ elif api_endpoint.lower() == "local-llm":
94
+ response = chat_with_local_llm(input_data, prompt, temp, system_message)
95
+ elif api_endpoint.lower() == "huggingface":
96
+ response = chat_with_huggingface(api_key, input_data, prompt, temp) # , system_message)
97
+ elif api_endpoint.lower() == "ollama":
98
+ response = chat_with_ollama(input_data, prompt, temp, system_message)
99
+ elif api_endpoint.lower() == "aphrodite":
100
+ response = chat_with_aphrodite(input_data, prompt, temp, system_message)
101
+ else:
102
+ raise ValueError(f"Unsupported API endpoint: {api_endpoint}")
103
+
104
+ return response
105
+
106
+ except Exception as e:
107
+ logging.error(f"Error in chat function: {str(e)}")
108
+ return f"An error occurred: {str(e)}"
109
+
110
+
111
+ def save_chat_history_to_db_wrapper(chatbot, conversation_id, media_content):
112
+ logging.info(f"Attempting to save chat history. Media content type: {type(media_content)}")
113
+ try:
114
+ # Extract the media_id and media_name from the media_content
115
+ media_id = None
116
+ media_name = None
117
+ if isinstance(media_content, dict):
118
+ logging.debug(f"Media content keys: {media_content.keys()}")
119
+ if 'content' in media_content:
120
+ try:
121
+ content = media_content['content']
122
+ if isinstance(content, str):
123
+ content_json = json.loads(content)
124
+ elif isinstance(content, dict):
125
+ content_json = content
126
+ else:
127
+ raise ValueError(f"Unexpected content type: {type(content)}")
128
+
129
+ # Use the webpage_url as the media_id
130
+ media_id = content_json.get('webpage_url')
131
+ # Use the title as the media_name
132
+ media_name = content_json.get('title')
133
+
134
+ logging.info(f"Extracted media_id: {media_id}, media_name: {media_name}")
135
+ except json.JSONDecodeError:
136
+ logging.error("Failed to decode JSON from media_content['content']")
137
+ except Exception as e:
138
+ logging.error(f"Error processing media_content: {str(e)}")
139
+ else:
140
+ logging.warning("'content' key not found in media_content")
141
+ else:
142
+ logging.warning(f"media_content is not a dictionary. Type: {type(media_content)}")
143
+
144
+ if media_id is None:
145
+ # If we couldn't find a media_id, we'll use a placeholder
146
+ media_id = "unknown_media"
147
+ logging.warning(f"Unable to extract media_id from media_content. Using placeholder: {media_id}")
148
+
149
+ if media_name is None:
150
+ media_name = "Unnamed Media"
151
+ logging.warning(f"Unable to extract media_name from media_content. Using placeholder: {media_name}")
152
+
153
+ # Generate a unique conversation name using media_id and current timestamp
154
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
155
+ conversation_name = f"Chat_{media_id}_{timestamp}"
156
+
157
+ new_conversation_id = save_chat_history_to_database(chatbot, conversation_id, media_id, media_name,
158
+ conversation_name)
159
+ return new_conversation_id, f"Chat history saved successfully as {conversation_name}!"
160
+ except Exception as e:
161
+ error_message = f"Failed to save chat history: {str(e)}"
162
+ logging.error(error_message, exc_info=True)
163
+ return conversation_id, error_message
164
+
165
+
166
+ def save_chat_history(history, conversation_id, media_content):
167
+ try:
168
+ content, conversation_name = generate_chat_history_content(history, conversation_id, media_content)
169
+
170
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
171
+ safe_conversation_name = re.sub(r'[^a-zA-Z0-9_-]', '_', conversation_name)
172
+ base_filename = f"{safe_conversation_name}_{timestamp}.json"
173
+
174
+ # Create a temporary file
175
+ with tempfile.NamedTemporaryFile(mode='w', delete=False, suffix='.json') as temp_file:
176
+ temp_file.write(content)
177
+ temp_file_path = temp_file.name
178
+
179
+ # Generate a unique filename
180
+ unique_filename = generate_unique_filename(os.path.dirname(temp_file_path), base_filename)
181
+ final_path = os.path.join(os.path.dirname(temp_file_path), unique_filename)
182
+
183
+ # Rename the temporary file to the unique filename
184
+ os.rename(temp_file_path, final_path)
185
+
186
+ return final_path
187
+ except Exception as e:
188
+ logging.error(f"Error saving chat history: {str(e)}")
189
+ return None
190
+
191
+
192
+ def generate_chat_history_content(history, conversation_id, media_content):
193
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
194
+
195
+ conversation_name = get_conversation_name(conversation_id)
196
+
197
+ if not conversation_name:
198
+ media_name = extract_media_name(media_content)
199
+ if media_name:
200
+ conversation_name = f"{media_name}-chat"
201
+ else:
202
+ conversation_name = f"chat-{timestamp}" # Fallback name
203
+
204
+ chat_data = {
205
+ "conversation_id": conversation_id,
206
+ "conversation_name": conversation_name,
207
+ "timestamp": timestamp,
208
+ "history": [
209
+ {
210
+ "role": "user" if i % 2 == 0 else "bot",
211
+ "content": msg[0] if isinstance(msg, tuple) else msg
212
+ }
213
+ for i, msg in enumerate(history)
214
+ ]
215
+ }
216
+
217
+ return json.dumps(chat_data, indent=2), conversation_name
218
+
219
+
220
+ def extract_media_name(media_content):
221
+ if isinstance(media_content, dict):
222
+ content = media_content.get('content', {})
223
+ if isinstance(content, str):
224
+ try:
225
+ content = json.loads(content)
226
+ except json.JSONDecodeError:
227
+ logging.warning("Failed to parse media_content JSON string")
228
+ return None
229
+
230
+ # Try to extract title from the content
231
+ if isinstance(content, dict):
232
+ return content.get('title') or content.get('name')
233
+
234
+ logging.warning(f"Unexpected media_content format: {type(media_content)}")
235
+ return None
236
+
237
+
238
+ def update_chat_content(selected_item, use_content, use_summary, use_prompt, item_mapping):
239
+ logging.debug(f"Debug - Update Chat Content - Selected Item: {selected_item}\n")
240
+ logging.debug(f"Debug - Update Chat Content - Use Content: {use_content}\n\n\n\n")
241
+ logging.debug(f"Debug - Update Chat Content - Use Summary: {use_summary}\n\n")
242
+ logging.debug(f"Debug - Update Chat Content - Use Prompt: {use_prompt}\n\n")
243
+ logging.debug(f"Debug - Update Chat Content - Item Mapping: {item_mapping}\n\n")
244
+
245
+ if selected_item and selected_item in item_mapping:
246
+ media_id = item_mapping[selected_item]
247
+ content = load_media_content(media_id)
248
+ selected_parts = []
249
+ if use_content and "content" in content:
250
+ selected_parts.append("content")
251
+ if use_summary and "summary" in content:
252
+ selected_parts.append("summary")
253
+ if use_prompt and "prompt" in content:
254
+ selected_parts.append("prompt")
255
+
256
+ # Modified debug print
257
+ if isinstance(content, dict):
258
+ print(f"Debug - Update Chat Content - Content keys: {list(content.keys())}")
259
+ for key, value in content.items():
260
+ print(f"Debug - Update Chat Content - {key} (first 500 char): {str(value)[:500]}\n\n\n\n")
261
+ else:
262
+ print(f"Debug - Update Chat Content - Content(first 500 char): {str(content)[:500]}\n\n\n\n")
263
+
264
+ print(f"Debug - Update Chat Content - Selected Parts: {selected_parts}")
265
+ return content, selected_parts
266
+ else:
267
+ print(f"Debug - Update Chat Content - No item selected or item not in mapping")
268
+ return {}, []
269
+
270
+
271
+ #
272
+ # End of Chat.py
273
+ ##########################################################################################################################
App_Function_Libraries/Chat_related_functions.py ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Chat_related_functions.py
2
+ # Contains functions related to chat
3
+ # WIP.
4
+ #
5
+ # Importing required libraries
6
+ import json
7
+ import os
8
+ from pathlib import Path
9
+ import json
10
+ #
11
+ ########################################################################################################################
12
+ # Set globals
13
+ CHARACTERS_FILE = Path('.', 'Helper_Scripts', 'Character_Cards', 'Characters.json')
14
+
15
+ def save_character(character_data):
16
+ if CHARACTERS_FILE.exists():
17
+ with CHARACTERS_FILE.open('r') as f:
18
+ characters = json.load(f)
19
+ else:
20
+ characters = {}
21
+
22
+ characters[character_data['name']] = character_data
23
+
24
+ with CHARACTERS_FILE.open('w') as f:
25
+ json.dump(characters, f, indent=2)
26
+
27
+
28
+ def load_characters():
29
+ if os.path.exists(CHARACTERS_FILE):
30
+ with open(CHARACTERS_FILE, 'r') as f:
31
+ return json.load(f)
32
+ return {}
33
+
34
+
35
+ def get_character_names():
36
+ characters = load_characters()
37
+ return list(characters.keys())
38
+
39
+
40
+
41
+
App_Function_Libraries/ChromaDB_Library.py ADDED
@@ -0,0 +1,225 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import configparser
2
+ import logging
3
+ import sqlite3
4
+ from typing import List, Dict, Any
5
+
6
+ import chromadb
7
+ import requests
8
+
9
+ from App_Function_Libraries.Chunk_Lib import improved_chunking_process
10
+
11
+ #######################################################################################################################
12
+ #
13
+ # Functions for ChromaDB
14
+
15
+ # Get ChromaDB settings
16
+ # Load configuration
17
+ config = configparser.ConfigParser()
18
+ config.read('config.txt')
19
+ chroma_db_path = config.get('Database', 'chroma_db_path', fallback='chroma_db')
20
+ chroma_client = chromadb.PersistentClient(path=chroma_db_path)
21
+
22
+ # Get embedding settings
23
+ embedding_provider = config.get('Embeddings', 'provider', fallback='openai')
24
+ embedding_model = config.get('Embeddings', 'model', fallback='text-embedding-3-small')
25
+ embedding_api_key = config.get('Embeddings', 'api_key', fallback='')
26
+ embedding_api_url = config.get('Embeddings', 'api_url', fallback='')
27
+
28
+ # Get chunking options
29
+ chunk_options = {
30
+ 'method': config.get('Chunking', 'method', fallback='words'),
31
+ 'max_size': config.getint('Chunking', 'max_size', fallback=400),
32
+ 'overlap': config.getint('Chunking', 'overlap', fallback=200),
33
+ 'adaptive': config.getboolean('Chunking', 'adaptive', fallback=False),
34
+ 'multi_level': config.getboolean('Chunking', 'multi_level', fallback=False),
35
+ 'language': config.get('Chunking', 'language', fallback='english')
36
+ }
37
+
38
+
39
+ def auto_update_chroma_embeddings(media_id: int, content: str):
40
+ """
41
+ Automatically update ChromaDB embeddings when a new item is ingested into the SQLite database.
42
+
43
+ :param media_id: The ID of the newly ingested media item
44
+ :param content: The content of the newly ingested media item
45
+ """
46
+ collection_name = f"media_{media_id}"
47
+
48
+ # Initialize or get the ChromaDB collection
49
+ collection = chroma_client.get_or_create_collection(name=collection_name)
50
+
51
+ # Check if embeddings already exist for this media_id
52
+ existing_embeddings = collection.get(ids=[f"{media_id}_chunk_{i}" for i in range(len(content))])
53
+
54
+ if existing_embeddings and len(existing_embeddings) > 0:
55
+ logging.info(f"Embeddings already exist for media ID {media_id}, skipping...")
56
+ else:
57
+ # Process and store content if embeddings do not already exist
58
+ process_and_store_content(content, collection_name, media_id)
59
+ logging.info(f"Updated ChromaDB embeddings for media ID: {media_id}")
60
+
61
+
62
+ # Function to process content, create chunks, embeddings, and store in ChromaDB and SQLite
63
+ def process_and_store_content(content: str, collection_name: str, media_id: int):
64
+ # Process the content into chunks
65
+ chunks = improved_chunking_process(content, chunk_options)
66
+ texts = [chunk['text'] for chunk in chunks]
67
+
68
+ # Generate embeddings for each chunk
69
+ embeddings = [create_embedding(text) for text in texts]
70
+
71
+ # Create unique IDs for each chunk using the media_id and chunk index
72
+ ids = [f"{media_id}_chunk_{i}" for i in range(len(texts))]
73
+
74
+ # Store the texts, embeddings, and IDs in ChromaDB
75
+ store_in_chroma(collection_name, texts, embeddings, ids)
76
+
77
+ # Store the chunks in SQLite FTS as well
78
+ from App_Function_Libraries.DB_Manager import db
79
+ with db.get_connection() as conn:
80
+ cursor = conn.cursor()
81
+ for text in texts:
82
+ cursor.execute("INSERT INTO media_fts (content) VALUES (?)", (text,))
83
+ conn.commit()
84
+
85
+
86
+ # Function to store documents and their embeddings in ChromaDB
87
+ def store_in_chroma(collection_name: str, texts: List[str], embeddings: List[List[float]], ids: List[str]):
88
+ collection = chroma_client.get_or_create_collection(name=collection_name)
89
+ collection.add(
90
+ documents=texts,
91
+ embeddings=embeddings,
92
+ ids=ids
93
+ )
94
+
95
+ # Function to perform vector search using ChromaDB
96
+ def vector_search(collection_name: str, query: str, k: int = 10) -> List[str]:
97
+ query_embedding = create_embedding(query)
98
+ collection = chroma_client.get_collection(name=collection_name)
99
+ results = collection.query(
100
+ query_embeddings=[query_embedding],
101
+ n_results=k
102
+ )
103
+ return results['documents'][0]
104
+
105
+
106
+ def create_embedding(text: str) -> List[float]:
107
+ if embedding_provider == 'openai':
108
+ import openai
109
+ openai.api_key = embedding_api_key
110
+ response = openai.Embedding.create(input=text, model=embedding_model)
111
+ return response['data'][0]['embedding']
112
+ elif embedding_provider == 'local':
113
+ # FIXME - This is a placeholder for API calls to a local embedding model
114
+ response = requests.post(
115
+ embedding_api_url,
116
+ json={"text": text, "model": embedding_model},
117
+ headers={"Authorization": f"Bearer {embedding_api_key}"}
118
+ )
119
+ return response.json()['embedding']
120
+ # FIXME - this seems correct, but idk....
121
+ elif embedding_provider == 'huggingface':
122
+ from transformers import AutoTokenizer, AutoModel
123
+ import torch
124
+
125
+ tokenizer = AutoTokenizer.from_pretrained(embedding_model)
126
+ model = AutoModel.from_pretrained(embedding_model)
127
+
128
+ inputs = tokenizer(text, return_tensors="pt", padding=True, truncation=True, max_length=512)
129
+ with torch.no_grad():
130
+ outputs = model(**inputs)
131
+
132
+ # Use the mean of the last hidden state as the sentence embedding
133
+ embeddings = outputs.last_hidden_state.mean(dim=1)
134
+ return embeddings[0].tolist() # Convert to list for consistency
135
+ else:
136
+ raise ValueError(f"Unsupported embedding provider: {embedding_provider}")
137
+
138
+
139
+ def create_all_embeddings(api_choice: str) -> str:
140
+ try:
141
+ global embedding_provider
142
+ embedding_provider = api_choice
143
+
144
+ all_content = get_all_content_from_database()
145
+
146
+ if not all_content:
147
+ return "No content found in the database."
148
+
149
+ texts_to_embed = []
150
+ embeddings_to_store = []
151
+ ids_to_store = []
152
+ collection_name = "all_content_embeddings"
153
+
154
+ # Initialize or get the ChromaDB collection
155
+ collection = chroma_client.get_or_create_collection(name=collection_name)
156
+
157
+ for content_item in all_content:
158
+ media_id = content_item['id']
159
+ text = content_item['content']
160
+
161
+ # Check if the embedding already exists in ChromaDB
162
+ embedding_exists = collection.get(ids=[f"doc_{media_id}"])
163
+
164
+ if embedding_exists:
165
+ logging.info(f"Embedding already exists for media ID {media_id}, skipping...")
166
+ continue # Skip if embedding already exists
167
+
168
+ # Create the embedding
169
+ embedding = create_embedding(text)
170
+
171
+ # Collect the text, embedding, and ID for batch storage
172
+ texts_to_embed.append(text)
173
+ embeddings_to_store.append(embedding)
174
+ ids_to_store.append(f"doc_{media_id}")
175
+
176
+ # Store all new embeddings in ChromaDB
177
+ if texts_to_embed and embeddings_to_store:
178
+ store_in_chroma(collection_name, texts_to_embed, embeddings_to_store, ids_to_store)
179
+
180
+ return "Embeddings created and stored successfully for all new content."
181
+ except Exception as e:
182
+ logging.error(f"Error during embedding creation: {str(e)}")
183
+ return f"Error: {str(e)}"
184
+
185
+
186
+ def get_all_content_from_database() -> List[Dict[str, Any]]:
187
+ """
188
+ Retrieve all media content from the database that requires embedding.
189
+
190
+ Returns:
191
+ List[Dict[str, Any]]: A list of dictionaries, each containing the media ID, content, title, and other relevant fields.
192
+ """
193
+ try:
194
+ from App_Function_Libraries.DB_Manager import db
195
+ with db.get_connection() as conn:
196
+ cursor = conn.cursor()
197
+ cursor.execute("""
198
+ SELECT id, content, title, author, type
199
+ FROM Media
200
+ WHERE is_trash = 0 -- Exclude items marked as trash
201
+ """)
202
+ media_items = cursor.fetchall()
203
+
204
+ # Convert the results into a list of dictionaries
205
+ all_content = [
206
+ {
207
+ 'id': item[0],
208
+ 'content': item[1],
209
+ 'title': item[2],
210
+ 'author': item[3],
211
+ 'type': item[4]
212
+ }
213
+ for item in media_items
214
+ ]
215
+
216
+ return all_content
217
+
218
+ except sqlite3.Error as e:
219
+ logging.error(f"Error retrieving all content from database: {e}")
220
+ from App_Function_Libraries.SQLite_DB import DatabaseError
221
+ raise DatabaseError(f"Error retrieving all content from database: {e}")
222
+
223
+ #
224
+ # End of Functions for ChromaDB
225
+ #######################################################################################################################
App_Function_Libraries/Chunk_Lib.py CHANGED
@@ -1,583 +1,587 @@
1
- # Chunk_Lib.py
2
- #########################################
3
- # Chunking Library
4
- # This library is used to perform chunking of input files.
5
- # Currently, uses naive approaches. Nothing fancy.
6
- #
7
- ####
8
- # Import necessary libraries
9
- import logging
10
- import re
11
-
12
- from typing import List, Optional, Tuple, Dict, Any
13
-
14
- from openai import OpenAI
15
- from tqdm import tqdm
16
- #
17
- # Import 3rd party
18
- from transformers import GPT2Tokenizer
19
- import nltk
20
- from nltk.tokenize import sent_tokenize, word_tokenize
21
- from sklearn.feature_extraction.text import TfidfVectorizer
22
- from sklearn.metrics.pairwise import cosine_similarity
23
- #
24
- # Import Local
25
- from App_Function_Libraries.Tokenization_Methods_Lib import openai_tokenize
26
- from App_Function_Libraries.Utils import load_comprehensive_config
27
-
28
-
29
- #
30
- #######################################################################################################################
31
- # Function Definitions
32
- #
33
-
34
- # FIXME - Make sure it only downloads if it already exists, and does a check first.
35
- # Ensure NLTK data is downloaded
36
- def ntlk_prep():
37
- nltk.download('punkt')
38
-
39
- # Load GPT2 tokenizer
40
- tokenizer = GPT2Tokenizer.from_pretrained("gpt2")
41
-
42
- # Load Config file for API keys
43
- config = load_comprehensive_config()
44
- openai_api_key = config.get('API', 'openai_api_key', fallback=None)
45
-
46
- def load_document(file_path):
47
- with open(file_path, 'r') as file:
48
- text = file.read()
49
- return re.sub('\\s+', ' ', text).strip()
50
-
51
-
52
- def improved_chunking_process(text: str, chunk_options: Dict[str, Any]) -> List[Dict[str, Any]]:
53
- chunk_method = chunk_options.get('method', 'words')
54
- max_chunk_size = chunk_options.get('max_size', 300)
55
- overlap = chunk_options.get('overlap', 0)
56
- language = chunk_options.get('language', 'english')
57
- adaptive = chunk_options.get('adaptive', False)
58
- multi_level = chunk_options.get('multi_level', False)
59
-
60
- if adaptive:
61
- max_chunk_size = adaptive_chunk_size(text, max_chunk_size)
62
-
63
- if multi_level:
64
- chunks = multi_level_chunking(text, chunk_method, max_chunk_size, overlap, language)
65
- else:
66
- if chunk_method == 'words':
67
- chunks = chunk_text_by_words(text, max_chunk_size, overlap)
68
- elif chunk_method == 'sentences':
69
- chunks = chunk_text_by_sentences(text, max_chunk_size, overlap, language)
70
- elif chunk_method == 'paragraphs':
71
- chunks = chunk_text_by_paragraphs(text, max_chunk_size, overlap)
72
- elif chunk_method == 'tokens':
73
- chunks = chunk_text_by_tokens(text, max_chunk_size, overlap)
74
- elif chunk_method == 'chapters':
75
- return chunk_ebook_by_chapters(text, chunk_options)
76
- else:
77
- # No chunking applied
78
- chunks = [text]
79
-
80
- return [{'text': chunk, 'metadata': get_chunk_metadata(chunk, text)} for chunk in chunks]
81
-
82
-
83
- def adaptive_chunk_size(text: str, base_size: int) -> int:
84
- # Simple adaptive logic: adjust chunk size based on text complexity
85
- avg_word_length = sum(len(word) for word in text.split()) / len(text.split())
86
- if avg_word_length > 6: # Arbitrary threshold for "complex" text
87
- return int(base_size * 0.8) # Reduce chunk size for complex text
88
- return base_size
89
-
90
-
91
- def multi_level_chunking(text: str, method: str, max_size: int, overlap: int, language: str) -> List[str]:
92
- # First level: chunk by paragraphs
93
- paragraphs = chunk_text_by_paragraphs(text, max_size * 2, overlap)
94
-
95
- # Second level: chunk each paragraph further
96
- chunks = []
97
- for para in paragraphs:
98
- if method == 'words':
99
- chunks.extend(chunk_text_by_words(para, max_size, overlap))
100
- elif method == 'sentences':
101
- chunks.extend(chunk_text_by_sentences(para, max_size, overlap, language))
102
- else:
103
- chunks.append(para)
104
-
105
- return chunks
106
-
107
-
108
- def chunk_text_by_words(text: str, max_words: int = 300, overlap: int = 0) -> List[str]:
109
- words = text.split()
110
- chunks = []
111
- for i in range(0, len(words), max_words - overlap):
112
- chunk = ' '.join(words[i:i + max_words])
113
- chunks.append(chunk)
114
- return post_process_chunks(chunks)
115
-
116
-
117
- def chunk_text_by_sentences(text: str, max_sentences: int = 10, overlap: int = 0, language: str = 'english') -> List[
118
- str]:
119
- nltk.download('punkt', quiet=True)
120
- sentences = nltk.sent_tokenize(text, language=language)
121
- chunks = []
122
- for i in range(0, len(sentences), max_sentences - overlap):
123
- chunk = ' '.join(sentences[i:i + max_sentences])
124
- chunks.append(chunk)
125
- return post_process_chunks(chunks)
126
-
127
-
128
- def chunk_text_by_paragraphs(text: str, max_paragraphs: int = 5, overlap: int = 0) -> List[str]:
129
- paragraphs = re.split(r'\n\s*\n', text)
130
- chunks = []
131
- for i in range(0, len(paragraphs), max_paragraphs - overlap):
132
- chunk = '\n\n'.join(paragraphs[i:i + max_paragraphs])
133
- chunks.append(chunk)
134
- return post_process_chunks(chunks)
135
-
136
-
137
- def chunk_text_by_tokens(text: str, max_tokens: int = 1000, overlap: int = 0) -> List[str]:
138
- # This is a simplified token-based chunking. For more accurate tokenization,
139
- # consider using a proper tokenizer like GPT-2 TokenizerFast
140
- words = text.split()
141
- chunks = []
142
- current_chunk = []
143
- current_token_count = 0
144
-
145
- for word in words:
146
- word_token_count = len(word) // 4 + 1 # Rough estimate of token count
147
- if current_token_count + word_token_count > max_tokens and current_chunk:
148
- chunks.append(' '.join(current_chunk))
149
- current_chunk = current_chunk[-overlap:] if overlap > 0 else []
150
- current_token_count = sum(len(w) // 4 + 1 for w in current_chunk)
151
-
152
- current_chunk.append(word)
153
- current_token_count += word_token_count
154
-
155
- if current_chunk:
156
- chunks.append(' '.join(current_chunk))
157
-
158
- return post_process_chunks(chunks)
159
-
160
-
161
- def post_process_chunks(chunks: List[str]) -> List[str]:
162
- return [chunk.strip() for chunk in chunks if chunk.strip()]
163
-
164
-
165
- def get_chunk_metadata(chunk: str, full_text: str, chunk_type: str = "generic", chapter_number: Optional[int] = None, chapter_pattern: Optional[str] = None) -> Dict[str, Any]:
166
- start_index = full_text.index(chunk)
167
- metadata = {
168
- 'start_index': start_index,
169
- 'end_index': start_index + len(chunk),
170
- 'word_count': len(chunk.split()),
171
- 'char_count': len(chunk),
172
- 'chunk_type': chunk_type
173
- }
174
- if chunk_type == "chapter":
175
- metadata['chapter_number'] = chapter_number
176
- metadata['chapter_pattern'] = chapter_pattern
177
- return metadata
178
-
179
-
180
- # Hybrid approach, chunk each sentence while ensuring total token size does not exceed a maximum number
181
- def chunk_text_hybrid(text, max_tokens=1000):
182
- sentences = nltk.tokenize.sent_tokenize(text)
183
- chunks = []
184
- current_chunk = []
185
- current_length = 0
186
-
187
- for sentence in sentences:
188
- tokens = tokenizer.encode(sentence)
189
- if current_length + len(tokens) <= max_tokens:
190
- current_chunk.append(sentence)
191
- current_length += len(tokens)
192
- else:
193
- chunks.append(' '.join(current_chunk))
194
- current_chunk = [sentence]
195
- current_length = len(tokens)
196
-
197
- if current_chunk:
198
- chunks.append(' '.join(current_chunk))
199
-
200
- return chunks
201
-
202
- # Thanks openai
203
- def chunk_on_delimiter(input_string: str,
204
- max_tokens: int,
205
- delimiter: str) -> List[str]:
206
- chunks = input_string.split(delimiter)
207
- combined_chunks, _, dropped_chunk_count = combine_chunks_with_no_minimum(
208
- chunks, max_tokens, chunk_delimiter=delimiter, add_ellipsis_for_overflow=True)
209
- if dropped_chunk_count > 0:
210
- print(f"Warning: {dropped_chunk_count} chunks were dropped due to exceeding the token limit.")
211
- combined_chunks = [f"{chunk}{delimiter}" for chunk in combined_chunks]
212
- return combined_chunks
213
-
214
- # ????FIXME
215
- def recursive_summarize_chunks(chunks, summarize_func, custom_prompt, temp=None, system_prompt=None):
216
- summarized_chunks = []
217
- current_summary = ""
218
-
219
- logging.debug(f"recursive_summarize_chunks: Summarizing {len(chunks)} chunks recursively...")
220
- logging.debug(f"recursive_summarize_chunks: temperature is @ {temp}")
221
- for i, chunk in enumerate(chunks):
222
- if i == 0:
223
- current_summary = summarize_func(chunk, custom_prompt, temp, system_prompt)
224
- else:
225
- combined_text = current_summary + "\n\n" + chunk
226
- current_summary = summarize_func(combined_text, custom_prompt, temp, system_prompt)
227
-
228
- summarized_chunks.append(current_summary)
229
-
230
- return summarized_chunks
231
-
232
-
233
- # Sample text for testing
234
- sample_text = """
235
- Natural language processing (NLP) is a subfield of linguistics, computer science, and artificial intelligence
236
- concerned with the interactions between computers and human language, in particular how to program computers
237
- to process and analyze large amounts of natural language data. The result is a computer capable of "understanding"
238
- the contents of documents, including the contextual nuances of the language within them. The technology can then
239
- accurately extract information and insights contained in the documents as well as categorize and organize the documents themselves.
240
-
241
- Challenges in natural language processing frequently involve speech recognition, natural language understanding,
242
- and natural language generation.
243
-
244
- Natural language processing has its roots in the 1950s. Already in 1950, Alan Turing published an article titled
245
- "Computing Machinery and Intelligence" which proposed what is now called the Turing test as a criterion of intelligence.
246
- """
247
-
248
- # Example usage of different chunking methods
249
- # print("Chunking by words:")
250
- # print(chunk_text_by_words(sample_text, max_words=50))
251
- #
252
- # print("\nChunking by sentences:")
253
- # print(chunk_text_by_sentences(sample_text, max_sentences=2))
254
- #
255
- # print("\nChunking by paragraphs:")
256
- # print(chunk_text_by_paragraphs(sample_text, max_paragraphs=1))
257
- #
258
- # print("\nChunking by tokens:")
259
- # print(chunk_text_by_tokens(sample_text, max_tokens=50))
260
- #
261
- # print("\nHybrid chunking:")
262
- # print(chunk_text_hybrid(sample_text, max_tokens=50))
263
-
264
-
265
-
266
- #######################################################################################################################
267
- #
268
- # Experimental Semantic Chunking
269
- #
270
-
271
- # Chunk text into segments based on semantic similarity
272
- def count_units(text, unit='tokens'):
273
- if unit == 'words':
274
- return len(text.split())
275
- elif unit == 'tokens':
276
- return len(word_tokenize(text))
277
- elif unit == 'characters':
278
- return len(text)
279
- else:
280
- raise ValueError("Invalid unit. Choose 'words', 'tokens', or 'characters'.")
281
-
282
-
283
- def semantic_chunking(text, max_chunk_size=2000, unit='words'):
284
- nltk.download('punkt', quiet=True)
285
- sentences = sent_tokenize(text)
286
- vectorizer = TfidfVectorizer()
287
- sentence_vectors = vectorizer.fit_transform(sentences)
288
-
289
- chunks = []
290
- current_chunk = []
291
- current_size = 0
292
-
293
- for i, sentence in enumerate(sentences):
294
- sentence_size = count_units(sentence, unit)
295
- if current_size + sentence_size > max_chunk_size and current_chunk:
296
- chunks.append(' '.join(current_chunk))
297
- overlap_size = count_units(' '.join(current_chunk[-3:]), unit) # Use last 3 sentences for overlap
298
- current_chunk = current_chunk[-3:] # Keep last 3 sentences for overlap
299
- current_size = overlap_size
300
-
301
- current_chunk.append(sentence)
302
- current_size += sentence_size
303
-
304
- if i + 1 < len(sentences):
305
- current_vector = sentence_vectors[i]
306
- next_vector = sentence_vectors[i + 1]
307
- similarity = cosine_similarity(current_vector, next_vector)[0][0]
308
- if similarity < 0.5 and current_size >= max_chunk_size // 2:
309
- chunks.append(' '.join(current_chunk))
310
- overlap_size = count_units(' '.join(current_chunk[-3:]), unit)
311
- current_chunk = current_chunk[-3:]
312
- current_size = overlap_size
313
-
314
- if current_chunk:
315
- chunks.append(' '.join(current_chunk))
316
-
317
- return chunks
318
-
319
-
320
- def semantic_chunk_long_file(file_path, max_chunk_size=1000, overlap=100):
321
- try:
322
- with open(file_path, 'r', encoding='utf-8') as file:
323
- content = file.read()
324
-
325
- chunks = semantic_chunking(content, max_chunk_size, overlap)
326
- return chunks
327
- except Exception as e:
328
- logging.error(f"Error chunking text file: {str(e)}")
329
- return None
330
- #######################################################################################################################
331
-
332
-
333
-
334
-
335
-
336
-
337
- #######################################################################################################################
338
- #
339
- # OpenAI Rolling Summarization
340
- #
341
-
342
- client = OpenAI(api_key=openai_api_key)
343
- def get_chat_completion(messages, model='gpt-4-turbo'):
344
- response = client.chat.completions.create(
345
- model=model,
346
- messages=messages,
347
- temperature=0,
348
- )
349
- return response.choices[0].message.content
350
-
351
-
352
- # This function combines text chunks into larger blocks without exceeding a specified token count.
353
- # It returns the combined chunks, their original indices, and the number of dropped chunks due to overflow.
354
- def combine_chunks_with_no_minimum(
355
- chunks: List[str],
356
- max_tokens: int,
357
- chunk_delimiter="\n\n",
358
- header: Optional[str] = None,
359
- add_ellipsis_for_overflow=False,
360
- ) -> Tuple[List[str], List[int]]:
361
- dropped_chunk_count = 0
362
- output = [] # list to hold the final combined chunks
363
- output_indices = [] # list to hold the indices of the final combined chunks
364
- candidate = (
365
- [] if header is None else [header]
366
- ) # list to hold the current combined chunk candidate
367
- candidate_indices = []
368
- for chunk_i, chunk in enumerate(chunks):
369
- chunk_with_header = [chunk] if header is None else [header, chunk]
370
- # FIXME MAKE NOT OPENAI SPECIFIC
371
- if len(openai_tokenize(chunk_delimiter.join(chunk_with_header))) > max_tokens:
372
- print(f"warning: chunk overflow")
373
- if (
374
- add_ellipsis_for_overflow
375
- # FIXME MAKE NOT OPENAI SPECIFIC
376
- and len(openai_tokenize(chunk_delimiter.join(candidate + ["..."]))) <= max_tokens
377
- ):
378
- candidate.append("...")
379
- dropped_chunk_count += 1
380
- continue # this case would break downstream assumptions
381
- # estimate token count with the current chunk added
382
- # FIXME MAKE NOT OPENAI SPECIFIC
383
- extended_candidate_token_count = len(openai_tokenize(chunk_delimiter.join(candidate + [chunk])))
384
- # If the token count exceeds max_tokens, add the current candidate to output and start a new candidate
385
- if extended_candidate_token_count > max_tokens:
386
- output.append(chunk_delimiter.join(candidate))
387
- output_indices.append(candidate_indices)
388
- candidate = chunk_with_header # re-initialize candidate
389
- candidate_indices = [chunk_i]
390
- # otherwise keep extending the candidate
391
- else:
392
- candidate.append(chunk)
393
- candidate_indices.append(chunk_i)
394
- # add the remaining candidate to output if it's not empty
395
- if (header is not None and len(candidate) > 1) or (header is None and len(candidate) > 0):
396
- output.append(chunk_delimiter.join(candidate))
397
- output_indices.append(candidate_indices)
398
- return output, output_indices, dropped_chunk_count
399
-
400
-
401
- def rolling_summarize(text: str,
402
- detail: float = 0,
403
- model: str = 'gpt-4-turbo',
404
- additional_instructions: Optional[str] = None,
405
- minimum_chunk_size: Optional[int] = 500,
406
- chunk_delimiter: str = ".",
407
- summarize_recursively=False,
408
- verbose=False):
409
- """
410
- Summarizes a given text by splitting it into chunks, each of which is summarized individually.
411
- The level of detail in the summary can be adjusted, and the process can optionally be made recursive.
412
-
413
- Parameters:
414
- - text (str): The text to be summarized.
415
- - detail (float, optional): A value between 0 and 1
416
- indicating the desired level of detail in the summary. 0 leads to a higher level summary, and 1 results in a more
417
- detailed summary. Defaults to 0.
418
- - additional_instructions (Optional[str], optional): Additional instructions to provide to the
419
- model for customizing summaries. - minimum_chunk_size (Optional[int], optional): The minimum size for text
420
- chunks. Defaults to 500.
421
- - chunk_delimiter (str, optional): The delimiter used to split the text into chunks. Defaults to ".".
422
- - summarize_recursively (bool, optional): If True, summaries are generated recursively, using previous summaries for context.
423
- - verbose (bool, optional): If True, prints detailed information about the chunking process.
424
- Returns:
425
- - str: The final compiled summary of the text.
426
-
427
- The function first determines the number of chunks by interpolating between a minimum and a maximum chunk count
428
- based on the `detail` parameter. It then splits the text into chunks and summarizes each chunk. If
429
- `summarize_recursively` is True, each summary is based on the previous summaries, adding more context to the
430
- summarization process. The function returns a compiled summary of all chunks.
431
- """
432
-
433
- # check detail is set correctly
434
- assert 0 <= detail <= 1
435
-
436
- # interpolate the number of chunks based to get specified level of detail
437
- max_chunks = len(chunk_on_delimiter(text, minimum_chunk_size, chunk_delimiter))
438
- min_chunks = 1
439
- num_chunks = int(min_chunks + detail * (max_chunks - min_chunks))
440
-
441
- # adjust chunk_size based on interpolated number of chunks
442
- # FIXME MAKE NOT OPENAI SPECIFIC
443
- document_length = len(openai_tokenize(text))
444
- chunk_size = max(minimum_chunk_size, document_length // num_chunks)
445
- text_chunks = chunk_on_delimiter(text, chunk_size, chunk_delimiter)
446
- if verbose:
447
- print(f"Splitting the text into {len(text_chunks)} chunks to be summarized.")
448
- # FIXME MAKE NOT OPENAI SPECIFIC
449
- print(f"Chunk lengths are {[len(openai_tokenize(x)) for x in text_chunks]}")
450
-
451
- # set system message - FIXME
452
- system_message_content = "Rewrite this text in summarized form."
453
- if additional_instructions is not None:
454
- system_message_content += f"\n\n{additional_instructions}"
455
-
456
- accumulated_summaries = []
457
- for i, chunk in enumerate(tqdm(text_chunks)):
458
- if summarize_recursively and accumulated_summaries:
459
- # Combine previous summary with current chunk for recursive summarization
460
- combined_text = accumulated_summaries[-1] + "\n\n" + chunk
461
- user_message_content = f"Previous summary and new content to summarize:\n\n{combined_text}"
462
- else:
463
- user_message_content = chunk
464
-
465
- messages = [
466
- {"role": "system", "content": system_message_content},
467
- {"role": "user", "content": user_message_content}
468
- ]
469
-
470
- response = get_chat_completion(messages, model=model)
471
- accumulated_summaries.append(response)
472
-
473
- final_summary = '\n\n'.join(accumulated_summaries)
474
- return final_summary
475
-
476
- #
477
- #
478
- #######################################################################################################################
479
- #
480
- # Ebook Chapter Chunking
481
-
482
-
483
- def chunk_ebook_by_chapters(text: str, chunk_options: Dict[str, Any]) -> List[Dict[str, Any]]:
484
- max_chunk_size = chunk_options.get('max_size', 300)
485
- overlap = chunk_options.get('overlap', 0)
486
- custom_pattern = chunk_options.get('custom_chapter_pattern', None)
487
-
488
- # List of chapter heading patterns to try, in order
489
- chapter_patterns = [
490
- custom_pattern,
491
- r'^#{1,2}\s+', # Markdown style: '# ' or '## '
492
- r'^Chapter\s+\d+', # 'Chapter ' followed by numbers
493
- r'^\d+\.\s+', # Numbered chapters: '1. ', '2. ', etc.
494
- r'^[A-Z\s]+$' # All caps headings
495
- ]
496
-
497
- chapter_positions = []
498
- used_pattern = None
499
-
500
- for pattern in chapter_patterns:
501
- if pattern is None:
502
- continue
503
- chapter_regex = re.compile(pattern, re.MULTILINE | re.IGNORECASE)
504
- chapter_positions = [match.start() for match in chapter_regex.finditer(text)]
505
- if chapter_positions:
506
- used_pattern = pattern
507
- break
508
-
509
- # If no chapters found, return the entire content as one chunk
510
- if not chapter_positions:
511
- return [{'text': text, 'metadata': get_chunk_metadata(text, text, chunk_type="whole_document")}]
512
-
513
- # Split content into chapters
514
- chunks = []
515
- for i in range(len(chapter_positions)):
516
- start = chapter_positions[i]
517
- end = chapter_positions[i + 1] if i + 1 < len(chapter_positions) else None
518
- chapter = text[start:end]
519
-
520
- # Apply overlap if specified
521
- if overlap > 0 and i > 0:
522
- overlap_start = max(0, start - overlap)
523
- chapter = text[overlap_start:end]
524
-
525
- chunks.append(chapter)
526
-
527
- # Post-process chunks
528
- processed_chunks = post_process_chunks(chunks)
529
-
530
- # Add metadata to chunks
531
- return [{'text': chunk, 'metadata': get_chunk_metadata(chunk, text, chunk_type="chapter", chapter_number=i + 1,
532
- chapter_pattern=used_pattern)}
533
- for i, chunk in enumerate(processed_chunks)]
534
-
535
-
536
- # # Example usage
537
- # if __name__ == "__main__":
538
- # sample_ebook_content = """
539
- # # Chapter 1: Introduction
540
- #
541
- # This is the introduction.
542
- #
543
- # ## Section 1.1
544
- #
545
- # Some content here.
546
- #
547
- # # Chapter 2: Main Content
548
- #
549
- # This is the main content.
550
- #
551
- # ## Section 2.1
552
- #
553
- # More content here.
554
- #
555
- # CHAPTER THREE
556
- #
557
- # This is the third chapter.
558
- #
559
- # 4. Fourth Chapter
560
- #
561
- # This is the fourth chapter.
562
- # """
563
- #
564
- # chunk_options = {
565
- # 'method': 'chapters',
566
- # 'max_size': 500,
567
- # 'overlap': 50,
568
- # 'custom_chapter_pattern': r'^CHAPTER\s+[A-Z]+' # Custom pattern for 'CHAPTER THREE' style
569
- # }
570
- #
571
- # chunked_chapters = improved_chunking_process(sample_ebook_content, chunk_options)
572
- #
573
- # for i, chunk in enumerate(chunked_chapters, 1):
574
- # print(f"Chunk {i}:")
575
- # print(chunk['text'])
576
- # print(f"Metadata: {chunk['metadata']}\n")
577
-
578
-
579
-
580
-
581
- #
582
- # End of Chunking Library
 
 
 
 
583
  #######################################################################################################################
 
1
+ # Chunk_Lib.py
2
+ #########################################
3
+ # Chunking Library
4
+ # This library is used to perform chunking of input files.
5
+ # Currently, uses naive approaches. Nothing fancy.
6
+ #
7
+ ####
8
+ # Import necessary libraries
9
+ import logging
10
+ import re
11
+
12
+ from typing import List, Optional, Tuple, Dict, Any
13
+
14
+ from openai import OpenAI
15
+ from tqdm import tqdm
16
+ #
17
+ # Import 3rd party
18
+ from transformers import GPT2Tokenizer
19
+ import nltk
20
+ from nltk.tokenize import sent_tokenize, word_tokenize
21
+ from sklearn.feature_extraction.text import TfidfVectorizer
22
+ from sklearn.metrics.pairwise import cosine_similarity
23
+ #
24
+ # Import Local
25
+ from App_Function_Libraries.Tokenization_Methods_Lib import openai_tokenize
26
+ from App_Function_Libraries.Utils import load_comprehensive_config
27
+
28
+
29
+ #
30
+ #######################################################################################################################
31
+ # Function Definitions
32
+ #
33
+
34
+ # FIXME - Make sure it only downloads if it already exists, and does a check first.
35
+ # Ensure NLTK data is downloaded
36
+ def ntlk_prep():
37
+ nltk.download('punkt')
38
+
39
+ # Load GPT2 tokenizer
40
+ tokenizer = GPT2Tokenizer.from_pretrained("gpt2")
41
+
42
+ # Load Config file for API keys
43
+ config = load_comprehensive_config()
44
+ openai_api_key = config.get('API', 'openai_api_key', fallback=None)
45
+
46
+ def load_document(file_path):
47
+ with open(file_path, 'r') as file:
48
+ text = file.read()
49
+ return re.sub('\\s+', ' ', text).strip()
50
+
51
+
52
+ def improved_chunking_process(text: str, chunk_options: Dict[str, Any]) -> List[Dict[str, Any]]:
53
+ chunk_method = chunk_options.get('method', 'words')
54
+ max_chunk_size = chunk_options.get('max_size', 300)
55
+ overlap = chunk_options.get('overlap', 0)
56
+ language = chunk_options.get('language', 'english')
57
+ adaptive = chunk_options.get('adaptive', False)
58
+ multi_level = chunk_options.get('multi_level', False)
59
+
60
+ if adaptive:
61
+ max_chunk_size = adaptive_chunk_size(text, max_chunk_size)
62
+
63
+ if multi_level:
64
+ chunks = multi_level_chunking(text, chunk_method, max_chunk_size, overlap, language)
65
+ else:
66
+ if chunk_method == 'words':
67
+ chunks = chunk_text_by_words(text, max_chunk_size, overlap)
68
+ elif chunk_method == 'sentences':
69
+ chunks = chunk_text_by_sentences(text, max_chunk_size, overlap, language)
70
+ elif chunk_method == 'paragraphs':
71
+ chunks = chunk_text_by_paragraphs(text, max_chunk_size, overlap)
72
+ elif chunk_method == 'tokens':
73
+ chunks = chunk_text_by_tokens(text, max_chunk_size, overlap)
74
+ elif chunk_method == 'chapters':
75
+ return chunk_ebook_by_chapters(text, chunk_options)
76
+ else:
77
+ # No chunking applied
78
+ chunks = [text]
79
+
80
+ return [{'text': chunk, 'metadata': get_chunk_metadata(chunk, text)} for chunk in chunks]
81
+
82
+
83
+ def adaptive_chunk_size(text: str, base_size: int) -> int:
84
+ # Simple adaptive logic: adjust chunk size based on text complexity
85
+ avg_word_length = sum(len(word) for word in text.split()) / len(text.split())
86
+ if avg_word_length > 6: # Arbitrary threshold for "complex" text
87
+ return int(base_size * 0.8) # Reduce chunk size for complex text
88
+ return base_size
89
+
90
+
91
+ def multi_level_chunking(text: str, method: str, max_size: int, overlap: int, language: str) -> List[str]:
92
+ # First level: chunk by paragraphs
93
+ paragraphs = chunk_text_by_paragraphs(text, max_size * 2, overlap)
94
+
95
+ # Second level: chunk each paragraph further
96
+ chunks = []
97
+ for para in paragraphs:
98
+ if method == 'words':
99
+ chunks.extend(chunk_text_by_words(para, max_size, overlap))
100
+ elif method == 'sentences':
101
+ chunks.extend(chunk_text_by_sentences(para, max_size, overlap, language))
102
+ else:
103
+ chunks.append(para)
104
+
105
+ return chunks
106
+
107
+
108
+ def chunk_text_by_words(text: str, max_words: int = 300, overlap: int = 0) -> List[str]:
109
+ words = text.split()
110
+ chunks = []
111
+ for i in range(0, len(words), max_words - overlap):
112
+ chunk = ' '.join(words[i:i + max_words])
113
+ chunks.append(chunk)
114
+ return post_process_chunks(chunks)
115
+
116
+
117
+ def chunk_text_by_sentences(text: str, max_sentences: int = 10, overlap: int = 0, language: str = 'english') -> List[
118
+ str]:
119
+ nltk.download('punkt', quiet=True)
120
+ sentences = nltk.sent_tokenize(text, language=language)
121
+ chunks = []
122
+ for i in range(0, len(sentences), max_sentences - overlap):
123
+ chunk = ' '.join(sentences[i:i + max_sentences])
124
+ chunks.append(chunk)
125
+ return post_process_chunks(chunks)
126
+
127
+
128
+ def chunk_text_by_paragraphs(text: str, max_paragraphs: int = 5, overlap: int = 0) -> List[str]:
129
+ paragraphs = re.split(r'\n\s*\n', text)
130
+ chunks = []
131
+ for i in range(0, len(paragraphs), max_paragraphs - overlap):
132
+ chunk = '\n\n'.join(paragraphs[i:i + max_paragraphs])
133
+ chunks.append(chunk)
134
+ return post_process_chunks(chunks)
135
+
136
+
137
+ def chunk_text_by_tokens(text: str, max_tokens: int = 1000, overlap: int = 0) -> List[str]:
138
+ # This is a simplified token-based chunking. For more accurate tokenization,
139
+ # consider using a proper tokenizer like GPT-2 TokenizerFast
140
+ words = text.split()
141
+ chunks = []
142
+ current_chunk = []
143
+ current_token_count = 0
144
+
145
+ for word in words:
146
+ word_token_count = len(word) // 4 + 1 # Rough estimate of token count
147
+ if current_token_count + word_token_count > max_tokens and current_chunk:
148
+ chunks.append(' '.join(current_chunk))
149
+ current_chunk = current_chunk[-overlap:] if overlap > 0 else []
150
+ current_token_count = sum(len(w) // 4 + 1 for w in current_chunk)
151
+
152
+ current_chunk.append(word)
153
+ current_token_count += word_token_count
154
+
155
+ if current_chunk:
156
+ chunks.append(' '.join(current_chunk))
157
+
158
+ return post_process_chunks(chunks)
159
+
160
+
161
+ def post_process_chunks(chunks: List[str]) -> List[str]:
162
+ return [chunk.strip() for chunk in chunks if chunk.strip()]
163
+
164
+
165
+ def get_chunk_metadata(chunk: str, full_text: str, chunk_type: str = "generic", chapter_number: Optional[int] = None, chapter_pattern: Optional[str] = None) -> Dict[str, Any]:
166
+ try:
167
+ start_index = full_text.index(chunk)
168
+ metadata = {
169
+ 'start_index': start_index,
170
+ 'end_index': start_index + len(chunk),
171
+ 'word_count': len(chunk.split()),
172
+ 'char_count': len(chunk),
173
+ 'chunk_type': chunk_type
174
+ }
175
+ if chunk_type == "chapter":
176
+ metadata['chapter_number'] = chapter_number
177
+ metadata['chapter_pattern'] = chapter_pattern
178
+ return metadata
179
+ except ValueError as e:
180
+ logging.error(f"Chunk not found in full_text: {chunk[:50]}... Full text length: {len(full_text)}")
181
+ raise
182
+
183
+
184
+ # Hybrid approach, chunk each sentence while ensuring total token size does not exceed a maximum number
185
+ def chunk_text_hybrid(text, max_tokens=1000):
186
+ sentences = nltk.tokenize.sent_tokenize(text)
187
+ chunks = []
188
+ current_chunk = []
189
+ current_length = 0
190
+
191
+ for sentence in sentences:
192
+ tokens = tokenizer.encode(sentence)
193
+ if current_length + len(tokens) <= max_tokens:
194
+ current_chunk.append(sentence)
195
+ current_length += len(tokens)
196
+ else:
197
+ chunks.append(' '.join(current_chunk))
198
+ current_chunk = [sentence]
199
+ current_length = len(tokens)
200
+
201
+ if current_chunk:
202
+ chunks.append(' '.join(current_chunk))
203
+
204
+ return chunks
205
+
206
+ # Thanks openai
207
+ def chunk_on_delimiter(input_string: str,
208
+ max_tokens: int,
209
+ delimiter: str) -> List[str]:
210
+ chunks = input_string.split(delimiter)
211
+ combined_chunks, _, dropped_chunk_count = combine_chunks_with_no_minimum(
212
+ chunks, max_tokens, chunk_delimiter=delimiter, add_ellipsis_for_overflow=True)
213
+ if dropped_chunk_count > 0:
214
+ print(f"Warning: {dropped_chunk_count} chunks were dropped due to exceeding the token limit.")
215
+ combined_chunks = [f"{chunk}{delimiter}" for chunk in combined_chunks]
216
+ return combined_chunks
217
+
218
+ # ????FIXME
219
+ def recursive_summarize_chunks(chunks, summarize_func, custom_prompt, temp=None, system_prompt=None):
220
+ summarized_chunks = []
221
+ current_summary = ""
222
+
223
+ logging.debug(f"recursive_summarize_chunks: Summarizing {len(chunks)} chunks recursively...")
224
+ logging.debug(f"recursive_summarize_chunks: temperature is @ {temp}")
225
+ for i, chunk in enumerate(chunks):
226
+ if i == 0:
227
+ current_summary = summarize_func(chunk, custom_prompt, temp, system_prompt)
228
+ else:
229
+ combined_text = current_summary + "\n\n" + chunk
230
+ current_summary = summarize_func(combined_text, custom_prompt, temp, system_prompt)
231
+
232
+ summarized_chunks.append(current_summary)
233
+
234
+ return summarized_chunks
235
+
236
+
237
+ # Sample text for testing
238
+ sample_text = """
239
+ Natural language processing (NLP) is a subfield of linguistics, computer science, and artificial intelligence
240
+ concerned with the interactions between computers and human language, in particular how to program computers
241
+ to process and analyze large amounts of natural language data. The result is a computer capable of "understanding"
242
+ the contents of documents, including the contextual nuances of the language within them. The technology can then
243
+ accurately extract information and insights contained in the documents as well as categorize and organize the documents themselves.
244
+
245
+ Challenges in natural language processing frequently involve speech recognition, natural language understanding,
246
+ and natural language generation.
247
+
248
+ Natural language processing has its roots in the 1950s. Already in 1950, Alan Turing published an article titled
249
+ "Computing Machinery and Intelligence" which proposed what is now called the Turing test as a criterion of intelligence.
250
+ """
251
+
252
+ # Example usage of different chunking methods
253
+ # print("Chunking by words:")
254
+ # print(chunk_text_by_words(sample_text, max_words=50))
255
+ #
256
+ # print("\nChunking by sentences:")
257
+ # print(chunk_text_by_sentences(sample_text, max_sentences=2))
258
+ #
259
+ # print("\nChunking by paragraphs:")
260
+ # print(chunk_text_by_paragraphs(sample_text, max_paragraphs=1))
261
+ #
262
+ # print("\nChunking by tokens:")
263
+ # print(chunk_text_by_tokens(sample_text, max_tokens=50))
264
+ #
265
+ # print("\nHybrid chunking:")
266
+ # print(chunk_text_hybrid(sample_text, max_tokens=50))
267
+
268
+
269
+
270
+ #######################################################################################################################
271
+ #
272
+ # Experimental Semantic Chunking
273
+ #
274
+
275
+ # Chunk text into segments based on semantic similarity
276
+ def count_units(text, unit='tokens'):
277
+ if unit == 'words':
278
+ return len(text.split())
279
+ elif unit == 'tokens':
280
+ return len(word_tokenize(text))
281
+ elif unit == 'characters':
282
+ return len(text)
283
+ else:
284
+ raise ValueError("Invalid unit. Choose 'words', 'tokens', or 'characters'.")
285
+
286
+
287
+ def semantic_chunking(text, max_chunk_size=2000, unit='words'):
288
+ nltk.download('punkt', quiet=True)
289
+ sentences = sent_tokenize(text)
290
+ vectorizer = TfidfVectorizer()
291
+ sentence_vectors = vectorizer.fit_transform(sentences)
292
+
293
+ chunks = []
294
+ current_chunk = []
295
+ current_size = 0
296
+
297
+ for i, sentence in enumerate(sentences):
298
+ sentence_size = count_units(sentence, unit)
299
+ if current_size + sentence_size > max_chunk_size and current_chunk:
300
+ chunks.append(' '.join(current_chunk))
301
+ overlap_size = count_units(' '.join(current_chunk[-3:]), unit) # Use last 3 sentences for overlap
302
+ current_chunk = current_chunk[-3:] # Keep last 3 sentences for overlap
303
+ current_size = overlap_size
304
+
305
+ current_chunk.append(sentence)
306
+ current_size += sentence_size
307
+
308
+ if i + 1 < len(sentences):
309
+ current_vector = sentence_vectors[i]
310
+ next_vector = sentence_vectors[i + 1]
311
+ similarity = cosine_similarity(current_vector, next_vector)[0][0]
312
+ if similarity < 0.5 and current_size >= max_chunk_size // 2:
313
+ chunks.append(' '.join(current_chunk))
314
+ overlap_size = count_units(' '.join(current_chunk[-3:]), unit)
315
+ current_chunk = current_chunk[-3:]
316
+ current_size = overlap_size
317
+
318
+ if current_chunk:
319
+ chunks.append(' '.join(current_chunk))
320
+
321
+ return chunks
322
+
323
+
324
+ def semantic_chunk_long_file(file_path, max_chunk_size=1000, overlap=100):
325
+ try:
326
+ with open(file_path, 'r', encoding='utf-8') as file:
327
+ content = file.read()
328
+
329
+ chunks = semantic_chunking(content, max_chunk_size, overlap)
330
+ return chunks
331
+ except Exception as e:
332
+ logging.error(f"Error chunking text file: {str(e)}")
333
+ return None
334
+ #######################################################################################################################
335
+
336
+
337
+
338
+
339
+
340
+
341
+ #######################################################################################################################
342
+ #
343
+ # OpenAI Rolling Summarization
344
+ #
345
+
346
+ client = OpenAI(api_key=openai_api_key)
347
+ def get_chat_completion(messages, model='gpt-4-turbo'):
348
+ response = client.chat.completions.create(
349
+ model=model,
350
+ messages=messages,
351
+ temperature=0,
352
+ )
353
+ return response.choices[0].message.content
354
+
355
+
356
+ # This function combines text chunks into larger blocks without exceeding a specified token count.
357
+ # It returns the combined chunks, their original indices, and the number of dropped chunks due to overflow.
358
+ def combine_chunks_with_no_minimum(
359
+ chunks: List[str],
360
+ max_tokens: int,
361
+ chunk_delimiter="\n\n",
362
+ header: Optional[str] = None,
363
+ add_ellipsis_for_overflow=False,
364
+ ) -> Tuple[List[str], List[int]]:
365
+ dropped_chunk_count = 0
366
+ output = [] # list to hold the final combined chunks
367
+ output_indices = [] # list to hold the indices of the final combined chunks
368
+ candidate = (
369
+ [] if header is None else [header]
370
+ ) # list to hold the current combined chunk candidate
371
+ candidate_indices = []
372
+ for chunk_i, chunk in enumerate(chunks):
373
+ chunk_with_header = [chunk] if header is None else [header, chunk]
374
+ # FIXME MAKE NOT OPENAI SPECIFIC
375
+ if len(openai_tokenize(chunk_delimiter.join(chunk_with_header))) > max_tokens:
376
+ print(f"warning: chunk overflow")
377
+ if (
378
+ add_ellipsis_for_overflow
379
+ # FIXME MAKE NOT OPENAI SPECIFIC
380
+ and len(openai_tokenize(chunk_delimiter.join(candidate + ["..."]))) <= max_tokens
381
+ ):
382
+ candidate.append("...")
383
+ dropped_chunk_count += 1
384
+ continue # this case would break downstream assumptions
385
+ # estimate token count with the current chunk added
386
+ # FIXME MAKE NOT OPENAI SPECIFIC
387
+ extended_candidate_token_count = len(openai_tokenize(chunk_delimiter.join(candidate + [chunk])))
388
+ # If the token count exceeds max_tokens, add the current candidate to output and start a new candidate
389
+ if extended_candidate_token_count > max_tokens:
390
+ output.append(chunk_delimiter.join(candidate))
391
+ output_indices.append(candidate_indices)
392
+ candidate = chunk_with_header # re-initialize candidate
393
+ candidate_indices = [chunk_i]
394
+ # otherwise keep extending the candidate
395
+ else:
396
+ candidate.append(chunk)
397
+ candidate_indices.append(chunk_i)
398
+ # add the remaining candidate to output if it's not empty
399
+ if (header is not None and len(candidate) > 1) or (header is None and len(candidate) > 0):
400
+ output.append(chunk_delimiter.join(candidate))
401
+ output_indices.append(candidate_indices)
402
+ return output, output_indices, dropped_chunk_count
403
+
404
+
405
+ def rolling_summarize(text: str,
406
+ detail: float = 0,
407
+ model: str = 'gpt-4-turbo',
408
+ additional_instructions: Optional[str] = None,
409
+ minimum_chunk_size: Optional[int] = 500,
410
+ chunk_delimiter: str = ".",
411
+ summarize_recursively=False,
412
+ verbose=False):
413
+ """
414
+ Summarizes a given text by splitting it into chunks, each of which is summarized individually.
415
+ The level of detail in the summary can be adjusted, and the process can optionally be made recursive.
416
+
417
+ Parameters:
418
+ - text (str): The text to be summarized.
419
+ - detail (float, optional): A value between 0 and 1
420
+ indicating the desired level of detail in the summary. 0 leads to a higher level summary, and 1 results in a more
421
+ detailed summary. Defaults to 0.
422
+ - additional_instructions (Optional[str], optional): Additional instructions to provide to the
423
+ model for customizing summaries. - minimum_chunk_size (Optional[int], optional): The minimum size for text
424
+ chunks. Defaults to 500.
425
+ - chunk_delimiter (str, optional): The delimiter used to split the text into chunks. Defaults to ".".
426
+ - summarize_recursively (bool, optional): If True, summaries are generated recursively, using previous summaries for context.
427
+ - verbose (bool, optional): If True, prints detailed information about the chunking process.
428
+ Returns:
429
+ - str: The final compiled summary of the text.
430
+
431
+ The function first determines the number of chunks by interpolating between a minimum and a maximum chunk count
432
+ based on the `detail` parameter. It then splits the text into chunks and summarizes each chunk. If
433
+ `summarize_recursively` is True, each summary is based on the previous summaries, adding more context to the
434
+ summarization process. The function returns a compiled summary of all chunks.
435
+ """
436
+
437
+ # check detail is set correctly
438
+ assert 0 <= detail <= 1
439
+
440
+ # interpolate the number of chunks based to get specified level of detail
441
+ max_chunks = len(chunk_on_delimiter(text, minimum_chunk_size, chunk_delimiter))
442
+ min_chunks = 1
443
+ num_chunks = int(min_chunks + detail * (max_chunks - min_chunks))
444
+
445
+ # adjust chunk_size based on interpolated number of chunks
446
+ # FIXME MAKE NOT OPENAI SPECIFIC
447
+ document_length = len(openai_tokenize(text))
448
+ chunk_size = max(minimum_chunk_size, document_length // num_chunks)
449
+ text_chunks = chunk_on_delimiter(text, chunk_size, chunk_delimiter)
450
+ if verbose:
451
+ print(f"Splitting the text into {len(text_chunks)} chunks to be summarized.")
452
+ # FIXME MAKE NOT OPENAI SPECIFIC
453
+ print(f"Chunk lengths are {[len(openai_tokenize(x)) for x in text_chunks]}")
454
+
455
+ # set system message - FIXME
456
+ system_message_content = "Rewrite this text in summarized form."
457
+ if additional_instructions is not None:
458
+ system_message_content += f"\n\n{additional_instructions}"
459
+
460
+ accumulated_summaries = []
461
+ for i, chunk in enumerate(tqdm(text_chunks)):
462
+ if summarize_recursively and accumulated_summaries:
463
+ # Combine previous summary with current chunk for recursive summarization
464
+ combined_text = accumulated_summaries[-1] + "\n\n" + chunk
465
+ user_message_content = f"Previous summary and new content to summarize:\n\n{combined_text}"
466
+ else:
467
+ user_message_content = chunk
468
+
469
+ messages = [
470
+ {"role": "system", "content": system_message_content},
471
+ {"role": "user", "content": user_message_content}
472
+ ]
473
+
474
+ response = get_chat_completion(messages, model=model)
475
+ accumulated_summaries.append(response)
476
+
477
+ final_summary = '\n\n'.join(accumulated_summaries)
478
+ return final_summary
479
+
480
+ #
481
+ #
482
+ #######################################################################################################################
483
+ #
484
+ # Ebook Chapter Chunking
485
+
486
+
487
+ def chunk_ebook_by_chapters(text: str, chunk_options: Dict[str, Any]) -> List[Dict[str, Any]]:
488
+ max_chunk_size = chunk_options.get('max_size', 300)
489
+ overlap = chunk_options.get('overlap', 0)
490
+ custom_pattern = chunk_options.get('custom_chapter_pattern', None)
491
+
492
+ # List of chapter heading patterns to try, in order
493
+ chapter_patterns = [
494
+ custom_pattern,
495
+ r'^#{1,2}\s+', # Markdown style: '# ' or '## '
496
+ r'^Chapter\s+\d+', # 'Chapter ' followed by numbers
497
+ r'^\d+\.\s+', # Numbered chapters: '1. ', '2. ', etc.
498
+ r'^[A-Z\s]+$' # All caps headings
499
+ ]
500
+
501
+ chapter_positions = []
502
+ used_pattern = None
503
+
504
+ for pattern in chapter_patterns:
505
+ if pattern is None:
506
+ continue
507
+ chapter_regex = re.compile(pattern, re.MULTILINE | re.IGNORECASE)
508
+ chapter_positions = [match.start() for match in chapter_regex.finditer(text)]
509
+ if chapter_positions:
510
+ used_pattern = pattern
511
+ break
512
+
513
+ # If no chapters found, return the entire content as one chunk
514
+ if not chapter_positions:
515
+ return [{'text': text, 'metadata': get_chunk_metadata(text, text, chunk_type="whole_document")}]
516
+
517
+ # Split content into chapters
518
+ chunks = []
519
+ for i in range(len(chapter_positions)):
520
+ start = chapter_positions[i]
521
+ end = chapter_positions[i + 1] if i + 1 < len(chapter_positions) else None
522
+ chapter = text[start:end]
523
+
524
+ # Apply overlap if specified
525
+ if overlap > 0 and i > 0:
526
+ overlap_start = max(0, start - overlap)
527
+ chapter = text[overlap_start:end]
528
+
529
+ chunks.append(chapter)
530
+
531
+ # Post-process chunks
532
+ processed_chunks = post_process_chunks(chunks)
533
+
534
+ # Add metadata to chunks
535
+ return [{'text': chunk, 'metadata': get_chunk_metadata(chunk, text, chunk_type="chapter", chapter_number=i + 1,
536
+ chapter_pattern=used_pattern)}
537
+ for i, chunk in enumerate(processed_chunks)]
538
+
539
+
540
+ # # Example usage
541
+ # if __name__ == "__main__":
542
+ # sample_ebook_content = """
543
+ # # Chapter 1: Introduction
544
+ #
545
+ # This is the introduction.
546
+ #
547
+ # ## Section 1.1
548
+ #
549
+ # Some content here.
550
+ #
551
+ # # Chapter 2: Main Content
552
+ #
553
+ # This is the main content.
554
+ #
555
+ # ## Section 2.1
556
+ #
557
+ # More content here.
558
+ #
559
+ # CHAPTER THREE
560
+ #
561
+ # This is the third chapter.
562
+ #
563
+ # 4. Fourth Chapter
564
+ #
565
+ # This is the fourth chapter.
566
+ # """
567
+ #
568
+ # chunk_options = {
569
+ # 'method': 'chapters',
570
+ # 'max_size': 500,
571
+ # 'overlap': 50,
572
+ # 'custom_chapter_pattern': r'^CHAPTER\s+[A-Z]+' # Custom pattern for 'CHAPTER THREE' style
573
+ # }
574
+ #
575
+ # chunked_chapters = improved_chunking_process(sample_ebook_content, chunk_options)
576
+ #
577
+ # for i, chunk in enumerate(chunked_chapters, 1):
578
+ # print(f"Chunk {i}:")
579
+ # print(chunk['text'])
580
+ # print(f"Metadata: {chunk['metadata']}\n")
581
+
582
+
583
+
584
+
585
+ #
586
+ # End of Chunking Library
587
  #######################################################################################################################
App_Function_Libraries/DB_Manager.py ADDED
@@ -0,0 +1,472 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import configparser
2
+ import logging
3
+ import os
4
+ from contextlib import contextmanager
5
+ from time import sleep
6
+ from typing import Tuple
7
+ import sqlite3
8
+ # 3rd-Party Libraries
9
+ from elasticsearch import Elasticsearch
10
+
11
+ ############################################################################################################
12
+ #
13
+ # This file contains the DatabaseManager class, which is responsible for managing the database connection, i.e. either SQLite or Elasticsearch.
14
+
15
+ ####
16
+ # The DatabaseManager class provides the following methods:
17
+ # - add_media: Add a new media item to the database
18
+ # - fetch_items_by_keyword: Fetch media items from the database based on a keyword
19
+ # - fetch_item_details: Fetch details of a specific media item from the database
20
+ # - update_media_content: Update the content of a specific media item in the database
21
+ # - search_and_display_items: Search for media items in the database and display the results
22
+ # - close_connection: Close the database connection
23
+ ####
24
+
25
+ # Import your existing SQLite functions
26
+ from SQLite_DB import (
27
+ update_media_content as sqlite_update_media_content,
28
+ list_prompts as sqlite_list_prompts,
29
+ search_and_display as sqlite_search_and_display,
30
+ fetch_prompt_details as sqlite_fetch_prompt_details,
31
+ keywords_browser_interface as sqlite_keywords_browser_interface,
32
+ add_keyword as sqlite_add_keyword,
33
+ delete_keyword as sqlite_delete_keyword,
34
+ export_keywords_to_csv as sqlite_export_keywords_to_csv,
35
+ ingest_article_to_db as sqlite_ingest_article_to_db,
36
+ add_media_to_database as sqlite_add_media_to_database,
37
+ import_obsidian_note_to_db as sqlite_import_obsidian_note_to_db,
38
+ add_prompt as sqlite_add_prompt,
39
+ delete_chat_message as sqlite_delete_chat_message,
40
+ update_chat_message as sqlite_update_chat_message,
41
+ add_chat_message as sqlite_add_chat_message,
42
+ get_chat_messages as sqlite_get_chat_messages,
43
+ search_chat_conversations as sqlite_search_chat_conversations,
44
+ create_chat_conversation as sqlite_create_chat_conversation,
45
+ save_chat_history_to_database as sqlite_save_chat_history_to_database,
46
+ view_database as sqlite_view_database,
47
+ get_transcripts as sqlite_get_transcripts,
48
+ get_trashed_items as sqlite_get_trashed_items,
49
+ user_delete_item as sqlite_user_delete_item,
50
+ empty_trash as sqlite_empty_trash,
51
+ create_automated_backup as sqlite_create_automated_backup,
52
+ add_or_update_prompt as sqlite_add_or_update_prompt,
53
+ load_prompt_details as sqlite_load_prompt_details,
54
+ load_preset_prompts as sqlite_load_preset_prompts,
55
+ insert_prompt_to_db as sqlite_insert_prompt_to_db,
56
+ delete_prompt as sqlite_delete_prompt,
57
+ search_and_display_items as sqlite_search_and_display_items,
58
+ get_conversation_name as sqlite_get_conversation_name,
59
+ add_media_with_keywords as sqlite_add_media_with_keywords,
60
+ check_media_and_whisper_model as sqlite_check_media_and_whisper_model,
61
+ DatabaseError
62
+ )
63
+
64
+ class Database:
65
+ def __init__(self, db_path=None):
66
+ self.db_path = db_path or os.getenv('DB_NAME', 'media_summary.db')
67
+ self.pool = []
68
+ self.pool_size = 10
69
+
70
+ @contextmanager
71
+ def get_connection(self):
72
+ retry_count = 5
73
+ retry_delay = 1
74
+ conn = None
75
+ while retry_count > 0:
76
+ try:
77
+ conn = self.pool.pop() if self.pool else sqlite3.connect(self.db_path, check_same_thread=False)
78
+ yield conn
79
+ self.pool.append(conn)
80
+ return
81
+ except sqlite3.OperationalError as e:
82
+ if 'database is locked' in str(e):
83
+ logging.warning(f"Database is locked, retrying in {retry_delay} seconds...")
84
+ retry_count -= 1
85
+ sleep(retry_delay)
86
+ else:
87
+ raise DatabaseError(f"Database error: {e}")
88
+ except Exception as e:
89
+ raise DatabaseError(f"Unexpected error: {e}")
90
+ finally:
91
+ # Ensure the connection is returned to the pool even on failure
92
+ if conn and conn not in self.pool:
93
+ self.pool.append(conn)
94
+ raise DatabaseError("Database is locked and retries have been exhausted")
95
+
96
+ def execute_query(self, query: str, params: Tuple = ()) -> None:
97
+ with self.get_connection() as conn:
98
+ try:
99
+ cursor = conn.cursor()
100
+ cursor.execute(query, params)
101
+ conn.commit()
102
+ except sqlite3.Error as e:
103
+ raise DatabaseError(f"Database error: {e}, Query: {query}")
104
+
105
+ def close_all_connections(self):
106
+ for conn in self.pool:
107
+ conn.close()
108
+ self.pool.clear()
109
+
110
+ def get_db_config():
111
+ config = configparser.ConfigParser()
112
+ config.read('config.txt')
113
+ return {
114
+ 'type': config['Database']['type'],
115
+ 'sqlite_path': config.get('Database', 'sqlite_path', fallback='media_summary.db'),
116
+ 'elasticsearch_host': config.get('Database', 'elasticsearch_host', fallback='localhost'),
117
+ 'elasticsearch_port': config.getint('Database', 'elasticsearch_port', fallback=9200)
118
+ }
119
+
120
+ db_config = get_db_config()
121
+ db_type = db_config['type']
122
+
123
+ if db_type == 'sqlite':
124
+ # Use the config path if provided, otherwise fall back to default
125
+ db = Database(db_config.get('sqlite_path'))
126
+ elif db_type == 'elasticsearch':
127
+ es = Elasticsearch([{
128
+ 'host': db_config['elasticsearch_host'],
129
+ 'port': db_config['elasticsearch_port']
130
+ }])
131
+ else:
132
+ raise ValueError(f"Unsupported database type: {db_type}")
133
+
134
+ db_path = db_config['sqlite_path']
135
+
136
+ # Update this path to the directory where you want to store the database backups
137
+ backup_dir = os.environ.get('DB_BACKUP_DIR', 'path/to/backup/directory')
138
+
139
+
140
+
141
+
142
+ if db_type == 'sqlite':
143
+ conn = sqlite3.connect(db_config['sqlite_path'])
144
+ cursor = conn.cursor()
145
+ elif db_type == 'elasticsearch':
146
+ es = Elasticsearch([{
147
+ 'host': db_config['elasticsearch_host'],
148
+ 'port': db_config['elasticsearch_port']
149
+ }])
150
+ else:
151
+ raise ValueError(f"Unsupported database type: {db_type}")
152
+
153
+ ############################################################################################################
154
+ #
155
+ # DB-Searching functions
156
+
157
+ def view_database(*args, **kwargs):
158
+ if db_type == 'sqlite':
159
+ return sqlite_view_database(*args, **kwargs)
160
+ elif db_type == 'elasticsearch':
161
+ # Implement Elasticsearch version
162
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
163
+
164
+ def search_and_display_items(*args, **kwargs):
165
+ if db_type == 'sqlite':
166
+ return sqlite_search_and_display_items(*args, **kwargs)
167
+ elif db_type == 'elasticsearch':
168
+ # Implement Elasticsearch version
169
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
170
+
171
+ def search_and_display(*args, **kwargs):
172
+ if db_type == 'sqlite':
173
+ return sqlite_search_and_display(*args, **kwargs)
174
+ elif db_type == 'elasticsearch':
175
+ # Implement Elasticsearch version
176
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
177
+
178
+ #
179
+ # End of DB-Searching functions
180
+ ############################################################################################################
181
+
182
+ ############################################################################################################
183
+ #
184
+ # Transcript-related Functions
185
+
186
+ def get_transcripts(*args, **kwargs):
187
+ if db_type == 'sqlite':
188
+ return sqlite_get_transcripts(*args, **kwargs)
189
+ elif db_type == 'elasticsearch':
190
+ # Implement Elasticsearch version
191
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
192
+
193
+ #
194
+ # End of Transcript-related Functions
195
+ ############################################################################################################
196
+
197
+ ############################################################################################################
198
+ #
199
+ # DB-Ingestion functions
200
+
201
+ def add_media_to_database(*args, **kwargs):
202
+ if db_type == 'sqlite':
203
+ return sqlite_add_media_to_database(*args, **kwargs)
204
+ elif db_type == 'elasticsearch':
205
+ # Implement Elasticsearch version
206
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
207
+
208
+
209
+ def import_obsidian_note_to_db(*args, **kwargs):
210
+ if db_type == 'sqlite':
211
+ return sqlite_import_obsidian_note_to_db(*args, **kwargs)
212
+ elif db_type == 'elasticsearch':
213
+ # Implement Elasticsearch version
214
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
215
+
216
+ def update_media_content(*args, **kwargs):
217
+ if db_type == 'sqlite':
218
+ return sqlite_update_media_content(*args, **kwargs)
219
+ elif db_type == 'elasticsearch':
220
+ # Implement Elasticsearch version
221
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
222
+
223
+ def add_media_with_keywords(*args, **kwargs):
224
+ if db_type == 'sqlite':
225
+ return sqlite_add_media_with_keywords(*args, **kwargs)
226
+ elif db_type == 'elasticsearch':
227
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
228
+
229
+ def check_media_and_whisper_model(*args, **kwargs):
230
+ if db_type == 'sqlite':
231
+ return sqlite_check_media_and_whisper_model(*args, **kwargs)
232
+ elif db_type == 'elasticsearch':
233
+ raise NotImplementedError("Elasticsearch version of check_media_and_whisper_model not yet implemented")
234
+
235
+ def ingest_article_to_db(url, title, author, content, keywords, summary, ingestion_date, custom_prompt):
236
+ if db_type == 'sqlite':
237
+ return sqlite_ingest_article_to_db(url, title, author, content, keywords, summary, ingestion_date, custom_prompt)
238
+ elif db_type == 'elasticsearch':
239
+ # Implement Elasticsearch version
240
+ raise NotImplementedError("Elasticsearch version of ingest_article_to_db not yet implemented")
241
+ else:
242
+ raise ValueError(f"Unsupported database type: {db_type}")
243
+
244
+ #
245
+ # End of DB-Ingestion functions
246
+ ############################################################################################################
247
+
248
+
249
+ ############################################################################################################
250
+ #
251
+ # Prompt-related functions
252
+
253
+ def list_prompts(*args, **kwargs):
254
+ if db_type == 'sqlite':
255
+ return sqlite_list_prompts(*args, **kwargs)
256
+ elif db_type == 'elasticsearch':
257
+ # Implement Elasticsearch version
258
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
259
+
260
+
261
+ def fetch_prompt_details(*args, **kwargs):
262
+ if db_type == 'sqlite':
263
+ return sqlite_fetch_prompt_details(*args, **kwargs)
264
+ elif db_type == 'elasticsearch':
265
+ # Implement Elasticsearch version
266
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
267
+
268
+ def add_prompt(*args, **kwargs):
269
+ if db_type == 'sqlite':
270
+ return sqlite_add_prompt(*args, **kwargs)
271
+ elif db_type == 'elasticsearch':
272
+ # Implement Elasticsearch version
273
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
274
+
275
+
276
+ def add_or_update_prompt(*args, **kwargs):
277
+ if db_type == 'sqlite':
278
+ return sqlite_add_or_update_prompt(*args, **kwargs)
279
+ elif db_type == 'elasticsearch':
280
+ # Implement Elasticsearch version
281
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
282
+
283
+ def load_prompt_details(*args, **kwargs):
284
+ if db_type == 'sqlite':
285
+ return sqlite_load_prompt_details(*args, **kwargs)
286
+ elif db_type == 'elasticsearch':
287
+ # Implement Elasticsearch version
288
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
289
+
290
+ def load_preset_prompts(*args, **kwargs):
291
+ if db_type == 'sqlite':
292
+ return sqlite_load_preset_prompts(*args, **kwargs)
293
+ elif db_type == 'elasticsearch':
294
+ # Implement Elasticsearch version
295
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
296
+
297
+ def insert_prompt_to_db(*args, **kwargs):
298
+ if db_type == 'sqlite':
299
+ return sqlite_insert_prompt_to_db(*args, **kwargs)
300
+ elif db_type == 'elasticsearch':
301
+ # Implement Elasticsearch version
302
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
303
+
304
+ def delete_prompt(*args, **kwargs):
305
+ if db_type == 'sqlite':
306
+ return sqlite_delete_prompt(*args, **kwargs)
307
+ elif db_type == 'elasticsearch':
308
+ # Implement Elasticsearch version
309
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
310
+
311
+ #
312
+ # End of Prompt-related functions
313
+ ############################################################################################################
314
+
315
+ ############################################################################################################
316
+ #
317
+ # Keywords-related Functions
318
+
319
+ def keywords_browser_interface(*args, **kwargs):
320
+ if db_type == 'sqlite':
321
+ return sqlite_keywords_browser_interface(*args, **kwargs)
322
+ elif db_type == 'elasticsearch':
323
+ # Implement Elasticsearch version
324
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
325
+
326
+ def add_keyword(*args, **kwargs):
327
+ if db_type == 'sqlite':
328
+ with db.get_connection() as conn:
329
+ cursor = conn.cursor()
330
+ return sqlite_add_keyword(*args, **kwargs)
331
+ elif db_type == 'elasticsearch':
332
+ # Implement Elasticsearch version
333
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
334
+
335
+ def delete_keyword(*args, **kwargs):
336
+ if db_type == 'sqlite':
337
+ return sqlite_delete_keyword(*args, **kwargs)
338
+ elif db_type == 'elasticsearch':
339
+ # Implement Elasticsearch version
340
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
341
+
342
+ def export_keywords_to_csv(*args, **kwargs):
343
+ if db_type == 'sqlite':
344
+ return sqlite_export_keywords_to_csv(*args, **kwargs)
345
+ elif db_type == 'elasticsearch':
346
+ # Implement Elasticsearch version
347
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
348
+
349
+ #
350
+ # End of Keywords-related Functions
351
+ ############################################################################################################
352
+
353
+ ############################################################################################################
354
+ #
355
+ # Chat-related Functions
356
+
357
+ def delete_chat_message(*args, **kwargs):
358
+ if db_type == 'sqlite':
359
+ return sqlite_delete_chat_message(*args, **kwargs)
360
+ elif db_type == 'elasticsearch':
361
+ # Implement Elasticsearch version
362
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
363
+
364
+ def update_chat_message(*args, **kwargs):
365
+ if db_type == 'sqlite':
366
+ return sqlite_update_chat_message(*args, **kwargs)
367
+ elif db_type == 'elasticsearch':
368
+ # Implement Elasticsearch version
369
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
370
+
371
+ def add_chat_message(*args, **kwargs):
372
+ if db_type == 'sqlite':
373
+ return sqlite_add_chat_message(*args, **kwargs)
374
+ elif db_type == 'elasticsearch':
375
+ # Implement Elasticsearch version
376
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
377
+
378
+ def get_chat_messages(*args, **kwargs):
379
+ if db_type == 'sqlite':
380
+ return sqlite_get_chat_messages(*args, **kwargs)
381
+ elif db_type == 'elasticsearch':
382
+ # Implement Elasticsearch version
383
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
384
+
385
+ def search_chat_conversations(*args, **kwargs):
386
+ if db_type == 'sqlite':
387
+ return sqlite_search_chat_conversations(*args, **kwargs)
388
+ elif db_type == 'elasticsearch':
389
+ # Implement Elasticsearch version
390
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
391
+
392
+ def create_chat_conversation(*args, **kwargs):
393
+ if db_type == 'sqlite':
394
+ return sqlite_create_chat_conversation(*args, **kwargs)
395
+ elif db_type == 'elasticsearch':
396
+ # Implement Elasticsearch version
397
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
398
+
399
+ def save_chat_history_to_database(*args, **kwargs):
400
+ if db_type == 'sqlite':
401
+ return sqlite_save_chat_history_to_database(*args, **kwargs)
402
+ elif db_type == 'elasticsearch':
403
+ # Implement Elasticsearch version
404
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
405
+
406
+ def get_conversation_name(*args, **kwargs):
407
+ if db_type == 'sqlite':
408
+ return sqlite_get_conversation_name(*args, **kwargs)
409
+ elif db_type == 'elasticsearch':
410
+ # Implement Elasticsearch version
411
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
412
+
413
+ #
414
+ # End of Chat-related Functions
415
+ ############################################################################################################
416
+
417
+ ############################################################################################################
418
+ #
419
+ # Trash-related Functions
420
+
421
+ def get_trashed_items(*args, **kwargs):
422
+ if db_type == 'sqlite':
423
+ return sqlite_get_trashed_items(*args, **kwargs)
424
+ elif db_type == 'elasticsearch':
425
+ # Implement Elasticsearch version
426
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
427
+
428
+ def user_delete_item(*args, **kwargs):
429
+ if db_type == 'sqlite':
430
+ return sqlite_user_delete_item(*args, **kwargs)
431
+ elif db_type == 'elasticsearch':
432
+ # Implement Elasticsearch version
433
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
434
+
435
+ def empty_trash(*args, **kwargs):
436
+ if db_type == 'sqlite':
437
+ return sqlite_empty_trash(*args, **kwargs)
438
+ elif db_type == 'elasticsearch':
439
+ # Implement Elasticsearch version
440
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
441
+
442
+ #
443
+ # End of Trash-related Functions
444
+ ############################################################################################################
445
+
446
+ ############################################################################################################
447
+ #
448
+ # DB-Backup Functions
449
+
450
+ def create_automated_backup(*args, **kwargs):
451
+ if db_type == 'sqlite':
452
+ return sqlite_create_automated_backup(*args, **kwargs)
453
+ elif db_type == 'elasticsearch':
454
+ # Implement Elasticsearch version
455
+ raise NotImplementedError("Elasticsearch version of add_media_with_keywords not yet implemented")
456
+
457
+ #
458
+ # End of DB-Backup Functions
459
+ ############################################################################################################
460
+
461
+ ############################################################################################################
462
+ #
463
+ # Function to close the database connection for SQLite
464
+
465
+ def close_connection():
466
+ if db_type == 'sqlite':
467
+ db.close_all_connections()
468
+ # Elasticsearch doesn't need explicit closing
469
+
470
+ #
471
+ # End of file
472
+ ############################################################################################################
App_Function_Libraries/Diarization_Lib.py CHANGED
@@ -33,7 +33,7 @@ import yaml
33
 
34
  def load_pipeline_from_pretrained(path_to_config: str | Path) -> SpeakerDiarization:
35
  path_to_config = Path(path_to_config).resolve()
36
- print(f"Loading pyannote pipeline from {path_to_config}...")
37
 
38
  if not path_to_config.exists():
39
  raise FileNotFoundError(f"Config file not found: {path_to_config}")
@@ -45,11 +45,6 @@ def load_pipeline_from_pretrained(path_to_config: str | Path) -> SpeakerDiarizat
45
  # Store current working directory
46
  cwd = Path.cwd().resolve()
47
 
48
- # Change to the directory containing the config file
49
- cd_to = path_to_config.parent.resolve()
50
- print(f"Changing working directory to {cd_to}")
51
- os.chdir(cd_to)
52
-
53
  try:
54
  # Create a SpeakerDiarization pipeline
55
  pipeline = SpeakerDiarization()
 
33
 
34
  def load_pipeline_from_pretrained(path_to_config: str | Path) -> SpeakerDiarization:
35
  path_to_config = Path(path_to_config).resolve()
36
+ logging.debug(f"Loading pyannote pipeline from {path_to_config}...")
37
 
38
  if not path_to_config.exists():
39
  raise FileNotFoundError(f"Config file not found: {path_to_config}")
 
45
  # Store current working directory
46
  cwd = Path.cwd().resolve()
47
 
 
 
 
 
 
48
  try:
49
  # Create a SpeakerDiarization pipeline
50
  pipeline = SpeakerDiarization()
App_Function_Libraries/Gradio_Related.py CHANGED
The diff for this file is too large to render. See raw diff
 
App_Function_Libraries/Gradio_UI/Audio_ingestion_tab.py ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Audio_ingestion_tab.py
2
+ # Description: Gradio UI for ingesting audio files into the database
3
+ #
4
+ # Imports
5
+ #
6
+ # External Imports
7
+ import gradio as gr
8
+ #
9
+ # Local Imports
10
+ from App_Function_Libraries.Audio_Files import process_audio_files
11
+ from App_Function_Libraries.DB_Manager import load_preset_prompts
12
+ from App_Function_Libraries.Gradio_UI.Chat_ui import update_user_prompt
13
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import whisper_models
14
+ #
15
+ #######################################################################################################################
16
+ # Functions:
17
+
18
+ def create_audio_processing_tab():
19
+ with gr.TabItem("Audio File Transcription + Summarization"):
20
+ gr.Markdown("# Transcribe & Summarize Audio Files from URLs or Local Files!")
21
+ with gr.Row():
22
+ with gr.Column():
23
+ audio_url_input = gr.Textbox(label="Audio File URL(s)", placeholder="Enter the URL(s) of the audio file(s), one per line")
24
+ audio_file_input = gr.File(label="Upload Audio File", file_types=["audio/*"])
25
+
26
+ use_cookies_input = gr.Checkbox(label="Use cookies for authenticated download", value=False)
27
+ cookies_input = gr.Textbox(
28
+ label="Audio Download Cookies",
29
+ placeholder="Paste your cookies here (JSON format)",
30
+ lines=3,
31
+ visible=False
32
+ )
33
+
34
+ use_cookies_input.change(
35
+ fn=lambda x: gr.update(visible=x),
36
+ inputs=[use_cookies_input],
37
+ outputs=[cookies_input]
38
+ )
39
+
40
+ diarize_input = gr.Checkbox(label="Enable Speaker Diarization", value=False)
41
+ whisper_model_input = gr.Dropdown(choices=whisper_models, value="medium", label="Whisper Model")
42
+
43
+ with gr.Row():
44
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
45
+ value=False,
46
+ visible=True)
47
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
48
+ value=False,
49
+ visible=True)
50
+ with gr.Row():
51
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
52
+ choices=load_preset_prompts(),
53
+ visible=False)
54
+ with gr.Row():
55
+ custom_prompt_input = gr.Textbox(label="Custom Prompt",
56
+ placeholder="Enter custom prompt here",
57
+ lines=3,
58
+ visible=False)
59
+ with gr.Row():
60
+ system_prompt_input = gr.Textbox(label="System Prompt",
61
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
62
+ **Bulleted Note Creation Guidelines**
63
+
64
+ **Headings**:
65
+ - Based on referenced topics, not categories like quotes or terms
66
+ - Surrounded by **bold** formatting
67
+ - Not listed as bullet points
68
+ - No space between headings and list items underneath
69
+
70
+ **Emphasis**:
71
+ - **Important terms** set in bold font
72
+ - **Text ending in a colon**: also bolded
73
+
74
+ **Review**:
75
+ - Ensure adherence to specified format
76
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
77
+ """,
78
+ lines=3,
79
+ visible=False)
80
+
81
+ custom_prompt_checkbox.change(
82
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
83
+ inputs=[custom_prompt_checkbox],
84
+ outputs=[custom_prompt_input, system_prompt_input]
85
+ )
86
+ preset_prompt_checkbox.change(
87
+ fn=lambda x: gr.update(visible=x),
88
+ inputs=[preset_prompt_checkbox],
89
+ outputs=[preset_prompt]
90
+ )
91
+
92
+ def update_prompts(preset_name):
93
+ prompts = update_user_prompt(preset_name)
94
+ return (
95
+ gr.update(value=prompts["user_prompt"], visible=True),
96
+ gr.update(value=prompts["system_prompt"], visible=True)
97
+ )
98
+
99
+ preset_prompt.change(
100
+ update_prompts,
101
+ inputs=preset_prompt,
102
+ outputs=[custom_prompt_input, system_prompt_input]
103
+ )
104
+
105
+ api_name_input = gr.Dropdown(
106
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
107
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
108
+ value=None,
109
+ label="API for Summarization (Optional)"
110
+ )
111
+ api_key_input = gr.Textbox(label="API Key (if required)", placeholder="Enter your API key here", type="password")
112
+ custom_keywords_input = gr.Textbox(label="Custom Keywords", placeholder="Enter custom keywords, comma-separated")
113
+ keep_original_input = gr.Checkbox(label="Keep original audio file", value=False)
114
+
115
+ chunking_options_checkbox = gr.Checkbox(label="Show Chunking Options", value=False)
116
+ with gr.Row(visible=False) as chunking_options_box:
117
+ gr.Markdown("### Chunking Options")
118
+ with gr.Column():
119
+ chunk_method = gr.Dropdown(choices=['words', 'sentences', 'paragraphs', 'tokens'], label="Chunking Method")
120
+ max_chunk_size = gr.Slider(minimum=100, maximum=1000, value=300, step=50, label="Max Chunk Size")
121
+ chunk_overlap = gr.Slider(minimum=0, maximum=100, value=0, step=10, label="Chunk Overlap")
122
+ use_adaptive_chunking = gr.Checkbox(label="Use Adaptive Chunking")
123
+ use_multi_level_chunking = gr.Checkbox(label="Use Multi-level Chunking")
124
+ chunk_language = gr.Dropdown(choices=['english', 'french', 'german', 'spanish'], label="Chunking Language")
125
+
126
+ chunking_options_checkbox.change(
127
+ fn=lambda x: gr.update(visible=x),
128
+ inputs=[chunking_options_checkbox],
129
+ outputs=[chunking_options_box]
130
+ )
131
+
132
+ process_audio_button = gr.Button("Process Audio File(s)")
133
+
134
+ with gr.Column():
135
+ audio_progress_output = gr.Textbox(label="Progress")
136
+ audio_transcription_output = gr.Textbox(label="Transcription")
137
+ audio_summary_output = gr.Textbox(label="Summary")
138
+ download_transcription = gr.File(label="Download All Transcriptions as JSON")
139
+ download_summary = gr.File(label="Download All Summaries as Text")
140
+
141
+ process_audio_button.click(
142
+ fn=process_audio_files,
143
+ inputs=[audio_url_input, audio_file_input, whisper_model_input, api_name_input, api_key_input,
144
+ use_cookies_input, cookies_input, keep_original_input, custom_keywords_input, custom_prompt_input,
145
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking, use_multi_level_chunking,
146
+ chunk_language, diarize_input],
147
+ outputs=[audio_progress_output, audio_transcription_output, audio_summary_output]
148
+ )
149
+
150
+ #
151
+ # End of Audio_ingestion_tab.py
152
+ #######################################################################################################################
App_Function_Libraries/Gradio_UI/Chat_ui.py ADDED
@@ -0,0 +1,1017 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Chat_ui.py
2
+ # Description: Chat interface functions for Gradio
3
+ #
4
+ # Imports
5
+ import html
6
+ import json
7
+ import logging
8
+ import os
9
+ import sqlite3
10
+ from datetime import datetime
11
+ from pathlib import Path
12
+ #
13
+ # External Imports
14
+ import gradio as gr
15
+ #
16
+ # Local Imports
17
+ from App_Function_Libraries.Chat import chat, save_chat_history, update_chat_content, save_chat_history_to_db_wrapper
18
+ from App_Function_Libraries.DB_Manager import add_chat_message, search_chat_conversations, create_chat_conversation, \
19
+ get_chat_messages, update_chat_message, delete_chat_message, load_preset_prompts, db
20
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import update_dropdown, update_user_prompt
21
+
22
+
23
+ #
24
+ #
25
+ ########################################################################################################################
26
+ #
27
+ # Functions:
28
+
29
+
30
+ def show_edit_message(selected):
31
+ if selected:
32
+ return gr.update(value=selected[0], visible=True), gr.update(value=selected[1], visible=True), gr.update(
33
+ visible=True)
34
+ return gr.update(visible=False), gr.update(visible=False), gr.update(visible=False)
35
+
36
+
37
+ def show_delete_message(selected):
38
+ if selected:
39
+ return gr.update(value=selected[1], visible=True), gr.update(visible=True)
40
+ return gr.update(visible=False), gr.update(visible=False)
41
+
42
+
43
+ def debug_output(media_content, selected_parts):
44
+ print(f"Debug - Media Content: {media_content}")
45
+ print(f"Debug - Selected Parts: {selected_parts}")
46
+ return ""
47
+
48
+
49
+ def update_selected_parts(use_content, use_summary, use_prompt):
50
+ selected_parts = []
51
+ if use_content:
52
+ selected_parts.append("content")
53
+ if use_summary:
54
+ selected_parts.append("summary")
55
+ if use_prompt:
56
+ selected_parts.append("prompt")
57
+ print(f"Debug - Update Selected Parts: {selected_parts}")
58
+ return selected_parts
59
+
60
+
61
+ # Old update_user_prompt shim for backwards compatibility
62
+ def get_system_prompt(preset_name):
63
+ # For backwards compatibility
64
+ prompts = update_user_prompt(preset_name)
65
+ return prompts["system_prompt"]
66
+
67
+ def clear_chat():
68
+ # Return empty list for chatbot and None for conversation_id
69
+ return gr.update(value=[]), None
70
+
71
+
72
+ # FIXME - add additional features....
73
+ def chat_wrapper(message, history, media_content, selected_parts, api_endpoint, api_key, custom_prompt, conversation_id,
74
+ save_conversation, temperature, system_prompt, max_tokens=None, top_p=None, frequency_penalty=None,
75
+ presence_penalty=None, stop_sequence=None):
76
+ try:
77
+ if save_conversation:
78
+ if conversation_id is None:
79
+ # Create a new conversation
80
+ media_id = media_content.get('id', None)
81
+ conversation_name = f"Chat about {media_content.get('title', 'Unknown Media')} - {datetime.now().strftime('%Y-%m-%d %H:%M:%S')}"
82
+ conversation_id = create_chat_conversation(media_id, conversation_name)
83
+
84
+ # Add user message to the database
85
+ user_message_id = add_chat_message(conversation_id, "user", message)
86
+
87
+ # Include the selected parts and custom_prompt only for the first message
88
+ if not history and selected_parts:
89
+ message_body = "\n".join(selected_parts)
90
+ full_message = f"{custom_prompt}\n\n{message}\n\n{message_body}"
91
+ elif custom_prompt:
92
+ full_message = f"{custom_prompt}\n\n{message}"
93
+ else:
94
+ full_message = message
95
+
96
+ # Generate bot response
97
+ bot_message = chat(full_message, history, media_content, selected_parts, api_endpoint, api_key, custom_prompt,
98
+ temperature, system_prompt)
99
+
100
+ if save_conversation:
101
+ # Add assistant message to the database
102
+ add_chat_message(conversation_id, "assistant", bot_message)
103
+
104
+ # Update history
105
+ history.append((message, bot_message))
106
+
107
+ return bot_message, history, conversation_id
108
+ except Exception as e:
109
+ logging.error(f"Error in chat wrapper: {str(e)}")
110
+ return "An error occurred.", history, conversation_id
111
+
112
+
113
+ def search_conversations(query):
114
+ try:
115
+ conversations = search_chat_conversations(query)
116
+ if not conversations:
117
+ print(f"Debug - Search Conversations - No results found for query: {query}")
118
+ return gr.update(choices=[])
119
+
120
+ conversation_options = [
121
+ (f"{c['conversation_name']} (Media: {c['media_title']}, ID: {c['id']})", c['id'])
122
+ for c in conversations
123
+ ]
124
+ print(f"Debug - Search Conversations - Options: {conversation_options}")
125
+ return gr.update(choices=conversation_options)
126
+ except Exception as e:
127
+ print(f"Debug - Search Conversations - Error: {str(e)}")
128
+ return gr.update(choices=[])
129
+
130
+
131
+ def load_conversation(conversation_id):
132
+ if not conversation_id:
133
+ return [], None
134
+
135
+ messages = get_chat_messages(conversation_id)
136
+ history = [
137
+ (msg['message'], None) if msg['sender'] == 'user' else (None, msg['message'])
138
+ for msg in messages
139
+ ]
140
+ return history, conversation_id
141
+
142
+
143
+ def update_message_in_chat(message_id, new_text, history):
144
+ update_chat_message(message_id, new_text)
145
+ updated_history = [(msg1, msg2) if msg1[1] != message_id and msg2[1] != message_id
146
+ else ((new_text, msg1[1]) if msg1[1] == message_id else (new_text, msg2[1]))
147
+ for msg1, msg2 in history]
148
+ return updated_history
149
+
150
+
151
+ def delete_message_from_chat(message_id, history):
152
+ delete_chat_message(message_id)
153
+ updated_history = [(msg1, msg2) for msg1, msg2 in history if msg1[1] != message_id and msg2[1] != message_id]
154
+ return updated_history
155
+
156
+
157
+ def create_chat_interface():
158
+ custom_css = """
159
+ .chatbot-container .message-wrap .message {
160
+ font-size: 14px !important;
161
+ }
162
+ """
163
+ with gr.TabItem("Remote LLM Chat (Horizontal)"):
164
+ gr.Markdown("# Chat with a designated LLM Endpoint, using your selected item as starting context")
165
+ chat_history = gr.State([])
166
+ media_content = gr.State({})
167
+ selected_parts = gr.State([])
168
+ conversation_id = gr.State(None)
169
+
170
+ with gr.Row():
171
+ with gr.Column(scale=1):
172
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
173
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title",
174
+ label="Search By")
175
+ search_button = gr.Button("Search")
176
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
177
+ item_mapping = gr.State({})
178
+ with gr.Row():
179
+ use_content = gr.Checkbox(label="Use Content")
180
+ use_summary = gr.Checkbox(label="Use Summary")
181
+ use_prompt = gr.Checkbox(label="Use Prompt")
182
+ save_conversation = gr.Checkbox(label="Save Conversation", value=False, visible=True)
183
+ with gr.Row():
184
+ temperature = gr.Slider(label="Temperature", minimum=0.00, maximum=1.0, step=0.05, value=0.7)
185
+ with gr.Row():
186
+ conversation_search = gr.Textbox(label="Search Conversations")
187
+ with gr.Row():
188
+ search_conversations_btn = gr.Button("Search Conversations")
189
+ with gr.Row():
190
+ previous_conversations = gr.Dropdown(label="Select Conversation", choices=[], interactive=True)
191
+ with gr.Row():
192
+ load_conversations_btn = gr.Button("Load Selected Conversation")
193
+
194
+ api_endpoint = gr.Dropdown(label="Select API Endpoint",
195
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek",
196
+ "Mistral", "OpenRouter",
197
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama",
198
+ "HuggingFace"])
199
+ api_key = gr.Textbox(label="API Key (if required)", type="password")
200
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
201
+ value=False,
202
+ visible=True)
203
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
204
+ value=False,
205
+ visible=True)
206
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
207
+ choices=load_preset_prompts(),
208
+ visible=False)
209
+ user_prompt = gr.Textbox(label="Custom Prompt",
210
+ placeholder="Enter custom prompt here",
211
+ lines=3,
212
+ visible=False)
213
+ system_prompt_input = gr.Textbox(label="System Prompt",
214
+ value="You are a helpful AI assitant",
215
+ lines=3,
216
+ visible=False)
217
+ with gr.Column():
218
+ chatbot = gr.Chatbot(height=600, elem_classes="chatbot-container")
219
+ msg = gr.Textbox(label="Enter your message")
220
+ submit = gr.Button("Submit")
221
+ clear_chat_button = gr.Button("Clear Chat")
222
+
223
+ edit_message_id = gr.Number(label="Message ID to Edit", visible=False)
224
+ edit_message_text = gr.Textbox(label="Edit Message", visible=False)
225
+ update_message_button = gr.Button("Update Message", visible=False)
226
+
227
+ delete_message_id = gr.Number(label="Message ID to Delete", visible=False)
228
+ delete_message_button = gr.Button("Delete Message", visible=False)
229
+
230
+ save_chat_history_to_db = gr.Button("Save Chat History to DataBase")
231
+ save_chat_history_as_file = gr.Button("Save Chat History as File")
232
+ download_file = gr.File(label="Download Chat History")
233
+ save_status = gr.Textbox(label="Save Status", interactive=False)
234
+
235
+ # Restore original functionality
236
+ search_button.click(
237
+ fn=update_dropdown,
238
+ inputs=[search_query_input, search_type_input],
239
+ outputs=[items_output, item_mapping]
240
+ )
241
+
242
+ def save_chat_wrapper(history, conversation_id, media_content):
243
+ file_path = save_chat_history(history, conversation_id, media_content)
244
+ if file_path:
245
+ return file_path, f"Chat history saved successfully as {os.path.basename(file_path)}!"
246
+ else:
247
+ return None, "Error saving chat history. Please check the logs and try again."
248
+
249
+ save_chat_history_as_file.click(
250
+ save_chat_wrapper,
251
+ inputs=[chatbot, conversation_id, media_content],
252
+ outputs=[download_file, save_status]
253
+ )
254
+
255
+ def update_prompts(preset_name):
256
+ prompts = update_user_prompt(preset_name)
257
+ return (
258
+ gr.update(value=prompts["user_prompt"], visible=True),
259
+ gr.update(value=prompts["system_prompt"], visible=True)
260
+ )
261
+
262
+ def clear_chat():
263
+ return [], None # Return empty list for chatbot and None for conversation_id
264
+
265
+ clear_chat_button.click(
266
+ clear_chat,
267
+ outputs=[chatbot, conversation_id]
268
+ )
269
+ preset_prompt.change(
270
+ update_prompts,
271
+ inputs=preset_prompt,
272
+ outputs=[user_prompt, system_prompt_input]
273
+ )
274
+ custom_prompt_checkbox.change(
275
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
276
+ inputs=[custom_prompt_checkbox],
277
+ outputs=[user_prompt, system_prompt_input]
278
+ )
279
+ preset_prompt_checkbox.change(
280
+ fn=lambda x: gr.update(visible=x),
281
+ inputs=[preset_prompt_checkbox],
282
+ outputs=[preset_prompt]
283
+ )
284
+
285
+ submit.click(
286
+ chat_wrapper,
287
+ inputs=[msg, chatbot, media_content, selected_parts, api_endpoint, api_key, user_prompt,
288
+ conversation_id, save_conversation, temperature, system_prompt_input],
289
+ outputs=[msg, chatbot, conversation_id]
290
+ ).then( # Clear the message box after submission
291
+ lambda x: gr.update(value=""),
292
+ inputs=[chatbot],
293
+ outputs=[msg]
294
+ ).then( # Clear the user prompt after the first message
295
+ lambda: (gr.update(value=""), gr.update(value="")),
296
+ outputs=[user_prompt, system_prompt_input]
297
+ )
298
+
299
+ items_output.change(
300
+ update_chat_content,
301
+ inputs=[items_output, use_content, use_summary, use_prompt, item_mapping],
302
+ outputs=[media_content, selected_parts]
303
+ )
304
+ use_content.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
305
+ outputs=[selected_parts])
306
+ use_summary.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
307
+ outputs=[selected_parts])
308
+ use_prompt.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
309
+ outputs=[selected_parts])
310
+ items_output.change(debug_output, inputs=[media_content, selected_parts], outputs=[])
311
+
312
+ search_conversations_btn.click(
313
+ search_conversations,
314
+ inputs=[conversation_search],
315
+ outputs=[previous_conversations]
316
+ )
317
+
318
+ load_conversations_btn.click(
319
+ clear_chat,
320
+ outputs=[chatbot, chat_history]
321
+ ).then(
322
+ load_conversation,
323
+ inputs=[previous_conversations],
324
+ outputs=[chatbot, conversation_id]
325
+ )
326
+
327
+ previous_conversations.change(
328
+ load_conversation,
329
+ inputs=[previous_conversations],
330
+ outputs=[chat_history]
331
+ )
332
+
333
+ update_message_button.click(
334
+ update_message_in_chat,
335
+ inputs=[edit_message_id, edit_message_text, chat_history],
336
+ outputs=[chatbot]
337
+ )
338
+
339
+ delete_message_button.click(
340
+ delete_message_from_chat,
341
+ inputs=[delete_message_id, chat_history],
342
+ outputs=[chatbot]
343
+ )
344
+
345
+ save_chat_history_as_file.click(
346
+ save_chat_history,
347
+ inputs=[chatbot, conversation_id],
348
+ outputs=[download_file]
349
+ )
350
+
351
+ save_chat_history_to_db.click(
352
+ save_chat_history_to_db_wrapper,
353
+ inputs=[chatbot, conversation_id, media_content],
354
+ outputs=[conversation_id, gr.Textbox(label="Save Status")]
355
+ )
356
+
357
+ chatbot.select(show_edit_message, None, [edit_message_text, edit_message_id, update_message_button])
358
+ chatbot.select(show_delete_message, None, [delete_message_id, delete_message_button])
359
+
360
+
361
+ def create_chat_interface_stacked():
362
+ custom_css = """
363
+ .chatbot-container .message-wrap .message {
364
+ font-size: 14px !important;
365
+ }
366
+ """
367
+ with gr.TabItem("Remote LLM Chat - Stacked"):
368
+ gr.Markdown("# Stacked Chat")
369
+ chat_history = gr.State([])
370
+ media_content = gr.State({})
371
+ selected_parts = gr.State([])
372
+ conversation_id = gr.State(None)
373
+
374
+ with gr.Row():
375
+ with gr.Column():
376
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
377
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title",
378
+ label="Search By")
379
+ search_button = gr.Button("Search")
380
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
381
+ item_mapping = gr.State({})
382
+ with gr.Row():
383
+ use_content = gr.Checkbox(label="Use Content")
384
+ use_summary = gr.Checkbox(label="Use Summary")
385
+ use_prompt = gr.Checkbox(label="Use Prompt")
386
+ save_conversation = gr.Checkbox(label="Save Conversation", value=False, visible=True)
387
+ temp = gr.Slider(label="Temperature", minimum=0.00, maximum=1.0, step=0.05, value=0.7)
388
+ with gr.Row():
389
+ conversation_search = gr.Textbox(label="Search Conversations")
390
+ with gr.Row():
391
+ previous_conversations = gr.Dropdown(label="Select Conversation", choices=[], interactive=True)
392
+ with gr.Row():
393
+ search_conversations_btn = gr.Button("Search Conversations")
394
+ load_conversations_btn = gr.Button("Load Selected Conversation")
395
+ with gr.Column():
396
+ api_endpoint = gr.Dropdown(label="Select API Endpoint",
397
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek",
398
+ "OpenRouter", "Mistral", "Llama.cpp", "Kobold", "Ooba", "Tabbyapi",
399
+ "VLLM", "ollama", "HuggingFace"])
400
+ api_key = gr.Textbox(label="API Key (if required)", type="password")
401
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
402
+ choices=load_preset_prompts(),
403
+ visible=True)
404
+ system_prompt = gr.Textbox(label="System Prompt",
405
+ value="You are a helpful AI assistant.",
406
+ lines=3,
407
+ visible=True)
408
+ user_prompt = gr.Textbox(label="Custom User Prompt",
409
+ placeholder="Enter custom prompt here",
410
+ lines=3,
411
+ visible=True)
412
+ gr.Markdown("Scroll down for the chat window...")
413
+ with gr.Row():
414
+ with gr.Column(scale=1):
415
+ chatbot = gr.Chatbot(height=600, elem_classes="chatbot-container")
416
+ msg = gr.Textbox(label="Enter your message")
417
+ with gr.Row():
418
+ with gr.Column():
419
+ submit = gr.Button("Submit")
420
+ clear_chat_button = gr.Button("Clear Chat")
421
+
422
+ edit_message_id = gr.Number(label="Message ID to Edit", visible=False)
423
+ edit_message_text = gr.Textbox(label="Edit Message", visible=False)
424
+ update_message_button = gr.Button("Update Message", visible=False)
425
+
426
+ delete_message_id = gr.Number(label="Message ID to Delete", visible=False)
427
+ delete_message_button = gr.Button("Delete Message", visible=False)
428
+ save_chat_history_to_db = gr.Button("Save Chat History to DataBase")
429
+ save_chat_history_as_file = gr.Button("Save Chat History as File")
430
+ with gr.Column():
431
+ download_file = gr.File(label="Download Chat History")
432
+
433
+ # Restore original functionality
434
+ search_button.click(
435
+ fn=update_dropdown,
436
+ inputs=[search_query_input, search_type_input],
437
+ outputs=[items_output, item_mapping]
438
+ )
439
+
440
+ def update_prompts(preset_name):
441
+ prompts = update_user_prompt(preset_name)
442
+ return (
443
+ gr.update(value=prompts["user_prompt"], visible=True),
444
+ gr.update(value=prompts["system_prompt"], visible=True)
445
+ )
446
+
447
+ clear_chat_button.click(
448
+ clear_chat,
449
+ outputs=[chatbot, conversation_id]
450
+ )
451
+ preset_prompt.change(
452
+ update_prompts,
453
+ inputs=preset_prompt,
454
+ outputs=[user_prompt, system_prompt]
455
+ )
456
+
457
+ submit.click(
458
+ chat_wrapper,
459
+ inputs=[msg, chatbot, media_content, selected_parts, api_endpoint, api_key, user_prompt,
460
+ conversation_id, save_conversation, temp, system_prompt],
461
+ outputs=[msg, chatbot, conversation_id]
462
+ ).then( # Clear the message box after submission
463
+ lambda x: gr.update(value=""),
464
+ inputs=[chatbot],
465
+ outputs=[msg]
466
+ ).then( # Clear the user prompt after the first message
467
+ lambda: gr.update(value=""),
468
+ outputs=[user_prompt, system_prompt]
469
+ )
470
+
471
+ items_output.change(
472
+ update_chat_content,
473
+ inputs=[items_output, use_content, use_summary, use_prompt, item_mapping],
474
+ outputs=[media_content, selected_parts]
475
+ )
476
+ use_content.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
477
+ outputs=[selected_parts])
478
+ use_summary.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
479
+ outputs=[selected_parts])
480
+ use_prompt.change(update_selected_parts, inputs=[use_content, use_summary, use_prompt],
481
+ outputs=[selected_parts])
482
+ items_output.change(debug_output, inputs=[media_content, selected_parts], outputs=[])
483
+
484
+ search_conversations_btn.click(
485
+ search_conversations,
486
+ inputs=[conversation_search],
487
+ outputs=[previous_conversations]
488
+ )
489
+
490
+ load_conversations_btn.click(
491
+ clear_chat,
492
+ outputs=[chatbot, chat_history]
493
+ ).then(
494
+ load_conversation,
495
+ inputs=[previous_conversations],
496
+ outputs=[chatbot, conversation_id]
497
+ )
498
+
499
+ previous_conversations.change(
500
+ load_conversation,
501
+ inputs=[previous_conversations],
502
+ outputs=[chat_history]
503
+ )
504
+
505
+ update_message_button.click(
506
+ update_message_in_chat,
507
+ inputs=[edit_message_id, edit_message_text, chat_history],
508
+ outputs=[chatbot]
509
+ )
510
+
511
+ delete_message_button.click(
512
+ delete_message_from_chat,
513
+ inputs=[delete_message_id, chat_history],
514
+ outputs=[chatbot]
515
+ )
516
+
517
+ save_chat_history_as_file.click(
518
+ save_chat_history,
519
+ inputs=[chatbot, conversation_id],
520
+ outputs=[download_file]
521
+ )
522
+
523
+ save_chat_history_to_db.click(
524
+ save_chat_history_to_db_wrapper,
525
+ inputs=[chatbot, conversation_id, media_content],
526
+ outputs=[conversation_id, gr.Textbox(label="Save Status")]
527
+ )
528
+
529
+ chatbot.select(show_edit_message, None, [edit_message_text, edit_message_id, update_message_button])
530
+ chatbot.select(show_delete_message, None, [delete_message_id, delete_message_button])
531
+
532
+
533
+ # FIXME - System prompts
534
+ def create_chat_interface_multi_api():
535
+ custom_css = """
536
+ .chatbot-container .message-wrap .message {
537
+ font-size: 14px !important;
538
+ }
539
+ .chat-window {
540
+ height: 400px;
541
+ overflow-y: auto;
542
+ }
543
+ """
544
+ with gr.TabItem("One Prompt - Multiple APIs"):
545
+ gr.Markdown("# One Prompt but Multiple API Chat Interface")
546
+
547
+ with gr.Row():
548
+ with gr.Column(scale=1):
549
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
550
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title",
551
+ label="Search By")
552
+ search_button = gr.Button("Search")
553
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
554
+ item_mapping = gr.State({})
555
+ with gr.Row():
556
+ use_content = gr.Checkbox(label="Use Content")
557
+ use_summary = gr.Checkbox(label="Use Summary")
558
+ use_prompt = gr.Checkbox(label="Use Prompt")
559
+ with gr.Column():
560
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt", choices=load_preset_prompts(), visible=True)
561
+ system_prompt = gr.Textbox(label="System Prompt", value="You are a helpful AI assistant.", lines=5)
562
+ user_prompt = gr.Textbox(label="Modify Prompt", lines=5, value=".")
563
+
564
+ with gr.Row():
565
+ chatbots = []
566
+ api_endpoints = []
567
+ api_keys = []
568
+ temperatures = []
569
+ for i in range(3):
570
+ with gr.Column():
571
+ gr.Markdown(f"### Chat Window {i + 1}")
572
+ api_endpoint = gr.Dropdown(label=f"API Endpoint {i + 1}",
573
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq",
574
+ "DeepSeek", "Mistral", "OpenRouter", "Llama.cpp", "Kobold",
575
+ "Ooba",
576
+ "Tabbyapi", "VLLM", "ollama", "HuggingFace"])
577
+ api_key = gr.Textbox(label=f"API Key {i + 1} (if required)", type="password")
578
+ temperature = gr.Slider(label=f"Temperature {i + 1}", minimum=0.0, maximum=1.0, step=0.05,
579
+ value=0.7)
580
+ chatbot = gr.Chatbot(height=800, elem_classes="chat-window")
581
+ chatbots.append(chatbot)
582
+ api_endpoints.append(api_endpoint)
583
+ api_keys.append(api_key)
584
+ temperatures.append(temperature)
585
+
586
+ with gr.Row():
587
+ msg = gr.Textbox(label="Enter your message", scale=4)
588
+ submit = gr.Button("Submit", scale=1)
589
+ # FIXME - clear chat
590
+ # clear_chat_button = gr.Button("Clear Chat")
591
+ #
592
+ # clear_chat_button.click(
593
+ # clear_chat,
594
+ # outputs=[chatbot]
595
+ # )
596
+
597
+ # State variables
598
+ chat_history = [gr.State([]) for _ in range(3)]
599
+ media_content = gr.State({})
600
+ selected_parts = gr.State([])
601
+ conversation_id = gr.State(None)
602
+
603
+ # Event handlers
604
+ search_button.click(
605
+ fn=update_dropdown,
606
+ inputs=[search_query_input, search_type_input],
607
+ outputs=[items_output, item_mapping]
608
+ )
609
+
610
+ preset_prompt.change(update_user_prompt, inputs=preset_prompt, outputs=user_prompt)
611
+
612
+ def chat_wrapper_multi(message, custom_prompt, system_prompt, *args):
613
+ chat_histories = args[:3]
614
+ chatbots = args[3:6]
615
+ api_endpoints = args[6:9]
616
+ api_keys = args[9:12]
617
+ temperatures = args[12:15]
618
+ media_content = args[15]
619
+ selected_parts = args[16]
620
+
621
+ new_chat_histories = []
622
+ new_chatbots = []
623
+
624
+ for i in range(3):
625
+ # Call chat_wrapper with dummy values for conversation_id and save_conversation
626
+ bot_message, new_history, _ = chat_wrapper(
627
+ message, chat_histories[i], media_content, selected_parts,
628
+ api_endpoints[i], api_keys[i], custom_prompt, None, # None for conversation_id
629
+ False, # False for save_conversation
630
+ temperature=temperatures[i],
631
+ system_prompt=system_prompt
632
+ )
633
+
634
+ new_chatbot = chatbots[i] + [(message, bot_message)]
635
+
636
+ new_chat_histories.append(new_history)
637
+ new_chatbots.append(new_chatbot)
638
+
639
+ return [gr.update(value="")] + new_chatbots + new_chat_histories
640
+
641
+ # In the create_chat_interface_multi_api function:
642
+ submit.click(
643
+ chat_wrapper_multi,
644
+ inputs=[msg, user_prompt,
645
+ system_prompt] + chat_history + chatbots + api_endpoints + api_keys + temperatures +
646
+ [media_content, selected_parts],
647
+ outputs=[msg] + chatbots + chat_history
648
+ ).then(
649
+ lambda: (gr.update(value=""), gr.update(value="")),
650
+ outputs=[msg, user_prompt]
651
+ )
652
+
653
+ items_output.change(
654
+ update_chat_content,
655
+ inputs=[items_output, use_content, use_summary, use_prompt, item_mapping],
656
+ outputs=[media_content, selected_parts]
657
+ )
658
+
659
+ for checkbox in [use_content, use_summary, use_prompt]:
660
+ checkbox.change(
661
+ update_selected_parts,
662
+ inputs=[use_content, use_summary, use_prompt],
663
+ outputs=[selected_parts]
664
+ )
665
+
666
+
667
+ def create_chat_interface_four():
668
+ custom_css = """
669
+ .chatbot-container .message-wrap .message {
670
+ font-size: 14px !important;
671
+ }
672
+ .chat-window {
673
+ height: 400px;
674
+ overflow-y: auto;
675
+ }
676
+ """
677
+ with gr.TabItem("Four Independent API Chats"):
678
+ gr.Markdown("# Four Independent API Chat Interfaces")
679
+
680
+ with gr.Row():
681
+ with gr.Column():
682
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt", choices=load_preset_prompts(), visible=True)
683
+ user_prompt = gr.Textbox(label="Modify Prompt", lines=3, value=".")
684
+ with gr.Column():
685
+ gr.Markdown("Scroll down for the chat windows...")
686
+ chat_interfaces = []
687
+ for row in range(2):
688
+ with gr.Row():
689
+ for col in range(2):
690
+ i = row * 2 + col
691
+ with gr.Column():
692
+ gr.Markdown(f"### Chat Window {i + 1}")
693
+ api_endpoint = gr.Dropdown(label=f"API Endpoint {i + 1}",
694
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq",
695
+ "DeepSeek", "Mistral", "OpenRouter", "Llama.cpp", "Kobold",
696
+ "Ooba",
697
+ "Tabbyapi", "VLLM", "ollama", "HuggingFace"])
698
+ api_key = gr.Textbox(label=f"API Key {i + 1} (if required)", type="password")
699
+ temperature = gr.Slider(label=f"Temperature {i + 1}", minimum=0.0, maximum=1.0, step=0.05,
700
+ value=0.7)
701
+ chatbot = gr.Chatbot(height=400, elem_classes="chat-window")
702
+ msg = gr.Textbox(label=f"Enter your message for Chat {i + 1}")
703
+ submit = gr.Button(f"Submit to Chat {i + 1}")
704
+
705
+ chat_interfaces.append({
706
+ 'api_endpoint': api_endpoint,
707
+ 'api_key': api_key,
708
+ 'temperature': temperature,
709
+ 'chatbot': chatbot,
710
+ 'msg': msg,
711
+ 'submit': submit,
712
+ 'chat_history': gr.State([])
713
+ })
714
+
715
+ preset_prompt.change(update_user_prompt, inputs=preset_prompt, outputs=user_prompt)
716
+
717
+ def chat_wrapper_single(message, chat_history, api_endpoint, api_key, temperature, user_prompt):
718
+ logging.debug(f"Chat Wrapper Single - Message: {message}, Chat History: {chat_history}")
719
+ new_msg, new_history, _ = chat_wrapper(
720
+ message, chat_history, {}, [], # Empty media_content and selected_parts
721
+ api_endpoint, api_key, user_prompt, None, # No conversation_id
722
+ False, # Not saving conversation
723
+ temperature=temperature, system_prompt=""
724
+ )
725
+ chat_history.append((message, new_msg))
726
+ return "", chat_history, chat_history
727
+
728
+ for interface in chat_interfaces:
729
+ logging.debug(f"Chat Interface - Clicked Submit for Chat {interface['chatbot']}"),
730
+ interface['submit'].click(
731
+ chat_wrapper_single,
732
+ inputs=[
733
+ interface['msg'],
734
+ interface['chat_history'],
735
+ interface['api_endpoint'],
736
+ interface['api_key'],
737
+ interface['temperature'],
738
+ user_prompt
739
+ ],
740
+ outputs=[
741
+ interface['msg'],
742
+ interface['chatbot'],
743
+ interface['chat_history']
744
+ ]
745
+ )
746
+
747
+
748
+ def chat_wrapper_single(message, chat_history, chatbot, api_endpoint, api_key, temperature, media_content,
749
+ selected_parts, conversation_id, save_conversation, user_prompt):
750
+ new_msg, new_history, new_conv_id = chat_wrapper(
751
+ message, chat_history, media_content, selected_parts,
752
+ api_endpoint, api_key, user_prompt, conversation_id,
753
+ save_conversation, temperature, system_prompt=""
754
+ )
755
+
756
+ if new_msg:
757
+ updated_chatbot = chatbot + [(message, new_msg)]
758
+ else:
759
+ updated_chatbot = chatbot
760
+
761
+ return new_msg, updated_chatbot, new_history, new_conv_id
762
+
763
+
764
+ # FIXME - Finish implementing functions + testing/valdidation
765
+ def create_chat_management_tab():
766
+ with gr.TabItem("Chat Management"):
767
+ gr.Markdown("# Chat Management")
768
+
769
+ with gr.Row():
770
+ search_query = gr.Textbox(label="Search Conversations")
771
+ search_button = gr.Button("Search")
772
+
773
+ conversation_list = gr.Dropdown(label="Select Conversation", choices=[])
774
+ conversation_mapping = gr.State({})
775
+
776
+ with gr.Tabs():
777
+ with gr.TabItem("Edit"):
778
+ chat_content = gr.TextArea(label="Chat Content (JSON)", lines=20, max_lines=50)
779
+ save_button = gr.Button("Save Changes")
780
+
781
+ with gr.TabItem("Preview"):
782
+ chat_preview = gr.HTML(label="Chat Preview")
783
+ result_message = gr.Markdown("")
784
+
785
+ def search_conversations(query):
786
+ conversations = search_chat_conversations(query)
787
+ choices = [f"{conv['conversation_name']} (Media: {conv['media_title']}, ID: {conv['id']})" for conv in
788
+ conversations]
789
+ mapping = {choice: conv['id'] for choice, conv in zip(choices, conversations)}
790
+ return gr.update(choices=choices), mapping
791
+
792
+ def load_conversations(selected, conversation_mapping):
793
+ logging.info(f"Selected: {selected}")
794
+ logging.info(f"Conversation mapping: {conversation_mapping}")
795
+
796
+ try:
797
+ if selected and selected in conversation_mapping:
798
+ conversation_id = conversation_mapping[selected]
799
+ messages = get_chat_messages(conversation_id)
800
+ conversation_data = {
801
+ "conversation_id": conversation_id,
802
+ "messages": messages
803
+ }
804
+ json_content = json.dumps(conversation_data, indent=2)
805
+
806
+ # Create HTML preview
807
+ html_preview = "<div style='max-height: 500px; overflow-y: auto;'>"
808
+ for msg in messages:
809
+ sender_style = "background-color: #e6f3ff;" if msg[
810
+ 'sender'] == 'user' else "background-color: #f0f0f0;"
811
+ html_preview += f"<div style='margin-bottom: 10px; padding: 10px; border-radius: 5px; {sender_style}'>"
812
+ html_preview += f"<strong>{msg['sender']}:</strong> {html.escape(msg['message'])}<br>"
813
+ html_preview += f"<small>Timestamp: {msg['timestamp']}</small>"
814
+ html_preview += "</div>"
815
+ html_preview += "</div>"
816
+
817
+ logging.info("Returning json_content and html_preview")
818
+ return json_content, html_preview
819
+ else:
820
+ logging.warning("No conversation selected or not in mapping")
821
+ return "", "<p>No conversation selected</p>"
822
+ except Exception as e:
823
+ logging.error(f"Error in load_conversations: {str(e)}")
824
+ return f"Error: {str(e)}", "<p>Error loading conversation</p>"
825
+
826
+ def validate_conversation_json(content):
827
+ try:
828
+ data = json.loads(content)
829
+ if not isinstance(data, dict):
830
+ return False, "Invalid JSON structure: root should be an object"
831
+ if "conversation_id" not in data or not isinstance(data["conversation_id"], int):
832
+ return False, "Missing or invalid conversation_id"
833
+ if "messages" not in data or not isinstance(data["messages"], list):
834
+ return False, "Missing or invalid messages array"
835
+ for msg in data["messages"]:
836
+ if not all(key in msg for key in ["sender", "message"]):
837
+ return False, "Invalid message structure: missing required fields"
838
+ return True, data
839
+ except json.JSONDecodeError as e:
840
+ return False, f"Invalid JSON: {str(e)}"
841
+
842
+ def save_conversation(selected, conversation_mapping, content):
843
+ if not selected or selected not in conversation_mapping:
844
+ return "Please select a conversation before saving.", "<p>No changes made</p>"
845
+
846
+ conversation_id = conversation_mapping[selected]
847
+ is_valid, result = validate_conversation_json(content)
848
+
849
+ if not is_valid:
850
+ return f"Error: {result}", "<p>No changes made due to error</p>"
851
+
852
+ conversation_data = result
853
+ if conversation_data["conversation_id"] != conversation_id:
854
+ return "Error: Conversation ID mismatch.", "<p>No changes made due to ID mismatch</p>"
855
+
856
+ try:
857
+ with db.get_connection() as conn:
858
+ conn.execute("BEGIN TRANSACTION")
859
+ cursor = conn.cursor()
860
+
861
+ # Backup original conversation
862
+ cursor.execute("SELECT * FROM ChatMessages WHERE conversation_id = ?", (conversation_id,))
863
+ original_messages = cursor.fetchall()
864
+ backup_data = json.dumps({"conversation_id": conversation_id, "messages": original_messages})
865
+
866
+ # You might want to save this backup_data somewhere
867
+
868
+ # Delete existing messages
869
+ cursor.execute("DELETE FROM ChatMessages WHERE conversation_id = ?", (conversation_id,))
870
+
871
+ # Insert updated messages
872
+ for message in conversation_data["messages"]:
873
+ cursor.execute('''
874
+ INSERT INTO ChatMessages (conversation_id, sender, message, timestamp)
875
+ VALUES (?, ?, ?, COALESCE(?, CURRENT_TIMESTAMP))
876
+ ''', (conversation_id, message["sender"], message["message"], message.get("timestamp")))
877
+
878
+ conn.commit()
879
+
880
+ # Create updated HTML preview
881
+ html_preview = "<div style='max-height: 500px; overflow-y: auto;'>"
882
+ for msg in conversation_data["messages"]:
883
+ sender_style = "background-color: #e6f3ff;" if msg[
884
+ 'sender'] == 'user' else "background-color: #f0f0f0;"
885
+ html_preview += f"<div style='margin-bottom: 10px; padding: 10px; border-radius: 5px; {sender_style}'>"
886
+ html_preview += f"<strong>{msg['sender']}:</strong> {html.escape(msg['message'])}<br>"
887
+ html_preview += f"<small>Timestamp: {msg.get('timestamp', 'N/A')}</small>"
888
+ html_preview += "</div>"
889
+ html_preview += "</div>"
890
+
891
+ return "Conversation updated successfully.", html_preview
892
+ except sqlite3.Error as e:
893
+ conn.rollback()
894
+ logging.error(f"Database error in save_conversation: {e}")
895
+ return f"Error updating conversation: {str(e)}", "<p>Error occurred while saving</p>"
896
+ except Exception as e:
897
+ conn.rollback()
898
+ logging.error(f"Unexpected error in save_conversation: {e}")
899
+ return f"Unexpected error: {str(e)}", "<p>Unexpected error occurred</p>"
900
+
901
+ def parse_formatted_content(formatted_content):
902
+ lines = formatted_content.split('\n')
903
+ conversation_id = int(lines[0].split(': ')[1])
904
+ timestamp = lines[1].split(': ')[1]
905
+ history = []
906
+ current_role = None
907
+ current_content = None
908
+ for line in lines[3:]:
909
+ if line.startswith("Role: "):
910
+ if current_role is not None:
911
+ history.append({"role": current_role, "content": ["", current_content]})
912
+ current_role = line.split(': ')[1]
913
+ elif line.startswith("Content: "):
914
+ current_content = line.split(': ', 1)[1]
915
+ if current_role is not None:
916
+ history.append({"role": current_role, "content": ["", current_content]})
917
+ return json.dumps({
918
+ "conversation_id": conversation_id,
919
+ "timestamp": timestamp,
920
+ "history": history
921
+ }, indent=2)
922
+
923
+ search_button.click(
924
+ search_conversations,
925
+ inputs=[search_query],
926
+ outputs=[conversation_list, conversation_mapping]
927
+ )
928
+
929
+ conversation_list.change(
930
+ load_conversations,
931
+ inputs=[conversation_list, conversation_mapping],
932
+ outputs=[chat_content, chat_preview]
933
+ )
934
+
935
+ save_button.click(
936
+ save_conversation,
937
+ inputs=[conversation_list, conversation_mapping, chat_content],
938
+ outputs=[result_message, chat_preview]
939
+ )
940
+
941
+ return search_query, search_button, conversation_list, conversation_mapping, chat_content, save_button, result_message, chat_preview
942
+
943
+
944
+ # FIXME - busted and incomplete
945
+ # Mock function to simulate LLM processing
946
+ def process_with_llm(workflow, context, prompt):
947
+ return f"LLM output for {workflow} with context: {context[:30]}... and prompt: {prompt[:30]}..."
948
+
949
+
950
+ # Load workflows from a JSON file
951
+ json_path = Path('./Helper_Scripts/Workflows/Workflows.json')
952
+ with json_path.open('r') as f:
953
+ workflows = json.load(f)
954
+
955
+
956
+ # FIXME - broken Completely. Doesn't work.
957
+ def chat_workflows_tab():
958
+ with gr.TabItem("Chat Workflows"):
959
+ with gr.Blocks() as chat_workflows_block:
960
+ gr.Markdown("# Workflows using LLMs")
961
+
962
+ workflow_selector = gr.Dropdown(label="Select Workflow", choices=[wf['name'] for wf in workflows])
963
+ context_input = gr.Textbox(label="Context", lines=5)
964
+
965
+ # Create lists to hold UI components
966
+ prompt_inputs = []
967
+ process_buttons = []
968
+ output_boxes = []
969
+ max_prompts = max(len(wf['prompts']) for wf in workflows)
970
+
971
+ # Pre-create the maximum number of prompt sections
972
+ for i in range(max_prompts):
973
+ prompt_input = gr.Textbox(label=f"Prompt {i + 1}", lines=2, visible=False)
974
+ output_box = gr.Textbox(label=f"Output {i + 1}", lines=5, visible=False)
975
+ process_button = gr.Button(f"Process Prompt {i + 1}", visible=False)
976
+
977
+ prompt_inputs.append(prompt_input)
978
+ output_boxes.append(output_box)
979
+ process_buttons.append(process_button)
980
+
981
+ process_button.click(
982
+ fn=lambda context, prompt, workflow_name, step=i: process(context, prompt, workflow_name, step),
983
+ inputs=[context_input, prompt_input, workflow_selector],
984
+ outputs=[output_box]
985
+ )
986
+
987
+ def process(context, prompt, workflow_name, step):
988
+ selected_workflow = next(wf for wf in workflows if wf['name'] == workflow_name)
989
+ # Update context with previous outputs
990
+ for j in range(step):
991
+ context += f"\n\n{output_boxes[j].value}"
992
+ result = process_with_llm(selected_workflow['name'], context, prompt)
993
+ return result
994
+
995
+ def update_prompt_sections(workflow_name):
996
+ selected_workflow = next(wf for wf in workflows if wf['name'] == workflow_name)
997
+ num_prompts = len(selected_workflow['prompts'])
998
+
999
+ for i in range(max_prompts):
1000
+ if i < num_prompts:
1001
+ prompt_inputs[i].visible = True
1002
+ prompt_inputs[i].value = selected_workflow['prompts'][i]
1003
+ process_buttons[i].visible = True
1004
+ output_boxes[i].visible = True
1005
+ else:
1006
+ prompt_inputs[i].visible = False
1007
+ process_buttons[i].visible = False
1008
+ output_boxes[i].visible = False
1009
+
1010
+ # Bind the workflow selector to update the UI
1011
+ workflow_selector.change(update_prompt_sections, inputs=[workflow_selector], outputs=[])
1012
+
1013
+ return chat_workflows_block
1014
+
1015
+ #
1016
+ # End of Chat_ui.py
1017
+ #######################################################################################################################
App_Function_Libraries/Gradio_UI/Explain_summarize_tab.py ADDED
@@ -0,0 +1,193 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Explain_summarize_tab.py
2
+ # Gradio UI for explaining and summarizing text
3
+ #
4
+ # Imports
5
+ import logging
6
+ #
7
+ # External Imports
8
+ import gradio as gr
9
+ #
10
+ # Local Imports
11
+ from App_Function_Libraries.Local_Summarization_Lib import summarize_with_llama, summarize_with_kobold, \
12
+ summarize_with_oobabooga, summarize_with_tabbyapi, summarize_with_vllm, summarize_with_local_llm, \
13
+ summarize_with_ollama
14
+ from App_Function_Libraries.Summarization_General_Lib import summarize_with_openai, summarize_with_anthropic, \
15
+ summarize_with_cohere, summarize_with_groq, summarize_with_openrouter, summarize_with_deepseek, \
16
+ summarize_with_huggingface
17
+ #
18
+ #
19
+ ############################################################################################################
20
+ #
21
+ # Functions:
22
+
23
+ def create_summarize_explain_tab():
24
+ with gr.TabItem("Explain/Summarize Text"):
25
+ gr.Markdown("# Explain or Summarize Text without ingesting it into the DB")
26
+ with gr.Row():
27
+ with gr.Column():
28
+ text_to_work_input = gr.Textbox(label="Text to be Explained or Summarized",
29
+ placeholder="Enter the text you want explained or summarized here",
30
+ lines=20)
31
+ with gr.Row():
32
+ explanation_checkbox = gr.Checkbox(label="Explain Text", value=True)
33
+ summarization_checkbox = gr.Checkbox(label="Summarize Text", value=True)
34
+ api_endpoint = gr.Dropdown(
35
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral",
36
+ "OpenRouter",
37
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
38
+ value=None,
39
+ label="API for Summarization (Optional)"
40
+ )
41
+ api_key_input = gr.Textbox(label="API Key (if required)", placeholder="Enter your API key here",
42
+ type="password")
43
+ explain_summarize_button = gr.Button("Explain/Summarize")
44
+
45
+ with gr.Column():
46
+ summarization_output = gr.Textbox(label="Summary:", lines=20)
47
+ explanation_output = gr.Textbox(label="Explanation:", lines=50)
48
+
49
+ explain_summarize_button.click(
50
+ fn=summarize_explain_text,
51
+ inputs=[text_to_work_input, api_endpoint, api_key_input, summarization_checkbox, explanation_checkbox],
52
+ outputs=[summarization_output, explanation_output]
53
+ )
54
+
55
+
56
+ def summarize_explain_text(message, api_endpoint, api_key, summarization, explanation):
57
+ summarization_response = None
58
+ explanation_response = None
59
+ temp = 0.7
60
+ try:
61
+ logging.info(f"Debug - summarize_explain_text Function - Message: {message}")
62
+ logging.info(f"Debug - summarize_explain_text Function - API Endpoint: {api_endpoint}")
63
+
64
+ # Prepare the input for the API
65
+ input_data = f"User: {message}\n"
66
+ # Print first 500 chars
67
+ logging.info(f"Debug - Chat Function - Input Data: {input_data[:500]}...")
68
+ logging.debug(f"Debug - Chat Function - API Key: {api_key[:10]}")
69
+ user_prompt = " "
70
+ if not api_endpoint:
71
+ return "Please select an API endpoint", "Please select an API endpoint"
72
+ try:
73
+ if summarization:
74
+ system_prompt = """<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
75
+ **Bulleted Note Creation Guidelines**
76
+
77
+ **Headings**:
78
+ - Based on referenced topics, not categories like quotes or terms
79
+ - Surrounded by **bold** formatting
80
+ - Not listed as bullet points
81
+ - No space between headings and list items underneath
82
+
83
+ **Emphasis**:
84
+ - **Important terms** set in bold font
85
+ - **Text ending in a colon**: also bolded
86
+
87
+ **Review**:
88
+ - Ensure adherence to specified format
89
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]"""
90
+
91
+ # Use the existing API request code based on the selected endpoint
92
+ logging.info(f"Debug - Chat Function - API Endpoint: {api_endpoint}")
93
+ if api_endpoint.lower() == 'openai':
94
+ summarization_response = summarize_with_openai(api_key, input_data, user_prompt, temp,
95
+ system_prompt)
96
+ elif api_endpoint.lower() == "anthropic":
97
+ summarization_response = summarize_with_anthropic(api_key, input_data, user_prompt, temp,
98
+ system_prompt)
99
+ elif api_endpoint.lower() == "cohere":
100
+ summarization_response = summarize_with_cohere(api_key, input_data, user_prompt, temp,
101
+ system_prompt)
102
+ elif api_endpoint.lower() == "groq":
103
+ summarization_response = summarize_with_groq(api_key, input_data, user_prompt, temp, system_prompt)
104
+ elif api_endpoint.lower() == "openrouter":
105
+ summarization_response = summarize_with_openrouter(api_key, input_data, user_prompt, temp,
106
+ system_prompt)
107
+ elif api_endpoint.lower() == "deepseek":
108
+ summarization_response = summarize_with_deepseek(api_key, input_data, user_prompt, temp,
109
+ system_prompt)
110
+ elif api_endpoint.lower() == "llama.cpp":
111
+ summarization_response = summarize_with_llama(input_data, user_prompt, temp, system_prompt)
112
+ elif api_endpoint.lower() == "kobold":
113
+ summarization_response = summarize_with_kobold(input_data, api_key, user_prompt, temp,
114
+ system_prompt)
115
+ elif api_endpoint.lower() == "ooba":
116
+ summarization_response = summarize_with_oobabooga(input_data, api_key, user_prompt, temp,
117
+ system_prompt)
118
+ elif api_endpoint.lower() == "tabbyapi":
119
+ summarization_response = summarize_with_tabbyapi(input_data, user_prompt, temp, system_prompt)
120
+ elif api_endpoint.lower() == "vllm":
121
+ summarization_response = summarize_with_vllm(input_data, user_prompt, system_prompt)
122
+ elif api_endpoint.lower() == "local-llm":
123
+ summarization_response = summarize_with_local_llm(input_data, user_prompt, temp, system_prompt)
124
+ elif api_endpoint.lower() == "huggingface":
125
+ summarization_response = summarize_with_huggingface(api_key, input_data, user_prompt,
126
+ temp) # , system_prompt)
127
+ elif api_endpoint.lower() == "ollama":
128
+ summarization_response = summarize_with_ollama(input_data, user_prompt, temp, system_prompt)
129
+ else:
130
+ raise ValueError(f"Unsupported API endpoint: {api_endpoint}")
131
+ except Exception as e:
132
+ logging.error(f"Error in summarization: {str(e)}")
133
+ response1 = f"An error occurred during summarization: {str(e)}"
134
+
135
+ try:
136
+ if explanation:
137
+ system_prompt = """You are a professional teacher. Please explain the content presented in an easy to digest fashion so that a non-specialist may understand it."""
138
+ # Use the existing API request code based on the selected endpoint
139
+ logging.info(f"Debug - Chat Function - API Endpoint: {api_endpoint}")
140
+ if api_endpoint.lower() == 'openai':
141
+ explanation_response = summarize_with_openai(api_key, input_data, user_prompt, temp, system_prompt)
142
+ elif api_endpoint.lower() == "anthropic":
143
+ explanation_response = summarize_with_anthropic(api_key, input_data, user_prompt, temp,
144
+ system_prompt)
145
+ elif api_endpoint.lower() == "cohere":
146
+ explanation_response = summarize_with_cohere(api_key, input_data, user_prompt, temp, system_prompt)
147
+ elif api_endpoint.lower() == "groq":
148
+ explanation_response = summarize_with_groq(api_key, input_data, user_prompt, temp, system_prompt)
149
+ elif api_endpoint.lower() == "openrouter":
150
+ explanation_response = summarize_with_openrouter(api_key, input_data, user_prompt, temp,
151
+ system_prompt)
152
+ elif api_endpoint.lower() == "deepseek":
153
+ explanation_response = summarize_with_deepseek(api_key, input_data, user_prompt, temp,
154
+ system_prompt)
155
+ elif api_endpoint.lower() == "llama.cpp":
156
+ explanation_response = summarize_with_llama(input_data, user_prompt, temp, system_prompt)
157
+ elif api_endpoint.lower() == "kobold":
158
+ explanation_response = summarize_with_kobold(input_data, api_key, user_prompt, temp, system_prompt)
159
+ elif api_endpoint.lower() == "ooba":
160
+ explanation_response = summarize_with_oobabooga(input_data, api_key, user_prompt, temp,
161
+ system_prompt)
162
+ elif api_endpoint.lower() == "tabbyapi":
163
+ explanation_response = summarize_with_tabbyapi(input_data, user_prompt, temp, system_prompt)
164
+ elif api_endpoint.lower() == "vllm":
165
+ explanation_response = summarize_with_vllm(input_data, user_prompt, system_prompt)
166
+ elif api_endpoint.lower() == "local-llm":
167
+ explanation_response = summarize_with_local_llm(input_data, user_prompt, temp, system_prompt)
168
+ elif api_endpoint.lower() == "huggingface":
169
+ explanation_response = summarize_with_huggingface(api_key, input_data, user_prompt,
170
+ temp) # , system_prompt)
171
+ elif api_endpoint.lower() == "ollama":
172
+ explanation_response = summarize_with_ollama(input_data, user_prompt, temp, system_prompt)
173
+ else:
174
+ raise ValueError(f"Unsupported API endpoint: {api_endpoint}")
175
+ except Exception as e:
176
+ logging.error(f"Error in summarization: {str(e)}")
177
+ response2 = f"An error occurred during summarization: {str(e)}"
178
+
179
+ if summarization_response:
180
+ response1 = f"Summary: {summarization_response}"
181
+ else:
182
+ response1 = "Summary: No summary requested"
183
+
184
+ if explanation_response:
185
+ response2 = f"Explanation: {explanation_response}"
186
+ else:
187
+ response2 = "Explanation: No explanation requested"
188
+
189
+ return response1, response2
190
+
191
+ except Exception as e:
192
+ logging.error(f"Error in chat function: {str(e)}")
193
+ return f"An error occurred: {str(e)}"
App_Function_Libraries/Gradio_UI/Export_Functionality.py ADDED
@@ -0,0 +1,314 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Export_Functionality.py
2
+ # Functionality for exporting items as markdown files
3
+ import os
4
+ import json
5
+ import math
6
+ import logging
7
+ import shutil
8
+ import tempfile
9
+ from typing import List, Dict, Optional, Tuple
10
+ import gradio as gr
11
+ from App_Function_Libraries.DB_Manager import DatabaseError, create_automated_backup, db_path, backup_dir
12
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import fetch_item_details, fetch_items_by_keyword, browse_items
13
+
14
+ logger = logging.getLogger(__name__)
15
+
16
+ def export_item_as_markdown(media_id: int) -> Tuple[Optional[str], str]:
17
+ try:
18
+ content, prompt, summary = fetch_item_details(media_id)
19
+ title = f"Item {media_id}" # You might want to fetch the actual title
20
+ markdown_content = f"# {title}\n\n## Prompt\n{prompt}\n\n## Summary\n{summary}\n\n## Content\n{content}"
21
+
22
+ filename = f"export_item_{media_id}.md"
23
+ with open(filename, "w", encoding='utf-8') as f:
24
+ f.write(markdown_content)
25
+
26
+ logger.info(f"Successfully exported item {media_id} to {filename}")
27
+ return filename, f"Successfully exported item {media_id} to {filename}"
28
+ except Exception as e:
29
+ error_message = f"Error exporting item {media_id}: {str(e)}"
30
+ logger.error(error_message)
31
+ return None, error_message
32
+
33
+
34
+ def export_items_by_keyword(keyword: str) -> str:
35
+ try:
36
+ items = fetch_items_by_keyword(keyword)
37
+ if not items:
38
+ logger.warning(f"No items found for keyword: {keyword}")
39
+ return None
40
+
41
+ # Create a temporary directory to store individual markdown files
42
+ with tempfile.TemporaryDirectory() as temp_dir:
43
+ folder_name = f"export_keyword_{keyword}"
44
+ export_folder = os.path.join(temp_dir, folder_name)
45
+ os.makedirs(export_folder)
46
+
47
+ for item in items:
48
+ content, prompt, summary = fetch_item_details(item['id'])
49
+ markdown_content = f"# {item['title']}\n\n## Prompt\n{prompt}\n\n## Summary\n{summary}\n\n## Content\n{content}"
50
+
51
+ # Create individual markdown file for each item
52
+ file_name = f"{item['id']}_{item['title'][:50]}.md" # Limit filename length
53
+ file_path = os.path.join(export_folder, file_name)
54
+ with open(file_path, "w", encoding='utf-8') as f:
55
+ f.write(markdown_content)
56
+
57
+ # Create a zip file containing all markdown files
58
+ zip_filename = f"{folder_name}.zip"
59
+ shutil.make_archive(os.path.join(temp_dir, folder_name), 'zip', export_folder)
60
+
61
+ # Move the zip file to a location accessible by Gradio
62
+ final_zip_path = os.path.join(os.getcwd(), zip_filename)
63
+ shutil.move(os.path.join(temp_dir, zip_filename), final_zip_path)
64
+
65
+ logger.info(f"Successfully exported {len(items)} items for keyword '{keyword}' to {zip_filename}")
66
+ return final_zip_path
67
+ except Exception as e:
68
+ logger.error(f"Error exporting items for keyword '{keyword}': {str(e)}")
69
+ return None
70
+
71
+
72
+ def export_selected_items(selected_items: List[Dict]) -> Tuple[Optional[str], str]:
73
+ try:
74
+ logger.debug(f"Received selected_items: {selected_items}")
75
+ if not selected_items:
76
+ logger.warning("No items selected for export")
77
+ return None, "No items selected for export"
78
+
79
+ markdown_content = "# Selected Items\n\n"
80
+ for item in selected_items:
81
+ logger.debug(f"Processing item: {item}")
82
+ try:
83
+ # Check if 'value' is a string (JSON) or already a dictionary
84
+ if isinstance(item, str):
85
+ item_data = json.loads(item)
86
+ elif isinstance(item, dict) and 'value' in item:
87
+ item_data = item['value'] if isinstance(item['value'], dict) else json.loads(item['value'])
88
+ else:
89
+ item_data = item
90
+
91
+ logger.debug(f"Item data after processing: {item_data}")
92
+
93
+ if 'id' not in item_data:
94
+ logger.error(f"'id' not found in item data: {item_data}")
95
+ continue
96
+
97
+ content, prompt, summary = fetch_item_details(item_data['id'])
98
+ markdown_content += f"## {item_data.get('title', 'Item {}'.format(item_data['id']))}\n\n### Prompt\n{prompt}\n\n### Summary\n{summary}\n\n### Content\n{content}\n\n---\n\n"
99
+ except Exception as e:
100
+ logger.error(f"Error processing item {item}: {str(e)}")
101
+ markdown_content += f"## Error\n\nUnable to process this item.\n\n---\n\n"
102
+
103
+ filename = "export_selected_items.md"
104
+ with open(filename, "w", encoding='utf-8') as f:
105
+ f.write(markdown_content)
106
+
107
+ logger.info(f"Successfully exported {len(selected_items)} selected items to {filename}")
108
+ return filename, f"Successfully exported {len(selected_items)} items to {filename}"
109
+ except Exception as e:
110
+ error_message = f"Error exporting selected items: {str(e)}"
111
+ logger.error(error_message)
112
+ return None, error_message
113
+
114
+
115
+ def display_search_results_export_tab(search_query: str, search_type: str, page: int = 1, items_per_page: int = 10):
116
+ logger.info(f"Searching with query: '{search_query}', type: '{search_type}', page: {page}")
117
+ try:
118
+ results = browse_items(search_query, search_type)
119
+ logger.info(f"browse_items returned {len(results)} results")
120
+
121
+ if not results:
122
+ return [], f"No results found for query: '{search_query}'", 1, 1
123
+
124
+ total_pages = math.ceil(len(results) / items_per_page)
125
+ start_index = (page - 1) * items_per_page
126
+ end_index = start_index + items_per_page
127
+ paginated_results = results[start_index:end_index]
128
+
129
+ checkbox_data = [
130
+ {
131
+ "name": f"Name: {item[1]}\nURL: {item[2]}",
132
+ "value": {"id": item[0], "title": item[1], "url": item[2]}
133
+ }
134
+ for item in paginated_results
135
+ ]
136
+
137
+ logger.info(f"Returning {len(checkbox_data)} items for checkbox (page {page} of {total_pages})")
138
+ return checkbox_data, f"Found {len(results)} results (showing page {page} of {total_pages})", page, total_pages
139
+
140
+ except DatabaseError as e:
141
+ error_message = f"Error in display_search_results_export_tab: {str(e)}"
142
+ logger.error(error_message)
143
+ return [], error_message, 1, 1
144
+ except Exception as e:
145
+ error_message = f"Unexpected error in display_search_results_export_tab: {str(e)}"
146
+ logger.error(error_message)
147
+ return [], error_message, 1, 1
148
+
149
+
150
+ def create_export_tab():
151
+ with gr.Tab("Search and Export"):
152
+ with gr.Row():
153
+ with gr.Column():
154
+ gr.Markdown("# Search and Export Items")
155
+ gr.Markdown("Search for items and export them as markdown files")
156
+ gr.Markdown("You can also export items by keyword")
157
+ search_query = gr.Textbox(label="Search Query")
158
+ search_type = gr.Radio(["Title", "URL", "Keyword", "Content"], label="Search By")
159
+ search_button = gr.Button("Search")
160
+
161
+ with gr.Column():
162
+ prev_button = gr.Button("Previous Page")
163
+ next_button = gr.Button("Next Page")
164
+
165
+ current_page = gr.State(1)
166
+ total_pages = gr.State(1)
167
+
168
+ search_results = gr.CheckboxGroup(label="Search Results", choices=[])
169
+ export_selected_button = gr.Button("Export Selected Items")
170
+
171
+ keyword_input = gr.Textbox(label="Enter keyword for export")
172
+ export_by_keyword_button = gr.Button("Export items by keyword")
173
+
174
+ export_output = gr.File(label="Download Exported File")
175
+ error_output = gr.Textbox(label="Status/Error Messages", interactive=False)
176
+
177
+ def search_and_update(query, search_type, page):
178
+ results, message, current, total = display_search_results_export_tab(query, search_type, page)
179
+ logger.debug(f"search_and_update results: {results}")
180
+ return results, message, current, total, gr.update(choices=results)
181
+
182
+ search_button.click(
183
+ fn=search_and_update,
184
+ inputs=[search_query, search_type, current_page],
185
+ outputs=[search_results, error_output, current_page, total_pages, search_results],
186
+ show_progress="full"
187
+ )
188
+
189
+
190
+ def update_page(current, total, direction):
191
+ new_page = max(1, min(total, current + direction))
192
+ return new_page
193
+
194
+ prev_button.click(
195
+ fn=update_page,
196
+ inputs=[current_page, total_pages, gr.State(-1)],
197
+ outputs=[current_page]
198
+ ).then(
199
+ fn=search_and_update,
200
+ inputs=[search_query, search_type, current_page],
201
+ outputs=[search_results, error_output, current_page, total_pages],
202
+ show_progress=True
203
+ )
204
+
205
+ next_button.click(
206
+ fn=update_page,
207
+ inputs=[current_page, total_pages, gr.State(1)],
208
+ outputs=[current_page]
209
+ ).then(
210
+ fn=search_and_update,
211
+ inputs=[search_query, search_type, current_page],
212
+ outputs=[search_results, error_output, current_page, total_pages],
213
+ show_progress=True
214
+ )
215
+
216
+ def handle_export_selected(selected_items):
217
+ logger.debug(f"Exporting selected items: {selected_items}")
218
+ return export_selected_items(selected_items)
219
+
220
+ export_selected_button.click(
221
+ fn=handle_export_selected,
222
+ inputs=[search_results],
223
+ outputs=[export_output, error_output],
224
+ show_progress="full"
225
+ )
226
+
227
+ export_by_keyword_button.click(
228
+ fn=export_items_by_keyword,
229
+ inputs=[keyword_input],
230
+ outputs=[export_output, error_output],
231
+ show_progress="full"
232
+ )
233
+
234
+ def handle_item_selection(selected_items):
235
+ logger.debug(f"Selected items: {selected_items}")
236
+ if not selected_items:
237
+ return None, "No item selected"
238
+
239
+ try:
240
+ # Assuming selected_items is a list of dictionaries
241
+ selected_item = selected_items[0]
242
+ logger.debug(f"First selected item: {selected_item}")
243
+
244
+ # Check if 'value' is a string (JSON) or already a dictionary
245
+ if isinstance(selected_item['value'], str):
246
+ item_data = json.loads(selected_item['value'])
247
+ else:
248
+ item_data = selected_item['value']
249
+
250
+ logger.debug(f"Item data: {item_data}")
251
+
252
+ item_id = item_data['id']
253
+ return export_item_as_markdown(item_id)
254
+ except Exception as e:
255
+ error_message = f"Error processing selected item: {str(e)}"
256
+ logger.error(error_message)
257
+ return None, error_message
258
+
259
+ search_results.select(
260
+ fn=handle_item_selection,
261
+ inputs=[search_results],
262
+ outputs=[export_output, error_output],
263
+ show_progress="full"
264
+ )
265
+
266
+
267
+
268
+ def create_backup():
269
+ backup_file = create_automated_backup(db_path, backup_dir)
270
+ return f"Backup created: {backup_file}"
271
+
272
+ def list_backups():
273
+ backups = [f for f in os.listdir(backup_dir) if f.endswith('.db')]
274
+ return "\n".join(backups)
275
+
276
+ def restore_backup(backup_name):
277
+ backup_path = os.path.join(backup_dir, backup_name)
278
+ if os.path.exists(backup_path):
279
+ shutil.copy2(backup_path, db_path)
280
+ return f"Database restored from {backup_name}"
281
+ else:
282
+ return "Backup file not found"
283
+
284
+
285
+ def create_backup_tab():
286
+ with gr.Tab("Create Backup"):
287
+ gr.Markdown("# Create a backup of the database")
288
+ with gr.Row():
289
+ with gr.Column():
290
+ create_button = gr.Button("Create Backup")
291
+ create_output = gr.Textbox(label="Result")
292
+ with gr.Column():
293
+ create_button.click(create_backup, inputs=[], outputs=create_output)
294
+
295
+ def create_view_backups_tab():
296
+ with gr.TabItem("View Backups"):
297
+ gr.Markdown("# Browse available backups")
298
+ with gr.Row():
299
+ with gr.Column():
300
+ view_button = gr.Button("View Backups")
301
+ with gr.Column():
302
+ backup_list = gr.Textbox(label="Available Backups")
303
+ view_button.click(list_backups, inputs=[], outputs=backup_list)
304
+
305
+
306
+ def create_restore_backup_tab():
307
+ with gr.TabItem("Restore Backup"):
308
+ gr.Markdown("# Restore a backup of the database")
309
+ with gr.Column():
310
+ backup_input = gr.Textbox(label="Backup Filename")
311
+ restore_button = gr.Button("Restore")
312
+ with gr.Column():
313
+ restore_output = gr.Textbox(label="Result")
314
+ restore_button.click(restore_backup, inputs=[backup_input], outputs=restore_output)
App_Function_Libraries/Gradio_UI/Gradio_Shared.py ADDED
@@ -0,0 +1,284 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Gradio_Shared.py
2
+ # Gradio UI functions that are shared across multiple tabs
3
+ #
4
+ # Imports
5
+ import logging
6
+ import sqlite3
7
+ import traceback
8
+ from functools import wraps
9
+ from typing import List, Tuple
10
+ #
11
+ # External Imports
12
+ import gradio as gr
13
+ #
14
+ # Local Imports
15
+ from App_Function_Libraries.DB_Manager import list_prompts, db, search_and_display, fetch_prompt_details
16
+ from App_Function_Libraries.SQLite_DB import DatabaseError
17
+ from App_Function_Libraries.Utils import format_transcription
18
+ #
19
+ ##############################################################################################################
20
+ #
21
+ # Functions:
22
+
23
+ whisper_models = ["small", "medium", "small.en", "medium.en", "medium", "large", "large-v1", "large-v2", "large-v3",
24
+ "distil-large-v2", "distil-medium.en", "distil-small.en"]
25
+
26
+ # Sample data
27
+ prompts_category_1 = [
28
+ "What are the key points discussed in the video?",
29
+ "Summarize the main arguments made by the speaker.",
30
+ "Describe the conclusions of the study presented."
31
+ ]
32
+
33
+ prompts_category_2 = [
34
+ "How does the proposed solution address the problem?",
35
+ "What are the implications of the findings?",
36
+ "Can you explain the theory behind the observed phenomenon?"
37
+ ]
38
+
39
+ all_prompts = prompts_category_1 + prompts_category_2
40
+
41
+
42
+
43
+ #FIXME - SQL Functions that need to be addressed/added to DB manager
44
+ def search_media(query, fields, keyword, page):
45
+ try:
46
+ results = search_and_display(query, fields, keyword, page)
47
+ return results
48
+ except Exception as e:
49
+ logger = logging.getLogger()
50
+ logger.error(f"Error searching media: {e}")
51
+ return str(e)
52
+
53
+ def fetch_items_by_title_or_url(search_query: str, search_type: str):
54
+ try:
55
+ with db.get_connection() as conn:
56
+ cursor = conn.cursor()
57
+ if search_type == 'Title':
58
+ cursor.execute("SELECT id, title, url FROM Media WHERE title LIKE ?", (f'%{search_query}%',))
59
+ elif search_type == 'URL':
60
+ cursor.execute("SELECT id, title, url FROM Media WHERE url LIKE ?", (f'%{search_query}%',))
61
+ results = cursor.fetchall()
62
+ return results
63
+ except sqlite3.Error as e:
64
+ raise DatabaseError(f"Error fetching items by {search_type}: {e}")
65
+
66
+ def fetch_items_by_keyword(search_query: str):
67
+ try:
68
+ with db.get_connection() as conn:
69
+ cursor = conn.cursor()
70
+ cursor.execute("""
71
+ SELECT m.id, m.title, m.url
72
+ FROM Media m
73
+ JOIN MediaKeywords mk ON m.id = mk.media_id
74
+ JOIN Keywords k ON mk.keyword_id = k.id
75
+ WHERE k.keyword LIKE ?
76
+ """, (f'%{search_query}%',))
77
+ results = cursor.fetchall()
78
+ return results
79
+ except sqlite3.Error as e:
80
+ raise DatabaseError(f"Error fetching items by keyword: {e}")
81
+
82
+ # FIXME - Raw SQL not using DB_Manager...
83
+ def fetch_items_by_content(search_query: str):
84
+ try:
85
+ with db.get_connection() as conn:
86
+ cursor = conn.cursor()
87
+ cursor.execute("SELECT id, title, url FROM Media WHERE content LIKE ?", (f'%{search_query}%',))
88
+ results = cursor.fetchall()
89
+ return results
90
+ except sqlite3.Error as e:
91
+ raise DatabaseError(f"Error fetching items by content: {e}")
92
+
93
+
94
+
95
+ # FIXME - RAW SQL not using DB_Manager...
96
+ def fetch_item_details_single(media_id: int):
97
+ try:
98
+ with db.get_connection() as conn:
99
+ cursor = conn.cursor()
100
+ cursor.execute("""
101
+ SELECT prompt, summary
102
+ FROM MediaModifications
103
+ WHERE media_id = ?
104
+ ORDER BY modification_date DESC
105
+ LIMIT 1
106
+ """, (media_id,))
107
+ prompt_summary_result = cursor.fetchone()
108
+ cursor.execute("SELECT content FROM Media WHERE id = ?", (media_id,))
109
+ content_result = cursor.fetchone()
110
+
111
+ prompt = prompt_summary_result[0] if prompt_summary_result else ""
112
+ summary = prompt_summary_result[1] if prompt_summary_result else ""
113
+ content = content_result[0] if content_result else ""
114
+
115
+ return prompt, summary, content
116
+ except sqlite3.Error as e:
117
+ raise Exception(f"Error fetching item details: {e}")
118
+
119
+
120
+ # FIXME - RAW SQL not using DB_Manager...
121
+ def fetch_item_details(media_id: int):
122
+ try:
123
+ with db.get_connection() as conn:
124
+ cursor = conn.cursor()
125
+ cursor.execute("""
126
+ SELECT prompt, summary
127
+ FROM MediaModifications
128
+ WHERE media_id = ?
129
+ ORDER BY modification_date DESC
130
+ LIMIT 1
131
+ """, (media_id,))
132
+ prompt_summary_result = cursor.fetchone()
133
+ cursor.execute("SELECT content FROM Media WHERE id = ?", (media_id,))
134
+ content_result = cursor.fetchone()
135
+
136
+ prompt = prompt_summary_result[0] if prompt_summary_result else ""
137
+ summary = prompt_summary_result[1] if prompt_summary_result else ""
138
+ content = content_result[0] if content_result else ""
139
+
140
+ return content, prompt, summary
141
+ except sqlite3.Error as e:
142
+ logging.error(f"Error fetching item details: {e}")
143
+ return "", "", "" # Return empty strings if there's an error
144
+
145
+ # Handle prompt selection
146
+ def handle_prompt_selection(prompt):
147
+ return f"You selected: {prompt}"
148
+
149
+
150
+ def update_user_prompt(preset_name):
151
+ details = fetch_prompt_details(preset_name)
152
+ if details:
153
+ # Return a dictionary with all details
154
+ return {
155
+ "title": details[0],
156
+ "details": details[1],
157
+ "system_prompt": details[2],
158
+ "user_prompt": details[3] if len(details) > 3 else ""
159
+ }
160
+ return {"title": "", "details": "", "system_prompt": "", "user_prompt": ""}
161
+
162
+ def browse_items(search_query, search_type):
163
+ if search_type == 'Keyword':
164
+ results = fetch_items_by_keyword(search_query)
165
+ elif search_type == 'Content':
166
+ results = fetch_items_by_content(search_query)
167
+ else:
168
+ results = fetch_items_by_title_or_url(search_query, search_type)
169
+ return results
170
+
171
+
172
+ def update_dropdown(search_query, search_type):
173
+ results = browse_items(search_query, search_type)
174
+ item_options = [f"{item[1]} ({item[2]})" for item in results]
175
+ new_item_mapping = {f"{item[1]} ({item[2]})": item[0] for item in results}
176
+ print(f"Debug - Update Dropdown - New Item Mapping: {new_item_mapping}")
177
+ return gr.update(choices=item_options), new_item_mapping
178
+
179
+
180
+
181
+ def get_media_id(selected_item, item_mapping):
182
+ return item_mapping.get(selected_item)
183
+
184
+
185
+ def update_detailed_view(item, item_mapping):
186
+ # Function to update the detailed view based on selected item
187
+ if item:
188
+ item_id = item_mapping.get(item)
189
+ if item_id:
190
+ content, prompt, summary = fetch_item_details(item_id)
191
+ if content or prompt or summary:
192
+ details_html = "<h4>Details:</h4>"
193
+ if prompt:
194
+ formatted_prompt = format_transcription(prompt)
195
+ details_html += f"<h4>Prompt:</h4>{formatted_prompt}</p>"
196
+ if summary:
197
+ formatted_summary = format_transcription(summary)
198
+ details_html += f"<h4>Summary:</h4>{formatted_summary}</p>"
199
+ # Format the transcription content for better readability
200
+ formatted_content = format_transcription(content)
201
+ #content_html = f"<h4>Transcription:</h4><div style='white-space: pre-wrap;'>{content}</div>"
202
+ content_html = f"<h4>Transcription:</h4><div style='white-space: pre-wrap;'>{formatted_content}</div>"
203
+ return details_html, content_html
204
+ else:
205
+ return "No details available.", "No details available."
206
+ else:
207
+ return "No item selected", "No item selected"
208
+ else:
209
+ return "No item selected", "No item selected"
210
+
211
+
212
+ def format_content(content):
213
+ # Format content using markdown
214
+ formatted_content = f"```\n{content}\n```"
215
+ return formatted_content
216
+
217
+
218
+ def update_prompt_dropdown():
219
+ prompt_names = list_prompts()
220
+ return gr.update(choices=prompt_names)
221
+
222
+
223
+ def display_prompt_details(selected_prompt):
224
+ if selected_prompt:
225
+ prompts = update_user_prompt(selected_prompt)
226
+ if prompts["title"]: # Check if we have any details
227
+ details_str = f"<h4>Details:</h4><p>{prompts['details']}</p>"
228
+ system_str = f"<h4>System:</h4><p>{prompts['system_prompt']}</p>"
229
+ user_str = f"<h4>User:</h4><p>{prompts['user_prompt']}</p>" if prompts['user_prompt'] else ""
230
+ return details_str + system_str + user_str
231
+ return "No details available."
232
+
233
+ def search_media_database(query: str) -> List[Tuple[int, str, str]]:
234
+ return browse_items(query, 'Title')
235
+
236
+
237
+ def load_media_content(media_id: int) -> dict:
238
+ try:
239
+ print(f"Debug - Load Media Content - Media ID: {media_id}")
240
+ item_details = fetch_item_details(media_id)
241
+ print(f"Debug - Load Media Content - Item Details: \n\n{item_details}\n\n\n\n")
242
+
243
+ if isinstance(item_details, tuple) and len(item_details) == 3:
244
+ content, prompt, summary = item_details
245
+ else:
246
+ print(f"Debug - Load Media Content - Unexpected item_details format: \n\n{item_details}\n\n\n\n")
247
+ content, prompt, summary = "", "", ""
248
+
249
+ return {
250
+ "content": content or "No content available",
251
+ "prompt": prompt or "No prompt available",
252
+ "summary": summary or "No summary available"
253
+ }
254
+ except Exception as e:
255
+ print(f"Debug - Load Media Content - Error: {str(e)}")
256
+ return {"content": "", "prompt": "", "summary": ""}
257
+
258
+
259
+ def error_handler(func):
260
+ @wraps(func)
261
+ def wrapper(*args, **kwargs):
262
+ try:
263
+ return func(*args, **kwargs)
264
+ except Exception as e:
265
+ error_message = f"Error in {func.__name__}: {str(e)}"
266
+ logging.error(f"{error_message}\n{traceback.format_exc()}")
267
+ return {"error": error_message, "details": traceback.format_exc()}
268
+ return wrapper
269
+
270
+
271
+ def create_chunking_inputs():
272
+ chunk_text_by_words_checkbox = gr.Checkbox(label="Chunk Text by Words", value=False, visible=True)
273
+ max_words_input = gr.Number(label="Max Words", value=300, precision=0, visible=True)
274
+ chunk_text_by_sentences_checkbox = gr.Checkbox(label="Chunk Text by Sentences", value=False, visible=True)
275
+ max_sentences_input = gr.Number(label="Max Sentences", value=10, precision=0, visible=True)
276
+ chunk_text_by_paragraphs_checkbox = gr.Checkbox(label="Chunk Text by Paragraphs", value=False, visible=True)
277
+ max_paragraphs_input = gr.Number(label="Max Paragraphs", value=5, precision=0, visible=True)
278
+ chunk_text_by_tokens_checkbox = gr.Checkbox(label="Chunk Text by Tokens", value=False, visible=True)
279
+ max_tokens_input = gr.Number(label="Max Tokens", value=1000, precision=0, visible=True)
280
+ gr_semantic_chunk_long_file = gr.Checkbox(label="Semantic Chunking by Sentence similarity", value=False, visible=True)
281
+ gr_semantic_chunk_long_file_size = gr.Number(label="Max Chunk Size", value=2000, visible=True)
282
+ gr_semantic_chunk_long_file_overlap = gr.Number(label="Max Chunk Overlap Size", value=100, visible=True)
283
+ return [chunk_text_by_words_checkbox, max_words_input, chunk_text_by_sentences_checkbox, max_sentences_input,
284
+ chunk_text_by_paragraphs_checkbox, max_paragraphs_input, chunk_text_by_tokens_checkbox, max_tokens_input]
App_Function_Libraries/Gradio_UI/Import_Functionality.py ADDED
@@ -0,0 +1,473 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Import_Functionality.py
2
+ # Functionality to import content into the DB
3
+ #
4
+ # Imports
5
+ from time import sleep
6
+ import logging
7
+ import re
8
+ import shutil
9
+ import tempfile
10
+ import os
11
+ import traceback
12
+ import zipfile
13
+ #
14
+ # External Imports
15
+ import gradio as gr
16
+ import pypandoc
17
+ #
18
+ # Local Imports
19
+ from App_Function_Libraries.DB_Manager import insert_prompt_to_db, load_preset_prompts, import_obsidian_note_to_db, \
20
+ add_media_to_database
21
+ from App_Function_Libraries.Prompt_Handling import import_prompt_from_file, import_prompts_from_zip#
22
+ from App_Function_Libraries.Summarization_General_Lib import perform_summarization
23
+
24
+ ###################################################################################################################
25
+ #
26
+ # Functions:
27
+
28
+ logger = logging.getLogger()
29
+
30
+
31
+ def import_data(file, title, author, keywords, custom_prompt, summary, auto_summarize, api_name, api_key):
32
+ if file is None:
33
+ return "No file uploaded. Please upload a file."
34
+
35
+ try:
36
+ logging.debug(f"File object type: {type(file)}")
37
+ logging.debug(f"File object attributes: {dir(file)}")
38
+
39
+ if hasattr(file, 'name'):
40
+ file_name = file.name
41
+ else:
42
+ file_name = 'unknown_file'
43
+
44
+ # Create a temporary file
45
+ with tempfile.NamedTemporaryFile(mode='w+', delete=False, suffix='.txt', encoding='utf-8') as temp_file:
46
+ if isinstance(file, str):
47
+ # If file is a string, it's likely file content
48
+ temp_file.write(file)
49
+ elif hasattr(file, 'read'):
50
+ # If file has a 'read' method, it's likely a file-like object
51
+ content = file.read()
52
+ if isinstance(content, bytes):
53
+ content = content.decode('utf-8')
54
+ temp_file.write(content)
55
+ else:
56
+ # If it's neither a string nor a file-like object, try converting it to a string
57
+ temp_file.write(str(file))
58
+
59
+ temp_file.seek(0)
60
+ file_content = temp_file.read()
61
+
62
+ logging.debug(f"File name: {file_name}")
63
+ logging.debug(f"File content (first 100 chars): {file_content[:100]}")
64
+
65
+ # Create info_dict
66
+ info_dict = {
67
+ 'title': title or 'Untitled',
68
+ 'uploader': author or 'Unknown',
69
+ }
70
+
71
+ # FIXME - Add chunking support... I added chapter chunking specifically for this...
72
+ # Create segments (assuming one segment for the entire content)
73
+ segments = [{'Text': file_content}]
74
+
75
+ # Process keywords
76
+ keyword_list = [kw.strip() for kw in keywords.split(',') if kw.strip()]
77
+
78
+ # Handle summarization
79
+ if auto_summarize and api_name and api_key:
80
+ summary = perform_summarization(api_name, file_content, custom_prompt, api_key)
81
+ elif not summary:
82
+ summary = "No summary provided"
83
+
84
+ # Add to database
85
+ add_media_to_database(
86
+ url=file_name, # Using filename as URL
87
+ info_dict=info_dict,
88
+ segments=segments,
89
+ summary=summary,
90
+ keywords=keyword_list,
91
+ custom_prompt_input=custom_prompt,
92
+ whisper_model="Imported", # Indicating this was an imported file
93
+ media_type="document"
94
+ )
95
+
96
+ # Clean up the temporary file
97
+ os.unlink(temp_file.name)
98
+
99
+ return f"File '{file_name}' successfully imported with title '{title}' and author '{author}'."
100
+ except Exception as e:
101
+ logging.error(f"Error importing file: {str(e)}")
102
+ return f"Error importing file: {str(e)}"
103
+
104
+
105
+
106
+
107
+
108
+ def process_obsidian_zip(zip_file):
109
+ with tempfile.TemporaryDirectory() as temp_dir:
110
+ try:
111
+ with zipfile.ZipFile(zip_file, 'r') as zip_ref:
112
+ zip_ref.extractall(temp_dir)
113
+
114
+ imported_files, total_files, errors = import_obsidian_vault(temp_dir)
115
+
116
+ return imported_files, total_files, errors
117
+ except zipfile.BadZipFile:
118
+ error_msg = "The uploaded file is not a valid zip file."
119
+ logger.error(error_msg)
120
+ return 0, 0, [error_msg]
121
+ except Exception as e:
122
+ error_msg = f"Error processing zip file: {str(e)}\n{traceback.format_exc()}"
123
+ logger.error(error_msg)
124
+ return 0, 0, [error_msg]
125
+ finally:
126
+ shutil.rmtree(temp_dir, ignore_errors=True)
127
+
128
+
129
+
130
+ def scan_obsidian_vault(vault_path):
131
+ markdown_files = []
132
+ for root, dirs, files in os.walk(vault_path):
133
+ for file in files:
134
+ if file.endswith('.md'):
135
+ markdown_files.append(os.path.join(root, file))
136
+ return markdown_files
137
+
138
+
139
+ def parse_obsidian_note(file_path):
140
+ with open(file_path, 'r', encoding='utf-8') as file:
141
+ content = file.read()
142
+
143
+ frontmatter = {}
144
+ frontmatter_match = re.match(r'^---\s*\n(.*?)\n---\s*\n', content, re.DOTALL)
145
+ if frontmatter_match:
146
+ frontmatter_text = frontmatter_match.group(1)
147
+ import yaml
148
+ frontmatter = yaml.safe_load(frontmatter_text)
149
+ content = content[frontmatter_match.end():]
150
+
151
+ tags = re.findall(r'#(\w+)', content)
152
+ links = re.findall(r'\[\[(.*?)\]\]', content)
153
+
154
+ return {
155
+ 'title': os.path.basename(file_path).replace('.md', ''),
156
+ 'content': content,
157
+ 'frontmatter': frontmatter,
158
+ 'tags': tags,
159
+ 'links': links,
160
+ 'file_path': file_path # Add this line
161
+ }
162
+
163
+ def create_import_single_prompt_tab():
164
+ with gr.TabItem("Import a Prompt"):
165
+ gr.Markdown("# Import a prompt into the database")
166
+
167
+ with gr.Row():
168
+ with gr.Column():
169
+ import_file = gr.File(label="Upload file for import", file_types=["txt", "md"])
170
+ title_input = gr.Textbox(label="Title", placeholder="Enter the title of the content")
171
+ author_input = gr.Textbox(label="Author", placeholder="Enter the author's name")
172
+ system_input = gr.Textbox(label="System", placeholder="Enter the system message for the prompt", lines=3)
173
+ user_input = gr.Textbox(label="User", placeholder="Enter the user message for the prompt", lines=3)
174
+ keywords_input = gr.Textbox(label="Keywords", placeholder="Enter keywords separated by commas")
175
+ import_button = gr.Button("Import Prompt")
176
+
177
+ with gr.Column():
178
+ import_output = gr.Textbox(label="Import Status")
179
+ save_button = gr.Button("Save to Database")
180
+ save_output = gr.Textbox(label="Save Status")
181
+
182
+ def handle_import(file):
183
+ result = import_prompt_from_file(file)
184
+ if isinstance(result, tuple) and len(result) == 5:
185
+ title, author, system, user, keywords = result
186
+ return gr.update(value="File successfully imported. You can now edit the content before saving."), \
187
+ gr.update(value=title), gr.update(value=author), gr.update(value=system), \
188
+ gr.update(value=user), gr.update(value=", ".join(keywords))
189
+ else:
190
+ return gr.update(value=result), gr.update(), gr.update(), gr.update(), gr.update(), gr.update()
191
+
192
+ import_button.click(
193
+ fn=handle_import,
194
+ inputs=[import_file],
195
+ outputs=[import_output, title_input, author_input, system_input, user_input, keywords_input]
196
+ )
197
+
198
+ def save_prompt_to_db(title, author, system, user, keywords):
199
+ keyword_list = [k.strip() for k in keywords.split(',') if k.strip()]
200
+ return insert_prompt_to_db(title, author, system, user, keyword_list)
201
+
202
+ save_button.click(
203
+ fn=save_prompt_to_db,
204
+ inputs=[title_input, author_input, system_input, user_input, keywords_input],
205
+ outputs=save_output
206
+ )
207
+
208
+ def update_prompt_dropdown():
209
+ return gr.update(choices=load_preset_prompts())
210
+
211
+ save_button.click(
212
+ fn=update_prompt_dropdown,
213
+ inputs=[],
214
+ outputs=[gr.Dropdown(label="Select Preset Prompt")]
215
+ )
216
+
217
+ def create_import_item_tab():
218
+ with gr.TabItem("Import Markdown/Text Files"):
219
+ gr.Markdown("# Import a markdown file or text file into the database")
220
+ gr.Markdown("...and have it tagged + summarized")
221
+ with gr.Row():
222
+ with gr.Column():
223
+ import_file = gr.File(label="Upload file for import", file_types=["txt", "md"])
224
+ title_input = gr.Textbox(label="Title", placeholder="Enter the title of the content")
225
+ author_input = gr.Textbox(label="Author", placeholder="Enter the author's name")
226
+ keywords_input = gr.Textbox(label="Keywords", placeholder="Enter keywords, comma-separated")
227
+ custom_prompt_input = gr.Textbox(label="Custom Prompt",
228
+ placeholder="Enter a custom prompt for summarization (optional)")
229
+ summary_input = gr.Textbox(label="Summary",
230
+ placeholder="Enter a summary or leave blank for auto-summarization", lines=3)
231
+ auto_summarize_checkbox = gr.Checkbox(label="Auto-summarize", value=False)
232
+ api_name_input = gr.Dropdown(
233
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
234
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
235
+ label="API for Auto-summarization"
236
+ )
237
+ api_key_input = gr.Textbox(label="API Key", type="password")
238
+ with gr.Column():
239
+ import_button = gr.Button("Import Data")
240
+ import_output = gr.Textbox(label="Import Status")
241
+
242
+ import_button.click(
243
+ fn=import_data,
244
+ inputs=[import_file, title_input, author_input, keywords_input, custom_prompt_input,
245
+ summary_input, auto_summarize_checkbox, api_name_input, api_key_input],
246
+ outputs=import_output
247
+ )
248
+
249
+
250
+ def create_import_multiple_prompts_tab():
251
+ with gr.TabItem("Import Multiple Prompts"):
252
+ gr.Markdown("# Import multiple prompts into the database")
253
+ gr.Markdown("Upload a zip file containing multiple prompt files (txt or md)")
254
+
255
+ with gr.Row():
256
+ with gr.Column():
257
+ zip_file = gr.File(label="Upload zip file for import", file_types=["zip"])
258
+ import_button = gr.Button("Import Prompts")
259
+ prompts_dropdown = gr.Dropdown(label="Select Prompt to Edit", choices=[])
260
+ title_input = gr.Textbox(label="Title", placeholder="Enter the title of the content")
261
+ author_input = gr.Textbox(label="Author", placeholder="Enter the author's name")
262
+ system_input = gr.Textbox(label="System", placeholder="Enter the system message for the prompt",
263
+ lines=3)
264
+ user_input = gr.Textbox(label="User", placeholder="Enter the user message for the prompt", lines=3)
265
+ keywords_input = gr.Textbox(label="Keywords", placeholder="Enter keywords separated by commas")
266
+
267
+ with gr.Column():
268
+ import_output = gr.Textbox(label="Import Status")
269
+ save_button = gr.Button("Save to Database")
270
+ save_output = gr.Textbox(label="Save Status")
271
+ prompts_display = gr.Textbox(label="Identified Prompts")
272
+
273
+ def handle_zip_import(zip_file):
274
+ result = import_prompts_from_zip(zip_file)
275
+ if isinstance(result, list):
276
+ prompt_titles = [prompt['title'] for prompt in result]
277
+ return gr.update(
278
+ value="Zip file successfully imported. Select a prompt to edit from the dropdown."), prompt_titles, gr.update(
279
+ value="\n".join(prompt_titles)), result
280
+ else:
281
+ return gr.update(value=result), [], gr.update(value=""), []
282
+
283
+ def handle_prompt_selection(selected_title, prompts):
284
+ selected_prompt = next((prompt for prompt in prompts if prompt['title'] == selected_title), None)
285
+ if selected_prompt:
286
+ return (
287
+ selected_prompt['title'],
288
+ selected_prompt.get('author', ''),
289
+ selected_prompt['system'],
290
+ selected_prompt.get('user', ''),
291
+ ", ".join(selected_prompt.get('keywords', []))
292
+ )
293
+ else:
294
+ return "", "", "", "", ""
295
+
296
+ zip_import_state = gr.State([])
297
+
298
+ import_button.click(
299
+ fn=handle_zip_import,
300
+ inputs=[zip_file],
301
+ outputs=[import_output, prompts_dropdown, prompts_display, zip_import_state]
302
+ )
303
+
304
+ prompts_dropdown.change(
305
+ fn=handle_prompt_selection,
306
+ inputs=[prompts_dropdown, zip_import_state],
307
+ outputs=[title_input, author_input, system_input, user_input, keywords_input]
308
+ )
309
+
310
+ def save_prompt_to_db(title, author, system, user, keywords):
311
+ keyword_list = [k.strip() for k in keywords.split(',') if k.strip()]
312
+ return insert_prompt_to_db(title, author, system, user, keyword_list)
313
+
314
+ save_button.click(
315
+ fn=save_prompt_to_db,
316
+ inputs=[title_input, author_input, system_input, user_input, keywords_input],
317
+ outputs=save_output
318
+ )
319
+
320
+ def update_prompt_dropdown():
321
+ return gr.update(choices=load_preset_prompts())
322
+
323
+ save_button.click(
324
+ fn=update_prompt_dropdown,
325
+ inputs=[],
326
+ outputs=[gr.Dropdown(label="Select Preset Prompt")]
327
+ )
328
+
329
+
330
+ def create_import_obsidian_vault_tab():
331
+ with gr.TabItem("Import Obsidian Vault"):
332
+ gr.Markdown("## Import Obsidian Vault")
333
+ with gr.Row():
334
+ with gr.Column():
335
+ vault_path_input = gr.Textbox(label="Obsidian Vault Path (Local)")
336
+ vault_zip_input = gr.File(label="Upload Obsidian Vault (Zip)")
337
+ with gr.Column():
338
+ import_vault_button = gr.Button("Import Obsidian Vault")
339
+ import_status = gr.Textbox(label="Import Status", interactive=False)
340
+
341
+
342
+ def import_vault(vault_path, vault_zip):
343
+ if vault_zip:
344
+ imported, total, errors = process_obsidian_zip(vault_zip.name)
345
+ elif vault_path:
346
+ imported, total, errors = import_obsidian_vault(vault_path)
347
+ else:
348
+ return "Please provide either a local vault path or upload a zip file."
349
+
350
+ status = f"Imported {imported} out of {total} files.\n"
351
+ if errors:
352
+ status += f"Encountered {len(errors)} errors:\n" + "\n".join(errors)
353
+ return status
354
+
355
+
356
+ import_vault_button.click(
357
+ fn=import_vault,
358
+ inputs=[vault_path_input, vault_zip_input],
359
+ outputs=[import_status],
360
+ show_progress=True
361
+ )
362
+
363
+
364
+
365
+ # Using pypandoc to convert EPUB to Markdown
366
+ def create_import_book_tab():
367
+ with gr.TabItem("Import .epub/ebook Files"):
368
+ with gr.Row():
369
+ with gr.Column():
370
+ gr.Markdown("# Ingest an .epub file using pypandoc")
371
+ gr.Markdown("...and have it tagged + summarized")
372
+ gr.Markdown(
373
+ "How to remove DRM from your ebooks: https://www.reddit.com/r/Calibre/comments/1ck4w8e/2024_guide_on_removing_drm_from_kobo_kindle_ebooks/")
374
+ import_file = gr.File(label="Upload file for import", file_types=[".epub"])
375
+ title_input = gr.Textbox(label="Title", placeholder="Enter the title of the content")
376
+ author_input = gr.Textbox(label="Author", placeholder="Enter the author's name")
377
+ keywords_input = gr.Textbox(label="Keywords(like genre or publish year)",
378
+ placeholder="Enter keywords, comma-separated")
379
+ system_prompt_input = gr.Textbox(label="System Prompt",
380
+ lines=3,
381
+ value=""""
382
+ <s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
383
+ **Bulleted Note Creation Guidelines**
384
+
385
+ **Headings**:
386
+ - Based on referenced topics, not categories like quotes or terms
387
+ - Surrounded by **bold** formatting
388
+ - Not listed as bullet points
389
+ - No space between headings and list items underneath
390
+
391
+ **Emphasis**:
392
+ - **Important terms** set in bold font
393
+ - **Text ending in a colon**: also bolded
394
+
395
+ **Review**:
396
+ - Ensure adherence to specified format
397
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
398
+ """, )
399
+ custom_prompt_input = gr.Textbox(label="Custom User Prompt",
400
+ placeholder="Enter a custom user prompt for summarization (optional)")
401
+ auto_summarize_checkbox = gr.Checkbox(label="Auto-summarize", value=False)
402
+ api_name_input = gr.Dropdown(
403
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral",
404
+ "OpenRouter",
405
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
406
+ label="API for Auto-summarization"
407
+ )
408
+ api_key_input = gr.Textbox(label="API Key", type="password")
409
+ import_button = gr.Button("Import eBook")
410
+ with gr.Column():
411
+ with gr.Row():
412
+ import_output = gr.Textbox(label="Import Status")
413
+
414
+ def import_epub(epub_file, title, author, keywords, system_prompt, user_prompt, auto_summarize, api_name,
415
+ api_key):
416
+ try:
417
+ # Create a temporary directory to store the converted file
418
+ with tempfile.TemporaryDirectory() as temp_dir:
419
+ epub_path = epub_file.name
420
+ md_path = os.path.join(temp_dir, "converted.md")
421
+
422
+ # Use pypandoc to convert EPUB to Markdown
423
+ output = pypandoc.convert_file(epub_path, 'md', outputfile=md_path)
424
+
425
+ if output != "":
426
+ return f"Error converting EPUB: {output}"
427
+
428
+ # Read the converted markdown content
429
+ with open(md_path, "r", encoding="utf-8") as md_file:
430
+ content = md_file.read()
431
+
432
+ # Now process the content as you would with a text file
433
+ return import_data(content, title, author, keywords, system_prompt,
434
+ user_prompt, auto_summarize, api_name, api_key)
435
+ except Exception as e:
436
+ return f"Error processing EPUB: {str(e)}"
437
+
438
+ import_button.click(
439
+ fn=import_epub,
440
+ inputs=[import_file, title_input, author_input, keywords_input, system_prompt_input,
441
+ custom_prompt_input, auto_summarize_checkbox, api_name_input, api_key_input],
442
+ outputs=import_output
443
+ )
444
+
445
+ def import_obsidian_vault(vault_path, progress=gr.Progress()):
446
+ try:
447
+ from App_Function_Libraries.Gradio_UI.Export_Functionality import scan_obsidian_vault
448
+ markdown_files = scan_obsidian_vault(vault_path)
449
+ total_files = len(markdown_files)
450
+ imported_files = 0
451
+ errors = []
452
+
453
+ for i, file_path in enumerate(markdown_files):
454
+ try:
455
+ note_data = parse_obsidian_note(file_path)
456
+ success, error_msg = import_obsidian_note_to_db(note_data)
457
+ if success:
458
+ imported_files += 1
459
+ else:
460
+ errors.append(error_msg)
461
+ except Exception as e:
462
+ error_msg = f"Error processing {file_path}: {str(e)}"
463
+ logger.error(error_msg)
464
+ errors.append(error_msg)
465
+
466
+ progress((i + 1) / total_files, f"Imported {imported_files} of {total_files} files")
467
+ sleep(0.1) # Small delay to prevent UI freezing
468
+
469
+ return imported_files, total_files, errors
470
+ except Exception as e:
471
+ error_msg = f"Error scanning vault: {str(e)}\n{traceback.format_exc()}"
472
+ logger.error(error_msg)
473
+ return 0, 0, [error_msg]
App_Function_Libraries/Gradio_UI/Introduction_tab.py ADDED
@@ -0,0 +1,161 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Introduction_tab.py
2
+ # Gradio UI functions for the Introduction tab
3
+
4
+ # Imports
5
+ #
6
+ # External Imports
7
+ import gradio as gr
8
+
9
+ from App_Function_Libraries.DB_Manager import get_db_config
10
+
11
+
12
+ #
13
+ # Local Imports
14
+
15
+
16
+ def create_introduction_tab():
17
+ with (gr.TabItem("Introduction")):
18
+ db_config = get_db_config()
19
+ db_type = db_config['type']
20
+ gr.Markdown(f"# tldw: Your LLM-powered Research Multi-tool (Using {db_type.capitalize()} Database)")
21
+ with gr.Row():
22
+ with gr.Column():
23
+ gr.Markdown("""### What can it do?
24
+ - Transcribe and summarize videos from URLs/Local files
25
+ - Transcribe and Summarize Audio files/Podcasts (URL/local file)
26
+ - Summarize articles from URLs/Local notes
27
+ - Ingest and summarize books(epub/PDF)
28
+ - Ingest and summarize research papers (PDFs - WIP)
29
+ - Search and display ingested content + summaries
30
+ - Create and manage custom prompts
31
+ - Chat with an LLM of your choice to generate content using the selected item + Prompts
32
+ - Keyword support for content search and display
33
+ - Export keywords/items to markdown/CSV(csv is wip)
34
+ - Import existing notes from Obsidian to the database (Markdown/txt files or a zip containing a collection of files)
35
+ - View and manage chat history
36
+ - Writing Tools: Grammar & Style check, Tone Analyzer & Editor, more planned...
37
+ - RAG (Retrieval-Augmented Generation) support for content generation(think about asking questions about your entire library of items)
38
+ - More features planned...
39
+ - All powered by your choice of LLM.
40
+ - Currently supports: Local-LLM(llamafile-server), OpenAI, Anthropic, Cohere, Groq, DeepSeek, OpenRouter, Llama.cpp, Kobold, Ooba, Tabbyapi, VLLM and more to come...
41
+ - All data is stored locally in a SQLite database for easy access and management.
42
+ - No trackers (Gradio has some analytics but it's disabled here...)
43
+ - No ads, no tracking, no BS. Just you and your content.
44
+ - Open-source and free to use. Contributions welcome!
45
+ - If you have any thoughts or feedback, please let me know on github or via email.
46
+ """)
47
+ gr.Markdown(
48
+ """Follow this project at [tl/dw: Too Long, Didn't Watch - Your Personal Research Multi-Tool - GitHub](https://github.com/rmusser01/tldw)""")
49
+ with gr.Column():
50
+ gr.Markdown("""### How to use:
51
+ ##### Quick Start: Just click on the appropriate tab for what you're trying to do and fill in the required fields. Click "Process <video/audio/article/etc>" and wait for the results.
52
+ #### Simple Instructions
53
+ - Basic Usage:
54
+ - If you don't have an API key/don't know what an LLM is/don't know what an API key is, please look further down the page for information on getting started.
55
+ - If you want summaries/chat with an LLM, you'll need:
56
+ 1. An API key for the LLM API service you want to use, or,
57
+ 2. A local inference server running an LLM (like llamafile-server/llama.cpp - for instructions on how to do so see the projects README or below), or,
58
+ 3. A "local" inference server you have access to running an LLM.
59
+ - If you just want transcriptions you can ignore the above.
60
+ - Select the tab for the task you want to perform
61
+ - Fill in the required fields
62
+ - Click the "Process" button
63
+ - Wait for the results to appear
64
+ - Download the results if needed
65
+ - Repeat as needed
66
+ - As of writing this, the UI is still a work in progress.
67
+ - That being said, I plan to replace it all eventually. In the meantime, please have patience.
68
+ - The UI is divided into tabs for different tasks.
69
+ - Each tab has a set of fields that you can fill in to perform the task.
70
+ - Some fields are mandatory, some are optional.
71
+ - The fields are mostly self-explanatory, but I will try to add more detailed instructions as I go.
72
+ #### Detailed Usage:
73
+ - There are 8 Top-level tabs in the UI. Each tab has a specific set of tasks that you can perform by selecting one of the 'sub-tabs' made available by clicking on the top tab.
74
+ - The tabs are as follows:
75
+ 1. Transcription / Summarization / Ingestion - This tab is for processing videos, audio files, articles, books, and PDFs/office docs.
76
+ 2. Search / Detailed View - This tab is for searching and displaying content from the database. You can also view detailed information about the selected item.
77
+ 3. Chat with an LLM - This tab is for chatting with an LLM to generate content based on the selected item and prompts.
78
+ 4. Edit Existing Items - This tab is for editing existing items in the database (Prompts + ingested items).
79
+ 5. Writing Tools - This tab is for using various writing tools like Grammar & Style check, Tone Analyzer & Editor, etc.
80
+ 6. Keywords - This tab is for managing keywords for content search and display.
81
+ 7. Import/Export - This tab is for importing notes from Obsidian and exporting keywords/items to markdown/CSV.
82
+ 8. Utilities - This tab contains some random utilities that I thought might be useful.
83
+ - Each sub-tab is responsible for that set of functionality. This is reflected in the codebase as well, where I have split the functionality into separate files for each tab/larger goal.
84
+ """)
85
+ with gr.Row():
86
+ gr.Markdown("""### HELP! I don't know what any of this this shit is!
87
+ ### DON'T PANIC
88
+ #### Its ok, you're not alone, most people have no clue what any of this stuff is.
89
+ - So let's try and fix that.
90
+
91
+ #### Introduction to LLMs:
92
+ - Non-Technical introduction to Generative AI and LLMs: https://paruir.medium.com/understanding-generative-ai-and-llms-a-non-technical-overview-part-1-788c0eb0dd64
93
+ - Google's Intro to LLMs: https://developers.google.com/machine-learning/resources/intro-llms#llm_considerations
94
+ - LLMs 101(coming from a tech background): https://vinija.ai/models/LLM/
95
+ - LLM Fundamentals / LLM Scientist / LLM Engineer courses(Free): https://github.com/mlabonne/llm-course
96
+
97
+ #### Various Phrases & Terms to know
98
+ - **LLM** - Large Language Model - A type of neural network that can generate human-like text.
99
+ - **API** - Application Programming Interface - A set of rules and protocols that allows one software application to communicate with another.
100
+ * Think of it like a post address for a piece of software. You can send messages to and from it.
101
+ - **API Key** - A unique identifier that is used to authenticate a user, developer, or calling program to an API.
102
+ * Like the key to a post office box. You need it to access the contents.
103
+ - **GUI** - Graphical User Interface - the thing facilitating your interact with this application.
104
+ - **DB** - Database
105
+ - **Prompt Engineering** - The process of designing prompts that are used to guide the output of a language model. Is a meme but also very much not.
106
+ - **Quantization** - The process of converting a continuous range of values into a finite range of discrete values.
107
+ * https://github.com/ggerganov/llama.cpp/blob/cddae4884c853b1a7ab420458236d666e2e34423/examples/quantize/README.md#L27
108
+ - **GGUF Files** - GGUF is a binary format that is designed for fast loading and saving of models, and for ease of reading. Models are traditionally developed using PyTorch or another framework, and then converted to GGUF for use in GGML. https://github.com/ggerganov/ggml/blob/master/docs/gguf.md
109
+ - **Inference Engine** - A software system that is designed to execute a model that has been trained by a machine learning algorithm. Llama.cpp and Kobold.cpp are examples of inference engines.
110
+ - **Abliteration** - https://huggingface.co/blog/mlabonne/abliteration
111
+ """)
112
+ with gr.Row():
113
+ gr.Markdown("""### Ok cool, but how do I get started? I don't have an API key or a local server running...
114
+ #### Great, glad you asked! Getting Started:
115
+ - **Getting an API key for a commercial services provider:
116
+ - **OpenAI:**
117
+ * https://platform.openai.com/docs/quickstart
118
+ - **Anthropic:**
119
+ * https://docs.anthropic.com/en/api/getting-started
120
+ - **Cohere:**
121
+ * https://docs.cohere.com/
122
+ * They offer 1k free requests a month(up to 1million tokens total I think?), so you can try it out without paying.
123
+ - **Groq:**
124
+ * https://console.groq.com/keys
125
+ * Offer an account with free credits to try out their service. No idea how much you get.
126
+ - **DeepSeek:**
127
+ * https://platform.deepseek.com/ (Chinese-hosted/is in english)
128
+ - **OpenRouter:**
129
+ * https://openrouter.ai/
130
+ - **Mistral:**
131
+ * https://console.mistral.ai/
132
+ - **Choosing a Model to download**
133
+ - You'll first need to select a model you want to use with the server.
134
+ - Keep in mind that the model you select will determine the quality of the output you get, and that models run fastest when offloaded fully to your GPU.
135
+ * So this means that you can run a large model (Command-R) on CPU+System RAM, but you're gonna see a massive performance hit. Not saying its unusable, but it's not ideal.
136
+ * With that in mind, I would recommend an abliterated version of Meta's Llama3.1 model for most tasks. (Abliterated since it won't refuse requests)
137
+ * I say this because of the general quality of the model + it's context size.
138
+ * You can find the model here: https://huggingface.co/mlabonne/Meta-Llama-3.1-8B-Instruct-abliterated-GGUF
139
+ * And the Q8 quant(total size 8.6GB): https://huggingface.co/mlabonne/Meta-Llama-3.1-8B-Instruct-abliterated-GGUF/resolve/main/meta-llama-3.1-8b-instruct-abliterated.Q8_0.gguf?download=true
140
+ - **Local Inference Server:**
141
+ - **Llamafile-Server (wrapper for llama.cpp):**
142
+ * Run this script with the `--local_llm` argument next time, and you'll be walked through setting up a local instance of llamafile-server.
143
+ - **Llama.cpp Inference Engine:**
144
+ * Download the latest release for your platform here: https://github.com/ggerganov/llama.cpp/releases
145
+ * Windows: `llama-<release_number>-bin-win-cuda-cu<11.7.1 or 12.2.0 - version depends on installed cuda>-x64.zip`
146
+ * Run it: `llama-server.exe --model <path_to_model> -ctx 8192 -ngl 999`
147
+ - `-ctx 8192` sets the context size to 8192 tokens, `-ngl 999` sets the number of layers to offload to the GPU to 999. (essentially ensuring we only use our GPU and not CPU for processing)
148
+ * Macos: `llama-<release_number>-bin-macos-arm64.zip - for Apple Silicon / `llama-<release_number>-bin-macos-x64.zip` - for Intel Macs
149
+ * Run it: `llama-server --model <path_to_model> -ctx 8192 -ngl 999`
150
+ - `-ctx 8192` sets the context size to 8192 tokens, `-ngl 999` sets the number of layers to offload to the GPU to 999. (essentially ensuring we only use our GPU and not CPU for processing)
151
+ * Linux: You can probably figure it out.
152
+ - **Kobold.cpp Server:**
153
+ 1. Download from here: https://github.com/LostRuins/koboldcpp/releases/latest
154
+ 2. `Double click KoboldCPP.exe and select model OR run "KoboldCPP.exe --help" in CMD prompt to get command line arguments for more control.`
155
+ 3. `Generally you don't have to change much besides the Presets and GPU Layers. Run with CuBLAS or CLBlast for GPU acceleration.`
156
+ 4. `Select your GGUF or GGML model you downloaded earlier, and connect to the displayed URL once it finishes loading.`
157
+ - **Linux**
158
+ 1. `On Linux, we provide a koboldcpp-linux-x64 PyInstaller prebuilt binary on the releases page for modern systems. Simply download and run the binary.`
159
+ * Alternatively, you can also install koboldcpp to the current directory by running the following terminal command: `curl -fLo koboldcpp https://github.com/LostRuins/koboldcpp/releases/latest/download/koboldcpp-linux-x64 && chmod +x koboldcpp`
160
+ 2. When you can't use the precompiled binary directly, we provide an automated build script which uses conda to obtain all dependencies, and generates (from source) a ready-to-use a pyinstaller binary for linux users. Simply execute the build script with `./koboldcpp.sh dist` and run the generated binary.
161
+ """)
App_Function_Libraries/Gradio_UI/Keywords.py ADDED
@@ -0,0 +1,65 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Keywords.py
2
+ # Purpose: This file contains the functions to create the Keywords tab in the Gradio UI.
3
+ #
4
+ # The Keywords tab allows the user to add, delete, view, and export keywords from the database.
5
+ #
6
+ # Imports:
7
+
8
+ #
9
+ # External Imports
10
+ import gradio as gr
11
+ #
12
+ # Internal Imports
13
+ from App_Function_Libraries.DB_Manager import add_keyword, delete_keyword, keywords_browser_interface, export_keywords_to_csv
14
+ #
15
+ #
16
+ ######################################################################################################################
17
+ #
18
+ # Functions:
19
+
20
+
21
+ def create_export_keywords_tab():
22
+ with gr.Tab("Export Keywords"):
23
+ with gr.Row():
24
+ with gr.Column():
25
+ export_keywords_button = gr.Button("Export Keywords")
26
+ with gr.Column():
27
+ export_keywords_output = gr.File(label="Download Exported Keywords")
28
+ export_keywords_status = gr.Textbox(label="Export Status")
29
+
30
+ export_keywords_button.click(
31
+ fn=export_keywords_to_csv,
32
+ outputs=[export_keywords_status, export_keywords_output]
33
+ )
34
+
35
+ def create_view_keywords_tab():
36
+ with gr.TabItem("View Keywords"):
37
+ gr.Markdown("# Browse Keywords")
38
+ with gr.Column():
39
+ browse_output = gr.Markdown()
40
+ browse_button = gr.Button("View Existing Keywords")
41
+ browse_button.click(fn=keywords_browser_interface, outputs=browse_output)
42
+
43
+
44
+ def create_add_keyword_tab():
45
+ with gr.TabItem("Add Keywords"):
46
+ with gr.Row():
47
+ with gr.Column():
48
+ gr.Markdown("# Add Keywords to the Database")
49
+ add_input = gr.Textbox(label="Add Keywords (comma-separated)", placeholder="Enter keywords here...")
50
+ add_button = gr.Button("Add Keywords")
51
+ with gr.Row():
52
+ add_output = gr.Textbox(label="Result")
53
+ add_button.click(fn=add_keyword, inputs=add_input, outputs=add_output)
54
+
55
+
56
+ def create_delete_keyword_tab():
57
+ with gr.Tab("Delete Keywords"):
58
+ with gr.Row():
59
+ with gr.Column():
60
+ gr.Markdown("# Delete Keywords from the Database")
61
+ delete_input = gr.Textbox(label="Delete Keyword", placeholder="Enter keyword to delete here...")
62
+ delete_button = gr.Button("Delete Keyword")
63
+ with gr.Row():
64
+ delete_output = gr.Textbox(label="Result")
65
+ delete_button.click(fn=delete_keyword, inputs=delete_input, outputs=delete_output)
App_Function_Libraries/Gradio_UI/Llamafile_tab.py ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Llamafile_tab.py
2
+ # Description: Functions relating to the Llamafile tab
3
+ #
4
+ # Imports
5
+ import os
6
+ import glob
7
+ #
8
+ # External Imports
9
+ import gradio as gr
10
+ #
11
+ # Local Imports
12
+ from App_Function_Libraries.Llamafile import start_llamafile
13
+ #
14
+ #######################################################################################################################
15
+ #
16
+ # Functions:
17
+
18
+
19
+ def create_chat_with_llamafile_tab():
20
+ def get_model_files(directory):
21
+ pattern = os.path.join(directory, "*.{gguf,llamafile}")
22
+ return [os.path.basename(f) for f in glob.glob(pattern)]
23
+
24
+ def update_dropdowns():
25
+ current_dir_models = get_model_files(".")
26
+ parent_dir_models = get_model_files("..")
27
+ return (
28
+ {"choices": current_dir_models, "value": None},
29
+ {"choices": parent_dir_models, "value": None}
30
+ )
31
+
32
+ with gr.TabItem("Local LLM with Llamafile"):
33
+ gr.Markdown("# Settings for Llamafile")
34
+ with gr.Row():
35
+ with gr.Column():
36
+ am_noob = gr.Checkbox(label="Check this to enable sane defaults", value=False, visible=True)
37
+ # FIXME - these get deleted at some point?
38
+ advanced_mode_toggle = gr.Checkbox(label="Advanced Mode - Enable to show all settings", value=False)
39
+
40
+
41
+ with gr.Column():
42
+ # FIXME - make this actually work
43
+ model_checked = gr.Checkbox(label="Enable Setting Local LLM Model Path", value=False, visible=True)
44
+ current_dir_dropdown = gr.Dropdown(
45
+ label="Select Model from Current Directory (.)",
46
+ choices=[], # Start with an empty list
47
+ visible=True
48
+ )
49
+ parent_dir_dropdown = gr.Dropdown(
50
+ label="Select Model from Parent Directory (..)",
51
+ choices=[], # Start with an empty list
52
+ visible=True
53
+ )
54
+ refresh_button = gr.Button("Refresh Model Lists")
55
+ model_value = gr.Textbox(label="Selected Model File", value="", visible=True)
56
+ with gr.Row():
57
+ with gr.Column():
58
+ ngl_checked = gr.Checkbox(label="Enable Setting GPU Layers", value=False, visible=True)
59
+ ngl_value = gr.Number(label="Number of GPU Layers", value=None, precision=0, visible=True)
60
+ advanced_inputs = create_llamafile_advanced_inputs()
61
+ with gr.Column():
62
+ start_button = gr.Button("Start Llamafile")
63
+ stop_button = gr.Button("Stop Llamafile (doesn't work)")
64
+ output_display = gr.Markdown()
65
+
66
+
67
+ def update_model_value(current_dir_model, parent_dir_model):
68
+ if current_dir_model:
69
+ return current_dir_model
70
+ elif parent_dir_model:
71
+ return os.path.join("..", parent_dir_model)
72
+ else:
73
+ return ""
74
+
75
+ current_dir_dropdown.change(
76
+ fn=update_model_value,
77
+ inputs=[current_dir_dropdown, parent_dir_dropdown],
78
+ outputs=model_value
79
+ )
80
+ parent_dir_dropdown.change(
81
+ fn=update_model_value,
82
+ inputs=[current_dir_dropdown, parent_dir_dropdown],
83
+ outputs=model_value
84
+ )
85
+
86
+ refresh_button.click(
87
+ fn=update_dropdowns,
88
+ inputs=[],
89
+ outputs=[current_dir_dropdown, parent_dir_dropdown]
90
+ )
91
+
92
+ start_button.click(
93
+ fn=start_llamafile,
94
+ inputs=[am_noob, model_checked, model_value, ngl_checked, ngl_value] + advanced_inputs,
95
+ outputs=output_display
96
+ )
97
+
98
+
99
+ def create_llamafile_advanced_inputs():
100
+ verbose_checked = gr.Checkbox(label="Enable Verbose Output", value=False, visible=False)
101
+ threads_checked = gr.Checkbox(label="Set CPU Threads", value=False, visible=False)
102
+ threads_value = gr.Number(label="Number of CPU Threads", value=None, precision=0, visible=False)
103
+ http_threads_checked = gr.Checkbox(label="Set HTTP Server Threads", value=False, visible=False)
104
+ http_threads_value = gr.Number(label="Number of HTTP Server Threads", value=None, precision=0, visible=False)
105
+ hf_repo_checked = gr.Checkbox(label="Use Huggingface Repo Model", value=False, visible=False)
106
+ hf_repo_value = gr.Textbox(label="Huggingface Repo Name", value="", visible=False)
107
+ hf_file_checked = gr.Checkbox(label="Set Huggingface Model File", value=False, visible=False)
108
+ hf_file_value = gr.Textbox(label="Huggingface Model File", value="", visible=False)
109
+ ctx_size_checked = gr.Checkbox(label="Set Prompt Context Size", value=False, visible=False)
110
+ ctx_size_value = gr.Number(label="Prompt Context Size", value=8124, precision=0, visible=False)
111
+ host_checked = gr.Checkbox(label="Set IP to Listen On", value=False, visible=False)
112
+ host_value = gr.Textbox(label="Host IP Address", value="", visible=False)
113
+ port_checked = gr.Checkbox(label="Set Server Port", value=False, visible=False)
114
+ port_value = gr.Number(label="Port Number", value=None, precision=0, visible=False)
115
+
116
+ return [verbose_checked, threads_checked, threads_value, http_threads_checked, http_threads_value,
117
+ hf_repo_checked, hf_repo_value, hf_file_checked, hf_file_value, ctx_size_checked, ctx_size_value,
118
+ host_checked, host_value, port_checked, port_value]
119
+
120
+ #
121
+ # End of Llamafile_tab.py
122
+ #########################################################################################################################
App_Function_Libraries/Gradio_UI/Media_edit.py ADDED
@@ -0,0 +1,273 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Media_edit.py
2
+ # Functions for Gradio Media_Edit UI
3
+
4
+ # Imports
5
+ import logging
6
+ import uuid
7
+
8
+ # External Imports
9
+ import gradio as gr
10
+ #
11
+ # Local Imports
12
+ from App_Function_Libraries.DB_Manager import add_prompt, update_media_content, db, add_or_update_prompt, \
13
+ load_prompt_details
14
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import update_dropdown, update_prompt_dropdown
15
+ from App_Function_Libraries.SQLite_DB import fetch_item_details
16
+
17
+
18
+ def create_media_edit_tab():
19
+ with gr.TabItem("Edit Existing Items"):
20
+ gr.Markdown("# Search and Edit Media Items")
21
+
22
+ with gr.Row():
23
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
24
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title", label="Search By")
25
+ search_button = gr.Button("Search")
26
+
27
+ with gr.Row():
28
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
29
+ item_mapping = gr.State({})
30
+
31
+ content_input = gr.Textbox(label="Edit Content", lines=10)
32
+ prompt_input = gr.Textbox(label="Edit Prompt", lines=3)
33
+ summary_input = gr.Textbox(label="Edit Summary", lines=5)
34
+
35
+ update_button = gr.Button("Update Media Content")
36
+ status_message = gr.Textbox(label="Status", interactive=False)
37
+
38
+ search_button.click(
39
+ fn=update_dropdown,
40
+ inputs=[search_query_input, search_type_input],
41
+ outputs=[items_output, item_mapping]
42
+ )
43
+
44
+ def load_selected_media_content(selected_item, item_mapping):
45
+ if selected_item and item_mapping and selected_item in item_mapping:
46
+ media_id = item_mapping[selected_item]
47
+ # FIXME - fetch_item_details is not handled by DB_Manager!
48
+ content, prompt, summary = fetch_item_details(media_id)
49
+ return content, prompt, summary
50
+ return "No item selected or invalid selection", "", ""
51
+
52
+ items_output.change(
53
+ fn=load_selected_media_content,
54
+ inputs=[items_output, item_mapping],
55
+ outputs=[content_input, prompt_input, summary_input]
56
+ )
57
+
58
+ update_button.click(
59
+ fn=update_media_content,
60
+ inputs=[items_output, item_mapping, content_input, prompt_input, summary_input],
61
+ outputs=status_message
62
+ )
63
+
64
+
65
+ def create_media_edit_and_clone_tab():
66
+ with gr.TabItem("Clone and Edit Existing Items"):
67
+ gr.Markdown("# Search, Edit, and Clone Existing Items")
68
+
69
+ with gr.Row():
70
+ with gr.Column():
71
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
72
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title",
73
+ label="Search By")
74
+ with gr.Column():
75
+ search_button = gr.Button("Search")
76
+ clone_button = gr.Button("Clone Item")
77
+ save_clone_button = gr.Button("Save Cloned Item", visible=False)
78
+ with gr.Row():
79
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
80
+ item_mapping = gr.State({})
81
+
82
+ content_input = gr.Textbox(label="Edit Content", lines=10)
83
+ prompt_input = gr.Textbox(label="Edit Prompt", lines=3)
84
+ summary_input = gr.Textbox(label="Edit Summary", lines=5)
85
+ new_title_input = gr.Textbox(label="New Title (for cloning)", visible=False)
86
+ status_message = gr.Textbox(label="Status", interactive=False)
87
+
88
+ search_button.click(
89
+ fn=update_dropdown,
90
+ inputs=[search_query_input, search_type_input],
91
+ outputs=[items_output, item_mapping]
92
+ )
93
+
94
+ def load_selected_media_content(selected_item, item_mapping):
95
+ if selected_item and item_mapping and selected_item in item_mapping:
96
+ media_id = item_mapping[selected_item]
97
+ content, prompt, summary = fetch_item_details(media_id)
98
+ return content, prompt, summary, gr.update(visible=True), gr.update(visible=False)
99
+ return "No item selected or invalid selection", "", "", gr.update(visible=False), gr.update(visible=False)
100
+
101
+ items_output.change(
102
+ fn=load_selected_media_content,
103
+ inputs=[items_output, item_mapping],
104
+ outputs=[content_input, prompt_input, summary_input, clone_button, save_clone_button]
105
+ )
106
+
107
+ def prepare_for_cloning(selected_item):
108
+ return gr.update(value=f"Copy of {selected_item}", visible=True), gr.update(visible=True)
109
+
110
+ clone_button.click(
111
+ fn=prepare_for_cloning,
112
+ inputs=[items_output],
113
+ outputs=[new_title_input, save_clone_button]
114
+ )
115
+
116
+ def save_cloned_item(selected_item, item_mapping, content, prompt, summary, new_title):
117
+ if selected_item and item_mapping and selected_item in item_mapping:
118
+ original_media_id = item_mapping[selected_item]
119
+ try:
120
+ with db.get_connection() as conn:
121
+ cursor = conn.cursor()
122
+
123
+ # Fetch the original item's details
124
+ cursor.execute("SELECT type, url FROM Media WHERE id = ?", (original_media_id,))
125
+ original_type, original_url = cursor.fetchone()
126
+
127
+ # Generate a new unique URL
128
+ new_url = f"{original_url}_clone_{uuid.uuid4().hex[:8]}"
129
+
130
+ # Insert new item into Media table
131
+ cursor.execute("""
132
+ INSERT INTO Media (title, content, url, type)
133
+ VALUES (?, ?, ?, ?)
134
+ """, (new_title, content, new_url, original_type))
135
+
136
+ new_media_id = cursor.lastrowid
137
+
138
+ # Insert new item into MediaModifications table
139
+ cursor.execute("""
140
+ INSERT INTO MediaModifications (media_id, prompt, summary, modification_date)
141
+ VALUES (?, ?, ?, CURRENT_TIMESTAMP)
142
+ """, (new_media_id, prompt, summary))
143
+
144
+ # Copy keywords from the original item
145
+ cursor.execute("""
146
+ INSERT INTO MediaKeywords (media_id, keyword_id)
147
+ SELECT ?, keyword_id
148
+ FROM MediaKeywords
149
+ WHERE media_id = ?
150
+ """, (new_media_id, original_media_id))
151
+
152
+ # Update full-text search index
153
+ cursor.execute("""
154
+ INSERT INTO media_fts (rowid, title, content)
155
+ VALUES (?, ?, ?)
156
+ """, (new_media_id, new_title, content))
157
+
158
+ conn.commit()
159
+
160
+ return f"Cloned item saved successfully with ID: {new_media_id}", gr.update(
161
+ visible=False), gr.update(visible=False)
162
+ except Exception as e:
163
+ logging.error(f"Error saving cloned item: {e}")
164
+ return f"Error saving cloned item: {str(e)}", gr.update(visible=True), gr.update(visible=True)
165
+ else:
166
+ return "No item selected or invalid selection", gr.update(visible=True), gr.update(visible=True)
167
+
168
+ save_clone_button.click(
169
+ fn=save_cloned_item,
170
+ inputs=[items_output, item_mapping, content_input, prompt_input, summary_input, new_title_input],
171
+ outputs=[status_message, new_title_input, save_clone_button]
172
+ )
173
+
174
+
175
+ def create_prompt_edit_tab():
176
+ with gr.TabItem("Edit Prompts"):
177
+ with gr.Row():
178
+ with gr.Column():
179
+ prompt_dropdown = gr.Dropdown(
180
+ label="Select Prompt",
181
+ choices=[],
182
+ interactive=True
183
+ )
184
+ prompt_list_button = gr.Button("List Prompts")
185
+
186
+ with gr.Column():
187
+ title_input = gr.Textbox(label="Title", placeholder="Enter the prompt title")
188
+ description_input = gr.Textbox(label="Description", placeholder="Enter the prompt description", lines=3)
189
+ system_prompt_input = gr.Textbox(label="System Prompt", placeholder="Enter the system prompt", lines=3)
190
+ user_prompt_input = gr.Textbox(label="User Prompt", placeholder="Enter the user prompt", lines=3)
191
+ add_prompt_button = gr.Button("Add/Update Prompt")
192
+ add_prompt_output = gr.HTML()
193
+
194
+ # Event handlers
195
+ prompt_list_button.click(
196
+ fn=update_prompt_dropdown,
197
+ outputs=prompt_dropdown
198
+ )
199
+
200
+ add_prompt_button.click(
201
+ fn=add_or_update_prompt,
202
+ inputs=[title_input, description_input, system_prompt_input, user_prompt_input],
203
+ outputs=add_prompt_output
204
+ )
205
+
206
+ # Load prompt details when selected
207
+ prompt_dropdown.change(
208
+ fn=load_prompt_details,
209
+ inputs=[prompt_dropdown],
210
+ outputs=[title_input, description_input, system_prompt_input, user_prompt_input]
211
+ )
212
+
213
+
214
+ def create_prompt_clone_tab():
215
+ with gr.TabItem("Clone and Edit Prompts"):
216
+ with gr.Row():
217
+ with gr.Column():
218
+ gr.Markdown("# Clone and Edit Prompts")
219
+ prompt_dropdown = gr.Dropdown(
220
+ label="Select Prompt",
221
+ choices=[],
222
+ interactive=True
223
+ )
224
+ prompt_list_button = gr.Button("List Prompts")
225
+
226
+ with gr.Column():
227
+ title_input = gr.Textbox(label="Title", placeholder="Enter the prompt title")
228
+ description_input = gr.Textbox(label="Description", placeholder="Enter the prompt description", lines=3)
229
+ system_prompt_input = gr.Textbox(label="System Prompt", placeholder="Enter the system prompt", lines=3)
230
+ user_prompt_input = gr.Textbox(label="User Prompt", placeholder="Enter the user prompt", lines=3)
231
+ clone_prompt_button = gr.Button("Clone Selected Prompt")
232
+ save_cloned_prompt_button = gr.Button("Save Cloned Prompt", visible=False)
233
+ add_prompt_output = gr.HTML()
234
+
235
+ # Event handlers
236
+ prompt_list_button.click(
237
+ fn=update_prompt_dropdown,
238
+ outputs=prompt_dropdown
239
+ )
240
+
241
+ # Load prompt details when selected
242
+ prompt_dropdown.change(
243
+ fn=load_prompt_details,
244
+ inputs=[prompt_dropdown],
245
+ outputs=[title_input, description_input, system_prompt_input, user_prompt_input]
246
+ )
247
+
248
+ def prepare_for_cloning(selected_prompt):
249
+ if selected_prompt:
250
+ return gr.update(value=f"Copy of {selected_prompt}"), gr.update(visible=True)
251
+ return gr.update(), gr.update(visible=False)
252
+
253
+ clone_prompt_button.click(
254
+ fn=prepare_for_cloning,
255
+ inputs=[prompt_dropdown],
256
+ outputs=[title_input, save_cloned_prompt_button]
257
+ )
258
+
259
+ def save_cloned_prompt(title, description, system_prompt, user_prompt):
260
+ try:
261
+ result = add_prompt(title, description, system_prompt, user_prompt)
262
+ if result == "Prompt added successfully.":
263
+ return result, gr.update(choices=update_prompt_dropdown())
264
+ else:
265
+ return result, gr.update()
266
+ except Exception as e:
267
+ return f"Error saving cloned prompt: {str(e)}", gr.update()
268
+
269
+ save_cloned_prompt_button.click(
270
+ fn=save_cloned_prompt,
271
+ inputs=[title_input, description_input, system_prompt_input, user_prompt_input],
272
+ outputs=[add_prompt_output, prompt_dropdown]
273
+ )
App_Function_Libraries/Gradio_UI/PDF_ingestion_tab.py ADDED
@@ -0,0 +1,152 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # PDF_ingestion_tab.py
2
+ # Gradio UI for ingesting PDFs into the database
3
+ import os
4
+ import shutil
5
+ import tempfile
6
+
7
+ # Imports
8
+ #
9
+ # External Imports
10
+ import gradio as gr
11
+ #
12
+ # Local Imports
13
+ from App_Function_Libraries.DB_Manager import load_preset_prompts
14
+ from App_Function_Libraries.Gradio_UI.Chat_ui import update_user_prompt
15
+ from App_Function_Libraries.PDF_Ingestion_Lib import extract_metadata_from_pdf, extract_text_and_format_from_pdf, \
16
+ process_and_cleanup_pdf
17
+ #
18
+ #
19
+ ########################################################################################################################
20
+ #
21
+ # Functions:
22
+
23
+ def create_pdf_ingestion_tab():
24
+ with gr.TabItem("PDF Ingestion"):
25
+ # TODO - Add functionality to extract metadata from pdf as part of conversion process in marker
26
+ gr.Markdown("# Ingest PDF Files and Extract Metadata")
27
+ with gr.Row():
28
+ with gr.Column():
29
+ pdf_file_input = gr.File(label="Uploaded PDF File", file_types=[".pdf"], visible=False)
30
+ pdf_upload_button = gr.UploadButton("Click to Upload PDF", file_types=[".pdf"])
31
+ pdf_title_input = gr.Textbox(label="Title (Optional)")
32
+ pdf_author_input = gr.Textbox(label="Author (Optional)")
33
+ pdf_keywords_input = gr.Textbox(label="Keywords (Optional, comma-separated)")
34
+ with gr.Row():
35
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
36
+ value=False,
37
+ visible=True)
38
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
39
+ value=False,
40
+ visible=True)
41
+ with gr.Row():
42
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
43
+ choices=load_preset_prompts(),
44
+ visible=False)
45
+ with gr.Row():
46
+ custom_prompt_input = gr.Textbox(label="Custom Prompt",
47
+ placeholder="Enter custom prompt here",
48
+ lines=3,
49
+ visible=False)
50
+ with gr.Row():
51
+ system_prompt_input = gr.Textbox(label="System Prompt",
52
+ value="""
53
+ <s>You are a bulleted notes specialist.
54
+ [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
55
+ **Bulleted Note Creation Guidelines**
56
+
57
+ **Headings**:
58
+ - Based on referenced topics, not categories like quotes or terms
59
+ - Surrounded by **bold** formatting
60
+ - Not listed as bullet points
61
+ - No space between headings and list items underneath
62
+
63
+ **Emphasis**:
64
+ - **Important terms** set in bold font
65
+ - **Text ending in a colon**: also bolded
66
+
67
+ **Review**:
68
+ - Ensure adherence to specified format
69
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]""",
70
+ lines=3,
71
+ visible=False)
72
+
73
+ custom_prompt_checkbox.change(
74
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
75
+ inputs=[custom_prompt_checkbox],
76
+ outputs=[custom_prompt_input, system_prompt_input]
77
+ )
78
+ preset_prompt_checkbox.change(
79
+ fn=lambda x: gr.update(visible=x),
80
+ inputs=[preset_prompt_checkbox],
81
+ outputs=[preset_prompt]
82
+ )
83
+
84
+ def update_prompts(preset_name):
85
+ prompts = update_user_prompt(preset_name)
86
+ return (
87
+ gr.update(value=prompts["user_prompt"], visible=True),
88
+ gr.update(value=prompts["system_prompt"], visible=True)
89
+ )
90
+
91
+ preset_prompt.change(
92
+ update_prompts,
93
+ inputs=preset_prompt,
94
+ outputs=[custom_prompt_input, system_prompt_input]
95
+ )
96
+
97
+ pdf_ingest_button = gr.Button("Ingest PDF")
98
+
99
+ pdf_upload_button.upload(fn=lambda file: file, inputs=pdf_upload_button, outputs=pdf_file_input)
100
+ with gr.Column():
101
+ pdf_result_output = gr.Textbox(label="Result")
102
+
103
+ pdf_ingest_button.click(
104
+ fn=process_and_cleanup_pdf,
105
+ inputs=[pdf_file_input, pdf_title_input, pdf_author_input, pdf_keywords_input],
106
+ outputs=pdf_result_output
107
+ )
108
+
109
+
110
+ def test_pdf_ingestion(pdf_file):
111
+ if pdf_file is None:
112
+ return "No file uploaded", ""
113
+
114
+ try:
115
+ # Create a temporary directory
116
+ with tempfile.TemporaryDirectory() as temp_dir:
117
+ # Create a path for the temporary PDF file
118
+ temp_path = os.path.join(temp_dir, "temp.pdf")
119
+
120
+ # Copy the contents of the uploaded file to the temporary file
121
+ shutil.copy(pdf_file.name, temp_path)
122
+
123
+ # Extract text and convert to Markdown
124
+ markdown_text = extract_text_and_format_from_pdf(temp_path)
125
+
126
+ # Extract metadata from PDF
127
+ metadata = extract_metadata_from_pdf(temp_path)
128
+
129
+ # Use metadata for title and author if not provided
130
+ title = metadata.get('title', os.path.splitext(os.path.basename(pdf_file.name))[0])
131
+ author = metadata.get('author', 'Unknown')
132
+
133
+ result = f"PDF '{title}' by {author} processed successfully."
134
+ return result, markdown_text
135
+ except Exception as e:
136
+ return f"Error ingesting PDF: {str(e)}", ""
137
+
138
+ def create_pdf_ingestion_test_tab():
139
+ with gr.TabItem("Test PDF Ingestion"):
140
+ with gr.Row():
141
+ with gr.Column():
142
+ pdf_file_input = gr.File(label="Upload PDF for testing")
143
+ test_button = gr.Button("Test PDF Ingestion")
144
+ with gr.Column():
145
+ test_output = gr.Textbox(label="Test Result")
146
+ pdf_content_output = gr.Textbox(label="PDF Content", lines=200)
147
+ test_button.click(
148
+ fn=test_pdf_ingestion,
149
+ inputs=[pdf_file_input],
150
+ outputs=[test_output, pdf_content_output]
151
+ )
152
+
App_Function_Libraries/Gradio_UI/Podcast_tab.py ADDED
@@ -0,0 +1,164 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Podcast_tab.py
2
+ # Description: Gradio UI for ingesting podcasts into the database
3
+ #
4
+ # Imports
5
+ #
6
+ #
7
+ # External Imports
8
+ import gradio as gr
9
+ #
10
+ # Local Imports
11
+ from App_Function_Libraries.Audio_Files import process_podcast
12
+ from App_Function_Libraries.DB_Manager import load_preset_prompts
13
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import whisper_models, update_user_prompt
14
+
15
+
16
+ #
17
+ ########################################################################################################################
18
+ #
19
+ # Functions:
20
+
21
+
22
+ def create_podcast_tab():
23
+ with gr.TabItem("Podcast"):
24
+ gr.Markdown("# Podcast Transcription and Ingestion")
25
+ with gr.Row():
26
+ with gr.Column():
27
+ podcast_url_input = gr.Textbox(label="Podcast URL", placeholder="Enter the podcast URL here")
28
+ podcast_title_input = gr.Textbox(label="Podcast Title", placeholder="Will be auto-detected if possible")
29
+ podcast_author_input = gr.Textbox(label="Podcast Author", placeholder="Will be auto-detected if possible")
30
+
31
+ podcast_keywords_input = gr.Textbox(
32
+ label="Keywords",
33
+ placeholder="Enter keywords here (comma-separated, include series name if applicable)",
34
+ value="podcast,audio",
35
+ elem_id="podcast-keywords-input"
36
+ )
37
+
38
+ with gr.Row():
39
+ podcast_custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
40
+ value=False,
41
+ visible=True)
42
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
43
+ value=False,
44
+ visible=True)
45
+ with gr.Row():
46
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
47
+ choices=load_preset_prompts(),
48
+ visible=False)
49
+ with gr.Row():
50
+ podcast_custom_prompt_input = gr.Textbox(label="Custom Prompt",
51
+ placeholder="Enter custom prompt here",
52
+ lines=3,
53
+ visible=False)
54
+ with gr.Row():
55
+ system_prompt_input = gr.Textbox(label="System Prompt",
56
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
57
+ **Bulleted Note Creation Guidelines**
58
+
59
+ **Headings**:
60
+ - Based on referenced topics, not categories like quotes or terms
61
+ - Surrounded by **bold** formatting
62
+ - Not listed as bullet points
63
+ - No space between headings and list items underneath
64
+
65
+ **Emphasis**:
66
+ - **Important terms** set in bold font
67
+ - **Text ending in a colon**: also bolded
68
+
69
+ **Review**:
70
+ - Ensure adherence to specified format
71
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
72
+ """,
73
+ lines=3,
74
+ visible=False)
75
+
76
+ podcast_custom_prompt_checkbox.change(
77
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
78
+ inputs=[podcast_custom_prompt_checkbox],
79
+ outputs=[podcast_custom_prompt_input, system_prompt_input]
80
+ )
81
+ preset_prompt_checkbox.change(
82
+ fn=lambda x: gr.update(visible=x),
83
+ inputs=[preset_prompt_checkbox],
84
+ outputs=[preset_prompt]
85
+ )
86
+
87
+ def update_prompts(preset_name):
88
+ prompts = update_user_prompt(preset_name)
89
+ return (
90
+ gr.update(value=prompts["user_prompt"], visible=True),
91
+ gr.update(value=prompts["system_prompt"], visible=True)
92
+ )
93
+
94
+ preset_prompt.change(
95
+ update_prompts,
96
+ inputs=preset_prompt,
97
+ outputs=[podcast_custom_prompt_input, system_prompt_input]
98
+ )
99
+
100
+ podcast_api_name_input = gr.Dropdown(
101
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter", "Llama.cpp",
102
+ "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
103
+ value=None,
104
+ label="API Name for Summarization (Optional)"
105
+ )
106
+ podcast_api_key_input = gr.Textbox(label="API Key (if required)", type="password")
107
+ podcast_whisper_model_input = gr.Dropdown(choices=whisper_models, value="medium", label="Whisper Model")
108
+
109
+ keep_original_input = gr.Checkbox(label="Keep original audio file", value=False)
110
+ enable_diarization_input = gr.Checkbox(label="Enable speaker diarization", value=False)
111
+
112
+ use_cookies_input = gr.Checkbox(label="Use cookies for yt-dlp", value=False)
113
+ cookies_input = gr.Textbox(
114
+ label="yt-dlp Cookies",
115
+ placeholder="Paste your cookies here (JSON format)",
116
+ lines=3,
117
+ visible=False
118
+ )
119
+
120
+ use_cookies_input.change(
121
+ fn=lambda x: gr.update(visible=x),
122
+ inputs=[use_cookies_input],
123
+ outputs=[cookies_input]
124
+ )
125
+
126
+ chunking_options_checkbox = gr.Checkbox(label="Show Chunking Options", value=False)
127
+ with gr.Row(visible=False) as chunking_options_box:
128
+ gr.Markdown("### Chunking Options")
129
+ with gr.Column():
130
+ chunk_method = gr.Dropdown(choices=['words', 'sentences', 'paragraphs', 'tokens'], label="Chunking Method")
131
+ max_chunk_size = gr.Slider(minimum=100, maximum=1000, value=300, step=50, label="Max Chunk Size")
132
+ chunk_overlap = gr.Slider(minimum=0, maximum=100, value=0, step=10, label="Chunk Overlap")
133
+ use_adaptive_chunking = gr.Checkbox(label="Use Adaptive Chunking")
134
+ use_multi_level_chunking = gr.Checkbox(label="Use Multi-level Chunking")
135
+ chunk_language = gr.Dropdown(choices=['english', 'french', 'german', 'spanish'], label="Chunking Language")
136
+
137
+ chunking_options_checkbox.change(
138
+ fn=lambda x: gr.update(visible=x),
139
+ inputs=[chunking_options_checkbox],
140
+ outputs=[chunking_options_box]
141
+ )
142
+
143
+ podcast_process_button = gr.Button("Process Podcast")
144
+
145
+ with gr.Column():
146
+ podcast_progress_output = gr.Textbox(label="Progress")
147
+ podcast_error_output = gr.Textbox(label="Error Messages")
148
+ podcast_transcription_output = gr.Textbox(label="Transcription")
149
+ podcast_summary_output = gr.Textbox(label="Summary")
150
+ download_transcription = gr.File(label="Download Transcription as JSON")
151
+ download_summary = gr.File(label="Download Summary as Text")
152
+
153
+ podcast_process_button.click(
154
+ fn=process_podcast,
155
+ inputs=[podcast_url_input, podcast_title_input, podcast_author_input,
156
+ podcast_keywords_input, podcast_custom_prompt_input, podcast_api_name_input,
157
+ podcast_api_key_input, podcast_whisper_model_input, keep_original_input,
158
+ enable_diarization_input, use_cookies_input, cookies_input,
159
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
160
+ use_multi_level_chunking, chunk_language],
161
+ outputs=[podcast_progress_output, podcast_transcription_output, podcast_summary_output,
162
+ podcast_title_input, podcast_author_input, podcast_keywords_input, podcast_error_output,
163
+ download_transcription, download_summary]
164
+ )
App_Function_Libraries/Gradio_UI/Re_summarize_tab.py ADDED
@@ -0,0 +1,268 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Re_summarize_tab.py
2
+ # Gradio UI for Re-summarizing items in the database
3
+ #
4
+ # Imports
5
+ import json
6
+ import logging
7
+ #
8
+ # External Imports
9
+ import gradio as gr
10
+ #
11
+ # Local Imports
12
+ from App_Function_Libraries.Chunk_Lib import improved_chunking_process
13
+ from App_Function_Libraries.DB_Manager import update_media_content, load_preset_prompts
14
+ from App_Function_Libraries.Gradio_UI.Chat_ui import update_user_prompt
15
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import fetch_item_details, fetch_items_by_keyword, \
16
+ fetch_items_by_content, fetch_items_by_title_or_url
17
+ from App_Function_Libraries.Summarization_General_Lib import summarize_chunk
18
+ from App_Function_Libraries.Utils import load_comprehensive_config
19
+ #
20
+ #
21
+ ######################################################################################################################
22
+ #
23
+ # Functions:
24
+
25
+ def create_resummary_tab():
26
+ with gr.TabItem("Re-Summarize"):
27
+ gr.Markdown("# Re-Summarize Existing Content")
28
+ with gr.Row():
29
+ with gr.Column():
30
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
31
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title", label="Search By")
32
+ search_button = gr.Button("Search")
33
+
34
+ items_output = gr.Dropdown(label="Select Item", choices=[], interactive=True)
35
+ item_mapping = gr.State({})
36
+
37
+ with gr.Row():
38
+ api_name_input = gr.Dropdown(
39
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
40
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
41
+ value="Local-LLM", label="API Name")
42
+ api_key_input = gr.Textbox(label="API Key", placeholder="Enter your API key here", type="password")
43
+
44
+ chunking_options_checkbox = gr.Checkbox(label="Use Chunking", value=False)
45
+ with gr.Row(visible=False) as chunking_options_box:
46
+ chunk_method = gr.Dropdown(choices=['words', 'sentences', 'paragraphs', 'tokens', 'chapters'],
47
+ label="Chunking Method", value='words')
48
+ max_chunk_size = gr.Slider(minimum=100, maximum=1000, value=300, step=50, label="Max Chunk Size")
49
+ chunk_overlap = gr.Slider(minimum=0, maximum=100, value=0, step=10, label="Chunk Overlap")
50
+
51
+ with gr.Row():
52
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
53
+ value=False,
54
+ visible=True)
55
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
56
+ value=False,
57
+ visible=True)
58
+ with gr.Row():
59
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
60
+ choices=load_preset_prompts(),
61
+ visible=False)
62
+ with gr.Row():
63
+ custom_prompt_input = gr.Textbox(label="Custom Prompt",
64
+ placeholder="Enter custom prompt here",
65
+ lines=3,
66
+ visible=False)
67
+ with gr.Row():
68
+ system_prompt_input = gr.Textbox(label="System Prompt",
69
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
70
+ **Bulleted Note Creation Guidelines**
71
+
72
+ **Headings**:
73
+ - Based on referenced topics, not categories like quotes or terms
74
+ - Surrounded by **bold** formatting
75
+ - Not listed as bullet points
76
+ - No space between headings and list items underneath
77
+
78
+ **Emphasis**:
79
+ - **Important terms** set in bold font
80
+ - **Text ending in a colon**: also bolded
81
+
82
+ **Review**:
83
+ - Ensure adherence to specified format
84
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
85
+ """,
86
+ lines=3,
87
+ visible=False)
88
+
89
+ def update_prompts(preset_name):
90
+ prompts = update_user_prompt(preset_name)
91
+ return (
92
+ gr.update(value=prompts["user_prompt"], visible=True),
93
+ gr.update(value=prompts["system_prompt"], visible=True)
94
+ )
95
+
96
+ preset_prompt.change(
97
+ update_prompts,
98
+ inputs=preset_prompt,
99
+ outputs=[custom_prompt_input, system_prompt_input]
100
+ )
101
+
102
+ resummarize_button = gr.Button("Re-Summarize")
103
+ with gr.Column():
104
+ result_output = gr.Textbox(label="Result")
105
+
106
+ custom_prompt_checkbox.change(
107
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
108
+ inputs=[custom_prompt_checkbox],
109
+ outputs=[custom_prompt_input, system_prompt_input]
110
+ )
111
+ preset_prompt_checkbox.change(
112
+ fn=lambda x: gr.update(visible=x),
113
+ inputs=[preset_prompt_checkbox],
114
+ outputs=[preset_prompt]
115
+ )
116
+
117
+ # Connect the UI elements
118
+ search_button.click(
119
+ fn=update_resummarize_dropdown,
120
+ inputs=[search_query_input, search_type_input],
121
+ outputs=[items_output, item_mapping]
122
+ )
123
+
124
+ chunking_options_checkbox.change(
125
+ fn=lambda x: gr.update(visible=x),
126
+ inputs=[chunking_options_checkbox],
127
+ outputs=[chunking_options_box]
128
+ )
129
+
130
+ custom_prompt_checkbox.change(
131
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
132
+ inputs=[custom_prompt_checkbox],
133
+ outputs=[custom_prompt_input, system_prompt_input]
134
+ )
135
+
136
+ resummarize_button.click(
137
+ fn=resummarize_content_wrapper,
138
+ inputs=[items_output, item_mapping, api_name_input, api_key_input, chunking_options_checkbox, chunk_method,
139
+ max_chunk_size, chunk_overlap, custom_prompt_checkbox, custom_prompt_input],
140
+ outputs=result_output
141
+ )
142
+
143
+ return search_query_input, search_type_input, search_button, items_output, item_mapping, api_name_input, api_key_input, chunking_options_checkbox, chunking_options_box, chunk_method, max_chunk_size, chunk_overlap, custom_prompt_checkbox, custom_prompt_input, resummarize_button, result_output
144
+
145
+
146
+ def update_resummarize_dropdown(search_query, search_type):
147
+ if search_type in ['Title', 'URL']:
148
+ results = fetch_items_by_title_or_url(search_query, search_type)
149
+ elif search_type == 'Keyword':
150
+ results = fetch_items_by_keyword(search_query)
151
+ else: # Content
152
+ results = fetch_items_by_content(search_query)
153
+
154
+ item_options = [f"{item[1]} ({item[2]})" for item in results]
155
+ item_mapping = {f"{item[1]} ({item[2]})": item[0] for item in results}
156
+ logging.debug(f"item_options: {item_options}")
157
+ logging.debug(f"item_mapping: {item_mapping}")
158
+ return gr.update(choices=item_options), item_mapping
159
+
160
+
161
+ def resummarize_content_wrapper(selected_item, item_mapping, api_name, api_key=None, chunking_options_checkbox=None, chunk_method=None,
162
+ max_chunk_size=None, chunk_overlap=None, custom_prompt_checkbox=None, custom_prompt=None):
163
+ logging.debug(f"resummarize_content_wrapper called with item_mapping type: {type(item_mapping)}")
164
+ logging.debug(f"selected_item: {selected_item}")
165
+
166
+ if not selected_item or not api_name:
167
+ return "Please select an item and provide API details."
168
+
169
+ # Handle potential string representation of item_mapping
170
+ if isinstance(item_mapping, str):
171
+ try:
172
+ item_mapping = json.loads(item_mapping)
173
+ except json.JSONDecodeError:
174
+ return f"Error: item_mapping is a string but not valid JSON. Value: {item_mapping[:100]}..."
175
+
176
+ if not isinstance(item_mapping, dict):
177
+ return f"Error: item_mapping is not a dictionary or valid JSON string. Type: {type(item_mapping)}"
178
+
179
+ media_id = item_mapping.get(selected_item)
180
+ if not media_id:
181
+ return f"Invalid selection. Selected item: {selected_item}, Available items: {list(item_mapping.keys())[:5]}..."
182
+
183
+ content, old_prompt, old_summary = fetch_item_details(media_id)
184
+
185
+ if not content:
186
+ return "No content available for re-summarization."
187
+
188
+ # Prepare chunking options
189
+ chunk_options = {
190
+ 'method': chunk_method,
191
+ 'max_size': int(max_chunk_size) if max_chunk_size is not None else None,
192
+ 'overlap': int(chunk_overlap) if chunk_overlap is not None else None,
193
+ 'language': 'english',
194
+ 'adaptive': True,
195
+ 'multi_level': False,
196
+ } if chunking_options_checkbox else None
197
+
198
+ # Prepare summarization prompt
199
+ summarization_prompt = custom_prompt if custom_prompt_checkbox and custom_prompt else None
200
+
201
+ logging.debug(f"Calling resummarize_content with media_id: {media_id}")
202
+ # Call the resummarize_content function
203
+ result = resummarize_content(selected_item, item_mapping, content, api_name, api_key, chunk_options, summarization_prompt)
204
+
205
+ return result
206
+
207
+
208
+ # FIXME - should be moved...
209
+ def resummarize_content(selected_item, item_mapping, content, api_name, api_key=None, chunk_options=None, summarization_prompt=None):
210
+ logging.debug(f"resummarize_content called with selected_item: {selected_item}")
211
+ # Load configuration
212
+ config = load_comprehensive_config()
213
+
214
+ # Chunking logic
215
+ if chunk_options:
216
+ chunks = improved_chunking_process(content, chunk_options)
217
+ else:
218
+ chunks = [{'text': content, 'metadata': {}}]
219
+
220
+ # Use default prompt if not provided
221
+ if not summarization_prompt:
222
+ summarization_prompt = config.get('Prompts', 'default_summary_prompt', fallback="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
223
+ **Bulleted Note Creation Guidelines**
224
+
225
+ **Headings**:
226
+ - Based on referenced topics, not categories like quotes or terms
227
+ - Surrounded by **bold** formatting
228
+ - Not listed as bullet points
229
+ - No space between headings and list items underneath
230
+
231
+ **Emphasis**:
232
+ - **Important terms** set in bold font
233
+ - **Text ending in a colon**: also bolded
234
+
235
+ **Review**:
236
+ - Ensure adherence to specified format
237
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]""")
238
+
239
+ # Summarization logic
240
+ summaries = []
241
+ for chunk in chunks:
242
+ chunk_text = chunk['text']
243
+ try:
244
+ chunk_summary = summarize_chunk(api_name, chunk_text, summarization_prompt, api_key)
245
+ if chunk_summary:
246
+ summaries.append(chunk_summary)
247
+ else:
248
+ logging.warning(f"Summarization failed for chunk: {chunk_text[:100]}...")
249
+ except Exception as e:
250
+ logging.error(f"Error during summarization: {str(e)}")
251
+ return f"Error during summarization: {str(e)}"
252
+
253
+ if not summaries:
254
+ return "Summarization failed for all chunks."
255
+
256
+ new_summary = " ".join(summaries)
257
+
258
+ # Update the database with the new summary
259
+
260
+ try:
261
+ update_result = update_media_content(selected_item, item_mapping, content, summarization_prompt, new_summary)
262
+ if "successfully" in update_result.lower():
263
+ return f"Re-summarization complete. New summary: {new_summary}..."
264
+ else:
265
+ return f"Error during database update: {update_result}"
266
+ except Exception as e:
267
+ logging.error(f"Error updating database: {str(e)}")
268
+ return f"Error updating database: {str(e)}"
App_Function_Libraries/Gradio_UI/Search_Tab.py ADDED
@@ -0,0 +1,487 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Search_Tab.py
2
+ # Description: This file contains the code for the search tab in the Gradio UI
3
+ #
4
+ # Imports
5
+ import html
6
+ import logging
7
+ import sqlite3
8
+
9
+ #
10
+ # External Imports
11
+ import gradio as gr
12
+
13
+ from App_Function_Libraries.DB_Manager import view_database, search_and_display_items
14
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import update_dropdown, update_detailed_view
15
+ from App_Function_Libraries.RAG_Libary_2 import rag_search
16
+
17
+ #
18
+ # Local Imports
19
+ #
20
+ #
21
+ ###################################################################################################
22
+ #
23
+ # Functions:
24
+
25
+ logger = logging.getLogger()
26
+
27
+
28
+
29
+
30
+ # FIXME - SQL functions to be moved to DB_Manager
31
+ def search_prompts(query):
32
+ try:
33
+ conn = sqlite3.connect('prompts.db')
34
+ cursor = conn.cursor()
35
+ cursor.execute("SELECT name, details, system, user FROM Prompts WHERE name LIKE ? OR details LIKE ?",
36
+ (f"%{query}%", f"%{query}%"))
37
+ results = cursor.fetchall()
38
+ conn.close()
39
+ return results
40
+ except sqlite3.Error as e:
41
+ print(f"Error searching prompts: {e}")
42
+ return []
43
+
44
+
45
+
46
+
47
+
48
+
49
+
50
+
51
+
52
+
53
+
54
+
55
+ def create_rag_tab():
56
+ with gr.TabItem("RAG Search"):
57
+ gr.Markdown("# Retrieval-Augmented Generation (RAG) Search")
58
+
59
+ with gr.Row():
60
+ with gr.Column():
61
+ search_query = gr.Textbox(label="Enter your question", placeholder="What would you like to know?")
62
+ api_choice = gr.Dropdown(
63
+ choices=["Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter", "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
64
+ label="Select API for RAG",
65
+ value="OpenAI"
66
+ )
67
+ search_button = gr.Button("Search")
68
+
69
+ with gr.Column():
70
+ result_output = gr.Textbox(label="Answer", lines=10)
71
+ context_output = gr.Textbox(label="Context", lines=10, visible=False)
72
+
73
+ def perform_rag_search(query, api_choice):
74
+ result = rag_search(query, api_choice)
75
+ return result['answer'], result['context']
76
+
77
+ search_button.click(perform_rag_search, inputs=[search_query, api_choice], outputs=[result_output, context_output])
78
+
79
+ # FIXME - under construction
80
+ def create_embeddings_tab():
81
+ with gr.TabItem("Create Embeddings"):
82
+ gr.Markdown("# Create Embeddings for All Content")
83
+
84
+ with gr.Row():
85
+ with gr.Column():
86
+ embedding_api_choice = gr.Dropdown(
87
+ choices=["OpenAI", "Local", "HuggingFace"],
88
+ label="Select API for Embeddings",
89
+ value="OpenAI"
90
+ )
91
+ create_button = gr.Button("Create Embeddings")
92
+
93
+ with gr.Column():
94
+ status_output = gr.Textbox(label="Status", lines=10)
95
+
96
+ def create_embeddings(api_choice):
97
+ try:
98
+ # Assuming you have a function that handles the creation of embeddings
99
+ from App_Function_Libraries.ChromaDB_Library import create_all_embeddings
100
+ status = create_all_embeddings(api_choice)
101
+ return status
102
+ except Exception as e:
103
+ return f"Error: {str(e)}"
104
+
105
+ create_button.click(create_embeddings, inputs=[embedding_api_choice], outputs=status_output)
106
+
107
+
108
+
109
+
110
+ def create_search_tab():
111
+ with gr.TabItem("Search / Detailed View"):
112
+ with gr.Row():
113
+ with gr.Column():
114
+ gr.Markdown("# Search across all ingested items in the Database")
115
+ gr.Markdown(" by Title / URL / Keyword / or Content via SQLite Full-Text-Search")
116
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
117
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title", label="Search By")
118
+ search_button = gr.Button("Search")
119
+ items_output = gr.Dropdown(label="Select Item", choices=[])
120
+ item_mapping = gr.State({})
121
+ prompt_summary_output = gr.HTML(label="Prompt & Summary", visible=True)
122
+
123
+ search_button.click(
124
+ fn=update_dropdown,
125
+ inputs=[search_query_input, search_type_input],
126
+ outputs=[items_output, item_mapping]
127
+ )
128
+ with gr.Column():
129
+ content_output = gr.Markdown(label="Content", visible=True)
130
+ items_output.change(
131
+ fn=update_detailed_view,
132
+ inputs=[items_output, item_mapping],
133
+ outputs=[prompt_summary_output, content_output]
134
+ )
135
+
136
+
137
+ def display_search_results(query):
138
+ if not query.strip():
139
+ return "Please enter a search query."
140
+
141
+ results = search_prompts(query)
142
+
143
+ # Debugging: Print the results to the console to see what is being returned
144
+ print(f"Processed search results for query '{query}': {results}")
145
+
146
+ if results:
147
+ result_md = "## Search Results:\n"
148
+ for result in results:
149
+ # Debugging: Print each result to see its format
150
+ print(f"Result item: {result}")
151
+
152
+ if len(result) == 2:
153
+ name, details = result
154
+ result_md += f"**Title:** {name}\n\n**Description:** {details}\n\n---\n"
155
+
156
+ elif len(result) == 4:
157
+ name, details, system, user = result
158
+ result_md += f"**Title:** {name}\n\n"
159
+ result_md += f"**Description:** {details}\n\n"
160
+ result_md += f"**System Prompt:** {system}\n\n"
161
+ result_md += f"**User Prompt:** {user}\n\n"
162
+ result_md += "---\n"
163
+ else:
164
+ result_md += "Error: Unexpected result format.\n\n---\n"
165
+ return result_md
166
+ return "No results found."
167
+
168
+
169
+ def create_viewing_tab():
170
+ with gr.TabItem("View Database"):
171
+ gr.Markdown("# View Database Entries")
172
+ with gr.Row():
173
+ with gr.Column():
174
+ entries_per_page = gr.Dropdown(choices=[10, 20, 50, 100], label="Entries per Page", value=10)
175
+ page_number = gr.Number(value=1, label="Page Number", precision=0)
176
+ view_button = gr.Button("View Page")
177
+ next_page_button = gr.Button("Next Page")
178
+ previous_page_button = gr.Button("Previous Page")
179
+ with gr.Column():
180
+ results_display = gr.HTML()
181
+ pagination_info = gr.Textbox(label="Pagination Info", interactive=False)
182
+
183
+ def update_page(page, entries_per_page):
184
+ results, pagination, total_pages = view_database(page, entries_per_page)
185
+ next_disabled = page >= total_pages
186
+ prev_disabled = page <= 1
187
+ return results, pagination, page, gr.update(interactive=not next_disabled), gr.update(interactive=not prev_disabled)
188
+
189
+ def go_to_next_page(current_page, entries_per_page):
190
+ next_page = current_page + 1
191
+ return update_page(next_page, entries_per_page)
192
+
193
+ def go_to_previous_page(current_page, entries_per_page):
194
+ previous_page = max(1, current_page - 1)
195
+ return update_page(previous_page, entries_per_page)
196
+
197
+ view_button.click(
198
+ fn=update_page,
199
+ inputs=[page_number, entries_per_page],
200
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
201
+ )
202
+
203
+ next_page_button.click(
204
+ fn=go_to_next_page,
205
+ inputs=[page_number, entries_per_page],
206
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
207
+ )
208
+
209
+ previous_page_button.click(
210
+ fn=go_to_previous_page,
211
+ inputs=[page_number, entries_per_page],
212
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
213
+ )
214
+
215
+
216
+ def create_search_summaries_tab():
217
+ with gr.TabItem("Search/View Title+Summary "):
218
+ gr.Markdown("# Search across all ingested items in the Database and review their summaries")
219
+ gr.Markdown("Search by Title / URL / Keyword / or Content via SQLite Full-Text-Search")
220
+ with gr.Row():
221
+ with gr.Column():
222
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
223
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title",
224
+ label="Search By")
225
+ entries_per_page = gr.Dropdown(choices=[10, 20, 50, 100], label="Entries per Page", value=10)
226
+ page_number = gr.Number(value=1, label="Page Number", precision=0)
227
+ char_count_input = gr.Number(value=5000, label="Amount of characters to display from the main content",
228
+ precision=0)
229
+ with gr.Column():
230
+ search_button = gr.Button("Search")
231
+ next_page_button = gr.Button("Next Page")
232
+ previous_page_button = gr.Button("Previous Page")
233
+ pagination_info = gr.Textbox(label="Pagination Info", interactive=False)
234
+ search_results_output = gr.HTML()
235
+
236
+
237
+ def update_search_page(query, search_type, page, entries_per_page, char_count):
238
+ # Ensure char_count is a positive integer
239
+ char_count = max(1, int(char_count)) if char_count else 5000
240
+ results, pagination, total_pages = search_and_display_items(query, search_type, page, entries_per_page, char_count)
241
+ next_disabled = page >= total_pages
242
+ prev_disabled = page <= 1
243
+ return results, pagination, page, gr.update(interactive=not next_disabled), gr.update(
244
+ interactive=not prev_disabled)
245
+
246
+ def go_to_next_search_page(query, search_type, current_page, entries_per_page, char_count):
247
+ next_page = current_page + 1
248
+ return update_search_page(query, search_type, next_page, entries_per_page, char_count)
249
+
250
+ def go_to_previous_search_page(query, search_type, current_page, entries_per_page, char_count):
251
+ previous_page = max(1, current_page - 1)
252
+ return update_search_page(query, search_type, previous_page, entries_per_page, char_count)
253
+
254
+ search_button.click(
255
+ fn=update_search_page,
256
+ inputs=[search_query_input, search_type_input, page_number, entries_per_page, char_count_input],
257
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
258
+ )
259
+
260
+ next_page_button.click(
261
+ fn=go_to_next_search_page,
262
+ inputs=[search_query_input, search_type_input, page_number, entries_per_page, char_count_input],
263
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
264
+ )
265
+
266
+ previous_page_button.click(
267
+ fn=go_to_previous_search_page,
268
+ inputs=[search_query_input, search_type_input, page_number, entries_per_page, char_count_input],
269
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
270
+ )
271
+
272
+
273
+
274
+ def create_prompt_view_tab():
275
+ with gr.TabItem("View Prompt Database"):
276
+ gr.Markdown("# View Prompt Database Entries")
277
+ with gr.Row():
278
+ with gr.Column():
279
+ entries_per_page = gr.Dropdown(choices=[10, 20, 50, 100], label="Entries per Page", value=10)
280
+ page_number = gr.Number(value=1, label="Page Number", precision=0)
281
+ view_button = gr.Button("View Page")
282
+ next_page_button = gr.Button("Next Page")
283
+ previous_page_button = gr.Button("Previous Page")
284
+ with gr.Column():
285
+ pagination_info = gr.Textbox(label="Pagination Info", interactive=False)
286
+ results_display = gr.HTML()
287
+
288
+ # FIXME - SQL functions to be moved to DB_Manager
289
+ def view_database(page, entries_per_page):
290
+ offset = (page - 1) * entries_per_page
291
+ try:
292
+ with sqlite3.connect('prompts.db') as conn:
293
+ cursor = conn.cursor()
294
+ cursor.execute('''
295
+ SELECT p.name, p.details, p.system, p.user, GROUP_CONCAT(k.keyword, ', ') as keywords
296
+ FROM Prompts p
297
+ LEFT JOIN PromptKeywords pk ON p.id = pk.prompt_id
298
+ LEFT JOIN Keywords k ON pk.keyword_id = k.id
299
+ GROUP BY p.id
300
+ ORDER BY p.name
301
+ LIMIT ? OFFSET ?
302
+ ''', (entries_per_page, offset))
303
+ prompts = cursor.fetchall()
304
+
305
+ cursor.execute('SELECT COUNT(*) FROM Prompts')
306
+ total_prompts = cursor.fetchone()[0]
307
+
308
+ results = ""
309
+ for prompt in prompts:
310
+ # Escape HTML special characters and replace newlines with <br> tags
311
+ title = html.escape(prompt[0]).replace('\n', '<br>')
312
+ details = html.escape(prompt[1] or '').replace('\n', '<br>')
313
+ system_prompt = html.escape(prompt[2] or '')
314
+ user_prompt = html.escape(prompt[3] or '')
315
+ keywords = html.escape(prompt[4] or '').replace('\n', '<br>')
316
+
317
+ results += f"""
318
+ <div style="border: 1px solid #ddd; padding: 10px; margin-bottom: 20px;">
319
+ <div style="display: grid; grid-template-columns: 1fr 1fr; gap: 10px;">
320
+ <div><strong>Title:</strong> {title}</div>
321
+ <div><strong>Details:</strong> {details}</div>
322
+ </div>
323
+ <div style="margin-top: 10px;">
324
+ <strong>User Prompt:</strong>
325
+ <pre style="white-space: pre-wrap; word-wrap: break-word;">{user_prompt}</pre>
326
+ </div>
327
+ <div style="margin-top: 10px;">
328
+ <strong>System Prompt:</strong>
329
+ <pre style="white-space: pre-wrap; word-wrap: break-word;">{system_prompt}</pre>
330
+ </div>
331
+ <div style="margin-top: 10px;">
332
+ <strong>Keywords:</strong> {keywords}
333
+ </div>
334
+ </div>
335
+ """
336
+
337
+ total_pages = (total_prompts + entries_per_page - 1) // entries_per_page
338
+ pagination = f"Page {page} of {total_pages} (Total prompts: {total_prompts})"
339
+
340
+ return results, pagination, total_pages
341
+ except sqlite3.Error as e:
342
+ return f"<p>Error fetching prompts: {e}</p>", "Error", 0
343
+
344
+ def update_page(page, entries_per_page):
345
+ results, pagination, total_pages = view_database(page, entries_per_page)
346
+ next_disabled = page >= total_pages
347
+ prev_disabled = page <= 1
348
+ return results, pagination, page, gr.update(interactive=not next_disabled), gr.update(
349
+ interactive=not prev_disabled)
350
+
351
+ def go_to_next_page(current_page, entries_per_page):
352
+ next_page = current_page + 1
353
+ return update_page(next_page, entries_per_page)
354
+
355
+ def go_to_previous_page(current_page, entries_per_page):
356
+ previous_page = max(1, current_page - 1)
357
+ return update_page(previous_page, entries_per_page)
358
+
359
+ view_button.click(
360
+ fn=update_page,
361
+ inputs=[page_number, entries_per_page],
362
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
363
+ )
364
+
365
+ next_page_button.click(
366
+ fn=go_to_next_page,
367
+ inputs=[page_number, entries_per_page],
368
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
369
+ )
370
+
371
+ previous_page_button.click(
372
+ fn=go_to_previous_page,
373
+ inputs=[page_number, entries_per_page],
374
+ outputs=[results_display, pagination_info, page_number, next_page_button, previous_page_button]
375
+ )
376
+
377
+
378
+
379
+ def create_prompt_search_tab():
380
+ with gr.TabItem("Search Prompts"):
381
+ gr.Markdown("# Search and View Prompt Details")
382
+ gr.Markdown("Currently has all of the https://github.com/danielmiessler/fabric prompts already available")
383
+ with gr.Row():
384
+ with gr.Column():
385
+ search_query_input = gr.Textbox(label="Search Prompts", placeholder="Enter your search query...")
386
+ entries_per_page = gr.Dropdown(choices=[10, 20, 50, 100], label="Entries per Page", value=10)
387
+ page_number = gr.Number(value=1, label="Page Number", precision=0)
388
+ with gr.Column():
389
+ search_button = gr.Button("Search Prompts")
390
+ next_page_button = gr.Button("Next Page")
391
+ previous_page_button = gr.Button("Previous Page")
392
+ pagination_info = gr.Textbox(label="Pagination Info", interactive=False)
393
+ search_results_output = gr.HTML()
394
+
395
+ def search_and_display_prompts(query, page, entries_per_page):
396
+ offset = (page - 1) * entries_per_page
397
+ try:
398
+ # FIXME - SQL functions to be moved to DB_Manager
399
+ with sqlite3.connect('prompts.db') as conn:
400
+ cursor = conn.cursor()
401
+ cursor.execute('''
402
+ SELECT p.name, p.details, p.system, p.user, GROUP_CONCAT(k.keyword, ', ') as keywords
403
+ FROM Prompts p
404
+ LEFT JOIN PromptKeywords pk ON p.id = pk.prompt_id
405
+ LEFT JOIN Keywords k ON pk.keyword_id = k.id
406
+ WHERE p.name LIKE ? OR p.details LIKE ? OR p.system LIKE ? OR p.user LIKE ? OR k.keyword LIKE ?
407
+ GROUP BY p.id
408
+ ORDER BY p.name
409
+ LIMIT ? OFFSET ?
410
+ ''', (f'%{query}%', f'%{query}%', f'%{query}%', f'%{query}%', f'%{query}%', entries_per_page, offset))
411
+ prompts = cursor.fetchall()
412
+
413
+ cursor.execute('''
414
+ SELECT COUNT(DISTINCT p.id)
415
+ FROM Prompts p
416
+ LEFT JOIN PromptKeywords pk ON p.id = pk.prompt_id
417
+ LEFT JOIN Keywords k ON pk.keyword_id = k.id
418
+ WHERE p.name LIKE ? OR p.details LIKE ? OR p.system LIKE ? OR p.user LIKE ? OR k.keyword LIKE ?
419
+ ''', (f'%{query}%', f'%{query}%', f'%{query}%', f'%{query}%', f'%{query}%'))
420
+ total_prompts = cursor.fetchone()[0]
421
+
422
+ results = ""
423
+ for prompt in prompts:
424
+ title = html.escape(prompt[0]).replace('\n', '<br>')
425
+ details = html.escape(prompt[1] or '').replace('\n', '<br>')
426
+ system_prompt = html.escape(prompt[2] or '')
427
+ user_prompt = html.escape(prompt[3] or '')
428
+ keywords = html.escape(prompt[4] or '').replace('\n', '<br>')
429
+
430
+ results += f"""
431
+ <div style="border: 1px solid #ddd; padding: 10px; margin-bottom: 20px;">
432
+ <div style="display: grid; grid-template-columns: 1fr 1fr; gap: 10px;">
433
+ <div><strong>Title:</strong> {title}</div>
434
+ <div><strong>Details:</strong> {details}</div>
435
+ </div>
436
+ <div style="margin-top: 10px;">
437
+ <strong>User Prompt:</strong>
438
+ <pre style="white-space: pre-wrap; word-wrap: break-word;">{user_prompt}</pre>
439
+ </div>
440
+ <div style="margin-top: 10px;">
441
+ <strong>System Prompt:</strong>
442
+ <pre style="white-space: pre-wrap; word-wrap: break-word;">{system_prompt}</pre>
443
+ </div>
444
+ <div style="margin-top: 10px;">
445
+ <strong>Keywords:</strong> {keywords}
446
+ </div>
447
+ </div>
448
+ """
449
+
450
+ total_pages = (total_prompts + entries_per_page - 1) // entries_per_page
451
+ pagination = f"Page {page} of {total_pages} (Total prompts: {total_prompts})"
452
+
453
+ return results, pagination, total_pages
454
+ except sqlite3.Error as e:
455
+ return f"<p>Error searching prompts: {e}</p>", "Error", 0
456
+
457
+ def update_search_page(query, page, entries_per_page):
458
+ results, pagination, total_pages = search_and_display_prompts(query, page, entries_per_page)
459
+ next_disabled = page >= total_pages
460
+ prev_disabled = page <= 1
461
+ return results, pagination, page, gr.update(interactive=not next_disabled), gr.update(interactive=not prev_disabled)
462
+
463
+ def go_to_next_search_page(query, current_page, entries_per_page):
464
+ next_page = current_page + 1
465
+ return update_search_page(query, next_page, entries_per_page)
466
+
467
+ def go_to_previous_search_page(query, current_page, entries_per_page):
468
+ previous_page = max(1, current_page - 1)
469
+ return update_search_page(query, previous_page, entries_per_page)
470
+
471
+ search_button.click(
472
+ fn=update_search_page,
473
+ inputs=[search_query_input, page_number, entries_per_page],
474
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
475
+ )
476
+
477
+ next_page_button.click(
478
+ fn=go_to_next_search_page,
479
+ inputs=[search_query_input, page_number, entries_per_page],
480
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
481
+ )
482
+
483
+ previous_page_button.click(
484
+ fn=go_to_previous_search_page,
485
+ inputs=[search_query_input, page_number, entries_per_page],
486
+ outputs=[search_results_output, pagination_info, page_number, next_page_button, previous_page_button]
487
+ )
App_Function_Libraries/Gradio_UI/Transcript_comparison.py ADDED
@@ -0,0 +1,94 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Transcript_comparison.py
2
+ # Description: Gradio UI tab for comparing transcripts
3
+ #
4
+ # Imports
5
+ import logging
6
+
7
+ #
8
+ # External Imports
9
+ import gradio as gr
10
+
11
+ from App_Function_Libraries.DB_Manager import get_transcripts
12
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import browse_items
13
+ from App_Function_Libraries.Utils import format_transcription
14
+
15
+
16
+ #
17
+ # Local Imports
18
+
19
+ def get_transcript_options(media_id):
20
+ transcripts = get_transcripts(media_id)
21
+ return [f"{t[0]}: {t[1]} ({t[3]})" for t in transcripts]
22
+
23
+
24
+ def update_transcript_options(media_id):
25
+ options = get_transcript_options(media_id)
26
+ return gr.update(choices=options), gr.update(choices=options)
27
+
28
+ def compare_transcripts(media_id, transcript1_id, transcript2_id):
29
+ try:
30
+ transcripts = get_transcripts(media_id)
31
+ transcript1 = next((t for t in transcripts if t[0] == int(transcript1_id)), None)
32
+ transcript2 = next((t for t in transcripts if t[0] == int(transcript2_id)), None)
33
+
34
+ if not transcript1 or not transcript2:
35
+ return "One or both selected transcripts not found."
36
+
37
+ comparison = f"Transcript 1 (Model: {transcript1[1]}, Created: {transcript1[3]}):\n\n"
38
+ comparison += format_transcription(transcript1[2])
39
+ comparison += f"\n\nTranscript 2 (Model: {transcript2[1]}, Created: {transcript2[3]}):\n\n"
40
+ comparison += format_transcription(transcript2[2])
41
+
42
+ return comparison
43
+ except Exception as e:
44
+ logging.error(f"Error in compare_transcripts: {str(e)}")
45
+ return f"Error comparing transcripts: {str(e)}"
46
+
47
+
48
+ def create_compare_transcripts_tab():
49
+ with gr.TabItem("Compare Transcripts"):
50
+ gr.Markdown("# Compare Transcripts")
51
+
52
+ with gr.Row():
53
+ search_query_input = gr.Textbox(label="Search Query", placeholder="Enter your search query here...")
54
+ search_type_input = gr.Radio(choices=["Title", "URL", "Keyword", "Content"], value="Title", label="Search By")
55
+ search_button = gr.Button("Search")
56
+
57
+ with gr.Row():
58
+ media_id_output = gr.Dropdown(label="Select Media Item", choices=[], interactive=True)
59
+ media_mapping = gr.State({})
60
+
61
+ media_id_input = gr.Number(label="Media ID", visible=False)
62
+ transcript1_dropdown = gr.Dropdown(label="Transcript 1")
63
+ transcript2_dropdown = gr.Dropdown(label="Transcript 2")
64
+ compare_button = gr.Button("Compare Transcripts")
65
+ comparison_output = gr.Textbox(label="Comparison Result", lines=20)
66
+
67
+ def update_media_dropdown(search_query, search_type):
68
+ results = browse_items(search_query, search_type)
69
+ item_options = [f"{item[1]} ({item[2]})" for item in results]
70
+ new_item_mapping = {f"{item[1]} ({item[2]})": item[0] for item in results}
71
+ return gr.update(choices=item_options), new_item_mapping
72
+
73
+ search_button.click(
74
+ fn=update_media_dropdown,
75
+ inputs=[search_query_input, search_type_input],
76
+ outputs=[media_id_output, media_mapping]
77
+ )
78
+
79
+ def load_selected_media_id(selected_media, media_mapping):
80
+ if selected_media and media_mapping and selected_media in media_mapping:
81
+ media_id = media_mapping[selected_media]
82
+ return media_id
83
+ return None
84
+
85
+ media_id_output.change(
86
+ fn=load_selected_media_id,
87
+ inputs=[media_id_output, media_mapping],
88
+ outputs=[media_id_input]
89
+ )
90
+
91
+ media_id_input.change(update_transcript_options, inputs=[media_id_input],
92
+ outputs=[transcript1_dropdown, transcript2_dropdown])
93
+ compare_button.click(compare_transcripts, inputs=[media_id_input, transcript1_dropdown, transcript2_dropdown],
94
+ outputs=[comparison_output])
App_Function_Libraries/Gradio_UI/Trash.py ADDED
@@ -0,0 +1,134 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Trash.py
2
+ # Gradio UI for deleting items from the database
3
+ import html
4
+ import sqlite3
5
+
6
+ # Imports
7
+
8
+ # External Imports
9
+ import gradio as gr
10
+ #
11
+ # Local Imports
12
+ from App_Function_Libraries.DB_Manager import delete_prompt, empty_trash, get_trashed_items, user_delete_item
13
+
14
+
15
+ def delete_item(media_id, force):
16
+ return user_delete_item(media_id, force)
17
+
18
+ def list_trash():
19
+ items = get_trashed_items()
20
+ return "\n".join(
21
+ [f"ID: {item['id']}, Title: {item['title']}, Trashed on: {item['trash_date']}" for item in items])
22
+
23
+ def empty_trash_ui(days):
24
+ deleted, remaining = empty_trash(days)
25
+ return f"Deleted {deleted} items. {remaining} items remain in trash."
26
+
27
+ def create_view_trash_tab():
28
+ with gr.TabItem("View Trash"):
29
+ view_button = gr.Button("View Trash")
30
+ trash_list = gr.Textbox(label="Trashed Items")
31
+ view_button.click(list_trash, inputs=[], outputs=trash_list)
32
+
33
+
34
+
35
+
36
+ def search_prompts_for_deletion(query):
37
+ try:
38
+ with sqlite3.connect('prompts.db') as conn:
39
+ cursor = conn.cursor()
40
+ cursor.execute('''
41
+ SELECT id, name, details
42
+ FROM Prompts
43
+ WHERE name LIKE ? OR details LIKE ?
44
+ LIMIT 10
45
+ ''', (f'%{query}%', f'%{query}%'))
46
+ results = cursor.fetchall()
47
+
48
+ if not results:
49
+ return "No matching prompts found."
50
+
51
+ output = "<h3>Matching Prompts:</h3>"
52
+ for row in results:
53
+ output += f"<p><strong>ID:</strong> {row[0]} | <strong>Name:</strong> {html.escape(row[1])} | <strong>Details:</strong> {html.escape(row[2][:100])}...</p>"
54
+ return output
55
+ except sqlite3.Error as e:
56
+ return f"An error occurred while searching prompts: {e}"
57
+
58
+
59
+ def search_media_for_deletion(query):
60
+ try:
61
+ with sqlite3.connect('media.db') as conn:
62
+ cursor = conn.cursor()
63
+ cursor.execute('''
64
+ SELECT id, title, description
65
+ FROM media
66
+ WHERE title LIKE ? OR description LIKE ?
67
+ LIMIT 10
68
+ ''', (f'%{query}%', f'%{query}%'))
69
+ results = cursor.fetchall()
70
+
71
+ if not results:
72
+ return "No matching media found."
73
+
74
+ output = "<h3>Matching Media:</h3>"
75
+ for row in results:
76
+ output += f"<p><strong>ID:</strong> {row[0]} | <strong>Title:</strong> {html.escape(row[1])} | <strong>Description:</strong> {html.escape(row[2][:100])}...</p>"
77
+ return output
78
+ except sqlite3.Error as e:
79
+ return f"An error occurred while searching media: {e}"
80
+
81
+ def create_delete_trash_tab():
82
+ with gr.TabItem("Delete DB Item"):
83
+ gr.Markdown("# Search and Delete Items from Databases")
84
+
85
+ with gr.Row():
86
+ with gr.Column():
87
+ gr.Markdown("## Search and Delete Prompts")
88
+ prompt_search_input = gr.Textbox(label="Search Prompts")
89
+ prompt_search_button = gr.Button("Search Prompts")
90
+ prompt_search_results = gr.HTML()
91
+ prompt_id_input = gr.Number(label="Prompt ID")
92
+ prompt_delete_button = gr.Button("Delete Prompt")
93
+ prompt_delete_output = gr.Textbox(label="Delete Result")
94
+
95
+ with gr.Column():
96
+ gr.Markdown("## Search and Delete Media")
97
+ media_search_input = gr.Textbox(label="Search Media")
98
+ media_search_button = gr.Button("Search Media")
99
+ media_search_results = gr.HTML()
100
+ media_id_input = gr.Number(label="Media ID")
101
+ media_force_checkbox = gr.Checkbox(label="Force Delete")
102
+ media_delete_button = gr.Button("Delete Media")
103
+ media_delete_output = gr.Textbox(label="Delete Result")
104
+
105
+ prompt_search_button.click(
106
+ search_prompts_for_deletion,
107
+ inputs=[prompt_search_input],
108
+ outputs=prompt_search_results
109
+ )
110
+
111
+ prompt_delete_button.click(
112
+ delete_prompt,
113
+ inputs=[prompt_id_input],
114
+ outputs=prompt_delete_output
115
+ )
116
+
117
+ media_search_button.click(
118
+ search_media_for_deletion,
119
+ inputs=[media_search_input],
120
+ outputs=media_search_results
121
+ )
122
+
123
+ media_delete_button.click(
124
+ delete_item,
125
+ inputs=[media_id_input, media_force_checkbox],
126
+ outputs=media_delete_output
127
+ )
128
+
129
+ def create_empty_trash_tab():
130
+ with gr.TabItem("Empty Trash"):
131
+ days_input = gr.Slider(minimum=15, maximum=90, step=5, label="Delete items older than (days)")
132
+ empty_button = gr.Button("Empty Trash")
133
+ empty_output = gr.Textbox(label="Result")
134
+ empty_button.click(empty_trash_ui, inputs=[days_input], outputs=empty_output)
App_Function_Libraries/Gradio_UI/Utilities.py ADDED
@@ -0,0 +1,118 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import os
2
+ import shutil
3
+ import tempfile
4
+ from pathlib import Path
5
+
6
+ import gradio as gr
7
+ import yt_dlp
8
+
9
+ from App_Function_Libraries.Utils import sanitize_filename, downloaded_files
10
+
11
+
12
+ def create_utilities_yt_video_tab():
13
+ with gr.Tab("YouTube Video Downloader"):
14
+ with gr.Row():
15
+ with gr.Column():
16
+ gr.Markdown(
17
+ "<h3>Youtube Video Downloader</h3><p>This Input takes a Youtube URL as input and creates a webm file for you to download. </br><em>If you want a full-featured one:</em> <strong><em>https://github.com/StefanLobbenmeier/youtube-dl-gui</strong></em> or <strong><em>https://github.com/yt-dlg/yt-dlg</em></strong></p>")
18
+ youtube_url_input = gr.Textbox(label="YouTube URL", placeholder="Enter YouTube video URL here")
19
+ download_button = gr.Button("Download Video")
20
+ with gr.Column():
21
+ output_file = gr.File(label="Download Video")
22
+ output_message = gr.Textbox(label="Status")
23
+
24
+ download_button.click(
25
+ fn=gradio_download_youtube_video,
26
+ inputs=youtube_url_input,
27
+ outputs=[output_file, output_message]
28
+ )
29
+
30
+ def create_utilities_yt_audio_tab():
31
+ with gr.Tab("YouTube Audio Downloader"):
32
+ with gr.Row():
33
+ with gr.Column():
34
+ gr.Markdown(
35
+ "<h3>Youtube Audio Downloader</h3><p>This Input takes a Youtube URL as input and creates an audio file for you to download.</p>"
36
+ +"\n<em>If you want a full-featured one:</em> <strong><em>https://github.com/StefanLobbenmeier/youtube-dl-gui</strong></em>\n or \n<strong><em>https://github.com/yt-dlg/yt-dlg</em></strong></p>")
37
+ youtube_url_input_audio = gr.Textbox(label="YouTube URL", placeholder="Enter YouTube video URL here")
38
+ download_button_audio = gr.Button("Download Audio")
39
+ with gr.Column():
40
+ output_file_audio = gr.File(label="Download Audio")
41
+ output_message_audio = gr.Textbox(label="Status")
42
+
43
+ from App_Function_Libraries.Audio_Files import download_youtube_audio
44
+ download_button_audio.click(
45
+ fn=download_youtube_audio,
46
+ inputs=youtube_url_input_audio,
47
+ outputs=[output_file_audio, output_message_audio]
48
+ )
49
+
50
+ def create_utilities_yt_timestamp_tab():
51
+ with gr.Tab("YouTube Timestamp URL Generator"):
52
+ gr.Markdown("## Generate YouTube URL with Timestamp")
53
+ with gr.Row():
54
+ with gr.Column():
55
+ url_input = gr.Textbox(label="YouTube URL")
56
+ hours_input = gr.Number(label="Hours", value=0, minimum=0, precision=0)
57
+ minutes_input = gr.Number(label="Minutes", value=0, minimum=0, maximum=59, precision=0)
58
+ seconds_input = gr.Number(label="Seconds", value=0, minimum=0, maximum=59, precision=0)
59
+ generate_button = gr.Button("Generate URL")
60
+ with gr.Column():
61
+ output_url = gr.Textbox(label="Timestamped URL")
62
+
63
+ from App_Function_Libraries.Video_DL_Ingestion_Lib import generate_timestamped_url
64
+ generate_button.click(
65
+ fn=generate_timestamped_url,
66
+ inputs=[url_input, hours_input, minutes_input, seconds_input],
67
+ outputs=output_url
68
+ )
69
+
70
+
71
+ def gradio_download_youtube_video(url):
72
+ try:
73
+ # Determine ffmpeg path based on the operating system.
74
+ ffmpeg_path = './Bin/ffmpeg.exe' if os.name == 'nt' else 'ffmpeg'
75
+
76
+ # Create a temporary directory
77
+ with tempfile.TemporaryDirectory() as temp_dir:
78
+ # Extract information about the video
79
+ with yt_dlp.YoutubeDL({'quiet': True}) as ydl:
80
+ info_dict = ydl.extract_info(url, download=False)
81
+ sanitized_title = sanitize_filename(info_dict['title'])
82
+ original_ext = info_dict['ext']
83
+
84
+ # Setup the temporary filename
85
+ temp_file_path = Path(temp_dir) / f"{sanitized_title}.{original_ext}"
86
+
87
+ # Initialize yt-dlp with generic options and the output template
88
+ ydl_opts = {
89
+ 'format': 'bestvideo+bestaudio/best',
90
+ 'ffmpeg_location': ffmpeg_path,
91
+ 'outtmpl': str(temp_file_path),
92
+ 'noplaylist': True,
93
+ 'quiet': True
94
+ }
95
+
96
+ # Execute yt-dlp to download the video
97
+ with yt_dlp.YoutubeDL(ydl_opts) as ydl:
98
+ ydl.download([url])
99
+
100
+ # Final check to ensure file exists
101
+ if not temp_file_path.exists():
102
+ raise FileNotFoundError(f"Expected file was not found: {temp_file_path}")
103
+
104
+ # Create a persistent directory for the download if it doesn't exist
105
+ persistent_dir = Path("downloads")
106
+ persistent_dir.mkdir(exist_ok=True)
107
+
108
+ # Move the file from the temporary directory to the persistent directory
109
+ persistent_file_path = persistent_dir / f"{sanitized_title}.{original_ext}"
110
+ shutil.move(str(temp_file_path), str(persistent_file_path))
111
+
112
+ # Add the file to the list of downloaded files
113
+ downloaded_files.append(str(persistent_file_path))
114
+
115
+ return str(persistent_file_path), f"Video downloaded successfully: {sanitized_title}.{original_ext}"
116
+ except Exception as e:
117
+ return None, f"Error downloading video: {str(e)}"
118
+
App_Function_Libraries/Gradio_UI/Video_transcription_tab.py ADDED
@@ -0,0 +1,691 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Video_transcription_tab.py
2
+ # Description: This file contains the code for the video transcription tab in the Gradio UI.
3
+ #
4
+ # Imports
5
+ import json
6
+ import logging
7
+ import os
8
+ #
9
+ # External Imports
10
+ import gradio as gr
11
+ import yt_dlp
12
+ #
13
+ # Local Imports
14
+ from App_Function_Libraries.DB_Manager import load_preset_prompts, add_media_to_database
15
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import whisper_models, update_user_prompt
16
+ from App_Function_Libraries.Gradio_UI.Gradio_Shared import error_handler
17
+ from App_Function_Libraries.Summarization_General_Lib import perform_transcription, perform_summarization, \
18
+ save_transcription_and_summary
19
+ from App_Function_Libraries.Utils import convert_to_seconds, safe_read_file, format_transcription, \
20
+ create_download_directory, generate_unique_identifier, extract_text_from_segments
21
+ from App_Function_Libraries.Video_DL_Ingestion_Lib import parse_and_expand_urls, extract_metadata, download_video
22
+ #
23
+ ################################################################################################################################################################
24
+ #
25
+ # Functions:
26
+
27
+ def create_video_transcription_tab():
28
+ with (gr.TabItem("Video Transcription + Summarization")):
29
+ gr.Markdown("# Transcribe & Summarize Videos from URLs")
30
+ with gr.Row():
31
+ gr.Markdown("""Follow this project at [tldw - GitHub](https://github.com/rmusser01/tldw)""")
32
+ with gr.Row():
33
+ gr.Markdown(
34
+ """If you're wondering what all this is, please see the 'Introduction/Help' tab up above for more detailed information and how to obtain an API Key.""")
35
+ with gr.Row():
36
+ with gr.Column():
37
+ url_input = gr.Textbox(label="URL(s) (Mandatory)",
38
+ placeholder="Enter video URLs here, one per line. Supports YouTube, Vimeo, other video sites and Youtube playlists.",
39
+ lines=5)
40
+ video_file_input = gr.File(label="Upload Video File (Optional)", file_types=["video/*"])
41
+ diarize_input = gr.Checkbox(label="Enable Speaker Diarization", value=False)
42
+ whisper_model_input = gr.Dropdown(choices=whisper_models, value="medium", label="Whisper Model")
43
+
44
+ with gr.Row():
45
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
46
+ value=False,
47
+ visible=True)
48
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
49
+ value=False,
50
+ visible=True)
51
+ with gr.Row():
52
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
53
+ choices=load_preset_prompts(),
54
+ visible=False)
55
+ with gr.Row():
56
+ custom_prompt_input = gr.Textbox(label="Custom Prompt",
57
+ placeholder="Enter custom prompt here",
58
+ lines=3,
59
+ visible=False)
60
+ with gr.Row():
61
+ system_prompt_input = gr.Textbox(label="System Prompt",
62
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
63
+ **Bulleted Note Creation Guidelines**
64
+
65
+ **Headings**:
66
+ - Based on referenced topics, not categories like quotes or terms
67
+ - Surrounded by **bold** formatting
68
+ - Not listed as bullet points
69
+ - No space between headings and list items underneath
70
+
71
+ **Emphasis**:
72
+ - **Important terms** set in bold font
73
+ - **Text ending in a colon**: also bolded
74
+
75
+ **Review**:
76
+ - Ensure adherence to specified format
77
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
78
+ """,
79
+ lines=3,
80
+ visible=False,
81
+ interactive=True)
82
+ custom_prompt_checkbox.change(
83
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
84
+ inputs=[custom_prompt_checkbox],
85
+ outputs=[custom_prompt_input, system_prompt_input]
86
+ )
87
+ preset_prompt_checkbox.change(
88
+ fn=lambda x: gr.update(visible=x),
89
+ inputs=[preset_prompt_checkbox],
90
+ outputs=[preset_prompt]
91
+ )
92
+
93
+ def update_prompts(preset_name):
94
+ prompts = update_user_prompt(preset_name)
95
+ return (
96
+ gr.update(value=prompts["user_prompt"], visible=True),
97
+ gr.update(value=prompts["system_prompt"], visible=True)
98
+ )
99
+
100
+ preset_prompt.change(
101
+ update_prompts,
102
+ inputs=preset_prompt,
103
+ outputs=[custom_prompt_input, system_prompt_input]
104
+ )
105
+
106
+ api_name_input = gr.Dropdown(
107
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral",
108
+ "OpenRouter",
109
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
110
+ value=None, label="API Name (Mandatory)")
111
+ api_key_input = gr.Textbox(label="API Key (Mandatory)", placeholder="Enter your API key here",
112
+ type="password")
113
+ keywords_input = gr.Textbox(label="Keywords", placeholder="Enter keywords here (comma-separated)",
114
+ value="default,no_keyword_set")
115
+ batch_size_input = gr.Slider(minimum=1, maximum=10, value=1, step=1,
116
+ label="Batch Size (Number of videos to process simultaneously)")
117
+ timestamp_option = gr.Radio(choices=["Include Timestamps", "Exclude Timestamps"],
118
+ value="Include Timestamps", label="Timestamp Option")
119
+ keep_original_video = gr.Checkbox(label="Keep Original Video", value=False)
120
+ # First, create a checkbox to toggle the chunking options
121
+ chunking_options_checkbox = gr.Checkbox(label="Show Chunking Options", value=False)
122
+ summarize_recursively = gr.Checkbox(label="Enable Recursive Summarization", value=False)
123
+ use_cookies_input = gr.Checkbox(label="Use cookies for authenticated download", value=False)
124
+ use_time_input = gr.Checkbox(label="Use Start and End Time", value=False)
125
+
126
+ with gr.Row(visible=False) as time_input_box:
127
+ gr.Markdown("### Start and End time")
128
+ with gr.Column():
129
+ start_time_input = gr.Textbox(label="Start Time (Optional)",
130
+ placeholder="e.g., 1:30 or 90 (in seconds)")
131
+ end_time_input = gr.Textbox(label="End Time (Optional)",
132
+ placeholder="e.g., 5:45 or 345 (in seconds)")
133
+
134
+ use_time_input.change(
135
+ fn=lambda x: gr.update(visible=x),
136
+ inputs=[use_time_input],
137
+ outputs=[time_input_box]
138
+ )
139
+
140
+ cookies_input = gr.Textbox(
141
+ label="User Session Cookies",
142
+ placeholder="Paste your cookies here (JSON format)",
143
+ lines=3,
144
+ visible=False
145
+ )
146
+
147
+ use_cookies_input.change(
148
+ fn=lambda x: gr.update(visible=x),
149
+ inputs=[use_cookies_input],
150
+ outputs=[cookies_input]
151
+ )
152
+ # Then, create a Box to group the chunking options
153
+ with gr.Row(visible=False) as chunking_options_box:
154
+ gr.Markdown("### Chunking Options")
155
+ with gr.Column():
156
+ chunk_method = gr.Dropdown(choices=['words', 'sentences', 'paragraphs', 'tokens'],
157
+ label="Chunking Method")
158
+ max_chunk_size = gr.Slider(minimum=100, maximum=1000, value=300, step=50,
159
+ label="Max Chunk Size")
160
+ chunk_overlap = gr.Slider(minimum=0, maximum=100, value=0, step=10, label="Chunk Overlap")
161
+ use_adaptive_chunking = gr.Checkbox(
162
+ label="Use Adaptive Chunking (Adjust chunking based on text complexity)")
163
+ use_multi_level_chunking = gr.Checkbox(label="Use Multi-level Chunking")
164
+ chunk_language = gr.Dropdown(choices=['english', 'french', 'german', 'spanish'],
165
+ label="Chunking Language")
166
+
167
+ # Add JavaScript to toggle the visibility of the chunking options box
168
+ chunking_options_checkbox.change(
169
+ fn=lambda x: gr.update(visible=x),
170
+ inputs=[chunking_options_checkbox],
171
+ outputs=[chunking_options_box]
172
+ )
173
+ process_button = gr.Button("Process Videos")
174
+
175
+ with gr.Column():
176
+ progress_output = gr.Textbox(label="Progress")
177
+ error_output = gr.Textbox(label="Errors", visible=False)
178
+ results_output = gr.HTML(label="Results")
179
+ download_transcription = gr.File(label="Download All Transcriptions as JSON")
180
+ download_summary = gr.File(label="Download All Summaries as Text")
181
+
182
+ @error_handler
183
+ def process_videos_with_error_handling(inputs, start_time, end_time, diarize, whisper_model,
184
+ custom_prompt_checkbox, custom_prompt, chunking_options_checkbox,
185
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
186
+ use_multi_level_chunking, chunk_language, api_name,
187
+ api_key, keywords, use_cookies, cookies, batch_size,
188
+ timestamp_option, keep_original_video, summarize_recursively,
189
+ progress: gr.Progress = gr.Progress()) -> tuple:
190
+ try:
191
+ logging.info("Entering process_videos_with_error_handling")
192
+ logging.info(f"Received inputs: {inputs}")
193
+
194
+ if not inputs:
195
+ raise ValueError("No inputs provided")
196
+
197
+ logging.debug("Input(s) is(are) valid")
198
+
199
+ # Ensure batch_size is an integer
200
+ try:
201
+ batch_size = int(batch_size)
202
+ except (ValueError, TypeError):
203
+ batch_size = 1 # Default to processing one video at a time if invalid
204
+
205
+ # Separate URLs and local files
206
+ urls = [input for input in inputs if
207
+ isinstance(input, str) and input.startswith(('http://', 'https://'))]
208
+ local_files = [input for input in inputs if
209
+ isinstance(input, str) and not input.startswith(('http://', 'https://'))]
210
+
211
+ # Parse and expand URLs if there are any
212
+ expanded_urls = parse_and_expand_urls(urls) if urls else []
213
+
214
+ valid_local_files = []
215
+ invalid_local_files = []
216
+
217
+ for file_path in local_files:
218
+ if os.path.exists(file_path):
219
+ valid_local_files.append(file_path)
220
+ else:
221
+ invalid_local_files.append(file_path)
222
+ error_message = f"Local file not found: {file_path}"
223
+ logging.error(error_message)
224
+
225
+ if invalid_local_files:
226
+ logging.warning(f"Found {len(invalid_local_files)} invalid local file paths")
227
+ # FIXME - Add more complete error handling for invalid local files
228
+
229
+ all_inputs = expanded_urls + valid_local_files
230
+ logging.info(f"Total valid inputs to process: {len(all_inputs)} "
231
+ f"({len(expanded_urls)} URLs, {len(valid_local_files)} local files)")
232
+
233
+ all_inputs = expanded_urls + local_files
234
+ logging.info(f"Total inputs to process: {len(all_inputs)}")
235
+ results = []
236
+ errors = []
237
+ results_html = ""
238
+ all_transcriptions = {}
239
+ all_summaries = ""
240
+
241
+ for i in range(0, len(all_inputs), batch_size):
242
+ batch = all_inputs[i:i + batch_size]
243
+ batch_results = []
244
+
245
+ for input_item in batch:
246
+ try:
247
+ start_seconds = convert_to_seconds(start_time)
248
+ end_seconds = convert_to_seconds(end_time) if end_time else None
249
+
250
+ logging.info(f"Attempting to extract metadata for {input_item}")
251
+
252
+ if input_item.startswith(('http://', 'https://')):
253
+ logging.info(f"Attempting to extract metadata for URL: {input_item}")
254
+ video_metadata = extract_metadata(input_item, use_cookies, cookies)
255
+ if not video_metadata:
256
+ raise ValueError(f"Failed to extract metadata for {input_item}")
257
+ else:
258
+ logging.info(f"Processing local file: {input_item}")
259
+ video_metadata = {"title": os.path.basename(input_item), "url": input_item}
260
+
261
+ chunk_options = {
262
+ 'method': chunk_method,
263
+ 'max_size': max_chunk_size,
264
+ 'overlap': chunk_overlap,
265
+ 'adaptive': use_adaptive_chunking,
266
+ 'multi_level': use_multi_level_chunking,
267
+ 'language': chunk_language
268
+ } if chunking_options_checkbox else None
269
+
270
+ if custom_prompt_checkbox:
271
+ custom_prompt = custom_prompt
272
+ else:
273
+ custom_prompt = ("""
274
+ <s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
275
+ **Bulleted Note Creation Guidelines**
276
+
277
+ **Headings**:
278
+ - Based on referenced topics, not categories like quotes or terms
279
+ - Surrounded by **bold** formatting
280
+ - Not listed as bullet points
281
+ - No space between headings and list items underneath
282
+
283
+ **Emphasis**:
284
+ - **Important terms** set in bold font
285
+ - **Text ending in a colon**: also bolded
286
+
287
+ **Review**:
288
+ - Ensure adherence to specified format
289
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
290
+ """)
291
+
292
+ logging.debug("Gradio_Related.py: process_url_with_metadata being called")
293
+ result = process_url_with_metadata(
294
+ input_item, 2, whisper_model,
295
+ custom_prompt,
296
+ start_seconds, api_name, api_key,
297
+ False, False, False, False, 0.01, None, keywords, None, diarize,
298
+ end_time=end_seconds,
299
+ include_timestamps=(timestamp_option == "Include Timestamps"),
300
+ metadata=video_metadata,
301
+ use_chunking=chunking_options_checkbox,
302
+ chunk_options=chunk_options,
303
+ keep_original_video=keep_original_video,
304
+ current_whisper_model=whisper_model,
305
+ )
306
+
307
+ if result[0] is None:
308
+ error_message = "Processing failed without specific error"
309
+ batch_results.append(
310
+ (input_item, error_message, "Error", video_metadata, None, None))
311
+ errors.append(f"Error processing {input_item}: {error_message}")
312
+ else:
313
+ url, transcription, summary, json_file, summary_file, result_metadata = result
314
+ if transcription is None:
315
+ error_message = f"Processing failed for {input_item}: Transcription is None"
316
+ batch_results.append(
317
+ (input_item, error_message, "Error", result_metadata, None, None))
318
+ errors.append(error_message)
319
+ else:
320
+ batch_results.append(
321
+ (input_item, transcription, "Success", result_metadata, json_file,
322
+ summary_file))
323
+
324
+
325
+ except Exception as e:
326
+ error_message = f"Error processing {input_item}: {str(e)}"
327
+ logging.error(error_message, exc_info=True)
328
+ batch_results.append((input_item, error_message, "Error", {}, None, None))
329
+ errors.append(error_message)
330
+
331
+ results.extend(batch_results)
332
+ logging.debug(f"Processed {len(batch_results)} videos in batch")
333
+ if isinstance(progress, gr.Progress):
334
+ progress((i + len(batch)) / len(all_inputs),
335
+ f"Processed {i + len(batch)}/{len(all_inputs)} videos")
336
+
337
+ # Generate HTML for results
338
+ logging.debug(f"Generating HTML for {len(results)} results")
339
+ for url, transcription, status, metadata, json_file, summary_file in results:
340
+ if status == "Success":
341
+ title = metadata.get('title', 'Unknown Title')
342
+
343
+ # Check if transcription is a string (which it should be now)
344
+ if isinstance(transcription, str):
345
+ # Split the transcription into metadata and actual transcription
346
+ parts = transcription.split('\n\n', 1)
347
+ if len(parts) == 2:
348
+ metadata_text, transcription_text = parts
349
+ else:
350
+ metadata_text = "Metadata not found"
351
+ transcription_text = transcription
352
+ else:
353
+ metadata_text = "Metadata format error"
354
+ transcription_text = "Transcription format error"
355
+
356
+ summary = safe_read_file(summary_file) if summary_file else "No summary available"
357
+
358
+ # FIXME - Add to other functions that generate HTML
359
+ # Format the transcription
360
+ formatted_transcription = format_transcription(transcription_text)
361
+ # Format the summary
362
+ formatted_summary = format_transcription(summary)
363
+
364
+ results_html += f"""
365
+ <div class="result-box">
366
+ <gradio-accordion>
367
+ <gradio-accordion-item label="{title}">
368
+ <p><strong>URL:</strong> <a href="{url}" target="_blank">{url}</a></p>
369
+ <h4>Metadata:</h4>
370
+ <pre>{metadata_text}</pre>
371
+ <h4>Transcription:</h4>
372
+ <div class="transcription" style="white-space: pre-wrap; word-wrap: break-word;">
373
+ {formatted_transcription}
374
+ </div>
375
+ <h4>Summary:</h4>
376
+ <div class="summary">{formatted_summary}</div>
377
+ </gradio-accordion-item>
378
+ </gradio-accordion>
379
+ </div>
380
+ """
381
+ logging.debug(f"Transcription for {url}: {transcription[:200]}...")
382
+ all_transcriptions[url] = transcription
383
+ all_summaries += f"Title: {title}\nURL: {url}\n\n{metadata_text}\n\nTranscription:\n{transcription_text}\n\nSummary:\n{summary}\n\n---\n\n"
384
+ else:
385
+ results_html += f"""
386
+ <div class="result-box error">
387
+ <h3>Error processing {url}</h3>
388
+ <p>{transcription}</p>
389
+ </div>
390
+ """
391
+
392
+ # Save all transcriptions and summaries to files
393
+ logging.debug("Saving all transcriptions and summaries to files")
394
+ with open('all_transcriptions.json', 'w', encoding='utf-8') as f:
395
+ json.dump(all_transcriptions, f, indent=2, ensure_ascii=False)
396
+
397
+ with open('all_summaries.txt', 'w', encoding='utf-8') as f:
398
+ f.write(all_summaries)
399
+
400
+ error_summary = "\n".join(errors) if errors else "No errors occurred."
401
+
402
+ total_inputs = len(all_inputs)
403
+ return (
404
+ f"Processed {total_inputs} videos. {len(errors)} errors occurred.",
405
+ error_summary,
406
+ results_html,
407
+ 'all_transcriptions.json',
408
+ 'all_summaries.txt'
409
+ )
410
+ except Exception as e:
411
+ logging.error(f"Unexpected error in process_videos_with_error_handling: {str(e)}", exc_info=True)
412
+ return (
413
+ f"An unexpected error occurred: {str(e)}",
414
+ str(e),
415
+ "<div class='result-box error'><h3>Unexpected Error</h3><p>" + str(e) + "</p></div>",
416
+ None,
417
+ None
418
+ )
419
+
420
+ def process_videos_wrapper(url_input, video_file, start_time, end_time, diarize, whisper_model,
421
+ custom_prompt_checkbox, custom_prompt, chunking_options_checkbox,
422
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
423
+ use_multi_level_chunking, chunk_language, summarize_recursively, api_name,
424
+ api_key, keywords, use_cookies, cookies, batch_size,
425
+ timestamp_option, keep_original_video):
426
+ global result
427
+ try:
428
+ logging.info("process_videos_wrapper(): process_videos_wrapper called")
429
+
430
+ # Define file paths
431
+ transcriptions_file = os.path.join('all_transcriptions.json')
432
+ summaries_file = os.path.join('all_summaries.txt')
433
+
434
+ # Delete existing files if they exist
435
+ for file_path in [transcriptions_file, summaries_file]:
436
+ try:
437
+ if os.path.exists(file_path):
438
+ os.remove(file_path)
439
+ logging.info(f"Deleted existing file: {file_path}")
440
+ except Exception as e:
441
+ logging.warning(f"Failed to delete file {file_path}: {str(e)}")
442
+
443
+ # Handle both URL input and file upload
444
+ inputs = []
445
+ if url_input:
446
+ inputs.extend([url.strip() for url in url_input.split('\n') if url.strip()])
447
+ if video_file is not None:
448
+ # Assuming video_file is a file object with a 'name' attribute
449
+ inputs.append(video_file.name)
450
+
451
+ if not inputs:
452
+ raise ValueError("No input provided. Please enter URLs or upload a video file.")
453
+ try:
454
+ result = process_videos_with_error_handling(
455
+ inputs, start_time, end_time, diarize, whisper_model,
456
+ custom_prompt_checkbox, custom_prompt, chunking_options_checkbox,
457
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
458
+ use_multi_level_chunking, chunk_language, api_name,
459
+ api_key, keywords, use_cookies, cookies, batch_size,
460
+ timestamp_option, keep_original_video, summarize_recursively
461
+ )
462
+ except Exception as e:
463
+ logging.error(
464
+ f"process_videos_wrapper(): Error in process_videos_with_error_handling: {str(e)}",
465
+ exc_info=True)
466
+
467
+ logging.info("process_videos_wrapper(): process_videos_with_error_handling completed")
468
+
469
+ # Ensure that result is a tuple with 5 elements
470
+ if not isinstance(result, tuple) or len(result) != 5:
471
+ raise ValueError(
472
+ f"process_videos_wrapper(): Expected 5 outputs, but got {len(result) if isinstance(result, tuple) else 1}")
473
+
474
+ return result
475
+ except Exception as e:
476
+ logging.error(f"process_videos_wrapper(): Error in process_videos_wrapper: {str(e)}", exc_info=True)
477
+ # Return a tuple with 5 elements in case of any error
478
+ return (
479
+ # progress_output
480
+ f"process_videos_wrapper(): An error occurred: {str(e)}",
481
+ # error_output
482
+ str(e),
483
+ # results_output
484
+ f"<div class='error'>Error: {str(e)}</div>",
485
+ # download_transcription
486
+ None,
487
+ # download_summary
488
+ None
489
+ )
490
+
491
+ # FIXME - remove dead args for process_url_with_metadata
492
+ @error_handler
493
+ def process_url_with_metadata(input_item, num_speakers, whisper_model, custom_prompt, offset, api_name,
494
+ api_key,
495
+ vad_filter, download_video_flag, download_audio, rolling_summarization,
496
+ detail_level, question_box, keywords, local_file_path, diarize, end_time=None,
497
+ include_timestamps=True, metadata=None, use_chunking=False,
498
+ chunk_options=None, keep_original_video=False, current_whisper_model="Blank"):
499
+
500
+ try:
501
+ logging.info(f"Starting process_url_metadata for URL: {input_item}")
502
+ # Create download path
503
+ download_path = create_download_directory("Video_Downloads")
504
+ logging.info(f"Download path created at: {download_path}")
505
+
506
+ # Initialize info_dict
507
+ info_dict = {}
508
+
509
+ # Handle URL or local file
510
+ if os.path.isfile(input_item):
511
+ video_file_path = input_item
512
+ unique_id = generate_unique_identifier(input_item)
513
+ # Extract basic info from local file
514
+ info_dict = {
515
+ 'webpage_url': unique_id,
516
+ 'title': os.path.basename(input_item),
517
+ 'description': "Local file",
518
+ 'channel_url': None,
519
+ 'duration': None,
520
+ 'channel': None,
521
+ 'uploader': None,
522
+ 'upload_date': None
523
+ }
524
+ else:
525
+ # Extract video information
526
+ with yt_dlp.YoutubeDL({'quiet': True}) as ydl:
527
+ try:
528
+ full_info = ydl.extract_info(input_item, download=False)
529
+
530
+ # Create a safe subset of info to log
531
+ safe_info = {
532
+ 'title': full_info.get('title', 'No title'),
533
+ 'duration': full_info.get('duration', 'Unknown duration'),
534
+ 'upload_date': full_info.get('upload_date', 'Unknown upload date'),
535
+ 'uploader': full_info.get('uploader', 'Unknown uploader'),
536
+ 'view_count': full_info.get('view_count', 'Unknown view count')
537
+ }
538
+
539
+ logging.debug(f"Full info extracted for {input_item}: {safe_info}")
540
+ except Exception as e:
541
+ logging.error(f"Error extracting video info: {str(e)}")
542
+ return None, None, None, None, None, None
543
+
544
+ # Filter the required metadata
545
+ if full_info:
546
+ info_dict = {
547
+ 'webpage_url': full_info.get('webpage_url', input_item),
548
+ 'title': full_info.get('title'),
549
+ 'description': full_info.get('description'),
550
+ 'channel_url': full_info.get('channel_url'),
551
+ 'duration': full_info.get('duration'),
552
+ 'channel': full_info.get('channel'),
553
+ 'uploader': full_info.get('uploader'),
554
+ 'upload_date': full_info.get('upload_date')
555
+ }
556
+ logging.debug(f"Filtered info_dict: {info_dict}")
557
+ else:
558
+ logging.error("Failed to extract video information")
559
+ return None, None, None, None, None, None
560
+
561
+ # Download video/audio
562
+ logging.info("Downloading video/audio...")
563
+ video_file_path = download_video(input_item, download_path, full_info, download_video_flag,
564
+ current_whisper_model="Blank")
565
+ if not video_file_path:
566
+ logging.error(f"Failed to download video/audio from {input_item}")
567
+ return None, None, None, None, None, None
568
+
569
+ logging.info(f"Processing file: {video_file_path}")
570
+
571
+ # Perform transcription
572
+ logging.info("Starting transcription...")
573
+ audio_file_path, segments = perform_transcription(video_file_path, offset, whisper_model,
574
+ vad_filter, diarize)
575
+
576
+ if audio_file_path is None or segments is None:
577
+ logging.error("Transcription failed or segments not available.")
578
+ return None, None, None, None, None, None
579
+
580
+ logging.info(f"Transcription completed. Number of segments: {len(segments)}")
581
+
582
+ # Add metadata to segments
583
+ segments_with_metadata = {
584
+ "metadata": info_dict,
585
+ "segments": segments
586
+ }
587
+
588
+ # Save segments with metadata to JSON file
589
+ segments_json_path = os.path.splitext(audio_file_path)[0] + ".segments.json"
590
+ with open(segments_json_path, 'w') as f:
591
+ json.dump(segments_with_metadata, f, indent=2)
592
+
593
+ # FIXME - why isnt this working?
594
+ # Delete the .wav file after successful transcription
595
+ files_to_delete = [audio_file_path]
596
+ for file_path in files_to_delete:
597
+ if file_path and os.path.exists(file_path):
598
+ try:
599
+ os.remove(file_path)
600
+ logging.info(f"Successfully deleted file: {file_path}")
601
+ except Exception as e:
602
+ logging.warning(f"Failed to delete file {file_path}: {str(e)}")
603
+
604
+ # Delete the mp4 file after successful transcription if not keeping original audio
605
+ # Modify the file deletion logic to respect keep_original_video
606
+ if not keep_original_video:
607
+ files_to_delete = [audio_file_path, video_file_path]
608
+ for file_path in files_to_delete:
609
+ if file_path and os.path.exists(file_path):
610
+ try:
611
+ os.remove(file_path)
612
+ logging.info(f"Successfully deleted file: {file_path}")
613
+ except Exception as e:
614
+ logging.warning(f"Failed to delete file {file_path}: {str(e)}")
615
+ else:
616
+ logging.info(f"Keeping original video file: {video_file_path}")
617
+ logging.info(f"Keeping original audio file: {audio_file_path}")
618
+
619
+ # Process segments based on the timestamp option
620
+ if not include_timestamps:
621
+ segments = [{'Text': segment['Text']} for segment in segments]
622
+
623
+ logging.info(f"Segments processed for timestamp inclusion: {segments}")
624
+
625
+ # Extract text from segments
626
+ transcription_text = extract_text_from_segments(segments)
627
+
628
+ if transcription_text.startswith("Error:"):
629
+ logging.error(f"Failed to extract transcription: {transcription_text}")
630
+ return None, None, None, None, None, None
631
+
632
+ # Use transcription_text instead of segments for further processing
633
+ full_text_with_metadata = f"{json.dumps(info_dict, indent=2)}\n\n{transcription_text}"
634
+
635
+ logging.debug(f"Full text with metadata extracted: {full_text_with_metadata[:100]}...")
636
+
637
+ # Perform summarization if API is provided
638
+ summary_text = None
639
+ if api_name:
640
+ # API key resolution handled at base of function if none provided
641
+ api_key = api_key if api_key else None
642
+ logging.info(f"Starting summarization with {api_name}...")
643
+ summary_text = perform_summarization(api_name, full_text_with_metadata, custom_prompt, api_key)
644
+ if summary_text is None:
645
+ logging.error("Summarization failed.")
646
+ return None, None, None, None, None, None
647
+ logging.debug(f"Summarization completed: {summary_text[:100]}...")
648
+
649
+ # Save transcription and summary
650
+ logging.info("Saving transcription and summary...")
651
+ download_path = create_download_directory("Audio_Processing")
652
+ json_file_path, summary_file_path = save_transcription_and_summary(full_text_with_metadata,
653
+ summary_text,
654
+ download_path, info_dict)
655
+ logging.info(f"Transcription saved to: {json_file_path}")
656
+ logging.info(f"Summary saved to: {summary_file_path}")
657
+
658
+ # Prepare keywords for database
659
+ if isinstance(keywords, str):
660
+ keywords_list = [kw.strip() for kw in keywords.split(',') if kw.strip()]
661
+ elif isinstance(keywords, (list, tuple)):
662
+ keywords_list = keywords
663
+ else:
664
+ keywords_list = []
665
+ logging.info(f"Keywords prepared: {keywords_list}")
666
+
667
+ # Add to database
668
+ logging.info("Adding to database...")
669
+ add_media_to_database(info_dict['webpage_url'], info_dict, full_text_with_metadata, summary_text,
670
+ keywords_list, custom_prompt, whisper_model)
671
+ logging.info(f"Media added to database: {info_dict['webpage_url']}")
672
+
673
+ return info_dict[
674
+ 'webpage_url'], full_text_with_metadata, summary_text, json_file_path, summary_file_path, info_dict
675
+
676
+ except Exception as e:
677
+ logging.error(f"Error in process_url_with_metadata: {str(e)}", exc_info=True)
678
+ return None, None, None, None, None, None
679
+
680
+ process_button.click(
681
+ fn=process_videos_wrapper,
682
+ inputs=[
683
+ url_input, video_file_input, start_time_input, end_time_input, diarize_input, whisper_model_input,
684
+ custom_prompt_checkbox, custom_prompt_input, chunking_options_checkbox,
685
+ chunk_method, max_chunk_size, chunk_overlap, use_adaptive_chunking,
686
+ use_multi_level_chunking, chunk_language, summarize_recursively, api_name_input, api_key_input,
687
+ keywords_input, use_cookies_input, cookies_input, batch_size_input,
688
+ timestamp_option, keep_original_video
689
+ ],
690
+ outputs=[progress_output, error_output, results_output, download_transcription, download_summary]
691
+ )
App_Function_Libraries/Gradio_UI/Website_scraping_tab.py ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Website_scraping_tab.py
2
+ # Gradio UI for scraping websites
3
+
4
+ # Imports
5
+ #
6
+ # External Imports
7
+ import gradio as gr
8
+
9
+ from App_Function_Libraries.Article_Summarization_Lib import scrape_and_summarize_multiple
10
+ from App_Function_Libraries.DB_Manager import load_preset_prompts
11
+ from App_Function_Libraries.Gradio_UI.Chat_ui import update_user_prompt
12
+
13
+
14
+ #
15
+ # Local Imports
16
+ #
17
+ #
18
+ ########################################################################################################################
19
+ #
20
+ # Functions:
21
+
22
+
23
+ def create_website_scraping_tab():
24
+ with gr.TabItem("Website Scraping"):
25
+ gr.Markdown("# Scrape Websites & Summarize Articles using a Headless Chrome Browser!")
26
+ with gr.Row():
27
+ with gr.Column():
28
+ url_input = gr.Textbox(label="Article URLs", placeholder="Enter article URLs here, one per line", lines=5)
29
+ custom_article_title_input = gr.Textbox(label="Custom Article Titles (Optional, one per line)",
30
+ placeholder="Enter custom titles for the articles, one per line",
31
+ lines=5)
32
+ with gr.Row():
33
+ custom_prompt_checkbox = gr.Checkbox(label="Use a Custom Prompt",
34
+ value=False,
35
+ visible=True)
36
+ preset_prompt_checkbox = gr.Checkbox(label="Use a pre-set Prompt",
37
+ value=False,
38
+ visible=True)
39
+ with gr.Row():
40
+ preset_prompt = gr.Dropdown(label="Select Preset Prompt",
41
+ choices=load_preset_prompts(),
42
+ visible=False)
43
+ with gr.Row():
44
+ website_custom_prompt_input = gr.Textbox(label="Custom Prompt",
45
+ placeholder="Enter custom prompt here",
46
+ lines=3,
47
+ visible=False)
48
+ with gr.Row():
49
+ system_prompt_input = gr.Textbox(label="System Prompt",
50
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
51
+ **Bulleted Note Creation Guidelines**
52
+
53
+ **Headings**:
54
+ - Based on referenced topics, not categories like quotes or terms
55
+ - Surrounded by **bold** formatting
56
+ - Not listed as bullet points
57
+ - No space between headings and list items underneath
58
+
59
+ **Emphasis**:
60
+ - **Important terms** set in bold font
61
+ - **Text ending in a colon**: also bolded
62
+
63
+ **Review**:
64
+ - Ensure adherence to specified format
65
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
66
+ """,
67
+ lines=3,
68
+ visible=False)
69
+
70
+ custom_prompt_checkbox.change(
71
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
72
+ inputs=[custom_prompt_checkbox],
73
+ outputs=[website_custom_prompt_input, system_prompt_input]
74
+ )
75
+ preset_prompt_checkbox.change(
76
+ fn=lambda x: gr.update(visible=x),
77
+ inputs=[preset_prompt_checkbox],
78
+ outputs=[preset_prompt]
79
+ )
80
+
81
+ def update_prompts(preset_name):
82
+ prompts = update_user_prompt(preset_name)
83
+ return (
84
+ gr.update(value=prompts["user_prompt"], visible=True),
85
+ gr.update(value=prompts["system_prompt"], visible=True)
86
+ )
87
+
88
+ preset_prompt.change(
89
+ update_prompts,
90
+ inputs=preset_prompt,
91
+ outputs=[website_custom_prompt_input, system_prompt_input]
92
+ )
93
+
94
+ api_name_input = gr.Dropdown(
95
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
96
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"], value=None, label="API Name (Mandatory for Summarization)")
97
+ api_key_input = gr.Textbox(label="API Key (Mandatory if API Name is specified)",
98
+ placeholder="Enter your API key here; Ignore if using Local API or Built-in API", type="password")
99
+ keywords_input = gr.Textbox(label="Keywords", placeholder="Enter keywords here (comma-separated)",
100
+ value="default,no_keyword_set", visible=True)
101
+
102
+ scrape_button = gr.Button("Scrape and Summarize")
103
+ with gr.Column():
104
+ result_output = gr.Textbox(label="Result", lines=20)
105
+
106
+ scrape_button.click(
107
+ fn=scrape_and_summarize_multiple,
108
+ inputs=[url_input, website_custom_prompt_input, api_name_input, api_key_input, keywords_input,
109
+ custom_article_title_input, system_prompt_input],
110
+ outputs=result_output
111
+ )
112
+
113
+
App_Function_Libraries/Gradio_UI/Writing.py ADDED
@@ -0,0 +1,700 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ # Writing.py
2
+ # Description: This file contains the functions that are used for writing in the Gradio UI.
3
+ #
4
+ # Imports
5
+ import base64
6
+ from datetime import datetime as datetime
7
+ import logging
8
+ import json
9
+ import os
10
+ #
11
+ # External Imports
12
+ import gradio as gr
13
+ from PIL import Image
14
+ import textstat
15
+ #
16
+ # Local Imports
17
+ from App_Function_Libraries.Summarization_General_Lib import perform_summarization
18
+ from App_Function_Libraries.Chat import chat
19
+ #
20
+ ########################################################################################################################
21
+ #
22
+ # Functions:
23
+
24
+ def adjust_tone(text, concise, casual, api_name, api_key):
25
+ tones = [
26
+ {"tone": "concise", "weight": concise},
27
+ {"tone": "casual", "weight": casual},
28
+ {"tone": "professional", "weight": 1 - casual},
29
+ {"tone": "expanded", "weight": 1 - concise}
30
+ ]
31
+ tones = sorted(tones, key=lambda x: x['weight'], reverse=True)[:2]
32
+
33
+ tone_prompt = " and ".join([f"{t['tone']} (weight: {t['weight']:.2f})" for t in tones])
34
+
35
+ prompt = f"Rewrite the following text to match these tones: {tone_prompt}. Text: {text}"
36
+ # Performing tone adjustment request...
37
+ adjusted_text = perform_summarization(api_name, text, prompt, api_key)
38
+
39
+ return adjusted_text
40
+
41
+
42
+ def grammar_style_check(input_text, custom_prompt, api_name, api_key, system_prompt):
43
+ default_prompt = "Please analyze the following text for grammar and style. Offer suggestions for improvement and point out any misused words or incorrect spellings:\n\n"
44
+ full_prompt = custom_prompt if custom_prompt else default_prompt
45
+ full_text = full_prompt + input_text
46
+
47
+ return perform_summarization(api_name, full_text, custom_prompt, api_key, system_prompt)
48
+
49
+
50
+ def create_grammar_style_check_tab():
51
+ with gr.TabItem("Grammar and Style Check"):
52
+ with gr.Row():
53
+ with gr.Column():
54
+ gr.Markdown("# Grammar and Style Check")
55
+ gr.Markdown("This utility checks the grammar and style of the provided text by feeding it to an LLM and returning suggestions for improvement.")
56
+ input_text = gr.Textbox(label="Input Text", lines=10)
57
+ custom_prompt_checkbox = gr.Checkbox(label="Use Custom Prompt", value=False, visible=True)
58
+ system_prompt_input = gr.Textbox(label="System Prompt", placeholder="Please analyze the provided text for grammar and style. Offer any suggestions or points to improve you can identify. Additionally please point out any misuses of any words or incorrect spellings.", lines=5, visible=False)
59
+ custom_prompt_input = gr.Textbox(label="user Prompt",
60
+ value="""<s>You are a bulleted notes specialist. [INST]```When creating comprehensive bulleted notes, you should follow these guidelines: Use multiple headings based on the referenced topics, not categories like quotes or terms. Headings should be surrounded by bold formatting and not be listed as bullet points themselves. Leave no space between headings and their corresponding list items underneath. Important terms within the content should be emphasized by setting them in bold font. Any text that ends with a colon should also be bolded. Before submitting your response, review the instructions, and make any corrections necessary to adhered to the specified format. Do not reference these instructions within the notes.``` \nBased on the content between backticks create comprehensive bulleted notes.[/INST]
61
+ **Bulleted Note Creation Guidelines**
62
+
63
+ **Headings**:
64
+ - Based on referenced topics, not categories like quotes or terms
65
+ - Surrounded by **bold** formatting
66
+ - Not listed as bullet points
67
+ - No space between headings and list items underneath
68
+
69
+ **Emphasis**:
70
+ - **Important terms** set in bold font
71
+ - **Text ending in a colon**: also bolded
72
+
73
+ **Review**:
74
+ - Ensure adherence to specified format
75
+ - Do not reference these instructions in your response.</s>[INST] {{ .Prompt }} [/INST]
76
+ """,
77
+ lines=3,
78
+ visible=False)
79
+ custom_prompt_checkbox.change(
80
+ fn=lambda x: (gr.update(visible=x), gr.update(visible=x)),
81
+ inputs=[custom_prompt_checkbox],
82
+ outputs=[custom_prompt_input, system_prompt_input]
83
+ )
84
+ api_name_input = gr.Dropdown(
85
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
86
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
87
+ value=None,
88
+ label="API for Grammar Check"
89
+ )
90
+ api_key_input = gr.Textbox(label="API Key (if not set in config.txt)", placeholder="Enter your API key here",
91
+ type="password")
92
+ check_grammar_button = gr.Button("Check Grammar and Style")
93
+
94
+ with gr.Column():
95
+ gr.Markdown("# Resulting Suggestions")
96
+ gr.Markdown("(Keep in mind the API used can affect the quality of the suggestions)")
97
+
98
+ output_text = gr.Textbox(label="Grammar and Style Suggestions", lines=15)
99
+
100
+ check_grammar_button.click(
101
+ fn=grammar_style_check,
102
+ inputs=[input_text, custom_prompt_input, api_name_input, api_key_input, system_prompt_input],
103
+ outputs=output_text
104
+ )
105
+
106
+
107
+ def create_tone_adjustment_tab():
108
+ with gr.TabItem("Tone Analyzer & Editor"):
109
+ with gr.Row():
110
+ with gr.Column():
111
+ input_text = gr.Textbox(label="Input Text", lines=10)
112
+ concise_slider = gr.Slider(minimum=0, maximum=1, value=0.5, label="Concise vs Expanded")
113
+ casual_slider = gr.Slider(minimum=0, maximum=1, value=0.5, label="Casual vs Professional")
114
+ api_name_input = gr.Dropdown(
115
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
116
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM","ollama", "HuggingFace"],
117
+ value=None,
118
+ label="API for Grammar Check"
119
+ )
120
+ api_key_input = gr.Textbox(label="API Key (if not set in config.txt)", placeholder="Enter your API key here",
121
+ type="password")
122
+ adjust_btn = gr.Button("Adjust Tone")
123
+
124
+ with gr.Column():
125
+ output_text = gr.Textbox(label="Adjusted Text", lines=15)
126
+
127
+ adjust_btn.click(
128
+ adjust_tone,
129
+ inputs=[input_text, concise_slider, casual_slider],
130
+ outputs=output_text
131
+ )
132
+
133
+
134
+ persona_prompts = {
135
+ "Hemingway": "As Ernest Hemingway, known for concise and straightforward prose, provide feedback on the following text:",
136
+ "Shakespeare": "Channel William Shakespeare's poetic style and provide feedback on the following text:",
137
+ "Jane Austen": "Embodying Jane Austen's wit and social commentary, critique the following text:",
138
+ "Stephen King": "With Stephen King's flair for suspense and horror, analyze the following text:",
139
+ "J.K. Rowling": "As J.K. Rowling, creator of the magical world of Harry Potter, review the following text:"
140
+ }
141
+
142
+ def generate_writing_feedback(text, persona, aspect, api_name, api_key):
143
+ if isinstance(persona, dict): # If it's a character card
144
+ base_prompt = f"You are {persona['name']}. {persona['personality']}\n\nScenario: {persona['scenario']}\n\nRespond to the following message in character:"
145
+ else: # If it's a regular persona
146
+ base_prompt = persona_prompts.get(persona, f"As {persona}, provide feedback on the following text:")
147
+
148
+ if aspect != "Overall":
149
+ prompt = f"{base_prompt}\n\nFocus specifically on the {aspect.lower()} in the following text:\n\n{text}"
150
+ else:
151
+ prompt = f"{base_prompt}\n\n{text}"
152
+
153
+ return perform_summarization(api_name, text, prompt, api_key, system_message="You are a helpful AI assistant. You will respond to the user as if you were the persona declared in the user prompt.")
154
+
155
+ def generate_writing_prompt(persona, api_name, api_key):
156
+ prompt = f"Generate a writing prompt in the style of {persona}. The prompt should inspire a short story or scene that reflects {persona}'s typical themes and writing style."
157
+ #FIXME
158
+ return perform_summarization(api_name, prompt, "", api_key, system_message="You are a helpful AI assistant. You will respond to the user as if you were the persona declared in the user prompt." )
159
+
160
+ def calculate_readability(text):
161
+ ease = textstat.flesch_reading_ease(text)
162
+ grade = textstat.flesch_kincaid_grade(text)
163
+ return f"Readability: Flesch Reading Ease: {ease:.2f}, Flesch-Kincaid Grade Level: {grade:.2f}"
164
+
165
+
166
+ def generate_feedback_history_html(history):
167
+ html = "<h3>Recent Feedback History</h3>"
168
+ for entry in reversed(history):
169
+ html += f"<details><summary>{entry['persona']} Feedback</summary>"
170
+ html += f"<p><strong>Original Text:</strong> {entry['text'][:100]}...</p>"
171
+
172
+ feedback = entry.get('feedback')
173
+ if feedback:
174
+ html += f"<p><strong>Feedback:</strong> {feedback[:200]}...</p>"
175
+ else:
176
+ html += "<p><strong>Feedback:</strong> No feedback provided.</p>"
177
+
178
+ html += "</details>"
179
+ return html
180
+
181
+
182
+ # FIXME
183
+ def create_document_feedback_tab():
184
+ with gr.TabItem("Writing Feedback"):
185
+ with gr.Row():
186
+ with gr.Column(scale=2):
187
+ input_text = gr.Textbox(label="Your Writing", lines=10)
188
+ persona_dropdown = gr.Dropdown(
189
+ label="Select Persona",
190
+ choices=[
191
+ "Agatha Christie",
192
+ "Arthur Conan Doyle",
193
+ "Charles Bukowski",
194
+ "Charles Dickens",
195
+ "Chinua Achebe",
196
+ "Cormac McCarthy",
197
+ "David Foster Wallace",
198
+ "Edgar Allan Poe",
199
+ "F. Scott Fitzgerald",
200
+ "Flannery O'Connor",
201
+ "Franz Kafka",
202
+ "Fyodor Dostoevsky",
203
+ "Gabriel Garcia Marquez",
204
+ "George R.R. Martin",
205
+ "George Orwell",
206
+ "Haruki Murakami",
207
+ "Hemingway",
208
+ "Herman Melville",
209
+ "Isabel Allende",
210
+ "James Joyce",
211
+ "Jane Austen",
212
+ "J.K. Rowling",
213
+ "J.R.R. Tolkien",
214
+ "Jorge Luis Borges",
215
+ "Kurt Vonnegut",
216
+ "Leo Tolstoy",
217
+ "Margaret Atwood",
218
+ "Mark Twain",
219
+ "Mary Shelley",
220
+ "Milan Kundera",
221
+ "Naguib Mahfouz",
222
+ "Neil Gaiman",
223
+ "Octavia Butler",
224
+ "Philip K Dick",
225
+ "Ray Bradbury",
226
+ "Salman Rushdie",
227
+ "Shakespeare",
228
+ "Stephen King",
229
+ "Toni Morrison",
230
+ "T.S. Eliot",
231
+ "Ursula K. Le Guin",
232
+ "Virginia Woolf",
233
+ "Virginia Woolf",
234
+ "Zadie Smith"],
235
+ value="Hemingway"
236
+ )
237
+ custom_persona_name = gr.Textbox(label="Custom Persona Name")
238
+ custom_persona_description = gr.Textbox(label="Custom Persona Description", lines=3)
239
+ add_custom_persona_button = gr.Button("Add Custom Persona")
240
+ aspect_dropdown = gr.Dropdown(
241
+ label="Focus Feedback On",
242
+ choices=["Overall", "Grammar", "Word choice", "Structure of delivery", "Character Development", "Character Dialogue", "Descriptive Language", "Plot Structure"],
243
+ value="Overall"
244
+ )
245
+ api_name_input = gr.Dropdown(
246
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral", "OpenRouter",
247
+ "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
248
+ value=None,
249
+ label="API for Feedback"
250
+ )
251
+ api_key_input = gr.Textbox(label="API Key (if not set in config.txt)", type="password")
252
+ get_feedback_button = gr.Button("Get Feedback")
253
+ generate_prompt_button = gr.Button("Generate Writing Prompt")
254
+
255
+ with gr.Column(scale=2):
256
+ feedback_output = gr.Textbox(label="Feedback", lines=15)
257
+ readability_output = gr.Textbox(label="Readability Metrics")
258
+ feedback_history_display = gr.HTML(label="Feedback History")
259
+
260
+ with gr.Row():
261
+ compare_personas = gr.CheckboxGroup(
262
+ choices=[
263
+ "Agatha Christie",
264
+ "Arthur Conan Doyle",
265
+ "Charles Bukowski",
266
+ "Charles Dickens",
267
+ "Chinua Achebe",
268
+ "Cormac McCarthy",
269
+ "David Foster Wallace",
270
+ "Edgar Allan Poe",
271
+ "F. Scott Fitzgerald",
272
+ "Flannery O'Connor",
273
+ "Franz Kafka",
274
+ "Fyodor Dostoevsky",
275
+ "Gabriel Garcia Marquez",
276
+ "George R.R. Martin",
277
+ "George Orwell",
278
+ "Haruki Murakami",
279
+ "Hemingway",
280
+ "Herman Melville",
281
+ "Isabel Allende",
282
+ "James Joyce",
283
+ "Jane Austen",
284
+ "J.K. Rowling",
285
+ "J.R.R. Tolkien",
286
+ "Jorge Luis Borges",
287
+ "Kurt Vonnegut",
288
+ "Leo Tolstoy",
289
+ "Margaret Atwood",
290
+ "Mark Twain",
291
+ "Mary Shelley",
292
+ "Milan Kundera",
293
+ "Naguib Mahfouz",
294
+ "Neil Gaiman",
295
+ "Octavia Butler",
296
+ "Philip K Dick",
297
+ "Ray Bradbury",
298
+ "Salman Rushdie",
299
+ "Shakespeare",
300
+ "Stephen King",
301
+ "Toni Morrison",
302
+ "T.S. Eliot",
303
+ "Ursula K. Le Guin",
304
+ "Virginia Woolf",
305
+ "Virginia Woolf",
306
+ "Zadie Smith"],
307
+ label="Compare Multiple Persona's Feedback at Once"
308
+ )
309
+ with gr.Row():
310
+ compare_button = gr.Button("Compare Feedback")
311
+
312
+ feedback_history = gr.State([])
313
+
314
+ def add_custom_persona(name, description):
315
+ updated_choices = persona_dropdown.choices + [name]
316
+ persona_prompts[name] = f"As {name}, {description}, provide feedback on the following text:"
317
+ return gr.update(choices=updated_choices)
318
+
319
+ def update_feedback_history(current_text, persona, feedback):
320
+ # Ensure feedback_history.value is initialized and is a list
321
+ if feedback_history.value is None:
322
+ feedback_history.value = []
323
+
324
+ history = feedback_history.value
325
+
326
+ # Append the new entry to the history
327
+ history.append({"text": current_text, "persona": persona, "feedback": feedback})
328
+
329
+ # Keep only the last 5 entries in the history
330
+ feedback_history.value = history[-10:]
331
+
332
+ # Generate and return the updated HTML
333
+ return generate_feedback_history_html(feedback_history.value)
334
+
335
+ def compare_feedback(text, selected_personas, api_name, api_key):
336
+ results = []
337
+ for persona in selected_personas:
338
+ feedback = generate_writing_feedback(text, persona, "Overall", api_name, api_key)
339
+ results.append(f"### {persona}'s Feedback:\n{feedback}\n\n")
340
+ return "\n".join(results)
341
+
342
+ add_custom_persona_button.click(
343
+ fn=add_custom_persona,
344
+ inputs=[custom_persona_name, custom_persona_description],
345
+ outputs=persona_dropdown
346
+ )
347
+
348
+ get_feedback_button.click(
349
+ fn=lambda text, persona, aspect, api_name, api_key: (
350
+ generate_writing_feedback(text, persona, aspect, api_name, api_key),
351
+ calculate_readability(text),
352
+ update_feedback_history(text, persona, generate_writing_feedback(text, persona, aspect, api_name, api_key))
353
+ ),
354
+ inputs=[input_text, persona_dropdown, aspect_dropdown, api_name_input, api_key_input],
355
+ outputs=[feedback_output, readability_output, feedback_history_display]
356
+ )
357
+
358
+ compare_button.click(
359
+ fn=compare_feedback,
360
+ inputs=[input_text, compare_personas, api_name_input, api_key_input],
361
+ outputs=feedback_output
362
+ )
363
+
364
+ generate_prompt_button.click(
365
+ fn=generate_writing_prompt,
366
+ inputs=[persona_dropdown, api_name_input, api_key_input],
367
+ outputs=input_text
368
+ )
369
+
370
+ return input_text, feedback_output, readability_output, feedback_history_display
371
+
372
+
373
+ def create_creative_writing_tab():
374
+ with gr.TabItem("Creative Writing Assistant"):
375
+ gr.Markdown("# Utility to be added...")
376
+
377
+
378
+ #FIXME - change to use chat function
379
+ def chat_with_character(user_message, history, char_data, api_name_input, api_key):
380
+ if char_data is None:
381
+ return history, "Please import a character card first."
382
+
383
+ bot_message = generate_writing_feedback(user_message, char_data['name'], "Overall", api_name_input,
384
+ api_key)
385
+ history.append((user_message, bot_message))
386
+ return history, ""
387
+
388
+ def import_character_card(file):
389
+ if file is None:
390
+ logging.warning("No file provided for character card import")
391
+ return None
392
+ try:
393
+ if file.name.lower().endswith(('.png', '.webp')):
394
+ logging.info(f"Attempting to import character card from image: {file.name}")
395
+ json_data = extract_json_from_image(file)
396
+ if json_data:
397
+ logging.info("JSON data extracted from image, attempting to parse")
398
+ return import_character_card_json(json_data)
399
+ else:
400
+ logging.warning("No JSON data found in the image")
401
+ else:
402
+ logging.info(f"Attempting to import character card from JSON file: {file.name}")
403
+ content = file.read().decode('utf-8')
404
+ return import_character_card_json(content)
405
+ except Exception as e:
406
+ logging.error(f"Error importing character card: {e}")
407
+ return None
408
+
409
+
410
+ def import_character_card_json(json_content):
411
+ try:
412
+ # Remove any leading/trailing whitespace
413
+ json_content = json_content.strip()
414
+
415
+ # Log the first 100 characters of the content
416
+ logging.debug(f"JSON content (first 100 chars): {json_content[:100]}...")
417
+
418
+ card_data = json.loads(json_content)
419
+ logging.debug(f"Parsed JSON data keys: {list(card_data.keys())}")
420
+ if 'spec' in card_data and card_data['spec'] == 'chara_card_v2':
421
+ logging.info("Detected V2 character card")
422
+ return card_data['data']
423
+ else:
424
+ logging.info("Assuming V1 character card")
425
+ return card_data
426
+ except json.JSONDecodeError as e:
427
+ logging.error(f"JSON decode error: {e}")
428
+ logging.error(f"Problematic JSON content: {json_content[:500]}...")
429
+ except Exception as e:
430
+ logging.error(f"Unexpected error parsing JSON: {e}")
431
+ return None
432
+
433
+
434
+ def extract_json_from_image(image_file):
435
+ logging.debug(f"Attempting to extract JSON from image: {image_file.name}")
436
+ try:
437
+ with Image.open(image_file) as img:
438
+ logging.debug("Image opened successfully")
439
+ metadata = img.info
440
+ if 'chara' in metadata:
441
+ logging.debug("Found 'chara' in image metadata")
442
+ chara_content = metadata['chara']
443
+ logging.debug(f"Content of 'chara' metadata (first 100 chars): {chara_content[:100]}...")
444
+ try:
445
+ decoded_content = base64.b64decode(chara_content).decode('utf-8')
446
+ logging.debug(f"Decoded content (first 100 chars): {decoded_content[:100]}...")
447
+ return decoded_content
448
+ except Exception as e:
449
+ logging.error(f"Error decoding base64 content: {e}")
450
+
451
+ logging.debug("'chara' not found in metadata, checking for base64 encoded data")
452
+ raw_data = img.tobytes()
453
+ possible_json = raw_data.split(b'{', 1)[-1].rsplit(b'}', 1)[0]
454
+ if possible_json:
455
+ try:
456
+ decoded = base64.b64decode(possible_json).decode('utf-8')
457
+ if decoded.startswith('{') and decoded.endswith('}'):
458
+ logging.debug("Found and decoded base64 JSON data")
459
+ return '{' + decoded + '}'
460
+ except Exception as e:
461
+ logging.error(f"Error decoding base64 data: {e}")
462
+
463
+ logging.warning("No JSON data found in the image")
464
+ except Exception as e:
465
+ logging.error(f"Error extracting JSON from image: {e}")
466
+ return None
467
+
468
+ def load_chat_history(file):
469
+ try:
470
+ content = file.read().decode('utf-8')
471
+ chat_data = json.loads(content)
472
+ return chat_data['history'], chat_data['character']
473
+ except Exception as e:
474
+ logging.error(f"Error loading chat history: {e}")
475
+ return None, None
476
+
477
+ def create_character_card_interaction_tab():
478
+ with gr.TabItem("Chat with a Character Card"):
479
+ gr.Markdown("# Chat with a Character Card")
480
+ with gr.Row():
481
+ with gr.Column(scale=1):
482
+ character_card_upload = gr.File(label="Upload Character Card")
483
+ import_card_button = gr.Button("Import Character Card")
484
+ load_characters_button = gr.Button("Load Existing Characters")
485
+ from App_Function_Libraries.Chat_related_functions import get_character_names
486
+ character_dropdown = gr.Dropdown(label="Select Character", choices=get_character_names())
487
+ api_name_input = gr.Dropdown(
488
+ choices=[None, "Local-LLM", "OpenAI", "Anthropic", "Cohere", "Groq", "DeepSeek", "Mistral",
489
+ "OpenRouter", "Llama.cpp", "Kobold", "Ooba", "Tabbyapi", "VLLM", "ollama", "HuggingFace"],
490
+ value=None,
491
+ label="API for Interaction"
492
+ )
493
+ api_key_input = gr.Textbox(label="API Key (if not set in config.txt)",
494
+ placeholder="Enter your API key here", type="password")
495
+ temperature_slider = gr.Slider(minimum=0.0, maximum=2.0, value=0.7, step=0.05, label="Temperature")
496
+ import_chat_button = gr.Button("Import Chat History")
497
+ chat_file_upload = gr.File(label="Upload Chat History JSON", visible=False)
498
+
499
+
500
+ with gr.Column(scale=2):
501
+ chat_history = gr.Chatbot(label="Conversation")
502
+ user_input = gr.Textbox(label="Your message")
503
+ send_message_button = gr.Button("Send Message")
504
+ regenerate_button = gr.Button("Regenerate Last Message")
505
+ save_chat_button = gr.Button("Save This Chat")
506
+ save_status = gr.Textbox(label="Save Status", interactive=False)
507
+
508
+ character_data = gr.State(None)
509
+
510
+ def import_chat_history(file, current_history, char_data):
511
+ loaded_history, char_name = load_chat_history(file)
512
+ if loaded_history is None:
513
+ return current_history, char_data, "Failed to load chat history."
514
+
515
+ # Check if the loaded chat is for the current character
516
+ if char_data and char_data.get('name') != char_name:
517
+ return current_history, char_data, f"Warning: Loaded chat is for character '{char_name}', but current character is '{char_data.get('name')}'. Chat not imported."
518
+
519
+ # If no character is selected, try to load the character from the chat
520
+ if not char_data:
521
+ new_char_data = load_character(char_name)[0]
522
+ if new_char_data:
523
+ char_data = new_char_data
524
+ else:
525
+ return current_history, char_data, f"Warning: Character '{char_name}' not found. Please select the character manually."
526
+
527
+ return loaded_history, char_data, f"Chat history for '{char_name}' imported successfully."
528
+
529
+ def import_character(file):
530
+ card_data = import_character_card(file)
531
+ if card_data:
532
+ from App_Function_Libraries.Chat_related_functions import save_character
533
+ save_character(card_data)
534
+ return card_data, gr.update(choices=get_character_names())
535
+ else:
536
+ return None, gr.update()
537
+
538
+ def load_character(name):
539
+ from App_Function_Libraries.Chat_related_functions import load_characters
540
+ characters = load_characters()
541
+ char_data = characters.get(name)
542
+ if char_data:
543
+ first_message = char_data.get('first_mes', "Hello! I'm ready to chat.")
544
+ return char_data, [(None, first_message)] if first_message else []
545
+ return None, []
546
+
547
+ def character_chat_wrapper(message, history, char_data, api_endpoint, api_key, temperature):
548
+ logging.debug("Entered character_chat_wrapper")
549
+ if char_data is None:
550
+ return "Please select a character first.", history
551
+
552
+ # Prepare the character's background information
553
+ char_background = f"""
554
+ Name: {char_data.get('name', 'Unknown')}
555
+ Description: {char_data.get('description', 'N/A')}
556
+ Personality: {char_data.get('personality', 'N/A')}
557
+ Scenario: {char_data.get('scenario', 'N/A')}
558
+ """
559
+
560
+ # Prepare the system prompt for character impersonation
561
+ system_message = f"""You are roleplaying as the character described below. Respond to the user's messages in character, maintaining the personality and background provided. Do not break character or refer to yourself as an AI.
562
+
563
+ {char_background}
564
+
565
+ Additional instructions: {char_data.get('post_history_instructions', '')}
566
+ """
567
+
568
+ # Prepare media_content and selected_parts
569
+ media_content = {
570
+ 'id': char_data.get('name'),
571
+ 'title': char_data.get('name', 'Unknown Character'),
572
+ 'content': char_background,
573
+ 'description': char_data.get('description', ''),
574
+ 'personality': char_data.get('personality', ''),
575
+ 'scenario': char_data.get('scenario', '')
576
+ }
577
+ selected_parts = ['description', 'personality', 'scenario']
578
+
579
+ prompt = char_data.get('post_history_instructions', '')
580
+
581
+ # Prepare the input for the chat function
582
+ if not history:
583
+ full_message = f"{prompt}\n\n{message}" if prompt else message
584
+ else:
585
+ full_message = message
586
+
587
+ # Call the chat function
588
+ bot_message = chat(
589
+ message,
590
+ history,
591
+ media_content,
592
+ selected_parts,
593
+ api_endpoint,
594
+ api_key,
595
+ prompt,
596
+ temperature,
597
+ system_message
598
+ )
599
+
600
+ # Update history
601
+ history.append((message, bot_message))
602
+ return history
603
+
604
+ def save_chat_history(history, character_name):
605
+ # Create the Saved_Chats folder if it doesn't exist
606
+ save_directory = "Saved_Chats"
607
+ os.makedirs(save_directory, exist_ok=True)
608
+
609
+ timestamp = datetime.now().strftime("%Y%m%d_%H%M%S")
610
+ filename = f"chat_history_{character_name}_{timestamp}.json"
611
+ filepath = os.path.join(save_directory, filename)
612
+
613
+ chat_data = {
614
+ "character": character_name,
615
+ "timestamp": timestamp,
616
+ "history": history
617
+ }
618
+
619
+ try:
620
+ with open(filepath, 'w', encoding='utf-8') as f:
621
+ json.dump(chat_data, f, ensure_ascii=False, indent=2)
622
+ return filepath
623
+ except Exception as e:
624
+ return f"Error saving chat: {str(e)}"
625
+
626
+ def save_current_chat(history, char_data):
627
+ if not char_data or not history:
628
+ return "No chat to save or character not selected."
629
+
630
+ character_name = char_data.get('name', 'Unknown')
631
+ result = save_chat_history(history, character_name)
632
+ if result.startswith("Error"):
633
+ return result
634
+ return f"Chat saved successfully as {result}"
635
+
636
+ def regenerate_last_message(history, char_data, api_name, api_key, temperature):
637
+ if not history:
638
+ return history
639
+
640
+ last_user_message = history[-1][0]
641
+ new_history = history[:-1]
642
+
643
+ return character_chat_wrapper(last_user_message, new_history, char_data, api_name, api_key, temperature)
644
+
645
+ import_chat_button.click(
646
+ fn=lambda: gr.update(visible=True),
647
+ outputs=chat_file_upload
648
+ )
649
+
650
+ chat_file_upload.change(
651
+ fn=import_chat_history,
652
+ inputs=[chat_file_upload, chat_history, character_data],
653
+ outputs=[chat_history, character_data, save_status]
654
+ )
655
+
656
+ import_card_button.click(
657
+ fn=import_character,
658
+ inputs=[character_card_upload],
659
+ outputs=[character_data, character_dropdown]
660
+ )
661
+
662
+ load_characters_button.click(
663
+ fn=lambda: gr.update(choices=get_character_names()),
664
+ outputs=character_dropdown
665
+ )
666
+
667
+ character_dropdown.change(
668
+ fn=load_character,
669
+ inputs=[character_dropdown],
670
+ outputs=[character_data, chat_history]
671
+ )
672
+
673
+ send_message_button.click(
674
+ fn=character_chat_wrapper,
675
+ inputs=[user_input, chat_history, character_data, api_name_input, api_key_input, temperature_slider],
676
+ outputs=[chat_history]
677
+ ).then(lambda: "", outputs=user_input)
678
+
679
+ regenerate_button.click(
680
+ fn=regenerate_last_message,
681
+ inputs=[chat_history, character_data, api_name_input, api_key_input, temperature_slider],
682
+ outputs=[chat_history]
683
+ )
684
+
685
+ save_chat_button.click(
686
+ fn=save_current_chat,
687
+ inputs=[chat_history, character_data],
688
+ outputs=[save_status]
689
+ )
690
+
691
+ return character_data, chat_history, user_input
692
+
693
+
694
+ def create_mikupad_tab():
695
+ with gr.TabItem("Mikupad"):
696
+ gr.Markdown("I Wish. Gradio won't embed it successfully...")
697
+
698
+ #
699
+ # End of Writing.py
700
+ ########################################################################################################################
App_Function_Libraries/Gradio_UI/__init__.py ADDED
File without changes
App_Function_Libraries/Gradio_UI/__pycache__/Audio_ingestion_tab.cpython-312.pyc ADDED
Binary file (9.42 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Chat_ui.cpython-312.pyc ADDED
Binary file (49.3 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Explain_summarize_tab.cpython-312.pyc ADDED
Binary file (10.5 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Export_Functionality.cpython-312.pyc ADDED
Binary file (18.4 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Gradio_Shared.cpython-312.pyc ADDED
Binary file (13.7 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Import_Functionality.cpython-312.pyc ADDED
Binary file (27.5 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Introduction_tab.cpython-312.pyc ADDED
Binary file (15 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Keywords.cpython-312.pyc ADDED
Binary file (4.58 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Llamafile_tab.cpython-312.pyc ADDED
Binary file (6.52 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Media_edit.cpython-312.pyc ADDED
Binary file (14.3 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/PDF_ingestion_tab.cpython-312.pyc ADDED
Binary file (8.79 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Podcast_tab.cpython-312.pyc ADDED
Binary file (9.53 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Re_summarize_tab.cpython-312.pyc ADDED
Binary file (14.5 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Search_Tab.cpython-312.pyc ADDED
Binary file (26.2 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Trash.cpython-312.pyc ADDED
Binary file (7.73 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Utilities.cpython-312.pyc ADDED
Binary file (7.55 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Video_transcription_tab.cpython-312.pyc ADDED
Binary file (35.1 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Website_scraping_tab.cpython-312.pyc ADDED
Binary file (6.53 kB). View file
 
App_Function_Libraries/Gradio_UI/__pycache__/Writing.cpython-312.pyc ADDED
Binary file (33.7 kB). View file