Added S3 Functionality

config.py

@@ -0,0 +1,10 @@
+# AWS S3 Credentials
+AWS_ACCESS_KEY_ID = 'AKIA2YB4Z26VQCIW3DHO'  #'AKIA2YB4Z26VTD4PH6BX'
+AWS_SECRET_ACCESS_KEY = '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'  #'il0SbQjGbTrBhAkwSC2zvVCf7GqPpANmCoOold//'
+BUCKET_NAME = 'wt-video-dl'
+AWS_REGION = 'us-east-2'
+S3_OBJ_BASE_URL = f"https://{BUCKET_NAME}.s3.{AWS_REGION}.amazonaws.com"
+
+# OpenAI Credentials
+OPENAI_API_KEY = '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'
+OPENAI_API_URL = 'https://api.openai.com/v1/chat/completions'

main.py

@@ -8,14 +8,17 @@ from contextlib import asynccontextmanager
 
 from models import load_models
 from helperfunctions import *
+from s3_handler import S3Handler
 from media_download import YoutubeDownloader
 # from transcription import StableWhisper
 # from summarizer import Extract_Summary, AudioBookNarration
 # from audiobook import AudioBook
 
-# global MODELS
 
 
+# For Storing Models
+global MODELS
+
 ### API Configurations
 
 # Context Manager for FastAPI Start/Shutdown
@@ -43,6 +46,9 @@ app = FastAPI(lifespan=lifespan)
 # Output Directory for Files Storage
 output_folder = 'Output'
 
+# S3 Handler
+s3 = S3Handler()
+
 # Create a context variable to store the contexts for each user
 users_context = dict()
 
@@ -70,12 +76,23 @@ async def get_media_metadata(request: Request, url: str):
     user_ip = request.client.host
     user_id = generate_uuid(user_ip, url)
 
+    # User Folder Path
+    user_folder_path = os.path.join(output_folder, user_id)
+
     # Getting User's Youtube Downloader
-    youtube_downloader = YoutubeDownloader(url, output_folder)
+    youtube_downloader = YoutubeDownloader(url, user_folder_path)
 
     # Getting Youtube Media Info
     media_metadata = youtube_downloader.get_media_metadata()
 
+    # Storing User's Media Metadata to Directory
+    media_metadata_path = os.path.join(user_folder_path, 'media_metadata.json')
+    with open(media_metadata_path, "w") as outfile: 
+        json.dump(media_metadata, outfile)
+    
+    # Storing User's Media Metadata to S3
+    s3_path = s3.upload_file(user_id, 'media_metadata.json', media_metadata_path)
+
     # Getting Status
     status = 1 if media_metadata else 0
     
@@ -86,14 +103,25 @@ async def get_media_metadata(request: Request, url: str):
         # users_context[user_id]['media_metadata'] = media_metadata
         users_context[user_id]['url'] = url
     
-    return {'status': status, 'user_id': user_id, 'media_metadata': media_metadata}
+    return {'status': status, 'user_id': user_id, 'media_metadata': media_metadata, 'media_metadata_path': s3_path}
 
 
 @app.get("/get_media_formats")
 async def get_media_formats(user_id: str):
     
-    # Downloading Media for User
+    # Getting Media Formats for User
     media_formats = users_context[user_id]['downloader'].get_media_formats()
+
+    # User Folder Path
+    user_folder_path = os.path.join(output_folder, user_id)
+
+    # Storing User's Media Formats to Directory
+    media_formats_path = os.path.join(user_folder_path, 'media_formats.json')
+    with open(media_formats_path, "w") as outfile: 
+        json.dump(media_formats, outfile)
+    
+    # Storing User's Media Formats to S3
+    s3_path = s3.upload_file(user_id, 'media_formats.json', media_formats_path)
     
     # Getting Status
     status = 1 if media_formats else 0
@@ -102,7 +130,7 @@ async def get_media_formats(user_id: str):
         # Storing Media Info in the context for this user's session
         users_context[user_id]['media_formats'] = media_formats
     
-    return {'status': status, 'media_formats': media_formats}
+    return {'status': status, 'media_formats': media_formats, 'media_formats_path': s3_path}
 
 
 @app.get("/download_media")
@@ -110,6 +138,10 @@ async def download_media(user_id: str, media_type: str, media_format: str, media
     
     # Downloading Media for User
     media_path = users_context[user_id]['downloader'].download(media_type, media_format, media_quality)
+
+    # Storing User's Downloaded Media to S3
+    media_file = f"{media_type.lower()}_{media_quality.lower()}.{media_format.lower()}"
+    s3_path = s3.upload_file(user_id, media_file, media_path)
     
     # Getting Status
     status = 1 if media_path else 0
@@ -119,7 +151,7 @@ async def download_media(user_id: str, media_type: str, media_format: str, media
         users_context[user_id]['media_path'] = media_path
         users_context[user_id]['media_type'] = media_type
     
-    return {'status': status, 'media_path': media_path}
+    return {'status': status, 'media_path': s3_path}
 
 
 @app.get("/get_transcript")
@@ -141,12 +173,12 @@ async def get_transcript(user_id: str, subtitle_format: str = 'srt', word_level:
         # Downloading Audio for Transcription
         media_path = users_context[user_id]['downloader'].download('audio', 'mp3', '128kbps')
 
-        # Retrieving Audio
-
     # # Whisper based transcription
-    # stable_whisper_transcript = StableWhisper(media_path, output_folder, subtitle_format=subtitle_format, word_level=word_level)
+    # user_folder_path = os.path.join(output_folder, user_id)
+    # stable_whisper_transcript = StableWhisper(media_path, user_folder_path, subtitle_format=subtitle_format, word_level=word_level)
     # transcript = stable_whisper_transcript.generate_transcript()
     # transcript_path = stable_whisper_transcript.save_transcript()
+    # subtitles_path = stable_whisper_transcript.save_subtitles()
 
     temp_dir = 'temp'
     if word_level:
@@ -172,12 +204,19 @@ async def get_transcript(user_id: str, subtitle_format: str = 'srt', word_level:
 
 @app.get("/get_translation")
 async def get_translation(user_id: str, target_language: str = 'en'):
+
+    # If Transcript Available
+    if 'transcript' in users_context[user_id].keys():
     
-    # Retrieving the transcript from the context for this user's session
-    transcript = users_context[user_id]['transcript']
+        # Retrieving the transcript from the context for this user's session
+        transcript = users_context[user_id]['transcript']
+    
+    else:
+        return {'status': 0, 'message': 'Transcript not generated yet'}
 
-    # # # NLLB based Translation
-    # nllb_translator = Translation(transcript, transcript['language'], target_language, 'output_path')
+    # # NLLB based Translation
+    # user_folder_path = os.path.join(output_folder, user_id)
+    # nllb_translator = Translation(transcript, transcript['language'], target_language, user_folder_path)
     # translated_transcript = nllb_translator.get_translated_transcript()
     # translated_subtitles = nllb_translator.get_translated_subtitles()
 
@@ -207,8 +246,14 @@ async def get_translation(user_id: str, target_language: str = 'en'):
 async def get_summary(user_id: str, Summary_type: str, Summary_strategy: str, Target_Person_type: str,
                       Response_length:  str, Writing_style: str):
 
-    # Getting Transcript
-    text_input = users_context[user_id]['transcript']
+    # If Transcript Available
+    if 'transcript' in users_context[user_id].keys():
+    
+        # Retrieving the transcript from the context for this user's session
+        text_input = users_context[user_id]['transcript']
+    
+    else:
+        return {'status': 0, 'message': 'Transcript not generated yet'}
     
     # # Extracting Summary
     # summary_extractor = Extract_Summary(text_input=text_input)
@@ -239,8 +284,14 @@ async def get_summary(user_id: str, Summary_type: str, Summary_strategy: str, Ta
 async def get_key_info(user_id: str, Summary_type: str, Summary_strategy: str, Target_Person_type: str, 
                        Response_length: str, Writing_style: str):
 
-    # Getting Transcript
-    text_input = users_context[user_id]['transcript']
+    # If Transcript Available
+    if 'transcript' in users_context[user_id].keys():
+    
+        # Retrieving the transcript from the context for this user's session
+        text_input = users_context[user_id]['transcript']
+    
+    else:
+        return {'status': 0, 'message': 'Transcript not generated yet'}
     
     # # Extracting Summary
     # summary_extractor = Extract_Summary(text_input=text_input)
@@ -271,8 +322,14 @@ async def get_key_info(user_id: str, Summary_type: str, Summary_strategy: str, T
 async def get_audiobook(user_id: str, narration_style: str, speaker: str = "male", 
                         audio_format: str = "mp3", audio_quality: str = "128kbps"):
 
-    # Getting Transcript
-    text_input = users_context[user_id]['transcript']
+    # If Transcript Available
+    if 'transcript' in users_context[user_id].keys():
+    
+        # Retrieving the transcript from the context for this user's session
+        text_input = users_context[user_id]['transcript']
+    
+    else:
+        return {'status': 0, 'message': 'Transcript not generated yet'}
     
     # # Extracting Narration
 
@@ -325,10 +382,14 @@ async def get_rendered_video(user_id: str, video_format: str, video_quality: str
     # Burning Subtitles & Rendering Video
     rendered_video_path = burn_subtitles(media_path, subtitles_path)
 
+    # Storing User's Rendered Video to S3
+    media_file = f"subs_video_{video_quality.lower()}.{video_format.lower()}"
+    s3_path = s3.upload_file(user_id, media_file, media_path)
+
     # Getting Status
     status = 1 if rendered_video_path else 0
     
-    return {'status': status, "rendered_video_path": rendered_video_path}
+    return {'status': status, "rendered_video_path": s3_path}
 
 
 

s3_handler.py

@@ -0,0 +1,71 @@
+import os
+import boto3
+from botocore.exceptions import ClientError
+
+from config import AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, BUCKET_NAME, S3_OBJ_BASE_URL
+
+
+class S3Handler:
+
+    def __init__(self):
+        """
+        Establishes connection with the required s3 bucket
+        """
+        try:
+            self.s3 = boto3.client(
+                's3',
+                aws_access_key_id=AWS_ACCESS_KEY_ID,
+                aws_secret_access_key=AWS_SECRET_ACCESS_KEY
+            )
+        except ClientError as e:
+            print(f"Failed to connect to S3: {e}")
+
+    def upload_file(self, user_id, s3_path, file_path, **kwargs):
+        """
+        Uploads file to the s3
+        """
+        
+        try:
+            # Filename: File path to upload
+            # Bucket: Name of the bucket to upload the file.
+            # Key: Name of the key to upload to S3.
+            s3_user_path = user_id + '/' + s3_path
+            self.s3.upload_file( 
+                Filename=file_path, 
+                Bucket=BUCKET_NAME, 
+                Key=s3_user_path,
+                # ExtraArgs={'ACL': 'public-read'}
+            )
+            print(f"File uploaded to {BUCKET_NAME} as {s3_path}")
+            s3_obj_path = S3_OBJ_BASE_URL + '/' + user_id + '/' + s3_path
+            return s3_obj_path
+        except ClientError as e:
+            print(f"Failed to upload file to S3: {e}")
+
+    def download_file(self, user_id, s3_path, file_path, **kwargs):
+        """
+        Downloads file from s3
+        """
+        
+        try:
+            # Filename: Local File path to download to
+            # Bucket:  Name of the bucket to download the file from
+            # Key: Name of the file to download from the bucket
+            s3_user_path = user_id + '/' + s3_path
+            self.s3.download_file( 
+                Filename=file_path, 
+                Bucket=BUCKET_NAME, 
+                Key=s3_user_path
+            )
+            print(f"File downloaded from {BUCKET_NAME} as {file_path}")
+            s3_obj_path = S3_OBJ_BASE_URL + '/' + user_id + '/' + s3_path
+            return s3_obj_path
+        except ClientError as e:
+            print(f"Failed to download file from S3: {e}")
+
+
+# s3_path = "sample_video_srt_2.mp4"
+# file_path = "Output/video_1.mp4"
+
+# s3 = S3Handler()
+# s3.upload_file(s3_path, file_path, user_id="uzair") 

transcription.py

@@ -212,6 +212,16 @@ class StableWhisper(Whisper):
         '''
         Writes the transcript into file
         '''
+        # Writing to TXT file in UTF-8 format
+        file_path = f'{self.filename}.txt'
+        with open(file_path, 'w', encoding='utf-8') as file:
+            file.write(self.text)
+        return file_path
+
+    def save_subtitles(self):
+        '''
+        Writes the subtitles into file
+        '''
         # Writing according to the Format
         file_path = f'{self.filename}.{self.subtitle_format}'
         if self.subtitle_format == 'ass':