Update app.py

app.py

@@ -263,8 +263,7 @@ import wave
 import tempfile
 import numpy as np
 
-# Global variables to store file and line index
-file_index = 0
+# Global variables to store line index and lines
 line_index = 0
 lines = []
 
@@ -272,14 +271,13 @@ lines = []
 HF_TOKEN = os.getenv('HF_TOKEN')
 hf_writer = gr.HuggingFaceDatasetSaver(HF_TOKEN, "crowdsourced-calculator-demo")
 
-# Ensure the directory exists
-os.makedirs('./audio_samples', exist_ok=True)
-
 # Function to read lines from a file
 def read_lines_from_file(file_path):
-    global lines
+    global lines, line_index
     with open(file_path, 'r') as file:
         lines = file.readlines()
+    line_index = 0  # Reset line index when a new file is loaded
+    return lines[line_index].strip() if lines else "No lines found in the file."
 
 # Function to save audio to a WAV file
 def save_audio_to_file(audio):
@@ -302,57 +300,54 @@ def save_to_hf_dataset(text, audio_path):
         audio_data = f.read()
     hf_writer.save({"text": text, "audio": audio_data})
 
+# Function to capture and process the audio input
+def process_audio(audio):
+    global line_index, lines
+    try:
+        text_line = lines[line_index].strip()
+        file_path = save_audio_to_file(audio)
+        save_to_hf_dataset(text_line, file_path)
+        return f"Audio saved to {file_path} and uploaded to Hugging Face Dataset."
+    except Exception as e:
+        return f"Error saving audio: {str(e)}"
+
+# Function to handle navigation buttons
+def navigate_lines(button):
+    global line_index, lines
+    if button == 'forward':
+        line_index = min(line_index + 1, len(lines) - 1)
+    elif button == 'previous':
+        line_index = max(line_index - 1, 0)
+    return lines[line_index].strip()
+
 # Gradio interface function
 def audio_capture_interface():
-    global file_index, line_index, lines
-    
-    # Check for files in the directory
-    files = [f for f in os.listdir('./audio_samples') if os.path.isfile(os.path.join('./audio_samples', f))]
-    if not files:
-        return gr.Interface(fn=lambda: "No text files found in the directory.", inputs=None, outputs="text")
-    
-    read_lines_from_file(os.path.join('./audio_samples', files[file_index]))
-
-    # Define the interface components
-    audio_input = gr.Audio(source="microphone", type="numpy", label="Speak and click submit")
-    output_text = gr.Textbox(label="Status", placeholder="Status will appear here")
-
-    # Function to capture and process the audio input
-    def process_audio(audio):
-        global line_index, lines
-        
-        try:
-            text_line = lines[line_index].strip()
-            file_path = save_audio_to_file(audio)
-            save_to_hf_dataset(text_line, file_path)
-            return f"Audio saved to {file_path} and uploaded to Hugging Face Dataset."
-        except Exception as e:
-            return f"Error saving audio: {str(e)}"
-
-    # Function to handle navigation buttons
-    def navigate_lines(button):
-        global line_index, lines
-        
-        if button == 'forward':
-            line_index = min(line_index + 1, len(lines) - 1)
-        elif button == 'previous':
-            line_index = max(line_index - 1, 0)
-        
-        return lines[line_index].strip()
-
-    # Create the Gradio interface
     with gr.Blocks() as iface:
         with gr.Row():
-            text_display = gr.Textbox(label="Text", value=lines[line_index].strip(), interactive=False)
+            file_upload = gr.File(label="Upload a text file", file_types=["text"])
+            text_display = gr.Textbox(label="Text", value="Please upload a file to begin.", interactive=False)
+        with gr.Row():
+            audio_input = gr.Audio(source="microphone", type="numpy", label="Speak and click submit")
         with gr.Row():
-            audio_input.render()
+            status_output = gr.Textbox(label="Status", placeholder="Status will appear here")
         with gr.Row():
-            gr.Button("Previous").click(lambda: navigate_lines('previous'), None, text_display)
-            gr.Button("Forward").click(lambda: navigate_lines('forward'), None, text_display)
-            gr.Button("Submit").click(process_audio, inputs=audio_input, outputs=output_text)
+            gr.Button("Previous").click(lambda: navigate_lines('previous'), outputs=text_display)
+            gr.Button("Forward").click(lambda: navigate_lines('forward'), outputs=text_display)
+            gr.Button("Submit").click(process_audio, inputs=audio_input, outputs=status_output)
+
+        # Handle file upload
+        def upload_file(file):
+            if file:
+                file_path = file.name
+                text_display.update(read_lines_from_file(file_path))
+            else:
+                text_display.update("No file uploaded.")
+
+        file_upload.upload(upload_file, outputs=text_display)
 
     return iface
 
 # Launch the interface
 iface = audio_capture_interface()
 iface.launch()
+