Update app.py

app.py

@@ -1,13 +1,43 @@
 import gradio as gr
+import subprocess
 import requests
 from datetime import timedelta
+import os
 
 # -----------------------------
-# ASR model (NO TOKEN NEEDED)
+# Free ASR Model (No Token)
 # -----------------------------
 ASR_URL = "https://api-inference.huggingface.co/models/facebook/wav2vec2-large-960h"
 
 
+# -----------------------------
+# Extract audio from any video
+# -----------------------------
+def extract_audio(video_path):
+    try:
+        audio_path = "audio.wav"
+        cmd = [
+            "ffmpeg", "-y",
+            "-i", video_path,
+            "-ac", "1", "-ar", "16000",
+            audio_path
+        ]
+
+        subprocess.run(
+            cmd,
+            stdout=subprocess.PIPE,
+            stderr=subprocess.PIPE,
+            check=True
+        )
+
+        return audio_path, "Audio extracted successfully!"
+    except Exception as e:
+        return None, f"FFmpeg Error: {e}"
+
+
+# -----------------------------
+# Transcribe audio using free HF model
+# -----------------------------
 def transcribe_audio(audio_path):
     try:
         with open(audio_path, "rb") as f:
@@ -30,74 +60,81 @@ def transcribe_audio(audio_path):
 
 
 # -----------------------------
-# Manual SRT generator (NO srt library)
+# Create SRT without any library
 # -----------------------------
 def make_srt(text):
     try:
         words = text.split()
         lines = []
-        line = ""
+        chunk = ""
 
         for w in words:
-            if len(line.split()) < 8:   # 8 words per subtitle
-                line += w + " "
+            if len(chunk.split()) < 8:
+                chunk += w + " "
             else:
-                lines.append(line.strip())
-                line = w + " "
+                lines.append(chunk.strip())
+                chunk = w + " "
 
-        if line:
-            lines.append(line.strip())
+        if chunk:
+            lines.append(chunk.strip())
 
-        # Now build SRT manually
-        srt_output = ""
+        srt_out = ""
         for i, caption in enumerate(lines, start=1):
             start = timedelta(seconds=(i - 1) * 3)
             end = timedelta(seconds=i * 3)
 
-            srt_output += f"{i}\n"
-            srt_output += f"{str(start)[:-3].replace('.', ',')} --> {str(end)[:-3].replace('.', ',')}\n"
-            srt_output += f"{caption}\n\n"
-
-        out_file = "subtitles.srt"
-        with open(out_file, "w", encoding="utf-8") as f:
-            f.write(srt_output)
+            srt_out += f"{i}\n"
+            srt_out += f"{str(start)[:-3].replace('.', ',')} --> {str(end)[:-3].replace('.', ',')}\n"
+            srt_out += caption + "\n\n"
 
-        return out_file, "SRT created successfully!"
+        file = "subtitles.srt"
+        with open(file, "w", encoding="utf-8") as f:
+            f.write(srt_out)
 
+        return file, "SRT created!"
     except Exception as e:
-        return None, f"SRT creation error: {e}"
+        return None, f"SRT Error: {e}"
 
 
 # -----------------------------
-# Main Wrapper
+# Master function for Gradio UI
 # -----------------------------
-def process(audio):
+def process(video):
     try:
-        text, log1 = transcribe_audio(audio)
+        # Step 1: Extract audio
+        audio_path, log1 = extract_audio(video)
+        if not audio_path:
+            return None, log1
+
+        # Step 2: Transcribe
+        text, log2 = transcribe_audio(audio_path)
         if not text:
-            return None, f"ASR FAILED: {log1}"
+            return None, log2
 
-        srt_path, log2 = make_srt(text)
+        # Step 3: Create SRT
+        srt_file, log3 = make_srt(text)
+        if not srt_file:
+            return None, log3
 
-        return srt_path, f"{log1}\n{log2}"
+        logs = f"{log1}\n{log2}\n{log3}"
+        return srt_file, logs
 
     except Exception as e:
-        return None, f"PROCESS ERROR: {e}"
+        return None, f"Processing Error: {e}"
 
 
-# ------------------------------------------------
-# UI (Same as your previous version)
-# ------------------------------------------------
+# -----------------------------
+# Gradio App UI
+# -----------------------------
 with gr.Blocks() as app:
-    gr.Markdown("## 🎬 Auto Subtitle Generator (No Whisper, No Token, No srt Library)")
-
-    audio_in = gr.Audio(label="Upload Video or Audio")
-    btn = gr.Button("Generate SRT")
+    gr.Markdown("## 🎬 Smart Video Subtitle Generator (No Token, No Whisper)")
 
-    srt_out = gr.File(label="Download SRT")
-    dbg = gr.Textbox(label="Debug Log", lines=4)
+    video = gr.Video(label="Upload Any Video File")
+    btn = gr.Button("Generate Subtitles")
 
-    btn.click(process, inputs=audio_in, outputs=[srt_out, dbg])
+    srt_output = gr.File(label="Download SRT")
+    debug = gr.Textbox(label="Debug Log", lines=5)
 
+    btn.click(process, inputs=video, outputs=[srt_output, debug])
 
 app.launch()