Update app.py

app.py

@@ -10,6 +10,7 @@ import asyncio
 from image_fetcher import main
 from video import create_text_image
 from video2 import video_func
+import re
 
 app = Flask(__name__)
 
@@ -19,6 +20,35 @@ CLIPS_DIR= "/app/data/video/"
 os.makedirs(SAVE_DIR, exist_ok=True)
 
 
+
+def extract_english_paragraphs(text):
+    """
+    Extract paragraphs that contain only English text
+    """
+    paragraphs = text.split('\n\n')
+    english_paragraphs = []
+    
+    for para in paragraphs:
+        # Check if the paragraph contains only English characters
+        if not re.search(r'[^\x00-\x7F]', para):
+            english_paragraphs.append(para.strip())
+    
+    return '\n\n'.join(english_paragraphs)
+
+def extract_native_text(text):
+    pattern = r'[^\x00-\x7F]'
+    
+    # Search for the first non-English character
+    match = re.search(pattern, text)
+    
+    if match:
+        # Return everything from the first non-English character
+        return text[match.start():]
+    else:
+        # If no non-English characters found, return empty string
+        return ""
+
+
 @app.route("/")
 def home():
     return "Flask Video Generator is Running"
@@ -29,53 +59,94 @@ def home():
 def generate_video():
     try:
         data = request.get_json()
-        prompt = data.get("duration", '').strip()  # Note: This key is "duration" but seems like a typo; consider changing to "prompt" if possible
-        prompts = prompt.replace("**", "")
+        userprompt = data.get("duration", '').strip()
+        promt=extract_english_paragraphs(userprompt)
+        prompts=prompt.replace("**","")
         print(prompts)
         if prompts == '':
-            return jsonify({"error": "prompt must be provided"}), 400
-
-        # Helper to extract only the last paragraph (assumes paragraphs separated by blank lines)
-        def get_last_paragraph(text):
-            paragraphs = []
-            current = []
-            for line in text.splitlines():
-                if line.strip():
-                    current.append(line)
-                else:
-                    if current:
-                        paragraphs.append('\n'.join(current))
-                        current = []
-            if current:
-                paragraphs.append('\n'.join(current))
-            return paragraphs[-1] if paragraphs else ''  # Return the last one, or empty if none
-
-        last_text = get_last_paragraph(prompts)
-        if not last_text:
-            return jsonify({"error": "No valid text found in prompt"}), 400
-
-        lines = [last_text]  # Force to single block: only the latest text
-
-        # Generate unique id for this generation to avoid file conflicts
-        unique_id = str(uuid.uuid4())[:8]  # Short unique string
-
-        # Remove any existing files for this unique_id (optional, but safer)
-        try:
-            os.remove(os.path.join(CLIPS_DIR, f"clip{unique_id}.mp4"))
-            os.remove(os.path.join(AUDIO_DIR, f"audio{unique_id}.mp3"))
-        except Exception:
-            pass  # Ignore if files don't exist
-
-        video_link = os.path.join(CLIPS_DIR, f"clip{unique_id}.mp4")
-        video_func(unique_id, lines)  # Pass unique_id instead of 0
-
-        # For now, assuming single clip; if you add multi-clip support later, concatenate here
-        return send_file(video_link, mimetype='video/mp4')
-
+            return jsonify({"error": "prompts be must"}), 400
+        image_folder = "/tmp/images"
+        #line=prompts.splitlines()
+        #asyncio.run(main(line))
+        raw_lines = prompts.splitlines(keepends=False)
+        lines = []
+        
+        i = 0
+        while i < len(raw_lines):
+            line = raw_lines[i].strip()
+        
+            # Check if current line is a heading
+            if line.strip().startswith("#") and (line.endswith('?') or line.endswith(':')):
+                block = line  # Start block with heading
+                i += 1
+        
+                # Accumulate body lines until next heading or 5+ lines
+                paragraph_lines = []
+                while i < len(raw_lines):
+                    next_line = raw_lines[i].strip()
+        
+                    # Stop if next line is a heading
+                    if next_line.strip().startswith("#") and (next_line.endswith('?') or next_line.endswith(':')):
+                        break
+        
+                    paragraph_lines.append(next_line)
+                    i += 1
+        
+                    # If we've gathered enough lines for a slide, break to next
+                    if len(paragraph_lines) >= 5:
+                        break
+        
+                # Combine heading + paragraph
+                if paragraph_lines:
+                    block += '\n' + '\n'.join(paragraph_lines)
+        
+                lines.append(block)
+        
+            else:
+                # Group normal lines (not part of any heading)
+                block_lines = []
+                count = 0
+        
+                while i < len(raw_lines) and count < 5:
+                    next_line = raw_lines[i].strip()
+        
+                    # If this is a heading, break to handle it separately
+                    if next_line.strip().startswith("#") and (next_line.endswith('?') or next_line.endswith(':')):
+                        break
+        
+                    block_lines.append(next_line)
+                    i += 1
+                    count += 1
+        
+                if block_lines:
+                    lines.append('\n'.join(block_lines))
+        
+        # Print or use lines as slides
+        if len(lines)==1:
+            try:
+                os.remove("/app/data/video/clip0.mp4")
+                os.remove("/app/data/sound/audio0.mp3")
+            except Exception as e:
+                print(e)
+            #image_folder="/tmp/images"
+            #image_olst=[]
+            #create_text_image(lines[0],0,image_olst)
+            #image_files = sorted(glob.glob(os.path.join(image_folder, "*.png")))
+            #if not image_files:
+                #raise ValueError("No images found in folder!")
+            video_link = "/app/data/video/clip0.mp4"
+            video_func(0,lines)
+            #for img in image_files:
+                #os.remove(img)
+            return send_file(video_link, mimetype='video/mp4')
+            
+        else:
+            print("learn deeper")
     except Exception as e:
         traceback.print_exc()
         return jsonify({"error": str(e)}), 500
 
+
 if __name__ == "__main__":
     app.run(host="0.0.0.0", port=7860)