fix: Rewrite youtube parser to handle escaped markdown line-by-line

backend/import_youtube.py

@@ -18,32 +18,37 @@ def parse_youtube_md(filepath: str) -> list:
     with open(filepath, 'r') as f:
         content = f.read()
     
-    # Clean up escaped characters from markdown first
+    # Parse line-by-line looking for JSON objects
-    content = content.replace('\\`', '`').replace('\\-', '-').replace('\\_', '_')
+    # This handles the escaped markdown format in the source file
-    
-    # Find JSON block (between ```json and ```)
-    match = re.search(r'```json\s*\n?\s*(\[.*)', content, re.DOTALL)
-    if not match:
-        print("No JSON block found in file.")
-        return []
-    
-    json_str = match.group(1)
-    # Find the array - it may not be closed properly, so we find opening [ and match to end
-    # Try to parse as much valid JSON as possible
-    try:
-        # Try to find a complete JSON array
-        return json.loads(json_str.split('```')[0].strip())
-    except json.JSONDecodeError:
-        # If that fails, try line-by-line parsing
-        lines = json_str.split('\n')
     videos = []
-        for line in lines:
+    in_json = False
-            line = line.strip().rstrip(',')
+    
-            if line.startswith('{') and line.endswith('}'):
+    for line in content.split('\n'):
-                try:
+        line = line.strip()
-                    videos.append(json.loads(line))
+        
-                except json.JSONDecodeError:
+        # Detect start of JSON block (with or without escapes)
+        if 'json' in line.lower() and ('`' in line or '\\' in line):
+            in_json = True
             continue
+        
+        # Skip array markers
+        if line in ['[', '\\[', ']']:
+            continue
+            
+        # Process JSON objects
+        if in_json and line.startswith('{'):
+            # Clean the line
+            clean_line = line.rstrip(',').rstrip()
+            # Remove trailing markdown escapes
+            clean_line = clean_line.replace('\\_', '_').replace('\\-', '-')
+            
+            if clean_line.endswith('}'):
+                try:
+                    obj = json.loads(clean_line)
+                    videos.append(obj)
+                except json.JSONDecodeError as e:
+                    print(f"Parse error on line: {clean_line[:50]}... - {e}")
+    
     return videos
 
 def normalize_title(title: str) -> str: