Fix file handling and MIME type detection

app.py

@@ -26,9 +26,9 @@ Made with ❤️ using Docling and Gradio
 # Initialize the document parser
 parser = DocumentParser()
 
-def process_document(file_obj):
+def process_document(file_path):
     """Process uploaded document and return structured information"""
-    if file_obj is None:
+    if file_path is None:
         return (
             "Error: No file uploaded",
             pd.DataFrame(),
@@ -37,26 +37,9 @@ def process_document(file_obj):
             "Confidence Score: 0.0"
         )
 
-    temp_path = None
     try:
-        # Create temporary file with appropriate extension
-        original_filename = file_obj.name if hasattr(file_obj, 'name') else "uploaded_file.pdf"
-        extension = os.path.splitext(original_filename)[1].lower()
-        if not extension:
-            extension = '.pdf'  # Default to PDF if no extension
-
-        # Create temporary file and write content
-        with tempfile.NamedTemporaryFile(delete=False, suffix=extension) as tmp_file:
-            # Write the content
-            content = file_obj.read() if hasattr(file_obj, 'read') else file_obj
-            if isinstance(content, bytes):
-                tmp_file.write(content)
-            else:
-                tmp_file.write(content.encode('utf-8'))
-            temp_path = tmp_file.name
-
-        # Parse the document
-        result = parser.parse(temp_path)
+        # Parse the document directly using the file path
+        result = parser.parse(file_path)
         
         # Prepare the outputs
         metadata_df = pd.DataFrame([{
@@ -97,13 +80,6 @@ def process_document(file_obj):
             "No entities available",
             "Confidence Score: 0.0"
         )
-    finally:
-        # Clean up temporary file
-        if temp_path and os.path.exists(temp_path):
-            try:
-                os.unlink(temp_path)
-            except:
-                pass
 
 # Create Gradio interface
 with gr.Blocks(title=TITLE, theme=gr.themes.Soft()) as iface:
@@ -115,7 +91,7 @@ with gr.Blocks(title=TITLE, theme=gr.themes.Soft()) as iface:
             file_input = gr.File(
                 label="Upload Document",
                 file_types=[".pdf", ".docx", ".txt", ".html", ".md"],
-                type="filepath"  # Changed from binary to filepath
+                type="filepath"
             )
             submit_btn = gr.Button("Process Document", variant="primary")
         

dockling_parser/parser.py

@@ -18,7 +18,23 @@ class DocumentParser:
         'application/vnd.openxmlformats-officedocument.wordprocessingml.document': 'docx',
         'text/plain': 'txt',
         'text/html': 'html',
-        'text/markdown': 'md'
+        'text/markdown': 'md',
+        # Add common variations
+        'application/x-pdf': 'pdf',
+        'application/acrobat': 'pdf',
+        'application/msword': 'docx',
+        'text/x-markdown': 'md',
+        'text/x-html': 'html'
+    }
+
+    EXTENSION_TO_MIME = {
+        '.pdf': 'application/pdf',
+        '.docx': 'application/vnd.openxmlformats-officedocument.wordprocessingml.document',
+        '.txt': 'text/plain',
+        '.html': 'text/html',
+        '.htm': 'text/html',
+        '.md': 'text/markdown',
+        '.markdown': 'text/markdown'
     }
 
     def __init__(self, config: Optional[Dict[str, Any]] = None):
@@ -43,7 +59,14 @@ class DocumentParser:
         if not file_path.exists():
             raise FileNotFoundError(f"File not found: {file_path}")
 
-        mime_type = magic.from_file(str(file_path), mime=True)
+        # Try to determine format from extension first
+        extension = file_path.suffix.lower()
+        mime_type = self.EXTENSION_TO_MIME.get(extension)
+
+        # If extension not recognized, use magic
+        if not mime_type:
+            mime_type = magic.from_file(str(file_path), mime=True)
+
         if mime_type not in self.SUPPORTED_FORMATS:
             raise UnsupportedFormatError(f"Unsupported file format: {mime_type}")