Update app.py

app.py

@@ -4,7 +4,6 @@ import logging
 from fastapi import FastAPI, File, UploadFile
 from fastapi.middleware.cors import CORSMiddleware
 
-# Set up logging to see errors in the Hugging Face Log tab
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
@@ -19,17 +18,31 @@ app.add_middleware(
 )
 
 def normalize_color(color):
-    """Ensures color is always a 0-1 RGB dict for Figma"""
+    """Safely converts any PDF color format to Figma RGB format."""
     try:
-        if color is None: return {"r": 0, "g": 0, "b": 0}
-        # If color is an integer (from span['color'])
+        # If the PDF doesn't define a color for this element, return None
+        # This tells Figma not to apply a fill or stroke
+        if color is None:
+            return None
+            
+        # Handle integer colors (e.g., 16711680 for Red)
         if isinstance(color, int):
             color = fitz.utils.getColor(color)
         
+        # Ensure it's a list/tuple of numbers
+        if not isinstance(color, (list, tuple)):
+            return {"r": 0, "g": 0, "b": 0}
+
+        # Convert to 0.0 - 1.0 range based on component count
         if len(color) == 1: # Grayscale
-            return {"r": float(color[0]), "g": float(color[0]), "b": float(color[0])}
-        return {"r": float(color[0]), "g": float(color[1]), "b": float(color[2])}
-    except:
+            val = float(color[0])
+            return {"r": val, "g": val, "b": val}
+        elif len(color) >= 3: # RGB or CMYK (taking first 3)
+            return {"r": float(color[0]), "g": float(color[1]), "b": float(color[2])}
+        
+        return {"r": 0, "g": 0, "b": 0}
+    except Exception as e:
+        logger.error(f"Color error: {e}")
         return {"r": 0, "g": 0, "b": 0}
 
 @app.get("/")
@@ -39,7 +52,6 @@ async def root():
 @app.post("/convert")
 async def convert_pdf(file: UploadFile = File(...)):
     try:
-        logger.info(f"Processing file: {file.filename}")
         content = await file.read()
         doc = fitz.open(stream=content, filetype="pdf")
         pages_data = []
@@ -51,40 +63,39 @@ async def convert_pdf(file: UploadFile = File(...)):
                 "elements": []
             }
 
-            # 1. Extract Text and Images
-            # Using "dict" captures font size, location, and images
+            # 1. Text and Images
             raw_dict = page.get_text("dict")
             for block in raw_dict["blocks"]:
-                if block["type"] == 0:  # Text Block
-                    for line in block["lines"]:
-                        for span in line["spans"]:
+                if block.get("type") == 0:  # TEXT
+                    for line in block.get("lines", []):
+                        for span in line.get("spans", []):
                             page_dict["elements"].append({
                                 "type": "TEXT",
-                                "content": span["text"],
+                                "content": span.get("text", ""),
                                 "x": float(span["bbox"][0]),
                                 "y": float(span["bbox"][1]),
-                                "size": float(span["size"]),
-                                "color": normalize_color(span["color"])
+                                "size": float(span.get("size", 12)),
+                                "color": normalize_color(span.get("color"))
                             })
-                elif block["type"] == 1:  # Image Block
+                elif block.get("type") == 1:  # IMAGE
                     page_dict["elements"].append({
                         "type": "IMAGE",
                         "bytes": base64.b64encode(block["image"]).decode("utf-8"),
-                        "x": float(block["bbox"][0]),
+                        "x": float(block["bbox"][0]), 
                         "y": float(block["bbox"][1]),
                         "width": float(block["bbox"][2] - block["bbox"][0]),
                         "height": float(block["bbox"][3] - block["bbox"][1])
                     })
 
-            # 2. Extract Vector Drawings (Paths)
+            # 2. Vector Drawings
             for path in page.get_drawings():
                 svg_path = ""
-                for item in path["items"]:
-                    if item[0] == "l": # line
+                for item in path.get("items", []):
+                    if item[0] == "l": 
                         svg_path += f"M {item[1].x} {item[1].y} L {item[2].x} {item[2].y} "
-                    elif item[0] == "c": # curve
+                    elif item[0] == "c":
                         svg_path += f"M {item[1].x} {item[1].y} C {item[2].x} {item[2].y} {item[3].x} {item[3].y} {item[4].x} {item[4].y} "
-                    elif item[0] == "re": # rectangle
+                    elif item[0] == "re": 
                         r = item[1]
                         svg_path += f"M {r.x0} {r.y0} L {r.x1} {r.y0} L {r.x1} {r.y1} L {r.x0} {r.y1} Z "
                 
@@ -103,7 +114,7 @@ async def convert_pdf(file: UploadFile = File(...)):
         return {"pages": pages_data}
 
     except Exception as e:
-        logger.error(f"Conversion Error: {str(e)}")
+        logger.error(f"Global Error: {str(e)}")
         return {"error": str(e)}
 
 if __name__ == "__main__":