feat(tasks): extract and map embedded PDF metadata in upload_to_email

Implement extraction of embedded PDF metadata using pypdf, mapping
/Title, /Author, /Subject, /Keywords to filename, absender,
document_type, and tags respectively. JSON metadata takes precedence;
PDF metadata fills missing fields only.

Co-authored-by: christianlouis <361235+christianlouis@users.noreply.github.com>

app/tasks/upload_to_email.py

@@ -24,6 +24,15 @@ logger = logging.getLogger(__name__)
 # Constants
 _LOGO_FILENAME = "logo.png"
 
+# Mapping from PDF metadata keys (with leading slash stripped) to application-specific names.
+# This mirrors the inverse of the mapping used in embed_metadata_into_pdf.py.
+_PDF_METADATA_KEY_MAP = {
+    "Title": "filename",
+    "Author": "absender",
+    "Subject": "document_type",
+    "Keywords": "tags",
+}
+
 
 def get_email_template(template_name="default.html"):
     """
@@ -64,9 +73,12 @@ def extract_metadata_from_file(file_path):
     """
     Try to extract metadata from a file using several methods:
     1. Check for a .json metadata file with the same name
-    2. Extract metadata from PDF if it's embedded
+    2. Extract embedded metadata from PDF using pypdf
 
-    Returns a dictionary of metadata or None if not found
+    JSON metadata takes precedence; embedded PDF metadata fills in any missing
+    fields using the application's standard key mapping (e.g., /Title → filename).
+
+    Returns a dictionary of metadata or empty dict if not found.
     """
     metadata = {}
 
@@ -77,7 +89,6 @@ def extract_metadata_from_file(file_path):
             with open(metadata_path, "r", encoding="utf-8") as f:
                 metadata = json.load(f)
                 logger.info(f"Loaded metadata from external JSON file: {metadata_path}")
-                return metadata
         except Exception as e:
             logger.warning(f"Failed to load metadata from JSON file: {str(e)}")
 
@@ -88,11 +99,14 @@ def extract_metadata_from_file(file_path):
                 pdf_reader = pypdf.PdfReader(f)
                 pdf_metadata = pdf_reader.metadata
                 if pdf_metadata:
-                    # Convert metadata to a standard dictionary
                     for key, value in pdf_metadata.items():
                         # Remove the leading slash from PDF metadata keys (e.g., '/Title' -> 'Title')
                         clean_key = key[1:] if key.startswith("/") else key
-                        metadata[clean_key] = str(value)
+                        # Map to application-specific key names where possible
+                        mapped_key = _PDF_METADATA_KEY_MAP.get(clean_key, clean_key)
+                        # Only set if not already present (JSON metadata takes precedence)
+                        if mapped_key not in metadata:
+                            metadata[mapped_key] = str(value)
 
                     logger.info(f"Extracted embedded metadata from PDF: {file_path}")
         except Exception as e:

tests/test_upload_email.py

@@ -143,11 +143,72 @@ class TestExtractMetadataFromFile:
 
         result = extract_metadata_from_file(str(file_path))
 
-        # Check that the leading slash is stripped and keys/values match
-        assert result.get("Title") == "Test Title"
-        assert result.get("Author") == "Test Author"
-        assert result.get("Subject") == "Test Document"
-        assert result.get("Keywords") == "test, metadata, pypdf"
+        # Keys are mapped to application-specific names
+        assert result.get("filename") == "Test Title"
+        assert result.get("absender") == "Test Author"
+        assert result.get("document_type") == "Test Document"
+        assert result.get("tags") == "test, metadata, pypdf"
+
+    def test_extracts_embedded_metadata_from_pdf(self, tmp_path):
+        """Test that embedded PDF metadata is mapped to application-specific keys."""
+        import pypdf
+
+        file_path = tmp_path / "mapped.pdf"
+
+        writer = pypdf.PdfWriter()
+        writer.add_blank_page(width=100, height=100)
+        writer.add_metadata(
+            {
+                "/Title": "Invoice 2024",
+                "/Author": "Acme Corp",
+                "/Subject": "invoice",
+                "/Keywords": "finance, billing",
+            }
+        )
+        with open(file_path, "wb") as f:
+            writer.write(f)
+
+        result = extract_metadata_from_file(str(file_path))
+
+        # Verify the PDF-to-app key mapping
+        assert result["filename"] == "Invoice 2024"
+        assert result["absender"] == "Acme Corp"
+        assert result["document_type"] == "invoice"
+        assert result["tags"] == "finance, billing"
+
+    def test_pdf_metadata_does_not_overwrite_json(self, tmp_path):
+        """Test that JSON metadata takes precedence over embedded PDF metadata."""
+        import pypdf
+
+        file_path = tmp_path / "dual.pdf"
+
+        # Create a PDF with embedded metadata
+        writer = pypdf.PdfWriter()
+        writer.add_blank_page(width=100, height=100)
+        writer.add_metadata(
+            {
+                "/Title": "PDF Title",
+                "/Author": "PDF Author",
+                "/Subject": "PDF Subject",
+                "/Keywords": "pdf, keywords",
+            }
+        )
+        with open(file_path, "wb") as f:
+            writer.write(f)
+
+        # Create a companion JSON file that sets some overlapping fields
+        json_metadata = {"filename": "JSON Filename", "absender": "JSON Author"}
+        json_path = tmp_path / "dual.json"
+        json_path.write_text(json.dumps(json_metadata))
+
+        result = extract_metadata_from_file(str(file_path))
+
+        # JSON values must not be overwritten by PDF metadata
+        assert result["filename"] == "JSON Filename"
+        assert result["absender"] == "JSON Author"
+        # Fields missing from JSON are filled from PDF metadata
+        assert result["document_type"] == "PDF Subject"
+        assert result["tags"] == "pdf, keywords"
 
 
 @pytest.mark.unit