make meta from DOCXToDocument JSON serializable

haystack/components/converters/docx.py

@@ -47,13 +47,13 @@ class DOCXMetadata:
  category: str
  comments: str
  content_status: str
- created: Optional[datetime]
+ created: Optional[str]
  identifier: str
  keywords: str
  language: str
  last_modified_by: str
- last_printed: Optional[datetime]
- modified: Optional[datetime]
+ last_printed: Optional[str]
+ modified: Optional[str]
  revision: int
  subject: str
  title: str
@@ -192,13 +192,15 @@ def _get_docx_metadata(self, document: "DocxDocument") -> DOCXMetadata:
  category=document.core_properties.category,
  comments=document.core_properties.comments,
  content_status=document.core_properties.content_status,
- created=document.core_properties.created,
+ created=document.core_properties.created.isoformat() if document.core_properties.created else None,
  identifier=document.core_properties.identifier,
  keywords=document.core_properties.keywords,
  language=document.core_properties.language,
  last_modified_by=document.core_properties.last_modified_by,
- last_printed=document.core_properties.last_printed,
- modified=document.core_properties.modified,
+ last_printed=document.core_properties.last_printed.isoformat()
+ if document.core_properties.last_printed
+ else None,
+ modified=document.core_properties.modified.isoformat() if document.core_properties.modified else None,
  revision=document.core_properties.revision,
  subject=document.core_properties.subject,
  title=document.core_properties.title,

releasenotes/notes/docxmetadata-json-serializable-e4205ce70d042aed.yaml

@@ -0,0 +1,6 @@
+---
+fixes:
+ - |
+ The metadata produced by `DOCXToDocument` component is now JSON serializable.
+ Previously, it contained `datetime` objects automatically extracted from DOCX files, which are not JSON serializable.
+ Now, the `datetime` objects are converted to strings.

test/components/converters/test_docx_file_to_document.py

@@ -1,5 +1,5 @@
 import logging
-import datetime
+import json
 
 import pytest
 
@@ -34,21 +34,21 @@ def test_run(self, test_files_path, docx_converter):
  category="",
  comments="",
  content_status="",
- created=datetime.datetime(2024, 6, 9, 21, 17, tzinfo=datetime.timezone.utc),
+ created="2024-06-09T21:17:00+00:00",
  identifier="",
  keywords="",
  language="",
  last_modified_by="Carlos Fernández Lorán",
  last_printed=None,
- modified=datetime.datetime(2024, 6, 9, 21, 27, tzinfo=datetime.timezone.utc),
+ modified="2024-06-09T21:27:00+00:00",
  revision=2,
  subject="",
  title="",
  version="",
  ),
  }
 
- def test_run_with_meta_overwrites(self, test_files_path, docx_converter):
+ def test_run_with_additional_meta(self, test_files_path, docx_converter):
  paths = [test_files_path / "docx" / "sample_docx_1.docx"]
  output = docx_converter.run(sources=paths, meta={"language": "it", "author": "test_author"})
  doc = output["documents"][0]
@@ -59,13 +59,13 @@ def test_run_with_meta_overwrites(self, test_files_path, docx_converter):
  category="",
  comments="",
  content_status="",
- created=datetime.datetime(2024, 6, 9, 21, 17, tzinfo=datetime.timezone.utc),
+ created="2024-06-09T21:17:00+00:00",
  identifier="",
  keywords="",
  language="",
  last_modified_by="Carlos Fernández Lorán",
  last_printed=None,
- modified=datetime.datetime(2024, 6, 9, 21, 27, tzinfo=datetime.timezone.utc),
+ modified="2024-06-09T21:27:00+00:00",
  revision=2,
  subject="",
  title="",
@@ -82,7 +82,7 @@ def test_run_error_wrong_file_type(self, caplog, test_files_path, docx_converter
  assert "doc_1.txt and convert it" in caplog.text
  assert results["documents"] == []
 
- def test_run_error_non_existent_file(self, test_files_path, docx_converter, caplog):
+ def test_run_error_non_existent_file(self, docx_converter, caplog):
  """
  Test if the component correctly handles errors.
  """
@@ -121,13 +121,13 @@ def test_document_with_docx_metadata_to_dict(self):
  category="category",
  comments="comments",
  content_status="",
- created=datetime.datetime(2024, 6, 9, 21, 17, tzinfo=datetime.timezone.utc),
+ created="2024-06-09T21:17:00+00:00",
  identifier="",
  keywords="",
  language="",
  last_modified_by="Carlos Fernández Lorán",
  last_printed=None,
- modified=datetime.datetime(2024, 6, 9, 21, 27, tzinfo=datetime.timezone.utc),
+ modified="2024-06-09T21:27:00+00:00",
  revision=2,
  subject="",
  title="",
@@ -149,17 +149,21 @@ def test_document_with_docx_metadata_to_dict(self):
  "category": "category",
  "comments": "comments",
  "content_status": "",
- "created": datetime.datetime(2024, 6, 9, 21, 17, tzinfo=datetime.timezone.utc),
+ "created": "2024-06-09T21:17:00+00:00",
  "identifier": "",
  "keywords": "",
  "language": "",
  "last_modified_by": "Carlos Fernández Lorán",
  "last_printed": None,
- "modified": datetime.datetime(2024, 6, 9, 21, 27, tzinfo=datetime.timezone.utc),
+ "modified": "2024-06-09T21:27:00+00:00",
  "revision": 2,
  "subject": "",
  "title": "",
  "version": "",
  },
  },
  }
+
+ # check it is JSON serializable
+ json_str = json.dumps(doc.to_dict(flatten=False))
+ assert json.loads(json_str) == doc.to_dict(flatten=False)