fix: track filename in metadata for docx tables (#597)

* fix: track filename in metadata for docx tables

* bump version

* remove accidental commit

CHANGELOG.md

@@ -1,14 +1,9 @@
-## 0.6.7-dev5
+## 0.6.7-dev6
-
-### Enhancements
-
-* Add `page_name` to metadata. Currently used for the sheet name in XLSX documents.
-
-## 0.6.7-dev4
 
 ### Enhancements
 
 * Add `file_directory` to metadata
+* Add `page_name` to metadata. Currently used for the sheet name in XLSX documents.
 * Added a `--partition-strategy` parameter to unstructured-ingest so that users can specify
   partition strategy in CLI. For example, `--partition-strategy fast`.
 * Added metadata for filetype.
@@ -26,6 +21,7 @@
 * Makes `pytesseract` a function level import in `partition_pdf` so you can use the `"fast"`
   or `"hi_res"` strategies if `pytesseract` is not installed. Also adds the
   `required_dependencies` decorator for the `"hi_res"` and `"ocr_only"` strategies.
+* Fix to ensure `filename` is tracked in metadata for `docx` tables.
 
 ## 0.6.6
 

test_unstructured/partition/test_docx.py

@@ -7,6 +7,7 @@ from unstructured.documents.elements import (
     Address,
     ListItem,
     NarrativeText,
+    Table,
     Text,
     Title,
 )
@@ -97,3 +98,21 @@ def test_partition_docx_raises_with_both_specified(mock_document, tmpdir):
 def test_partition_docx_raises_with_neither():
     with pytest.raises(ValueError):
         partition_docx()
+
+
+def test_partition_docx_processes_table(filename="example-docs/fake_table.docx"):
+    elements = partition_docx(filename=filename)
+
+    assert isinstance(elements[0], Table)
+    assert (
+        elements[0].metadata.text_as_html
+        == """<table>
+<thead>
+<tr><th>Header Col 1   </th><th>Header Col 2  </th></tr>
+</thead>
+<tbody>
+<tr><td>Lorem ipsum    </td><td>A Link example</td></tr>
+</tbody>
+</table>"""
+    )
+    assert elements[0].metadata.filename == "fake_table.docx"

unstructured/__version__.py

@@ -1 +1 @@
-__version__ = "0.6.7-dev5"  # pragma: no cover
+__version__ = "0.6.7-dev6"  # pragma: no cover

unstructured/partition/docx.py

@@ -137,8 +137,10 @@ def partition_docx(
             text_table = _convert_table_to_text(table, as_html=False)
             element = Table(text_table)
             if element is not None:
-                element.metadata = ElementMetadata(filename=metadata_filename)
+                element.metadata = ElementMetadata(
-                element.metadata = ElementMetadata(text_as_html=html_table)
+                    text_as_html=html_table,
+                    filename=metadata_filename,
+                )
                 elements.append(element)
             table_index += 1
         elif element_item.tag.endswith("p"):