resolve PR review comments

test_unstructured/partition/xlsx/test_xlsx.py

@@ -230,10 +230,10 @@ def test_partition_xlsx_subtables(filename="example-docs/vodafone.xlsx"):
     assert sum(isinstance(element, Table) for element in elements) == 3
     assert len(elements) == 6
 
-    
+
 # NOTE (jennings) partition_xlsx returns a single TableElement,
 # so we can only detect the language of that single element
 def test_partition_xlsx_element_metadata_has_languages():
     filename = "example-docs/stanley-cups.xlsx"
     elements = partition_xlsx(filename=filename)
-    assert elements[0].metadata.languages == ["eng"]    
+    assert elements[0].metadata.languages == ["eng"]

unstructured/file_utils/metadata.py

@@ -1,7 +1,7 @@
 import datetime
 import io
 from dataclasses import dataclass, field
-from typing import IO, Any, Dict, Final, Iterable, Iterator, List, Optional, Union
+from typing import IO, Any, Dict, Final, Iterable, Iterator, List, Optional
 
 import docx
 import openpyxl
@@ -156,7 +156,7 @@ def _get_exif_datetime(exif_dict: Dict[str, Any], key: str) -> Optional[datetime
 
 
 def apply_lang_metadata(
-    elements: Union[Iterable[Element], List[Element]],
+    elements: Iterable[Element],
     languages: List[str],
     detect_language_per_element: bool = False,
 ) -> Iterator[Element]:
@@ -182,7 +182,7 @@ def apply_lang_metadata(
         and len(languages) == 1
         and detect_language_per_element is False
     ):
-        # -- apply detected languge to each metadata --
+        # -- apply detected language to each metadata --
         for e in elements:
             e.metadata.languages = detected_languages
             yield e

unstructured/partition/csv.py

@@ -83,9 +83,9 @@ def partition_csv(
     else:
         metadata = ElementMetadata()
 
-    return list(
-        apply_lang_metadata(
-            [Table(text=text, metadata=metadata)],
-            languages=languages,
-        ),
+    elements = apply_lang_metadata(
+        [Table(text=text, metadata=metadata)],
+        languages=languages,
     )
+
+    return list(elements)

unstructured/partition/docx.py

@@ -203,21 +203,19 @@ def partition_docx(
             'The language parameter must be a list of language codes as strings, ex. ["eng"]',
         )
 
-    elements = list(
-        apply_lang_metadata(
-            elements=_DocxPartitioner.iter_document_elements(
-                filename,
-                file,
-                metadata_filename,
-                include_page_breaks,
-                metadata_last_modified,
-            ),
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = _DocxPartitioner.iter_document_elements(
+        filename,
+        file,
+        metadata_filename,
+        include_page_breaks,
+        metadata_last_modified,
     )
-
-    return elements
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
+    )
+    return list(elements)
 
 
 class _DocxPartitioner:
@@ -292,8 +290,6 @@ def _iter_document_elements(self) -> Iterator[Element]:
         # -- functions like `._iter_paragraph_elements()` where the "just return when done"
         # -- characteristic of a generator avoids repeated code to form interim results into lists.
 
-        # full_text = ""
-
         for section_idx, section in enumerate(self._document.sections):
             yield from self._iter_section_page_breaks(section_idx, section)
             yield from self._iter_section_headers(section)
@@ -516,10 +512,11 @@ def page_is_odd() -> bool:
         # -- predict when two page breaks will be needed and emit one of them. The second will be
         # -- emitted by the rendered page-break to follow.
 
-        if start_type == WD_SECTION_START.EVEN_PAGE and not page_is_odd():
+        if start_type == WD_SECTION_START.EVEN_PAGE:  # noqa
             # -- on an even page we need two total, add one to supplement the rendered page break
             # -- to follow. There is no "first-document-page" special case because 1 is odd.
-            yield from self._increment_page_number()
+            if not page_is_odd():
+                yield from self._increment_page_number()
 
         elif start_type == WD_SECTION_START.ODD_PAGE:
             # -- the first page of the document is an implicit "new" odd-page, so no page-break --

unstructured/partition/email.py

@@ -465,10 +465,10 @@ def partition_email(
                     element.metadata.attached_to_filename = metadata_filename or filename
                     all_elements.append(element)
 
-    return list(
-        apply_lang_metadata(
-            elements=all_elements,
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = apply_lang_metadata(
+        elements=all_elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
     )
+
+    return list(elements)

unstructured/partition/epub.py

@@ -115,12 +115,10 @@ def partition_epub(
 
         elements.extend(section_elements)
 
-    elements = list(
-        apply_lang_metadata(
-            elements,
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = apply_lang_metadata(
+        elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
     )
 
-    return elements
+    return list(elements)

unstructured/partition/msg.py

@@ -135,13 +135,12 @@ def partition_msg(
                     element.metadata.attached_to_filename = metadata_filename or filename
                     elements.append(element)
 
-    return list(
-        apply_lang_metadata(
-            elements=elements,
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
     )
+    return list(elements)
 
 
 def build_msg_metadata(

unstructured/partition/pptx.py

@@ -103,19 +103,19 @@ def partition_pptx(
     source_file = file or filename
     assert source_file is not None
 
-    return list(
-        apply_lang_metadata(
-            elements=_PptxPartitioner.iter_presentation_elements(
-                source_file,
-                include_page_breaks,
-                include_slide_notes,
-                metadata_filename,
-                metadata_last_modified,
-            ),
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = _PptxPartitioner.iter_presentation_elements(
+        source_file,
+        include_page_breaks,
+        include_slide_notes,
+        metadata_filename,
+        metadata_last_modified,
     )
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
+    )
+    return list(elements)
 
 
 class _PptxPartitioner:  # pyright: ignore[reportUnusedClass]

unstructured/partition/text.py

@@ -273,14 +273,12 @@ def partition_text(
             element.metadata = copy.deepcopy(metadata)
             elements.append(element)
 
-    elements = list(
-        apply_lang_metadata(
-            elements=elements,
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
     )
-    return elements
+    return list(elements)
 
 
 def element_from_text(

unstructured/partition/tsv.py

@@ -79,9 +79,8 @@ def partition_tsv(
     else:
         metadata = ElementMetadata()
 
-    return list(
-        apply_lang_metadata(
-            [Table(text=text, metadata=metadata)],
-            languages=languages,
-        ),
+    elements = apply_lang_metadata(
+        [Table(text=text, metadata=metadata)],
+        languages=languages,
     )
+    return list(elements)

unstructured/partition/xlsx.py

@@ -173,12 +173,11 @@ def partition_xlsx(
                         element.metadata.languages = languages
                         elements.append(element)
 
-    return list(
-        apply_lang_metadata(
-            elements=elements,
-            languages=languages,
-        ),
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
     )
+    return list(elements)
 
 
 def _get_connected_components(

unstructured/partition/xml.py

@@ -177,10 +177,9 @@ def partition_xml(
                 element.metadata = copy.deepcopy(metadata)
                 elements.append(element)
 
-    return list(
-        apply_lang_metadata(
-            elements=elements,
-            languages=languages,
-            detect_language_per_element=detect_language_per_element,
-        ),
+    elements = apply_lang_metadata(
+        elements=elements,
+        languages=languages,
+        detect_language_per_element=detect_language_per_element,
     )
+    return list(elements)