refactor: rename get_word_bounding_box_from_element() to `get_words…

…_from_obj()`
Unstructured-IO · Oct 30, 2024 · d2332ca · d2332ca
1 parent c2cfd66
commit d2332ca
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 4 deletions.
diff --git a/unstructured/partition/pdf.py b/unstructured/partition/pdf.py
@@ -76,7 +76,7 @@
     clean_pdfminer_inner_elements,
     get_links_in_element,
     get_uris,
-    get_word_bounding_box_from_element,
+    get_words_from_obj,
     map_bbox_and_index,
     merge_inferred_with_extracted_layout,
 )
@@ -459,7 +459,7 @@ def _process_pdfminer_pages(
                     page_number,
                     annotation_threshold,
                 )
-                _, words = get_word_bounding_box_from_element(obj, height)
+                _, words = get_words_from_obj(obj, height)
                 for annot in annotations_within_element:
                     urls_metadata.append(map_bbox_and_index(words, annot))
 

diff --git a/unstructured/partition/pdf_image/pdfminer_processing.py b/unstructured/partition/pdf_image/pdfminer_processing.py
@@ -89,7 +89,7 @@ def process_data_with_pdfminer(
                     page_number,
                     annotation_threshold,
                 )
-                _, words = get_word_bounding_box_from_element(obj, height)
+                _, words = get_words_from_obj(obj, height)
                 for annot in annotations_within_element:
                     urls_metadata.append(map_bbox_and_index(words, annot))
 
@@ -516,7 +516,7 @@ def check_annotations_within_element(
     return annotations_within_element
 
 
-def get_word_bounding_box_from_element(
+def get_words_from_obj(
     obj: LTTextBox,
     height: float,
 ) -> tuple[list[LTChar], list[dict[str, Any]]]: