Merge pull request #35 from climatepolicyradar/feature/pods-1317-bugf…

…ix-for-hugging-face-sync Updating the conditional of whether to assign a page number.
climatepolicyradar · Jun 3, 2024 · 00d8182 · 00d8182
2 parents e68dda6 + 772b9af
commit 00d8182
Show file tree

Hide file tree

Showing 3 changed files with 4 additions and 3 deletions.
diff --git a/src/cpr_sdk/parser_models.py b/src/cpr_sdk/parser_models.py
@@ -415,10 +415,10 @@ def to_passage_level_json(self) -> list[dict[str, Any]]:
         ]
 
         for passage in passages_array:
-            page_number = passage.get("page_number")
+            page_number = passage.get("page_number", None)
             passage[PDF_PAGE_METADATA_KEY] = (
                 self.get_page_metadata_by_page_number(page_number)
-                if page_number
+                if page_number is not None
                 else None
             )
 

diff --git a/src/cpr_sdk/version.py b/src/cpr_sdk/version.py
@@ -1,6 +1,6 @@
 _MAJOR = "1"
 _MINOR = "1"
-_PATCH = "4"
+_PATCH = "5"
 _SUFFIX = ""
 
 VERSION_SHORT = "{0}.{1}".format(_MAJOR, _MINOR)

diff --git a/tests/test_parser_models.py b/tests/test_parser_models.py
@@ -204,6 +204,7 @@ def test_to_passage_level_json_method(
             )
 
             if passage["document_content_type"] == CONTENT_TYPE_PDF:
+                assert passage[PDF_PAGE_METADATA_KEY] is not None
                 assert set(passage["pdf_data"].keys()) == expected_pdf_data_fields
             elif passage["document_content_type"] == CONTENT_TYPE_HTML:
                 assert set(passage["html_data"].keys()) == expected_html_data_fields