Optimise tests

langchain-ai · Jan 17, 2025 · 830e10d · 830e10d
1 parent 23a73a9
commit 830e10d
Show file tree

Hide file tree

Showing 4 changed files with 51 additions and 8 deletions.
diff --git a/libs/community/langchain_community/document_loaders/parsers/pdf.py b/libs/community/langchain_community/document_loaders/parsers/pdf.py
@@ -466,7 +466,7 @@ class PyMuPDFParser(BaseBlobParser):
             parser = PyMuPDFParser(
                 # password = None,
                 mode = "single",
-                pages_delimitor = "\n\f",
+                pages_delimiter = "\n\f",
                 # extract_images = True,
                 # images_parser = TesseractBlobParser(),
                 # extract_tables="markdown",

diff --git a/libs/community/langchain_community/document_loaders/pdf.py b/libs/community/langchain_community/document_loaders/pdf.py
@@ -495,7 +495,7 @@ def __init__(
         pages_delimiter: str = _DEFAULT_PAGES_DELIMITER,
         extract_images: bool = False,
         images_parser: Optional[BaseImageBlobParser] = None,
-        images_inner_format:str="text",
+        images_inner_format: str = "text",
         extract_tables: Union[Literal["csv", "markdown", "html"], None] = None,
         headers: Optional[dict] = None,
         extract_tables_settings: Optional[dict[str, Any]] = None,

diff --git a/libs/community/tests/integration_tests/document_loaders/parsers/test_pdf_parsers.py b/libs/community/tests/integration_tests/document_loaders/parsers/test_pdf_parsers.py
@@ -143,13 +143,31 @@ def _analyze_image(self, img: "Image") -> str:
 
 
 @pytest.mark.parametrize(
-    "mode",
-    ["single", "page"],
+    "mode,image_parser",
+    [("single", EmptyImageBlobParser()), ("page", None)],
 )
 @pytest.mark.parametrize(
-    "image_parser",
-    [EmptyImageBlobParser(), None],
+    "parser_factory,params",
+    [
+        ("PyMuPDFParser", {}),
+    ],
 )
+@pytest.mark.requires("pillow")
+def test_mode_and_extract_images_variations(
+    parser_factory: str,
+    params: dict,
+    mode: str,
+    image_parser: BaseImageBlobParser,
+) -> None:
+    _test_matrix(
+        parser_factory,
+        params,
+        mode,
+        image_parser,
+        images_inner_format="text",
+    )
+
+
 @pytest.mark.parametrize(
     "images_inner_format",
     ["text", "markdown-img", "html-img"],
@@ -161,7 +179,24 @@ def _analyze_image(self, img: "Image") -> str:
     ],
 )
 @pytest.mark.requires("pillow")
-def test_mode_and_extract_images_variations(
+def test_mode_and_image_formats_variations(
+    parser_factory: str,
+    params: dict,
+    images_inner_format: str,
+) -> None:
+    mode = "single"
+    image_parser = EmptyImageBlobParser()
+
+    _test_matrix(
+        parser_factory,
+        params,
+        mode,
+        image_parser,
+        images_inner_format,
+    )
+
+
+def _test_matrix(
     parser_factory: str,
     params: dict,
     mode: str,

diff --git a/libs/community/tests/integration_tests/document_loaders/test_pdf.py b/libs/community/tests/integration_tests/document_loaders/test_pdf.py
@@ -226,7 +226,15 @@ def test_standard_parameters(
     assert len(docs) == 1
 
     file_path = Path(__file__).parent.parent / "examples/layout-parser-paper.pdf"
-    loader = loader_class(file_path, mode="page")
+    loader = loader_class(file_path,
+                          mode="page",
+                          page_delimiter="---",
+                          images_parser=None,
+                          images_inner_format="text",
+                          password=None,
+                          extract_tables=None,
+                          extract_tables_settings=None,
+                          )
     docs = loader.load()
     assert len(docs) == 16
     assert loader.web_path is None