luciano/markitdown
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

pre-commit changes

Browse source
This commit is contained in:
Mauro Druwel 2024-12-23 15:52:22 +01:00
parent 46d0829f78
commit 6ecc4f444f
1 changed files with 7 additions and 3 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

10
src/markitdown/_markitdown.py
View file

@ -699,7 +699,7 @@ class DocxConverter(HtmlConverter):
def sanitize_filename(self, name: str) -> str: def sanitize_filename(self, name: str) -> str:
"""Sanitizes a string to make it a valid file name.""" """Sanitizes a string to make it a valid file name."""
# Normalize whitespace # Normalize whitespace
name = re.sub(r'\s+', " ", name.strip()) name = re.sub(r"\s+", " ", name.strip())
# Replace invalid characters with underscores # Replace invalid characters with underscores
return re.sub(r'[\\/*?:"<>|]', "_", name) return re.sub(r'[\\/*?:"<>|]', "_", name)
@ -720,7 +720,9 @@ class DocxConverter(HtmlConverter):
# Return an empty src if saving fails # Return an empty src if saving fails
return {"src": ""} return {"src": ""}
def convert(self, local_path: str, **kwargs) -> Union[None, DocumentConverterResult]: def convert(
self, local_path: str, **kwargs
) -> Union[None, DocumentConverterResult]:
# Bail if not a DOCX # Bail if not a DOCX
extension = kwargs.get("file_extension", "") extension = kwargs.get("file_extension", "")
if extension.lower() != ".docx": if extension.lower() != ".docx":
@ -734,7 +736,9 @@ class DocxConverter(HtmlConverter):
mammoth_result = convert_to_html( mammoth_result = convert_to_html(
docx_file, docx_file,
style_map=style_map, style_map=style_map,
convert_image=images.inline(lambda img: self.convert_image(img, image_output_dir)), convert_image=images.inline(
lambda img: self.convert_image(img, image_output_dir)
),
) )
html_content = mammoth_result.value html_content = mammoth_result.value