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

chore: infer dtypes for columns as well, remove unneeded na_rep

Browse source
This commit is contained in:
Hew Li Yang 2025-04-03 11:45:58 +08:00
parent 767b8d611d
commit fae0faf8dd
1 changed files with 4 additions and 8 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

12
packages/markitdown/src/markitdown/converters/_xlsx_converter.py
View file

@ -77,9 +77,9 @@ class ExcelConverterBase(DocumentConverter):
# Because the latter does not replace NaT's # Because the latter does not replace NaT's
with pd.option_context("future.no_silent_downcasting", True): with pd.option_context("future.no_silent_downcasting", True):
sheet = sheet.fillna(na_rep, axis=1).infer_objects(copy=False) sheet = sheet.fillna(na_rep, axis=1).infer_objects(copy=False)
sheet.columns = sheet.columns.fillna(na_rep) sheet.columns = sheet.columns.fillna(na_rep).infer_objects(copy=False)
html_content = sheet.to_html(index=False, na_rep=na_rep) html_content = sheet.to_html(index=False)
md_content += ( md_content += (
self._html_converter.convert_string( self._html_converter.convert_string(
html_content, **kwargs html_content, **kwargs
@ -127,9 +127,7 @@ class XlsxConverter(ExcelConverterBase):
extension=".xlsx", extension=".xlsx",
feature="xlsx", feature="xlsx",
) )
) from _xlsx_dependency_exc_info[ ) from _xlsx_dependency_exc_info[1].with_traceback( # type: ignore[union-attr]
1
].with_traceback( # type: ignore[union-attr]
_xlsx_dependency_exc_info[2] _xlsx_dependency_exc_info[2]
) )
@ -178,9 +176,7 @@ class XlsConverter(ExcelConverterBase):
extension=".xls", extension=".xls",
feature="xls", feature="xls",
) )
) from _xls_dependency_exc_info[ ) from _xls_dependency_exc_info[1].with_traceback( # type: ignore[union-attr]
1
].with_traceback( # type: ignore[union-attr]
_xls_dependency_exc_info[2] _xls_dependency_exc_info[2]
) )