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

chore: consider header for column-wise drop

Browse source
This commit is contained in:
Hew Li Yang 2024-12-22 21:22:41 +08:00
parent 113f7748b7
commit 7b64e6ebfd
2 changed files with 7 additions and 2 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

9
src/markitdown/_markitdown.py
View file

@ -525,7 +525,7 @@ class XlsxConverter(HtmlConverter):
def _clean_colname(self, colname: Any) -> Any: def _clean_colname(self, colname: Any) -> Any:
# Remove Pandas header placeholders # Remove Pandas header placeholders
if isinstance(colname, str) and colname.startswith("Unnamed:"): if isinstance(colname, str) and colname.startswith("Unnamed:"):
return "" return None
return colname return colname
def convert( def convert(
@ -548,11 +548,16 @@ class XlsxConverter(HtmlConverter):
sheet = sheet.rename(columns=lambda col: self._clean_colname(col)) sheet = sheet.rename(columns=lambda col: self._clean_colname(col))
if drop_empty_cols: if drop_empty_cols:
sheet = sheet.dropna(axis=1, how="all") # also consider headers to be part of the column
sheet = sheet.loc[:, sheet.notna().any() | sheet.columns.notna()]
if drop_empty_rows: if drop_empty_rows:
sheet = sheet.dropna(axis=0, how="all") sheet = sheet.dropna(axis=0, how="all")
# convert remaining NaN's to empty string
# because .to_html(na_rep="") does not apply to headers
sheet.columns = sheet.columns.fillna(na_rep)
html_content = sheet.to_html(index=False, na_rep=na_rep) html_content = sheet.to_html(index=False, na_rep=na_rep)
md_content += self._convert(html_content).text_content.strip() + "\n\n" md_content += self._convert(html_content).text_content.strip() + "\n\n"

BIN
tests/test_files/test.xlsx vendored
View file

Binary file not shown.