remove pandas, use calamine + tabulate

microsoft · Jan 4, 2025 · bbacf89 · bbacf89
1 parent b953121
commit bbacf89
Show file tree

Hide file tree

Showing 3 changed files with 9 additions and 11 deletions.
diff --git a/pyproject.toml b/pyproject.toml
@@ -30,8 +30,8 @@ dependencies = [
   "markdownify",
   "numpy",
   "python-pptx",
-  "pandas",
   "python-calamine",
+  "tabulate",
   "pdfminer.six",
   "puremagic",
   "pydub",

diff --git a/src/markitdown/_markitdown.py b/src/markitdown/_markitdown.py
@@ -22,10 +22,11 @@
 import mammoth
 import markdownify
 import olefile
-import pandas as pd
 import pdfminer
 import pdfminer.high_level
 import pptx
+from python_calamine import load_workbook
+from tabulate import tabulate
 
 # File-format detection
 import puremagic
@@ -726,11 +727,14 @@ def convert(self, local_path, **kwargs) -> Union[None, DocumentConverterResult]:
         if extension.lower() not in [".xlsx", ".xls", ".xlsb", ".xlsm"]:
             return None
 
-        sheets = pd.read_excel(local_path, sheet_name=None, engine="calamine")
+        workbook = load_workbook(local_path)
         md_content = ""
-        for s in sheets:
+        for s in workbook.sheet_names:
+            sheet = workbook.get_sheet_by_name(s)
+            # TODO: Add argument to allow filtering empty row / columns
+            tabular_data = sheet.to_python(skip_empty_area=False)
             md_content += f"## {s}\n"
-            html_content = sheets[s].to_html(index=False)
+            html_content = tabulate(tabular_data, tablefmt="html")
             md_content += self._convert(html_content).text_content.strip() + "\n\n"
 
         return DocumentConverterResult(

diff --git a/tests/test_markitdown.py b/tests/test_markitdown.py
@@ -54,12 +54,6 @@
     "affc7dad-52dc-4b98-9b5d-51e65d8a8ad0",
 ]
 
-XLS_TEST_STRINGS = [
-    "## 09060124-b5e7-4717-9d07-3c046eb",
-    "6ff4173b-42a5-4784-9b19-f49caff4d93d",
-    "affc7dad-52dc-4b98-9b5d-51e65d8a8ad0",
-]
-
 DOCX_TEST_STRINGS = [
     "314b0a30-5b04-470b-b9f7-eed2c2bec74a",
     "49e168b7-d2ae-407f-a055-2167576f39a1",