orhan/test
1
1
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Feature/ Add xls support (#169)

Browse Source 
* add xlrd
* add xls converter with tests
This commit is contained in:
yeungadrian
2025-01-03 21:58:17 +00:00
committed by GitHub
parent d248621ba4
commit 08ed32869e
4 changed files with 39 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

BIN
tests/test_files/test.xls vendored Normal file
View File

Binary file not shown.

12
tests/test_markitdown.py
View File

@@ -54,6 +54,12 @@ XLSX_TEST_STRINGS = [
"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",
@@ -185,6 +191,12 @@ def test_markitdown_local() -> None:
result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test.xlsx"))
validate_strings(result, XLSX_TEST_STRINGS)
# Test XLS processing
result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test.xls"))
for test_string in XLS_TEST_STRINGS:
text_content = result.text_content.replace("\\", "")
assert test_string in text_content
# Test DOCX processing
result = markitdown.convert(os.path.join(TEST_FILES_DIR, "test.docx"))
validate_strings(result, DOCX_TEST_STRINGS)