danvics/haiku.rag
1
0
Fork 0
Code Issues Pull requests Projects Releases Packages Wiki Activity Actions

Use a neutral dotfile in the docling name examples

Browse source 
.gitignore reads as this repository's own file in the helper docstring,
the tests and the changelog entry.
This commit is contained in:
Yiorgis Gozadinos 2026-07-30 14:13:23 +03:00
parent 8e7e494fd1
commit e2b930d469
No known key found for this signature in database
3 changed files with 7 additions and 7 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

2
CHANGELOG.md
View file

@ -3,7 +3,7 @@
### Fixed ### Fixed
- Dotfiles (`.gitignore`, `.env`) are parsed as their actual format instead of a single unstructured text block. Docling ignores the extension of a name starting with a dot, so converters strip leading dots from the name they hand it. - Dotfiles are parsed as their actual format instead of a single unstructured text block. Docling ignores the extension of a name starting with a dot, so converters strip leading dots from the name they hand it.
## [0.71.0] - 2026-07-29 ## [0.71.0] - 2026-07-29

2
haiku_rag_slim/haiku/rag/converters/text_utils.py
View file

@ -10,7 +10,7 @@ def docling_safe_name(name: str) -> str:
"""Strip leading dots from a filename handed to docling. """Strip leading dots from a filename handed to docling.
Docling ignores the extension of any name starting with a dot, so a name Docling ignores the extension of any name starting with a dot, so a name
derived from a dotfile (".gitignore" -> ".gitignore.md") is classified as derived from a dotfile (".customrc" -> ".customrc.md") is classified as
an unknown format and parsed as one unstructured text block. an unknown format and parsed as one unstructured text block.
""" """
return name.lstrip(".") or name return name.lstrip(".") or name

10
tests/test_converters.py
View file

@ -168,7 +168,7 @@ class TestTextFileHandler:
def test_docling_safe_name(self): def test_docling_safe_name(self):
"""Leading dots are stripped, other names are untouched, and a name of """Leading dots are stripped, other names are untouched, and a name of
nothing but dots is kept rather than reduced to an empty filename.""" nothing but dots is kept rather than reduced to an empty filename."""
assert docling_safe_name(".gitignore.md") == "gitignore.md" assert docling_safe_name(".customrc.md") == "customrc.md"
assert docling_safe_name("notes.md") == "notes.md" assert docling_safe_name("notes.md") == "notes.md"
assert docling_safe_name("...") == "..." assert docling_safe_name("...") == "..."
@ -348,14 +348,14 @@ class TestTextToDoclingWithFormat:
@pytest.mark.asyncio @pytest.mark.asyncio
async def test_dotfile_derived_name_parses_as_markdown(self): async def test_dotfile_derived_name_parses_as_markdown(self):
"""Docling ignores the extension of a stream name starting with a dot, """Docling ignores the extension of a stream name starting with a dot,
so a name derived from a dotfile (".gitignore" -> ".gitignore.md") must so a name derived from a dotfile (".customrc" -> ".customrc.md") must
be normalized or the whole document collapses to one plain-text block. be normalized or the whole document collapses to one plain-text block.
""" """
config = AppConfig() config = AppConfig()
converter = DoclingLocalConverter(config) converter = DoclingLocalConverter(config)
text = "Overview.\n\n## History\n\n- First item\n- Second item" text = "Overview.\n\n## History\n\n- First item\n- Second item"
doc = await converter.convert_text(text, name=".gitignore.md") doc = await converter.convert_text(text, name=".customrc.md")
labels = [str(getattr(item, "label", "")) for item, _ in doc.iterate_items()] labels = [str(getattr(item, "label", "")) for item, _ in doc.iterate_items()]
assert "section_header" in labels assert "section_header" in labels
assert "list_item" in labels assert "list_item" in labels
@ -1343,9 +1343,9 @@ class TestDoclingServeConverter:
mock_client.__aexit__ = AsyncMock(return_value=None) mock_client.__aexit__ = AsyncMock(return_value=None)
mock_client_class.return_value = mock_client mock_client_class.return_value = mock_client
await converter.convert_text("# Test", name=".gitignore.md") await converter.convert_text("# Test", name=".customrc.md")
uploaded_name = mock_client.post.call_args.kwargs["files"]["files"][0] uploaded_name = mock_client.post.call_args.kwargs["files"]["files"][0]
assert uploaded_name == "gitignore.md" assert uploaded_name == "customrc.md"
@pytest.mark.asyncio @pytest.mark.asyncio
async def test_dotfile_path_uploads_with_detectable_name(self, converter, tmp_path): async def test_dotfile_path_uploads_with_detectable_name(self, converter, tmp_path):