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

Shrink the docling-serve context-expansion cassette

Browse source
This commit is contained in:
Yiorgis Gozadinos 2026-06-29 14:41:02 +03:00
parent afd6e3ae00
commit e4b1f16f97
No known key found for this signature in database
2 changed files with 32050 additions and 226160 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

258200
tests/cassettes/test_context_enhancement/test_expand_context_no_base64_images_docling_serve.yaml
View file

File diff suppressed because one or more lines are too long

10
tests/test_context_enhancement.py
View file

@ -558,19 +558,19 @@ async def test_expand_context_no_base64_images_docling_local(
@pytest.mark.vcr() @pytest.mark.vcr()
async def test_expand_context_no_base64_images_docling_serve(temp_db_path): async def test_expand_context_no_base64_images_docling_serve(
temp_db_path, doclaynet_first_page_pdf
):
"""Ensure expanded context from real PDF does not contain base64 image data. """Ensure expanded context from real PDF does not contain base64 image data.
Tests end-to-end with doclaynet.pdf using docling-serve converter. Tests end-to-end with a single page of doclaynet.pdf using docling-serve converter.
""" """
from pathlib import Path
config = AppConfig() config = AppConfig()
config.processing.converter = "docling-serve" config.processing.converter = "docling-serve"
config.processing.chunker = "docling-serve" config.processing.chunker = "docling-serve"
async with HaikuRAG(temp_db_path, config=config, create=True) as client: async with HaikuRAG(temp_db_path, config=config, create=True) as client:
pdf_path = Path(__file__).parent / "data" / "doclaynet.pdf" pdf_path = doclaynet_first_page_pdf
result = await client.create_document_from_source(pdf_path) result = await client.create_document_from_source(pdf_path)
doc = result if not isinstance(result, list) else result[0] doc = result if not isinstance(result, list) else result[0]
assert doc.id is not None assert doc.id is not None