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

Config options for generating page images

Browse source
This commit is contained in:
Yiorgis Gozadinos 2026-01-14 11:55:08 +02:00
parent 2930679dc1
commit 447e94c5db
No known key found for this signature in database
4 changed files with 9 additions and 1 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

1
haiku_rag_slim/haiku/rag/config/models.py
View file

@ -125,6 +125,7 @@ class ConversionOptions(BaseModel):
# Image options # Image options
images_scale: float = 2.0 images_scale: float = 2.0
generate_page_images: bool = True
generate_picture_images: bool = False generate_picture_images: bool = False
# VLM picture description # VLM picture description

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

@ -97,7 +97,7 @@ class DoclingLocalConverter(DocumentConverter):
do_ocr=opts.do_ocr, do_ocr=opts.do_ocr,
do_table_structure=opts.do_table_structure, do_table_structure=opts.do_table_structure,
images_scale=opts.images_scale, images_scale=opts.images_scale,
generate_page_images=True, generate_page_images=opts.generate_page_images,
generate_picture_images=opts.generate_picture_images or pic_desc.enabled, generate_picture_images=opts.generate_picture_images or pic_desc.enabled,
table_structure_options=TableStructureOptions( table_structure_options=TableStructureOptions(
do_cell_matching=opts.table_cell_matching, do_cell_matching=opts.table_cell_matching,

1
haiku_rag_slim/haiku/rag/converters/docling_serve.py
View file

@ -93,6 +93,7 @@ class DoclingServeConverter(DocumentConverter):
"table_mode": opts.table_mode, "table_mode": opts.table_mode,
"table_cell_matching": str(opts.table_cell_matching).lower(), "table_cell_matching": str(opts.table_cell_matching).lower(),
"images_scale": str(opts.images_scale), "images_scale": str(opts.images_scale),
"generate_page_images": str(opts.generate_page_images).lower(),
"generate_picture_images": str( "generate_picture_images": str(
opts.generate_picture_images or pic_desc.enabled opts.generate_picture_images or pic_desc.enabled
).lower(), ).lower(),

6
tests/test_converters.py
View file

@ -304,11 +304,15 @@ class TestDoclingLocalConverter:
config.processing.conversion_options.do_ocr = False config.processing.conversion_options.do_ocr = False
config.processing.conversion_options.table_mode = "fast" config.processing.conversion_options.table_mode = "fast"
config.processing.conversion_options.images_scale = 3.0 config.processing.conversion_options.images_scale = 3.0
config.processing.conversion_options.generate_page_images = False
converter = DoclingLocalConverter(config) converter = DoclingLocalConverter(config)
assert converter.config.processing.conversion_options.do_ocr is False assert converter.config.processing.conversion_options.do_ocr is False
assert converter.config.processing.conversion_options.table_mode == "fast" assert converter.config.processing.conversion_options.table_mode == "fast"
assert converter.config.processing.conversion_options.images_scale == 3.0 assert converter.config.processing.conversion_options.images_scale == 3.0
assert (
converter.config.processing.conversion_options.generate_page_images is False
)
@pytest.mark.asyncio @pytest.mark.asyncio
async def test_convert_pdf_without_picture_images(self, config): async def test_convert_pdf_without_picture_images(self, config):
@ -542,6 +546,7 @@ class TestDoclingServeConverter:
config.processing.conversion_options.table_cell_matching = False config.processing.conversion_options.table_cell_matching = False
config.processing.conversion_options.do_table_structure = False config.processing.conversion_options.do_table_structure = False
config.processing.conversion_options.images_scale = 3.0 config.processing.conversion_options.images_scale = 3.0
config.processing.conversion_options.generate_page_images = False
converter = DoclingServeConverter(config) converter = DoclingServeConverter(config)
doc_json = create_mock_docling_document("test") doc_json = create_mock_docling_document("test")
@ -567,6 +572,7 @@ class TestDoclingServeConverter:
assert data["table_cell_matching"] == "false" assert data["table_cell_matching"] == "false"
assert data["do_table_structure"] == "false" assert data["do_table_structure"] == "false"
assert data["images_scale"] == "3.0" assert data["images_scale"] == "3.0"
assert data["generate_page_images"] == "false"
@pytest.mark.asyncio @pytest.mark.asyncio
async def test_convert_text_connection_error(self, converter): async def test_convert_text_connection_error(self, converter):