mirror of
https://github.com/DS4SD/docling.git
synced 2025-12-08 20:58:11 +00:00
fix: Add unit tests (#51)
* add the pytests Signed-off-by: Peter Staar <taa@zurich.ibm.com> * renamed the test folder and added the toplevel test Signed-off-by: Peter Staar <taa@zurich.ibm.com> * updated the toplevel function test Signed-off-by: Peter Staar <taa@zurich.ibm.com> * need to start running all tests successfully Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added the reference converted documents Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added first test for json and md output Signed-off-by: Peter Staar <taa@zurich.ibm.com> * ran pre-commit Signed-off-by: Peter Staar <taa@zurich.ibm.com> * replaced deprecated json function with model_dump_json Signed-off-by: Peter Staar <taa@zurich.ibm.com> * replaced deprecated json function with model_dump_json Signed-off-by: Peter Staar <taa@zurich.ibm.com> * reformatted code Signed-off-by: Peter Staar <taa@zurich.ibm.com> * Fix backend tests Signed-off-by: Christoph Auer <cau@zurich.ibm.com> * commented out the drawing Signed-off-by: Peter Staar <taa@zurich.ibm.com> * ci: avoid duplicate runs Signed-off-by: Michele Dolfi <97102151+dolfim-ibm@users.noreply.github.com> * commented out json verification for now Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added verification of input cells Signed-off-by: Peter Staar <taa@zurich.ibm.com> * reformat code Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added test to verify the cells in the pages Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added test to verify the cells in the pages (2) Signed-off-by: Peter Staar <taa@zurich.ibm.com> * added test to verify the cells in the pages (3) Signed-off-by: Peter Staar <taa@zurich.ibm.com> * run all examples in CI Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * make sure examples return failures Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * raise a failure if examples fail Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * fix examples Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * run examples after tests Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * Add tests and update top_level_tests using only datamodels Signed-off-by: Christoph Auer <cau@zurich.ibm.com> * Remove unnecessary code Signed-off-by: Christoph Auer <cau@zurich.ibm.com> * Validate conversion status on e2e test Signed-off-by: Christoph Auer <cau@zurich.ibm.com> * package verify utils and add more tests Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * reduce docs in example, since they are already in the tests Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * skip batch_convert Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * pin docling-parse 1.1.2 Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> * updated the error messages Signed-off-by: Peter Staar <taa@zurich.ibm.com> * commented out the json verification for now Signed-off-by: Peter Staar <taa@zurich.ibm.com> * bumped GLM version Signed-off-by: Peter Staar <taa@zurich.ibm.com> * Fix lockfile Signed-off-by: Christoph Auer <cau@zurich.ibm.com> * Pin new docling-parse v1.1.3 Signed-off-by: Christoph Auer <cau@zurich.ibm.com> --------- Signed-off-by: Peter Staar <taa@zurich.ibm.com> Signed-off-by: Christoph Auer <cau@zurich.ibm.com> Signed-off-by: Michele Dolfi <97102151+dolfim-ibm@users.noreply.github.com> Signed-off-by: Michele Dolfi <dol@zurich.ibm.com> Co-authored-by: Christoph Auer <cau@zurich.ibm.com> Co-authored-by: Michele Dolfi <97102151+dolfim-ibm@users.noreply.github.com> Co-authored-by: Michele Dolfi <dol@zurich.ibm.com>
This commit is contained in:
Peter W. J. Staar
committed by
GitHub
GitHub
parent
256f4d504e
commit
48f4d1ba52
64
tests/test_backend_docling_parse.py
Normal file
64
tests/test_backend_docling_parse.py
Normal file
@@ -0,0 +1,64 @@
|
||||
from pathlib import Path
|
||||
|
||||
import pytest
|
||||
|
||||
from docling.backend.docling_parse_backend import (
|
||||
DoclingParseDocumentBackend,
|
||||
DoclingParsePageBackend,
|
||||
)
|
||||
from docling.datamodel.base_models import BoundingBox
|
||||
|
||||
|
||||
@pytest.fixture
|
||||
def test_doc_path():
|
||||
return Path("./tests/data/2206.01062.pdf")
|
||||
|
||||
|
||||
def test_text_cell_counts():
|
||||
pdf_doc = Path("./tests/data/redp5695.pdf")
|
||||
|
||||
doc_backend = DoclingParseDocumentBackend(pdf_doc, "123456xyz")
|
||||
|
||||
for page_index in range(0, doc_backend.page_count()):
|
||||
last_cell_count = None
|
||||
for i in range(10):
|
||||
page_backend: DoclingParsePageBackend = doc_backend.load_page(0)
|
||||
cells = list(page_backend.get_text_cells())
|
||||
|
||||
if last_cell_count is None:
|
||||
last_cell_count = len(cells)
|
||||
|
||||
if len(cells) != last_cell_count:
|
||||
assert (
|
||||
False
|
||||
), "Loading page multiple times yielded non-identical text cell counts"
|
||||
last_cell_count = len(cells)
|
||||
|
||||
|
||||
def test_get_text_from_rect(test_doc_path):
|
||||
doc_backend = DoclingParseDocumentBackend(test_doc_path, "123456xyz")
|
||||
page_backend: DoclingParsePageBackend = doc_backend.load_page(0)
|
||||
|
||||
# Get the title text of the DocLayNet paper
|
||||
textpiece = page_backend.get_text_in_rect(
|
||||
bbox=BoundingBox(l=102, t=77, r=511, b=124)
|
||||
)
|
||||
ref = "DocLayNet: A Large Human-Annotated Dataset for Document-Layout Analysis"
|
||||
|
||||
assert textpiece.strip() == ref
|
||||
|
||||
|
||||
def test_crop_page_image(test_doc_path):
|
||||
doc_backend = DoclingParseDocumentBackend(test_doc_path, "123456xyz")
|
||||
page_backend: DoclingParsePageBackend = doc_backend.load_page(0)
|
||||
|
||||
# Crop out "Figure 1" from the DocLayNet paper
|
||||
im = page_backend.get_page_image(
|
||||
scale=2, cropbox=BoundingBox(l=317, t=246, r=574, b=527)
|
||||
)
|
||||
# im.show()
|
||||
|
||||
|
||||
def test_num_pages(test_doc_path):
|
||||
doc_backend = DoclingParseDocumentBackend(test_doc_path, "123456xyz")
|
||||
doc_backend.page_count() == 9
|
||||
Reference in New Issue
Block a user