309ae0d57b
Merge in RR/cv-analysis from image-service-compat to master
Squashed commit of the following:
commit 397d12a96a6b78de762f7b3a80a72427f5f51e97
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Tue Aug 16 16:14:40 2022 +0200
update pdf2image, adjust response format for table-parsing & figure-detection
commit f2061bda8d25d64de974e97f36148dea29af50d9
Author: Julius Unverfehrt <julius.unverfehrt@iqser.com>
Date: Mon Aug 15 08:56:39 2022 +0200
add script to save figure detection data that can be used for image-service pipeline script
56 lines
1.6 KiB
Python
56 lines
1.6 KiB
Python
import fitz
|
|
import numpy as np
|
|
import pytest
|
|
|
|
from cv_analysis.server.pipeline import table_parsing_formatter, figure_detection_formatter, make_analysis_pipeline
|
|
from cv_analysis.utils.structures import Rectangle
|
|
|
|
|
|
def analysis_fn_mock(image: np.ndarray):
|
|
bbox = (0, 0, 42, 42)
|
|
return [Rectangle.from_xyxy(bbox)]
|
|
|
|
|
|
@pytest.fixture
|
|
def empty_pdf():
|
|
doc = fitz.open()
|
|
for n in range(1):
|
|
doc.new_page()
|
|
return doc.write()
|
|
|
|
|
|
@pytest.fixture
|
|
def expected_formatted_analysis_result(operation):
|
|
if operation == "table":
|
|
return [
|
|
{
|
|
"pageInfo": {"number": 0, "rotation": 0, "width": 595.0, "height": 842.0},
|
|
"tableCells": [{"x0": 0.0, "y0": 0.0, "x1": 15.12, "y1": 15.12, "width": 15.12, "height": 15.12}],
|
|
}
|
|
]
|
|
if operation == "figure":
|
|
return [
|
|
{
|
|
"pageInfo": {"number": 0, "rotation": 0, "width": 595.0, "height": 842.0},
|
|
"boundingBox": {"x0": 0.0, "y0": 0.0, "x1": 15.12, "y1": 15.12, "width": 15.12, "height": 15.12},
|
|
"alpha": False,
|
|
}
|
|
]
|
|
|
|
|
|
@pytest.fixture
|
|
def formatter(operation):
|
|
if operation == "table":
|
|
return table_parsing_formatter
|
|
elif operation == "figure":
|
|
return figure_detection_formatter
|
|
else:
|
|
raise
|
|
|
|
|
|
@pytest.mark.parametrize("operation", ["table", "figure"])
|
|
def test_analysis_pipeline(empty_pdf, formatter, expected_formatted_analysis_result):
|
|
analysis_pipeline = make_analysis_pipeline(analysis_fn_mock, formatter, dpi=200)
|
|
results = list(analysis_pipeline(empty_pdf))
|
|
assert list(results) == expected_formatted_analysis_result
|