qurator-spk/ocrd_calamari
1
0
Fork 0
mirror of https://github.com/mikegerber/ocrd_calamari.git synced 2025-10-31 15:54:13 +01:00
Code Releases Activity

✅ Convert to a pytest style test

Browse source
This commit is contained in:
Gerber, Mike 2019-12-05 13:38:45 +01:00
parent 2393edc645
commit e07b333db1
1 changed files with 46 additions and 43 deletions
Download patch file Download diff file Expand all files Collapse all files

89
test/test_recognize.py
View file

@ -3,63 +3,66 @@ import shutil
import subprocess import subprocess
import urllib.request import urllib.request
from test.base import TestCase, main, assets, skip import pytest
from ocrd.resolver import Resolver from ocrd.resolver import Resolver
from ocrd_calamari import CalamariRecognize from ocrd_calamari import CalamariRecognize
from test.base import main, assets
METS_KANT = assets.url_of('kant_aufklaerung_1784-page-block-line-word_glyph/data/mets.xml') METS_KANT = assets.url_of('kant_aufklaerung_1784-page-block-line-word_glyph/data/mets.xml')
WORKSPACE_DIR = '/tmp/test-ocrd-calamari' WORKSPACE_DIR = '/tmp/test-ocrd-calamari'
class TestCalamariRecognize(TestCase):
def setUp(self): @pytest.fixture
if os.path.exists(WORKSPACE_DIR): def workspace():
shutil.rmtree(WORKSPACE_DIR) if os.path.exists(WORKSPACE_DIR):
os.makedirs(WORKSPACE_DIR) shutil.rmtree(WORKSPACE_DIR)
os.makedirs(WORKSPACE_DIR)
def runTest(self): resolver = Resolver()
resolver = Resolver() workspace = resolver.workspace_from_url(METS_KANT, dst_dir=WORKSPACE_DIR)
workspace = resolver.workspace_from_url(METS_KANT, dst_dir=WORKSPACE_DIR)
# XXX Work around data bug(?): # XXX Work around data bug(?):
# PAGE-XML links to OCR-D-IMG/INPUT_0017.tif, but this is nothing core can download # PAGE-XML links to OCR-D-IMG/INPUT_0017.tif, but this is nothing core can download
os.makedirs(os.path.join(WORKSPACE_DIR, 'OCR-D-IMG')) os.makedirs(os.path.join(WORKSPACE_DIR, 'OCR-D-IMG'))
for f in ['INPUT_0017.tif', 'INPUT_0020.tif']: for f in ['INPUT_0017.tif', 'INPUT_0020.tif']:
urllib.request.urlretrieve( urllib.request.urlretrieve(
"https://github.com/OCR-D/assets/raw/master/data/kant_aufklaerung_1784/data/OCR-D-IMG/" + f, "https://github.com/OCR-D/assets/raw/master/data/kant_aufklaerung_1784/data/OCR-D-IMG/" + f,
os.path.join(WORKSPACE_DIR, 'OCR-D-IMG', f)) os.path.join(WORKSPACE_DIR, 'OCR-D-IMG', f))
# The binarization options I have are: return workspace
#
# a. ocrd_kraken which tries to install cltsm, whose installation is borken on my machine (protobuf)
# b. ocrd_olena which 1. I cannot fully install via pip and 2. whose dependency olena doesn't compile on my
# machine
# c. just fumble with the original files
#
# So I'm going for option c.
for f in ['INPUT_0017.tif', 'INPUT_0020.tif']:
ff = os.path.join(WORKSPACE_DIR, 'OCR-D-IMG', f)
subprocess.call(['convert', ff, '-colorspace', 'Gray', ff])
# XXX Should remove GT text to really test this
CalamariRecognize( def test_recognize(workspace):
workspace, # The binarization options I have are:
input_file_grp="OCR-D-GT-SEG-LINE", #
output_file_grp="OCR-D-OCR-CALAMARI", # a. ocrd_kraken which tries to install cltsm, whose installation is borken on my machine (protobuf)
parameter={ # b. ocrd_olena which 1. I cannot fully install via pip and 2. whose dependency olena doesn't compile on my
'checkpoint': os.path.join(os.getcwd(), 'gt4histocr-calamari/*.ckpt.json') # machine
} # c. just fumble with the original files
).process() #
workspace.save_mets() # So I'm going for option c.
for f in ['INPUT_0017.tif', 'INPUT_0020.tif']:
ff = os.path.join(WORKSPACE_DIR, 'OCR-D-IMG', f)
subprocess.call(['convert', ff, '-colorspace', 'Gray', ff])
# XXX Should remove GT text to really test this
CalamariRecognize(
workspace,
input_file_grp="OCR-D-GT-SEG-LINE",
output_file_grp="OCR-D-OCR-CALAMARI",
parameter={
'checkpoint': os.path.join(os.getcwd(), 'gt4histocr-calamari/*.ckpt.json')
}
).process()
workspace.save_mets()
page1 = os.path.join(workspace.directory, 'OCR-D-OCR-CALAMARI/OCR-D-OCR-CALAMARI_0001.xml')
assert os.path.exists(page1)
with open(page1, 'r', encoding='utf-8') as f:
assert 'verſchuldeten' in f.read()
page1 = os.path.join(workspace.directory, 'OCR-D-OCR-CALAMARI/OCR-D-OCR-CALAMARI_0001.xml')
self.assertTrue(os.path.exists(page1))
with open(page1, 'r', encoding='utf-8') as f:
self.assertIn('verſchuldeten', f.read())
if __name__ == '__main__': if __name__ == '__main__':
main() main()