qurator-spk/page2tsv
1
0
Fork 0
mirror of https://github.com/qurator-spk/page2tsv.git synced 2025-07-15 04:59:52 +02:00
Code Issues Releases Wiki Activity

add command line option for image scale factor

Browse source
This commit is contained in:
Kai Labusch 2020-01-10 13:04:07 +01:00
parent 311dac31ac
commit 2946909cf3
1 changed files with 4 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

6
cli.py
View file

@ -149,7 +149,8 @@ def ner(tsv, ner_rest_endpoint):
@click.option('--ner-rest-endpoint', type=str, default=None, @click.option('--ner-rest-endpoint', type=str, default=None,
help="REST endpoint of sbb_ner service. See https://github.com/qurator-spk/sbb_ner for details.") help="REST endpoint of sbb_ner service. See https://github.com/qurator-spk/sbb_ner for details.")
@click.option('--noproxy', type=bool, is_flag=True, help='disable proxy. default: enabled.') @click.option('--noproxy', type=bool, is_flag=True, help='disable proxy. default: enabled.')
def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, noproxy): @click.option('--scale-factor', type=float, default=0.5685, help='default: 0.5685')
def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, noproxy, scale_factor):
if noproxy: if noproxy:
os.environ['no_proxy'] = '*' os.environ['no_proxy'] = '*'
@ -173,7 +174,7 @@ def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, noproxy)
for coords in words.findall('.//{%s}Coords' % xmlns): for coords in words.findall('.//{%s}Coords' % xmlns):
# transform the OCR coordinates by 0.5685 to derive the correct coords for the web presentation image # transform the OCR coordinates by 0.5685 to derive the correct coords for the web presentation image
points = [int(0.5685 * float(pos)) for p in coords.attrib['points'].split(' ') for pos in p.split(',')] points = [int(scale_factor * float(pos)) for p in coords.attrib['points'].split(' ') for pos in p.split(',')]
x_points = [points[i] for i in range(0, len(points), 2)] x_points = [points[i] for i in range(0, len(points), 2)]
y_points = [points[i] for i in range(1, len(points), 2)] y_points = [points[i] for i in range(1, len(points), 2)]
@ -193,6 +194,7 @@ def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, noproxy)
'url_id', 'left', 'right', 'top', 'bottom']) 'url_id', 'left', 'right', 'top', 'bottom'])
if ner_rest_endpoint is not None: if ner_rest_endpoint is not None:
tsv = ner(tsv, ner_rest_endpoint) tsv = ner(tsv, ner_rest_endpoint)
tsv.to_csv(tsv_out_file, sep="\t", quoting=3, index=False, mode='a', header=False) tsv.to_csv(tsv_out_file, sep="\t", quoting=3, index=False, mode='a', header=False)