qurator-spk/page2tsv
1
0
Fork 0
mirror of https://github.com/qurator-spk/page2tsv.git synced 2025-07-19 06:59:52 +02:00
Code Issues Releases Wiki Activity

rename GND-ID column to more generic ID

Browse source
This commit is contained in:
Kai 2020-03-13 09:05:55 +01:00
parent 0cd9cd932a
commit b13dae29f5
1 changed files with 3 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

6
cli.py
View file

@ -140,7 +140,7 @@ def ner(tsv, ner_rest_endpoint):
tsv_result.append((0, ner_token, ner_tag, 'O', '-', row.url_id, row.left, row.right, row.top, tsv_result.append((0, ner_token, ner_tag, 'O', '-', row.url_id, row.left, row.right, row.top,
row.bottom)) row.bottom))
return pd.DataFrame(tsv_result, columns=['No.', 'TOKEN', 'NE-TAG', 'NE-EMB', 'GND-ID', 'url_id', return pd.DataFrame(tsv_result, columns=['No.', 'TOKEN', 'NE-TAG', 'NE-EMB', 'ID', 'url_id',
'left', 'right', 'top', 'bottom']), ner_result 'left', 'right', 'top', 'bottom']), ner_result
@ -166,7 +166,7 @@ def ned(tsv, ner_result, ned_rest_endpoint):
if eid in ned_result: if eid in ned_result:
candidates = ned_result[eid] candidates = ned_result[eid]
tsv.loc[rids, 'GND-ID'] = candidates[0][1]['wikidata'] tsv.loc[rids, 'ID'] = candidates[0][1]['wikidata']
rids = [] rids = []
entity = "" entity = ""
@ -197,7 +197,7 @@ def ned(tsv, ner_result, ned_rest_endpoint):
@click.option('--scale-factor', type=float, default=0.5685, help='default: 0.5685') @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, ned_rest_endpoint, noproxy, scale_factor): def page2tsv(page_xml_file, tsv_out_file, image_url, ner_rest_endpoint, ned_rest_endpoint, noproxy, scale_factor):
out_columns = ['No.', 'TOKEN', 'NE-TAG', 'NE-EMB', 'GND-ID', 'url_id', 'left', 'right', 'top', 'bottom'] out_columns = ['No.', 'TOKEN', 'NE-TAG', 'NE-EMB', 'ID', 'url_id', 'left', 'right', 'top', 'bottom']
if noproxy: if noproxy:
os.environ['no_proxy'] = '*' os.environ['no_proxy'] = '*'