qurator-spk/ocrd_repair_inconsistencies
1
0
Fork 0
mirror of https://github.com/qurator-spk/ocrd_repair_inconsistencies.git synced 2025-06-09 03:40:06 +02:00
Code Issues Releases Wiki Activity

relax concatenation test: text must be equal irrespective of tokenization/joiner

Browse source
This commit is contained in:
Robert Sachunsky 2019-11-29 17:19:56 +01:00
parent 6ee105b17c
commit f829015bb5
1 changed files with 5 additions and 3 deletions
Download patch file Download diff file Expand all files Collapse all files

8
ocrd_repair_inconsistencies/ocrd_repair_inconsistencies.py
View file

@ -126,11 +126,12 @@ def _fix_words(line, page_id, reverse=False):
key=lambda w: Polygon(polygon_from_points(w.get_Coords().points)).centroid.x)
sorted_words_text = get_text(sorted_words, ' ')
if sorted_words_text == line_text:
if (sorted_words_text == line_text or
sorted_words_text.replace(' ', '') == line_text.replace(' ', '')):
LOG.info('Fixing word order of page "%s" line "%s"', page_id, line.id)
line.set_Word(sorted_words)
else:
LOG.debug('Resorting lines of page "%s" region "%s" from %s to %s does not suffice to turn "%s" into "%s"',
LOG.debug('Resorting lines of page "%s" line "%s" from %s to %s does not suffice to turn "%s" into "%s"',
page_id, line.id,
str([word.id for word in words]),
str([word.id for word in sorted_words]),
@ -174,7 +175,8 @@ def _fix_lines(region, page_id, reverse=False):
key=lambda l: Polygon(polygon_from_points(l.get_Coords().points)).centroid.y)
sorted_lines_text = get_text(sorted_lines, '\n')
if sorted_lines_text == region_text:
if (sorted_lines_text == region_text or
sorted_lines_text.replace('\n', '') == region_text.replace('\n', '')):
LOG.info('Fixing line order of page "%s" region "%s"', page_id, region.id)
region.set_TextLine(sorted_lines)
else: