kai/invoice-master-poc-v2
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Add payment line parser and fix OCR override from payment_line

Browse Source 
- Add MachineCodeParser for Swedish invoice payment line parsing
- Fix OCR Reference extraction by normalizing account number spaces
- Add cross-validation tests for pipeline and field_extractor
- Update UI layout for compact upload and full-width results

Key changes:
- machine_code_parser.py: Handle spaces in Bankgiro numbers (e.g. "78 2 1 713")
- pipeline.py: OCR and Amount override from payment_line, BG/PG comparison only
- field_extractor.py: Improved invoice number normalization
- app.py: Responsive UI layout changes

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
This commit is contained in:
Yaojia Wang
2026-01-21 21:47:02 +01:00
parent e9460e9f34
commit 4ea4bc96d4
33 changed files with 7530 additions and 562 deletions
Download Patch File Download Diff File Expand all files Collapse all files

30
src/normalize/normalizer.py
View File

@@ -43,8 +43,8 @@ class FieldNormalizer:
# Remove zero-width characters
text = re.sub(r'[\u200b\u200c\u200d\ufeff]', '', text)
# Normalize different dash types to standard hyphen-minus (ASCII 45)
# en-dash (–, U+2013), em-dash (—, U+2014), minus sign (−, U+2212)
text = re.sub(r'[\u2013\u2014\u2212]', '-', text)
# en-dash (–, U+2013), em-dash (—, U+2014), minus sign (−, U+2212), middle dot (·, U+00B7)
text = re.sub(r'[\u2013\u2014\u2212\u00b7]', '-', text)
# Normalize whitespace
text = ' '.join(text.split())
return text.strip()
@@ -571,6 +571,15 @@ class FieldNormalizer:
# Short year with dot separator (e.g., 02.01.26)
eu_dot_short = parsed_date.strftime('%d.%m.%y')
# Short year with slash separator (e.g., 20/10/24) - DD/MM/YY format
eu_slash_short = parsed_date.strftime('%d/%m/%y')
# Short year with hyphen separator (e.g., 23-11-01) - common in Swedish invoices
yy_mm_dd_short = parsed_date.strftime('%y-%m-%d')
# Middle dot separator (OCR sometimes reads hyphens as middle dots)
iso_middot = parsed_date.strftime('%Y·%m·%d')
# Spaced formats (e.g., "2026 01 12", "26 01 12")
spaced_full = parsed_date.strftime('%Y %m %d')
spaced_short = parsed_date.strftime('%y %m %d')
@@ -581,10 +590,23 @@ class FieldNormalizer:
swedish_format_full = f"{parsed_date.day} {month_full} {parsed_date.year}"
swedish_format_abbrev = f"{parsed_date.day} {month_abbrev} {parsed_date.year}"
# Swedish month abbreviation with hyphen (e.g., "30-OKT-24", "30-okt-24")
month_abbrev_upper = month_abbrev.upper()
swedish_hyphen_short = f"{parsed_date.day:02d}-{month_abbrev_upper}-{parsed_date.strftime('%y')}"
swedish_hyphen_short_lower = f"{parsed_date.day:02d}-{month_abbrev}-{parsed_date.strftime('%y')}"
# Also without leading zero on day
swedish_hyphen_short_no_zero = f"{parsed_date.day}-{month_abbrev_upper}-{parsed_date.strftime('%y')}"
# Swedish month abbreviation with short year in different format (e.g., "SEP-24", "30 SEP 24")
month_year_only = f"{month_abbrev_upper}-{parsed_date.strftime('%y')}"
swedish_spaced = f"{parsed_date.day:02d} {month_abbrev_upper} {parsed_date.strftime('%y')}"
variants.extend([
iso, eu_slash, us_slash, eu_dot, iso_dot, compact, compact_short,
eu_dot_short, spaced_full, spaced_short,
swedish_format_full, swedish_format_abbrev
eu_dot_short, eu_slash_short, yy_mm_dd_short, iso_middot, spaced_full, spaced_short,
swedish_format_full, swedish_format_abbrev,
swedish_hyphen_short, swedish_hyphen_short_lower, swedish_hyphen_short_no_zero,
month_year_only, swedish_spaced
])
return list(set(v for v in variants if v))