Fix: don't pass raw text on when a zone transform fails

A date or numeric transform that found nothing returned the OCR text, so the
field got junk. Return nothing and report transform_failed, which also blocks
a combined field.
This commit is contained in:
stumpylog committed 2026-11-19 08:19:07 -08:00
1 parent fcc465ec2e
commit 0e7cbda3d9
2 files changed
+105 -21

No files matched your search

+17 -8
View File
@@ -276,7 +276,13 @@ def _process_template(
tmp_path,
)
extracted = extraction.value
reason = None if extracted else "no_value"
if extracted:
reason = None
elif extraction.raw_text:
# Text was read but the transform found nothing usable in it.
reason = "transform_failed"
else:
reason = "no_value"
if (
extracted is not None
and zone.validation_regex
@@ -558,12 +564,11 @@ def _extract_zone(
return ZoneExtraction(
raw_text=raw_text,
value=_apply_transform(
raw_text,
raw_text or "",
zone.transform,
_date_rules(zone.ocr_language, zone.date_format),
)
if raw_text
else None,
or None,
)
@@ -687,7 +692,11 @@ def _apply_transform(
transform: str,
dates: DateRules,
) -> str:
"""Apply post-processing transform to extracted text."""
"""Apply post-processing transform to extracted text.
A date or numeric transform that finds no date or digit returns an empty
string instead of passing the raw text on.
"""
text = text.strip()
if not text:
return text
@@ -696,14 +705,14 @@ def _apply_transform(
return text
elif transform == "date":
parsed = _parse_date(text, dates)
return parsed.isoformat() if parsed else text
return parsed.isoformat() if parsed else ""
elif transform == "uppercase":
return text.upper()
elif transform == "lowercase":
return text.lower()
elif transform == "numeric":
result = re.sub(r"[^\d.,\-]", "", text)
return result if result else text
numeric = re.sub(r"[^\d.,\-]", "", text)
return numeric if re.search(r"\d", numeric) else ""
elif transform == "strip_punctuation":
return text.strip(string.punctuation + " \t\r\n")
return text