Skip to content

AI turns documents or topics into real, native PowerPoint decks—with native shapes, transitions and animations, data-backed charts and tables on demand, audio narration from speaker notes, and support for your own .pptx templates. · by Hugo He

fix(image-search): required terms match whole Latin words

hugohe3d4ac79dfc0cf
diff --git a/skills/ppt-master/scripts/image_sources/provider_common.py b/skills/ppt-master/scripts/image_sources/provider_common.py
index d652d8af..8e4fdb32 100644
--- a/skills/ppt-master/scripts/image_sources/provider_common.py
+++ b/skills/ppt-master/scripts/image_sources/provider_common.py
@@ -386,6 +386,19 @@ def _normalize_match_text(text: str) -> str:
return _MATCH_SEPARATOR_RE.sub(" ", lowered).strip()
+def _contains_term(term: str, text: str) -> bool:
+ """Match one normalized term: whole words for Latin, spacing-free for CJK.
+
+ "iran" must not pass on "tirana"; "liberation monument" still matches a
+ run-together "liberationmonument" slug, and CJK ignores inserted spaces.
+ """
+ compact_term = term.replace(" ", "")
+ if term.isascii():
+ forms = "|".join(re.escape(form) for form in {term, compact_term})
+ return re.search(rf"(?<![a-z0-9])(?:{forms})(?![a-z0-9])", text) is not None
+ return term in text or compact_term in text.replace(" ", "")
+
+
def _term_group_alternatives(term_group: str) -> list[str]:
"""Split one required term group into alternatives.
@@ -414,16 +427,12 @@ def missing_required_terms(
return []
text = _normalize_match_text(_candidate_text(candidate))
- compact_text = text.replace(" ", "")
missing: list[str] = []
for group in required_terms:
alternatives = _term_group_alternatives(group)
if not alternatives:
continue
- matched = any(
- alt in text or alt.replace(" ", "") in compact_text
- for alt in alternatives
- )
+ matched = any(_contains_term(alt, text) for alt in alternatives)
if not matched:
missing.append(str(group))
return missing
@@ -480,13 +489,9 @@ def score_candidate(candidate: AssetCandidate, request: ImageSearchRequest) -> f
score = relevance * 10000.0
title_text = _normalize_match_text(candidate.title)
- compact_title = title_text.replace(" ", "")
for group in request.required_terms or ():
alternatives = _term_group_alternatives(group)
- if any(
- alt in title_text or alt.replace(" ", "") in compact_title
- for alt in alternatives
- ):
+ if any(_contains_term(alt, title_text) for alt in alternatives):
score += 1500.0
# Penalize infrastructure/transit metadata if the user didn't explicitly ask for it.
diff --git a/skills/ppt-master/scripts/tests/test_image_search.py b/skills/ppt-master/scripts/tests/test_image_search.py
index a9205ce7..4b14a7f4 100644
--- a/skills/ppt-master/scripts/tests/test_image_search.py
+++ b/skills/ppt-master/scripts/tests/test_image_search.py
@@ -257,3 +257,16 @@ class InfrastructurePenaltyTests(unittest.TestCase):
self._score("team", "team office"),
)
self.assertGreater(self._score("bus station", "team bus station"), 0)
+
+
+class RequiredTermsGateTests(unittest.TestCase):
+ def test_latin_terms_match_whole_words_and_cjk_ignores_spacing(self) -> None:
+ from image_sources.provider_common import AssetCandidate, missing_required_terms
+
+ def missing(title: str, terms: tuple[str, ...]) -> list[str]:
+ return missing_required_terms(AssetCandidate(provider="test", title=title), terms)
+
+ self.assertEqual(missing("Tirana, Albania", ("Iran",)), ["Iran"])
+ self.assertEqual(missing("Tehran, Iran", ("Iran",)), [])
+ self.assertEqual(missing("LiberationMonument at night", ("Liberation Monument",)), [])
+ self.assertEqual(missing("重庆 解放 碑", ("解放碑",)), [])