From 87b024b21f4cea7a7194b8518ef5ba5f8b7cfd32 Mon Sep 17 00:00:00 2001 From: Muhamed Fazal PS Date: Wed, 15 Jul 2026 14:28:42 +0530 Subject: [PATCH 1/2] Fix #80: Correct TYPE_CHECKING import path for Document The import at openagent_eval.providers.base.models does not exist; Document lives at openagent_eval.providers.models. Fixes #80 --- openagent_eval/providers/base/retriever.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/openagent_eval/providers/base/retriever.py b/openagent_eval/providers/base/retriever.py index 2089bac..7289933 100644 --- a/openagent_eval/providers/base/retriever.py +++ b/openagent_eval/providers/base/retriever.py @@ -11,7 +11,7 @@ from typing import TYPE_CHECKING if TYPE_CHECKING: - from openagent_eval.providers.base.models import Document + from openagent_eval.providers.models import Document class Retriever(ABC): From 78c92a4a4e406c5e254784fecccd5bbeafe2f630 Mon Sep 17 00:00:00 2001 From: Muhamed Fazal PS Date: Wed, 15 Jul 2026 15:17:31 +0530 Subject: [PATCH 2/2] Fix #122: Use \w instead of [a-zA-Z] for keyword extraction to support non-ASCII The regex r'\b[a-zA-Z]{4,}\b' only matches ASCII letters, so non-ASCII characters (accented letters, CJK, emoji, etc.) were silently excluded from content gap analysis. Changed to r'\b\w{4,}\b' which matches Unicode word characters in Python 3. Fixes #122 --- openagent_eval/diagnosis/chunking.py | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/openagent_eval/diagnosis/chunking.py b/openagent_eval/diagnosis/chunking.py index dedf887..964ba9c 100644 --- a/openagent_eval/diagnosis/chunking.py +++ b/openagent_eval/diagnosis/chunking.py @@ -159,7 +159,7 @@ def _check_content_gaps( "will", "about", "into", "your", "some", "than", } question_words = set( - w.lower() for w in re.findall(r"\b[a-zA-Z]{4,}\b", question) + w.lower() for w in re.findall(r"\b\w{4,}\b", question) ) - stop_words if not question_words: return issues