• 李文光's avatar
    fix(ai): LLM 直连超时/异常时回退本地结构化,不再 500 · c77014f9
    李文光 authored
    generate_jd_draft / analyze_resume / parse_pasted_jd / generate_v2_analysis 在调用 _chat_json
    超时(如 'Request timed out.')时原先会直接抛异常,导致 JD 生成/简历分析/粘贴解析/富报告
    报错。现在统一捕获 LLM 调用异常并回退本地结构化/规则/小节切分,补充超时回退单测。
    c77014f9
Name
Last commit
Last update
..
conftest.py Loading commit data...
test_ai.py Loading commit data...
test_auth.py Loading commit data...
test_external_claims.py Loading commit data...
test_ingest.py Loading commit data...
test_jd_import.py Loading commit data...
test_pool.py Loading commit data...
test_recruiting_rules_agent.py Loading commit data...
test_reports.py Loading commit data...
test_resume_parser.py Loading commit data...
test_resume_parser_llm.py Loading commit data...
test_risk_profile.py Loading commit data...
test_state.py Loading commit data...
test_static_security.py Loading commit data...