freedak
641e33b834
- 语义检索(FAISS + embedding)+ 精确查找(法规名+条号) - RAG 问答(SSE 流式,支持 thinking 折叠显示) - 法规浏览(原文阅读) - 历史记录(检索+对话持久化到 SQLite) - 设置页(系统提示词/模板/LLM 参数可配置) - 检索质量评估脚本 Generated with [Devin](https://devin.ai) Co-Authored-By: Devin <158243242+devin-ai-integration[bot]@users.noreply.github.com>
99 lines
3.2 KiB
Python
99 lines
3.2 KiB
Python
"""切片解析单测"""
|
|
import sys
|
|
from pathlib import Path
|
|
|
|
sys.path.insert(0, str(Path(__file__).parent.parent))
|
|
|
|
from scripts.parse_clause import (
|
|
parse_law_file,
|
|
extract_name_and_date,
|
|
load_region_mapping,
|
|
scan_category_dir,
|
|
)
|
|
|
|
|
|
# 测试数据路径(相对于项目根)
|
|
PACK_DIR = Path(__file__).parent.parent.parent / "law-pack-2026-07-01-markdown"
|
|
|
|
|
|
def test_extract_name_and_date():
|
|
"""测试文件名解析"""
|
|
name, date = extract_name_and_date("中华人民共和国民法典_20200528.md")
|
|
assert name == "中华人民共和国民法典"
|
|
assert date == "2020-05-28"
|
|
|
|
name, date = extract_name_and_date("不动产登记暂行条例_20240310.md")
|
|
assert name == "不动产登记暂行条例"
|
|
assert date == "2024-03-10"
|
|
|
|
# 无日期后缀
|
|
name, date = extract_name_and_date("某法规.md")
|
|
assert name == "某法规"
|
|
assert date is None
|
|
|
|
|
|
def test_parse_civil_code():
|
|
"""测试民法典切片"""
|
|
f = PACK_DIR / "法律" / "中华人民共和国民法典_20200528.md"
|
|
if not f.exists():
|
|
return # 数据不存在时跳过
|
|
clauses = parse_law_file(f, "法律")
|
|
assert len(clauses) > 1000 # 民法典 1260 条
|
|
assert clauses[0].clause_no == "第一条"
|
|
assert clauses[0].law_name == "中华人民共和国民法典"
|
|
assert clauses[0].category == "法律"
|
|
assert clauses[0].publish_date == "2020-05-28"
|
|
assert "保护民事主体" in clauses[0].content
|
|
|
|
|
|
def test_parse_admin_regulation():
|
|
"""测试行政法规切片"""
|
|
f = PACK_DIR / "行政法规" / "不动产登记暂行条例_20240310.md"
|
|
if not f.exists():
|
|
return
|
|
clauses = parse_law_file(f, "行政法规")
|
|
assert len(clauses) > 0
|
|
assert clauses[0].category == "行政法规"
|
|
|
|
|
|
def test_parse_local_regulation_with_mapping():
|
|
"""测试地方性法规 + 区域映射"""
|
|
mapping = load_region_mapping(PACK_DIR / "地方性法规区域映射.json")
|
|
f = PACK_DIR / "地方性法规" / "七台河市倭肯河流域水环境保护条例_20211224.md"
|
|
if not f.exists():
|
|
return
|
|
clauses = parse_law_file(f, "地方性法规", mapping)
|
|
assert len(clauses) > 0
|
|
assert clauses[0].province == "黑龙江省"
|
|
assert clauses[0].city == "七台河市"
|
|
assert clauses[0].region_level == "市级"
|
|
|
|
|
|
def test_null_mapping_skipped():
|
|
"""测试 null 映射文件跳过"""
|
|
mapping = load_region_mapping(PACK_DIR / "地方性法规区域映射.json")
|
|
f = PACK_DIR / "地方性法规" / "~$壮族自治区巴马盘阳河流域生态环境保护条例_20150527.md"
|
|
if not f.exists():
|
|
return
|
|
clauses = parse_law_file(f, "地方性法规", mapping)
|
|
assert len(clauses) == 0 # null 映射应跳过
|
|
|
|
|
|
if __name__ == "__main__":
|
|
# 手动运行(无 pytest 时)
|
|
tests = [
|
|
test_extract_name_and_date,
|
|
test_parse_civil_code,
|
|
test_parse_admin_regulation,
|
|
test_parse_local_regulation_with_mapping,
|
|
test_null_mapping_skipped,
|
|
]
|
|
for t in tests:
|
|
try:
|
|
t()
|
|
print(f"✅ {t.__name__} 通过")
|
|
except AssertionError as e:
|
|
print(f"❌ {t.__name__} 失败: {e}")
|
|
except Exception as e:
|
|
print(f"⚠️ {t.__name__} 跳过: {e}")
|