程序员转型练手项目：合同关键信息抽取器（完整代码 + 演示）

#!/usr/bin/env python3
"""合同关键信息抽取器：规则优先，结果落 CSV。仅依赖标准库。"""
import csv
import re
import sys
from pathlib import Path

# 字段名 -> (正则, 捕获组说明)。按合同常见写法设计，可按样本扩充。
RULES = {
    "party_a": r"甲\s*方[：:]\s*([^\n，。；]+)",
    "party_b": r"乙\s*方[：:]\s*([^\n，。；]+)",
    "amount": r"(?:合同)?(?:总)?金额[为：:]*\s*(人民币)?\s*([\d,]+(?:\.\d+)?)\s*元",
    "term": r"(?:合同)?期限[为：:]*\s*([^\n，。；]*?(?:年|月|日)[^\n，。；]*?)(?=[，。；\n])",
    "breach": r"违约责任[：:]\s*([^\n。]*。)",
}


def extract_field(text: str, pattern: str) -> str:
    m = re.search(pattern, text)
    if not m:
        return ""
    groups = [g for g in m.groups() if g]
    return groups[-1].strip() if groups else m.group(0).strip()


def extract_file(path: Path) -> dict:
    text = path.read_text(encoding="utf-8")
    row = {"file": path.name}
    miss = []
    for field, pattern in RULES.items():
        value = extract_field(text, pattern)
        row[field] = value
        if not value:
            miss.append(field)
    row["miss"] = "|".join(miss)
    return row


def main(src_dir: str, out_csv: str):
    rows = [extract_file(p) for p in sorted(Path(src_dir).glob("*.txt"))]
    fields = ["file"] + list(RULES) + ["miss"]
    with open(out_csv, "w", encoding="utf-8-sig", newline="") as f:
        writer = csv.DictWriter(f, fieldnames=fields)
        writer.writeheader()
        writer.writerows(rows)
    hit = sum(1 for r in rows if not r["miss"])
    print(f"共处理 {len(rows)} 份，全字段命中 {hit} 份")


if __name__ == "__main__":
    main(sys.argv[1] if len(sys.argv) > 1 else "contracts",
         sys.argv[2] if len(sys.argv) > 2 else "result.csv")

# make_sample.py：生成两份样例合同用于验证
from pathlib import Path

Path("contracts").mkdir(exist_ok=True)
Path("contracts/c1.txt").write_text(
    "甲方：云启科技有限公司\n乙方：某律师事务所\n"
    "合同总金额为人民币 1,200,000 元。\n"
    "合同期限为两年，自签署之日起算。\n"
    "违约责任：任何一方违约，应向对方支付合同总额 5% 的违约金。\n",
    encoding="utf-8")
Path("contracts/c2.txt").write_text(
    "甲方：恒通制造集团\n乙方：个人开发者\n"
    "合同期限为六个月。\n",
    encoding="utf-8")  # 故意缺金额与违约条款，验证 miss 列
print("样例已生成")

python3 make_sample.py && python3 extractor.py contracts result.csv

共处理 2 份，全字段命中 1 份