code/pipeline/prepare_demo_samples.py

from __future__ import annotations

import json
from pathlib import Path


CURRENT_FILE = Path(__file__).resolve()


def _find_inference_root() -> Path:
    for parent in CURRENT_FILE.parents:
        if (parent / "model").exists() and (parent / "code").exists() and (parent / "data").exists():
            return parent
    return CURRENT_FILE.parents[2]


INFERENCE_ROOT = _find_inference_root()
PROJECT_ROOT = INFERENCE_ROOT.parent.parent

SOURCE_FILE = PROJECT_ROOT / "LightXML_B" / "infer_outcome" / "full_v1_epoch9_test50_top10.json"
OUTPUT_FILE = INFERENCE_ROOT / "data" / "input" / "sample_raw_input_5.jsonl"

ZONE_KEYS = [
    "z1_header",
    "z2_facts",
    "z3_issues",
    "z5_result",
    "z6_dissent",
    "z7_other",
]

FALLBACK_SAMPLES = [
    {"id": "sample_1", "text": "原告主张被告未按约还款，请求判令偿还借款本息及相关费用。", "labels": []},
    {"id": "sample_2", "text": "双方就民间借贷关系发生争议，法院审查借条、转账记录后作出判决。", "labels": []},
    {"id": "sample_3", "text": "被告抗辩借款已部分清偿，原告提交证据不足以完全支持其主张。", "labels": []},
    {"id": "sample_4", "text": "本院认为借贷关系成立，逾期利息应按合同约定及法律规定计算。", "labels": []},
    {"id": "sample_5", "text": "判决被告于限定期限内支付本金、利息及案件受理费。", "labels": []},
]


def main() -> None:
    records: list[dict]

    if SOURCE_FILE.exists():
        items = json.loads(SOURCE_FILE.read_text(encoding="utf-8"))[:5]
        records = []
        for i, item in enumerate(items, start=1):
            zones = item.get("input", {})
            text = "\n".join([zones.get(k, "") for k in ZONE_KEYS if zones.get(k, "").strip()])
            records.append(
                {
                    "id": f"sample_{i}",
                    "text": text,
                    "labels": item.get("labels", []),
                }
            )
        print(f"[source] {SOURCE_FILE}")
    else:
        records = FALLBACK_SAMPLES
        print("[source] external infer_outcome not found, use built-in fallback samples")

    OUTPUT_FILE.parent.mkdir(parents=True, exist_ok=True)
    OUTPUT_FILE.write_text(
        "\n".join(json.dumps(record, ensure_ascii=False) for record in records),
        encoding="utf-8",
    )
    print(f"[out] {OUTPUT_FILE}")


if __name__ == "__main__":
    main()