Dify 企业级实验（10）：知识库持续更新闭环——数据飞轮怎么转起来？

- variable: title      # 文本，必填，文档标题

- variable: content    # 段落，必填，文档内容（最长 10000）

- variable: base_url   # 文本，必填，Dify 服务地址

def main(title: str, content: str) -> dict:

    import re

    text = content or ""

    # 完整性 40：长度

    score_complete = 40 if len(text) >= 150 else (20 if len(text) >= 80 else 0)

    # 格式 30：含 markdown 标题

    score_format = 30 if re.search(r"^#{1,3} ", text, re.M) else (15 if text.count("\n") >= 3 else 0)

    # 自包含 30：含具体数字/规格/明确陈述

    has_specific = bool(re.search(r"\d+", text)) or any(k in text for k in ["支持", "可以", "是", "提供", "需"])

    score_self = 30 if has_specific else 10

    score = score_complete + score_format + score_self

    valid = "true" if score >= 80 else "false"

    reasons = []

    if score_complete < 40: reasons.append("内容过短（<200字）")

    if score_format < 30: reasons.append("缺少标题结构")

    if score_self < 30: reasons.append("缺少具体规格/明确陈述")

    return {"score": str(score), "valid": valid,

            "reasons": "；".join(reasons) if reasons else "质量合格",

            "summary": f"标题：{title or '无'} | 评分 {score}/100"}

def main(title: str, content: str) -> dict:

    import json

    payload = {

        "name": title or "untitled.md",

        "text": content or "",

        "indexing_technique": "high_quality",

        "process_rule": {

            "mode": "custom",

            "rules": {

                "pre_processing_rules": [{"id": "remove_extra_spaces", "enabled": True}],

                "segmentation": {"separator": "\n\n", "max_tokens": 500}

            }

        }

    }

    return {"payload": json.dumps(payload, ensure_ascii=False)}

url: "{{#start.base_url#}}/v1/datasets/dfac575f-2490-4e1f-b730-76ccc8463c23/document/create_by_text"

method: POST

body: "{{#cd_body.payload#}}"   # JSON 字符串，Content-Type: application/json