fix:优化项目内容

backend/app/services/ai_service.py

@@ -3,7 +3,7 @@ import json
 import os
 import re
 from pathlib import Path
-from typing import Any, Dict, Tuple
+from typing import Any, Dict, Tuple, List
 
 from openai import AsyncOpenAI
 from openai import NotFoundError as OpenAINotFoundError
@@ -197,6 +197,65 @@ async def extract_invoice_metadata(image_bytes: bytes, mime: str = "image/jpeg")
     api_key = (config.get("api_key") or "").strip()
     if not api_key:
         return (None, None)
+
+
+async def extract_finance_tags(
+    content_text: str,
+    doc_type: str,
+    filename: str = "",
+) -> Tuple[List[str], Dict[str, Any]]:
+    """
+    从附件文本内容中抽取标签与结构化信息（JSON）。
+    返回 (tags, meta)。
+    """
+    config = _load_ai_config()
+    client = _client_from_config(config)
+    model = config.get("model_name") or "gpt-4o-mini"
+    temperature = float(config.get("temperature", 0.2))
+
+    prompt = (
+        "你是一名财务助理。请根据附件的文本内容，为它生成可检索的标签，并抽取关键字段。\n"
+        "只返回 JSON，不要任何解释文字。\n"
+        "输入信息：\n"
+        f"- 类型 doc_type: {doc_type}\n"
+        f"- 文件名 filename: {filename}\n"
+        "- 附件文本 content_text: (见下)\n\n"
+        "返回 JSON 格式：\n"
+        "{\n"
+        '  "tags": ["标签1","标签2"],\n'
+        '  "meta": {\n'
+        '    "counterparty": "对方单位/收款方/付款方（如能识别）或 null",\n'
+        '    "account": "账户/卡号后四位（如能识别）或 null",\n'
+        '    "amount": "金额数字字符串或 null",\n'
+        '    "date": "YYYY-MM-DD 或 null",\n'
+        '    "summary": "一句话摘要"\n'
+        "  }\n"
+        "}\n\n"
+        "content_text:\n"
+        f"{content_text[:12000]}\n"
+    )
+
+    completion = await client.chat.completions.create(
+        model=model,
+        response_format={"type": "json_object"},
+        messages=[{"role": "user", "content": prompt}],
+        temperature=temperature,
+        max_tokens=500,
+    )
+    content = completion.choices[0].message.content or "{}"
+    try:
+        data: Any = json.loads(content)
+    except Exception:
+        return ([], {"summary": "", "raw": content})
+
+    tags = data.get("tags") if isinstance(data, dict) else None
+    meta = data.get("meta") if isinstance(data, dict) else None
+    if not isinstance(tags, list):
+        tags = []
+    tags = [str(t).strip() for t in tags if str(t).strip()][:12]
+    if not isinstance(meta, dict):
+        meta = {}
+    return (tags, meta)
     try:
         client = _client_from_config(config)
         model = config.get("model_name") or "gpt-4o-mini"