feat: 添加文本摘要功能和Git提交信息获取

新增文本摘要功能接口和Git提交信息获取工具
在飞书通知中添加Git提交信息展示
重构标题打分接口以接收JSON格式请求

.gitignore

@@ -0,0 +1 @@
+__pycache__/

app.py

@@ -30,6 +30,8 @@ from utils.detection import (
 from utils.visualization import visualize_detected_tables, plot_results
 from utils.ocr import apply_ocr, save_csv
 from utils.title_scoring import score_article_titles
+from utils.summarization import summarize_text
+from utils.git import get_git_commit_message
 
 
 # 设置日志
@@ -58,10 +60,18 @@ def check_hf_token(token):
         # 向飞书 webhook 发送消息
         if WEB_HOOK:
             try:
+                # 获取Git commit message
+                commit_message = get_git_commit_message()
+
+                # 构建消息内容
+                message = f"你好，{user_info.get('name')}，我已启动"
+                if commit_message:
+                    message += f"\n\n最近的Git提交: {commit_message}"
+
                 webhook_headers = {"Content-Type": "application/json"}
                 webhook_data = {
                     "msg_type": "text",
-                    "content": {"text": f"你好，{user_info.get('name')}，我已启动"},
+                    "content": {"text": message},
                 }
                 webhook_response = requests.post(
                     WEB_HOOK, json=webhook_data, headers=webhook_headers
@@ -505,32 +515,57 @@ async def text_to_speech(
 
 
 @app.post("/text/score")
-async def score_titles(
-    titles: List[str] = Body(..., description="List of article titles to score")
-):
+async def score_titles(request_data: Dict[str, List[str]] = Body(..., description="Object containing titles list")):
     """
     为文章标题列表打分，返回标题和对应的分数。
 
     - **titles**: 文章标题列表
     """
     try:
+        # 获取标题列表
+        titles = request_data.get("titles", [])
+        
         # 调用打分函数
         scores = score_article_titles(titles)
-
+        
         # 组合标题和分数
         results = [
-            {"title": title, "score": score} for title, score in zip(titles, scores)
+            {"title": title, "score": score}
+            for title, score in zip(titles, scores)
         ]
-
+        
         # 按分数倒序排序
         results.sort(key=lambda x: x["score"], reverse=True)
-
+        
         return {"results": results}
     except Exception as e:
         logger.error(f"Title scoring error: {e}")
         raise HTTPException(status_code=500, detail=f"服务器内部错误: {e}")
 
 
+@app.post("/text/summary")
+async def generate_summary(
+    text: str = Body(..., description="Text to summarize"),
+    max_length: int = Body(300, description="Maximum length of the summary"),
+    min_length: int = Body(50, description="Minimum length of the summary")
+):
+    """
+    生成文本摘要。
+
+    - **text**: 需要摘要的文本（建议为英文）
+    - **max_length**: 摘要最大长度（token数）
+    - **min_length**: 摘要最小长度（token数）
+    """
+    try:
+        # 调用摘要函数
+        summary = summarize_text(text, max_length=max_length, min_length=min_length)
+        
+        return {"summary": summary}
+    except Exception as e:
+        logger.error(f"Summarization error: {e}")
+        raise HTTPException(status_code=500, detail=f"服务器内部错误: {e}")
+
+
 from fastapi.staticfiles import StaticFiles
 
 app.mount("/", StaticFiles(directory="static", html=True), name="web")

utils/git.py

@@ -0,0 +1,27 @@
+import os
+import subprocess
+import logging
+
+logger = logging.getLogger(__name__)
+
+def get_git_commit_message():
+    """
+    获取当前Git commit message
+    """
+    try:
+        # 执行git命令获取最近的commit message
+        result = subprocess.run(
+            ['git', 'log', '-1', '--pretty=%B'],
+            capture_output=True,
+            text=True,
+            cwd=os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+        )
+        if result.returncode == 0:
+            commit_message = result.stdout.strip()
+            return commit_message
+        else:
+            logger.error(f"获取Git commit message失败: {result.stderr}")
+            return None
+    except Exception as e:
+        logger.error(f"获取Git commit message出错: {e}")
+        return None

utils/summarization.py

@@ -0,0 +1,22 @@
+from transformers import pipeline
+
+def summarize_text(text, max_length=300, min_length=50):
+    """
+    使用 Falconsai/text_summarization 生成摘要
+    :param text: 输入的文章内容（建议为英文）
+    :param max_length: 摘要最大长度（token数，约等于汉字数*0.5）
+    :param min_length: 摘要最小长度
+    :return: 生成的摘要
+    """
+    # 加载Pipeline
+    summarizer = pipeline("summarization", model="Falconsai/text_summarization")
+    
+    # 生成摘要
+    result = summarizer(
+        text,
+        max_length=max_length,
+        min_length=min_length,
+        do_sample=False
+    )
+    
+    return result[0]['summary_text']