Update tools/read_file.py

tools/read_file.py

@@ -1,87 +1,41 @@
 import os
 import requests
 import tempfile
-from io import BytesIO
+import pandas as pd
+from io import StringIO, BytesIO
 
 def read_file(file_source: str) -> str:
     """
     Read a file from a URL or local path and return its content as text.
-    Supports CSV, Excel (.xlsx), and plain text files. For other formats (PDF, Word, images, audio)
-    use specialized tools. This tool is intended for tabular data and code/text files.
-    
-    Args:
-        file_source: A URL (http:// or https://) or a local file path.
-    
-    Returns:
-        The file content as text, or an error message.
+    Supports CSV, Excel (.xlsx), and plain text files.
     """
     try:
-        # 判断是否为 URL
         is_url = file_source.startswith(("http://", "https://"))
-        content = None
-        
         if is_url:
-            # 下载文件到临时文件
-            response = requests.get(file_source, timeout=15)
+            headers = {"Authorization": f"Bearer {os.getenv('HF_TOKEN')}"} if "huggingface.co" in file_source else {}
+            response = requests.get(file_source, headers=headers, timeout=15)
             response.raise_for_status()
-            # 根据 Content-Type 或 URL 后缀决定处理方式
             content_type = response.headers.get('Content-Type', '').lower()
-            # 保存到临时文件以便用 pandas 等库读取
-            suffix = ""
             if 'excel' in content_type or file_source.endswith('.xlsx'):
-                suffix = ".xlsx"
+                df = pd.read_excel(BytesIO(response.content))
+                return df.to_csv(index=False)[:8000]
             elif 'csv' in content_type or file_source.endswith('.csv'):
-                suffix = ".csv"
+                df = pd.read_csv(StringIO(response.text))
+                return df.to_csv(index=False)[:8000]
             else:
-                suffix = ".txt"
-            with tempfile.NamedTemporaryFile(delete=False, suffix=suffix) as tmp:
-                tmp.write(response.content)
-                tmp_path = tmp.name
-            # 使用本地读取逻辑处理临时文件
-            content = _read_local_file(tmp_path)
-            os.unlink(tmp_path)
+                return response.text[:8000]
         else:
-            # 本地文件路径
             if not os.path.exists(file_source):
                 return f"File not found: {file_source}"
-            content = _read_local_file(file_source)
-        
-        return content[:8000] if content else "File is empty or unsupported type."
-    
+            ext = os.path.splitext(file_source)[1].lower()
+            if ext == ".xlsx":
+                df = pd.read_excel(file_source)
+                return df.to_csv(index=False)[:8000]
+            elif ext == ".csv":
+                df = pd.read_csv(file_source)
+                return df.to_csv(index=False)[:8000]
+            else:
+                with open(file_source, "r", encoding="utf-8", errors="replace") as f:
+                    return f.read(8000)
     except Exception as e:
-        return f"Error reading file: {str(e)}"
-
-def _read_local_file(file_path: str) -> str:
-    """内部函数：读取本地文件，支持 CSV、Excel 和纯文本"""
-    ext = os.path.splitext(file_path)[1].lower()
-    
-    # Excel 文件
-    if ext == ".xlsx":
-        try:
-            import pandas as pd
-            # 读取所有工作表
-            sheets = pd.read_excel(file_path, sheet_name=None)
-            if not sheets:
-                return "No sheets found in Excel file."
-            # 将每个工作表转为 CSV 文本
-            result_parts = []
-            for sheet_name, df in sheets.items():
-                result_parts.append(f"## Sheet: {sheet_name}\n{df.to_csv(index=False)}")
-            return "\n\n".join(result_parts)
-        except ImportError:
-            return "pandas not installed. Cannot read Excel file."
-    
-    # CSV 文件
-    if ext == ".csv":
-        try:
-            import pandas as pd
-            df = pd.read_csv(file_path)
-            return df.to_csv(index=False)
-        except ImportError:
-            # 如果 pandas 不可用，降级为普通文本读取
-            with open(file_path, "r", encoding="utf-8", errors="replace") as f:
-                return f.read()
-    
-    # 其他文本文件（包括 .txt, .py, .json, .md 等）
-    with open(file_path, "r", encoding="utf-8", errors="replace") as f:
-        return f.read()
+        return f"Error reading file: {str(e)}"