feat: Editor + read pdf/docx...

metagpt/utils/common.py

@@ -12,6 +12,7 @@
 from __future__ import annotations
 
 import ast
+import asyncio
 import base64
 import contextlib
 import csv
@@ -852,7 +853,10 @@ async def get_mime_type(filename: str | Path, force_read: bool = False) -> str:
     }
 
     try:
-        stdout, _, _ = await shell_execute(f"file --mime-type {str(filename)}")
+        stdout, stderr, _ = await shell_execute(f"file --mime-type {str(filename)}")
+        if stderr:
+            logger.debug(f"file:{filename}, error:{stderr}")
+            return guess_mime_type
         ix = stdout.rfind(" ")
         mime_type = stdout[ix:].strip()
         if mime_type == "text/plain" and guess_mime_type in text_set:
@@ -1050,6 +1054,32 @@ def tool2name(cls, methods: List[str], entry) -> Dict[str, Any]:
     return mappings
 
 
+def run_coroutine_sync(coroutine, *args, **kwargs):
+    """
+    Runs a coroutine function synchronously by encapsulating its invocation as a non-coroutine function call.
+
+    Args:
+        coroutine: The coroutine function to be encapsulated.
+        *args: Positional arguments to be passed to the coroutine.
+        **kwargs: Keyword arguments to be passed to the coroutine.
+
+    Returns:
+        The return value of the coroutine.
+    """
+    try:
+        loop = asyncio.get_running_loop()
+    except RuntimeError:  # No running event loop
+        loop = None
+
+    if loop and loop.is_running():
+        # The event loop is already running
+        future = asyncio.run_coroutine_threadsafe(coroutine(*args, **kwargs), loop)
+        return future.result()
+    else:
+        # The event loop is not running
+        return asyncio.run(coroutine(*args, **kwargs))
+
+
 def log_time(method):
     """A time-consuming decorator for printing execution duration."""
 

metagpt/utils/repo_to_markdown.py

@@ -7,7 +7,7 @@ from __future__ import annotations
 
 import re
 from pathlib import Path
-from typing import Tuple
+from typing import Tuple, Union
 
 from gitignore_parser import parse_gitignore
 
@@ -82,7 +82,7 @@ async def _write_files(repo_path, gitignore_rules=None) -> str:
 
 
 async def _write_file(filename: Path, repo_path: Path) -> str:
-    is_text, mime_type = await _is_text_file(filename)
+    is_text, mime_type = await is_text_file(filename)
     if not is_text:
         logger.info(f"Ignore content: {filename}")
         return ""
@@ -100,7 +100,17 @@ async def _write_file(filename: Path, repo_path: Path) -> str:
         return ""
 
 
-async def _is_text_file(filename: Path) -> Tuple[bool, str]:
+async def is_text_file(filename: Union[str, Path]) -> Tuple[bool, str]:
+    """
+    Determines if the specified file is a text file based on its MIME type.
+
+    Args:
+        filename (Union[str, Path]): The path to the file.
+
+    Returns:
+        Tuple[bool, str]: A tuple where the first element indicates if the file is a text file
+        (True for text file, False otherwise), and the second element is the MIME type of the file.
+    """
     pass_set = {
         "application/json",
         "application/vnd.chipnuts.karaoke-mmd",
@@ -129,7 +139,7 @@ async def _is_text_file(filename: Path) -> Tuple[bool, str]:
         "image/vnd.microsoft.icon",
         "video/mp4",
     }
-    mime_type = await get_mime_type(filename, force_read=True)
+    mime_type = await get_mime_type(Path(filename), force_read=True)
     v = "text/" in mime_type or mime_type in pass_set
     if v:
         return True, mime_type