feat(cli): accept json5 config.

Author: Davidyz

docs/cli.md

@@ -202,6 +202,14 @@ default global configuration will be used. If `.git/` does not exist, VectorCode
 falls back to using the current working directory as the _project root_.
 
 ### Configuring VectorCode
+Since 0.6.4, VectorCode adapted a [json5 parser](https://github.com/dpranke/pyjson5) 
+for loading configuration. VectorCode will now look for `config.json5` in
+configuration directories, and if it doesn't find one, it'll look for
+`config.json` too. Regardless of the filename extension, the json5 syntax will
+be accepted. This allows you to leave trailing comma in the config file, as well
+as writing comments (`//`). This can be very useful if you're experimenting with
+the configs.
+
 The JSON configuration file may hold the following values:
 - `embedding_function`: string, one of the embedding functions supported by [Chromadb](https://www.trychroma.com/) 
   (find more [here](https://docs.trychroma.com/docs/embeddings/embedding-functions) and 

pyproject.toml

@@ -19,6 +19,7 @@ dependencies = [
     "wheel<0.46.0",
     "colorlog",
     "charset-normalizer>=3.4.1",
+    "json5",
 ]
 requires-python = ">=3.11,<3.14"
 readme = "README.md"

src/vectorcode/cli_utils.py

@@ -1,7 +1,6 @@
 import argparse
 import atexit
 import glob
-import json
 import logging
 import os
 import sys
@@ -11,6 +10,7 @@
 from pathlib import Path
 from typing import Any, Optional, Sequence, Union
 
+import json5
 import shtab
 
 from vectorcode import __version__
@@ -415,9 +415,17 @@ async def load_config_file(path: Optional[PathLike] = None):
     if os.path.isfile(path):
         logger.debug(f"Loading config from {path}")
         with open(path) as fin:
-            config = json.load(fin)
-        expand_envs_in_dict(config)
-        return await Config.import_from(config)
+            content = fin.read()
+        if content:
+            config = json5.loads(content)
+            if isinstance(config, dict):
+                expand_envs_in_dict(config)
+                return await Config.import_from(config)
+            else:
+                logger.error("Invalid configuration format!")
+                raise ValueError("Invalid configuration format!")
+        else:
+            logger.debug("Skipping empty json file.")
     else:
         logger.warning("Loading default config.")
     return Config()
@@ -464,10 +472,14 @@ async def get_project_config(project_root: PathLike) -> Config:
     """
     if not os.path.isabs(project_root):
         project_root = os.path.abspath(project_root)
-    local_config_path = os.path.join(project_root, ".vectorcode", "config.json")
-    if os.path.isfile(os.path.join(project_root, ".vectorcode", "config.json")):
-        config = await load_config_file(local_config_path)
-    else:
+    exts = ("json5", "json")
+    config = None
+    for ext in exts:
+        local_config_path = os.path.join(project_root, ".vectorcode", f"config.{ext}")
+        if os.path.isfile(local_config_path):
+            config = await load_config_file(local_config_path)
+            break
+    if config is None:
         config = await load_config_file()
     config.project_root = project_root
     return config

src/vectorcode/lsp_main.py

@@ -22,7 +22,7 @@
     cleanup_path,
     config_logging,
     find_project_root,
-    load_config_file,
+    get_project_config,
     parse_cli_args,
 )
 from vectorcode.common import get_client, get_collection, try_server
@@ -37,10 +37,7 @@
 async def make_caches(project_root: str):
     assert os.path.isabs(project_root)
     if cached_project_configs.get(project_root) is None:
-        config_file = os.path.join(project_root, ".vectorcode", "config.json")
-        if not os.path.isfile(config_file):
-            config_file = None
-        cached_project_configs[project_root] = await load_config_file(config_file)
+        cached_project_configs[project_root] = await get_project_config(project_root)
     config = cached_project_configs[project_root]
     config.project_root = project_root
     host, port = config.host, config.port

src/vectorcode/mcp_main.py

@@ -151,9 +151,7 @@ async def mcp_server():
         logger.info("Found project config: %s", local_config_dir)
         project_root = str(Path(local_config_dir).parent.resolve())
 
-        default_config = await load_config_file(
-            os.path.join(project_root, ".vectorcode", "config.json")
-        )
+        default_config = await get_project_config(project_root)
         default_config.project_root = project_root
         default_client = await get_client(default_config)
         try:

tests/test_cli_utils.py

@@ -1,4 +1,3 @@
-import json
 import os
 import tempfile
 from typing import Any, Dict
@@ -201,7 +200,18 @@ async def test_load_config_file_invalid_json():
         with open(config_path, "w") as f:
             f.write("invalid json")
 
-        with pytest.raises(json.JSONDecodeError):
+        with pytest.raises(ValueError):
+            await load_config_file(config_path)
+
+
+@pytest.mark.asyncio
+async def test_load_config_file_invalid_config():
+    with tempfile.TemporaryDirectory() as temp_dir:
+        config_path = os.path.join(temp_dir, "config.json")
+        with open(config_path, "w") as f:
+            f.write('"hello world"')
+
+        with pytest.raises(ValueError):
             await load_config_file(config_path)
 
 
@@ -225,8 +235,7 @@ async def test_load_config_file_empty_file():
         with open(config_path, "w") as f:
             f.write("")
 
-        with pytest.raises(json.JSONDecodeError):
-            await load_config_file(config_path)
+        assert await load_config_file(config_path) == Config()
 
 
 @pytest.mark.asyncio
@@ -399,8 +408,22 @@ async def test_get_project_config_local_config(tmp_path):
     assert config.host == "test_host"
     assert config.port == 9999
 
-    # Clean up the temporary directory
-    # shutil.rmtree(tmp_path)  # Use tmp_path fixture, no need to remove manually
+
+@pytest.mark.asyncio
+async def test_get_project_config_local_config_json5(tmp_path):
+    # Create a temporary directory and a .vectorcode subdirectory
+    project_root = tmp_path / "project"
+    vectorcode_dir = project_root / ".vectorcode"
+    vectorcode_dir.mkdir(parents=True)
+
+    # Create a config.json file inside .vectorcode with some custom settings
+    config_file = vectorcode_dir / "config.json5"
+    config_file.write_text('{"host": "test_host", "port": 9999}')
+
+    # Call get_project_config and check if it returns the custom settings
+    config = await get_project_config(project_root)
+    assert config.host == "test_host"
+    assert config.port == 9999
 
 
 def test_find_project_root_file_input(tmp_path):

tests/test_lsp.py

@@ -48,16 +48,16 @@ async def test_make_caches(tmp_path):
 
     with (
         patch(
-            "vectorcode.lsp_main.load_config_file", new_callable=AsyncMock
-        ) as mock_load_config_file,
+            "vectorcode.lsp_main.get_project_config", new_callable=AsyncMock
+        ) as mock_get_project_config,
         patch(
             "vectorcode.lsp_main.try_server", new_callable=AsyncMock
         ) as mock_try_server,
     ):
         mock_try_server.return_value = True
         await make_caches(project_root)
 
-        mock_load_config_file.assert_called_once()
+        mock_get_project_config.assert_called_once_with(project_root)
         assert project_root in cached_project_configs
 
 
@@ -69,7 +69,7 @@ async def test_make_caches_server_unavailable(tmp_path):
     config_file.write_text('{"host": "test_host", "port": 9999}')
 
     with (
-        patch("vectorcode.lsp_main.load_config_file", new_callable=AsyncMock),
+        patch("vectorcode.lsp_main.get_project_config", new_callable=AsyncMock),
         patch(
             "vectorcode.lsp_main.try_server", new_callable=AsyncMock
         ) as mock_try_server,