feat: Add language configuration and API validation (#186)

I've added a language configuration file `lang.config.json` to define selectable languages for content generation.

The API endpoints for chat and wiki caching have been updated to validate the selected language against this configuration. Requests with unsupported languages will now be rejected with an appropriate error message.

Key changes:
- Created `api/config/lang.json` with an initial set of supported languages.
- Modified `api/config.py` to load this configuration.
- Implemented language validation in:
  - `api/simple_chat.py` (chat_completions_stream)
  - `api/websocket_wiki.py` (handle_websocket_chat)
  - `api/api.py` (get_cached_wiki, store_wiki_cache, delete_wiki_cache)

Author: szl97

api/api.py

@@ -128,6 +128,10 @@ class AuthorizationConfig(BaseModel):
 
 from api.config import configs, WIKI_AUTH_MODE, WIKI_AUTH_CODE
 
+@app.get("/lang/config")
+async def get_lang_config():
+    return configs["lang_config"]
+
 @app.get("/auth/status")
 async def get_auth_status():
     """
@@ -444,6 +448,11 @@ async def get_cached_wiki(
     """
     Retrieves cached wiki data (structure and generated pages) for a repository.
     """
+    # Language validation
+    supported_langs = configs["lang_config"]["supported_languages"]
+    if not supported_langs.__contains__(language):
+        language = configs["lang_config"]["default"]
+
     logger.info(f"Attempting to retrieve wiki cache for {owner}/{repo} ({repo_type}), lang: {language}")
     cached_data = await read_wiki_cache(owner, repo, repo_type, language)
     if cached_data:
@@ -459,6 +468,12 @@ async def store_wiki_cache(request_data: WikiCacheRequest):
     """
     Stores generated wiki data (structure and pages) to the server-side cache.
     """
+    # Language validation
+    supported_langs = configs["lang_config"]["supported_languages"]
+
+    if not supported_langs.__contains__(request_data.language):
+        request_data.language = configs["lang_config"]["default"]
+
     logger.info(f"Attempting to save wiki cache for {request_data.owner}/{request_data.repo} ({request_data.repo_type}), lang: {request_data.language}")
     success = await save_wiki_cache(request_data)
     if success:
@@ -472,11 +487,16 @@ async def delete_wiki_cache(
     repo: str = Query(..., description="Repository name"),
     repo_type: str = Query(..., description="Repository type (e.g., github, gitlab)"),
     language: str = Query(..., description="Language of the wiki content"),
-    authorization_code: str = Query(..., description="Authorization code")
+    authorization_code: Optional[str] = Query(None, description="Authorization code")
 ):
     """
     Deletes a specific wiki cache from the file system.
     """
+    # Language validation
+    supported_langs = configs["lang_config"]["supported_languages"]
+    if not supported_langs.__contains__(language):
+        raise HTTPException(status_code=400, detail="Language is not supported")
+
     if WIKI_AUTH_MODE:
         logger.info("check the authorization code")
         if WIKI_AUTH_CODE != authorization_code:

api/config.py

@@ -179,6 +179,40 @@ def is_ollama_embedder():
 def load_repo_config():
     return load_json_config("repo.json")
 
+# Load language configuration
+def load_lang_config():
+    default_config = {
+        "supported_languages": {
+            "en": "English",
+            "ja": "Japanese (日本語)",
+            "zh": "Mandarin Chinese (中文)",
+            "es": "Spanish (Español)",
+            "kr": "Korean (한국어)",
+            "vi": "Vietnamese (Tiếng Việt)"
+        },
+        "default": "en"
+    }
+    try:
+        # If environment variable is set, use the directory specified by it
+        if CONFIG_DIR:
+            config_path = Path(CONFIG_DIR) / "lang.json"
+        else:
+            # Otherwise use default directory
+            config_path = Path(__file__).parent / "config" / "lang.json"
+
+        logger.info(f"Loading language configuration from {config_path}")
+
+        if not config_path.exists():
+            logger.warning(f"Language configuration file {config_path} does not exist")
+            return default_config
+        return load_json_config(config_path)
+    except json.JSONDecodeError as e:
+        logger.error(f"Error decoding JSON from language configuration file {config_path}: {str(e)}")
+        return default_config
+    except Exception as e:
+        logger.error(f"Error loading language configuration file {config_path}: {str(e)}")
+        return default_config
+
 # Default excluded directories and files
 DEFAULT_EXCLUDED_DIRS: List[str] = [
     # Virtual environments and package managers
@@ -227,6 +261,7 @@ def load_repo_config():
 generator_config = load_generator_config()
 embedder_config = load_embedder_config()
 repo_config = load_repo_config()
+lang_config = load_lang_config()
 
 # Update configuration
 if generator_config:
@@ -245,6 +280,11 @@ def load_repo_config():
         if key in repo_config:
             configs[key] = repo_config[key]
 
+# Update language configuration
+if lang_config:
+    configs["lang_config"] = lang_config
+
+
 def get_model_config(provider="google", model=None):
     """
     Get configuration for the specified provider and model

api/config/lang.json

@@ -0,0 +1,11 @@
+{
+  "supported_languages": {
+    "en": "English",
+    "ja": "Japanese (日本語)",
+    "zh": "Mandarin Chinese (中文)",
+    "es": "Spanish (Español)",
+    "kr": "Korean (한국어)",
+    "vi": "Vietnamese (Tiếng Việt)"
+  },
+  "default": "en"
+}

api/simple_chat.py

@@ -11,7 +11,7 @@
 from fastapi.responses import StreamingResponse
 from pydantic import BaseModel, Field
 
-from api.config import get_model_config
+from api.config import get_model_config, configs
 from api.data_pipeline import count_tokens, get_file_content
 from api.openai_client import OpenAIClient
 from api.openrouter_client import OpenRouterClient
@@ -245,15 +245,9 @@ async def chat_completions_stream(request: ChatCompletionRequest):
         repo_type = request.type
 
         # Get language information
-        language_code = request.language or "en"
-        language_name = {
-            "en": "English",
-            "ja": "Japanese (日本語)",
-            "zh": "Mandarin Chinese (中文)",
-            "es": "Spanish (Español)",
-            "kr": "Korean (한국어)",
-            "vi": "Vietnamese (Tiếng Việt)"
-        }.get(language_code, "English")
+        language_code = request.language or configs["lang_config"]["default"]
+        supported_langs = configs["lang_config"]["supported_languages"]
+        language_name = supported_langs.get(language_code, "English")
 
         # Create system prompt
         if is_deep_research:

api/websocket_wiki.py

@@ -9,7 +9,7 @@
 from fastapi import WebSocket, WebSocketDisconnect, HTTPException
 from pydantic import BaseModel, Field
 
-from api.config import get_model_config
+from api.config import get_model_config, configs
 from api.data_pipeline import count_tokens, get_file_content
 from api.openai_client import OpenAIClient
 from api.openrouter_client import OpenRouterClient
@@ -245,15 +245,9 @@ async def handle_websocket_chat(websocket: WebSocket):
         repo_type = request.type
 
         # Get language information
-        language_code = request.language or "en"
-        language_name = {
-            "en": "English",
-            "ja": "Japanese (日本語)",
-            "zh": "Mandarin Chinese (中文)",
-            "es": "Spanish (Español)",
-            "kr": "Korean (한국어)",
-            "vi": "Vietnamese (Tiếng Việt)"
-        }.get(language_code, "English")
+        language_code = request.language or configs["lang_config"]["default"]
+        supported_langs = configs["lang_config"]["supported_languages"]
+        language_name = supported_langs.get(language_code, "English")
 
         # Create system prompt
         if is_deep_research:

next.config.ts

@@ -59,6 +59,10 @@ const nextConfig: NextConfig = {
         source: '/api/auth/validate',
         destination: `${TARGET_SERVER_BASE_URL}/auth/validate`,
       },
+      {
+        source: '/api/lang/config',
+        destination: `${TARGET_SERVER_BASE_URL}/lang/config`,
+      },
     ];
   },
 };

src/app/page.tsx

@@ -44,7 +44,7 @@ const DEMO_SEQUENCE_CHART = `sequenceDiagram
 
 export default function Home() {
   const router = useRouter();
-  const { language, setLanguage, messages } = useLanguage();
+  const { language, setLanguage, messages, supportedLanguages } = useLanguage();
   const { projects, isLoading: projectsLoading } = useProcessedProjects();
 
   // Create a simple translation function
@@ -365,6 +365,7 @@ export default function Home() {
             repositoryInput={repositoryInput}
             selectedLanguage={selectedLanguage}
             setSelectedLanguage={setSelectedLanguage}
+            supportedLanguages={supportedLanguages}
             isComprehensiveView={isComprehensiveView}
             setIsComprehensiveView={setIsComprehensiveView}
             provider={provider}

src/components/ConfigurationModal.tsx

@@ -15,6 +15,7 @@ interface ConfigurationModalProps {
   // Language selection
   selectedLanguage: string;
   setSelectedLanguage: (value: string) => void;
+  supportedLanguages: Record<string, string>;
 
   // Wiki type options
   isComprehensiveView: boolean;
@@ -65,6 +66,7 @@ export default function ConfigurationModal({
   repositoryInput,
   selectedLanguage,
   setSelectedLanguage,
+  supportedLanguages,
   isComprehensiveView,
   setIsComprehensiveView,
   provider,
@@ -144,12 +146,9 @@ export default function ConfigurationModal({
                 onChange={(e) => setSelectedLanguage(e.target.value)}
                 className="input-japanese block w-full px-3 py-2 text-sm rounded-md bg-transparent text-[var(--foreground)] focus:outline-none focus:border-[var(--accent-primary)]"
               >
-                <option value="en">English</option>
-                <option value="ja">Japanese (日本語)</option>
-                <option value="zh">Mandarin (中文)</option>
-                <option value="es">Spanish (Español)</option>
-                <option value="kr">Korean (한국어)</option>
-                <option value="vi">Vietnamese (Tiếng Việt)</option>
+                {
+                  Object.entries(supportedLanguages).map(([key, value])=> <option key={key} value={key}>{value}</option>)
+                }
               </select>
             </div>