feat: security(CORS/API key), OpenAI-compatible endpoint, Paperless hook indexing; .env support

server/main.py

@@ -1,15 +1,31 @@
 from __future__ import annotations
 
-from fastapi import FastAPI, HTTPException
+from fastapi import FastAPI, HTTPException, Depends
+from fastapi.middleware.cors import CORSMiddleware
 from pydantic import BaseModel
 from typing import List, Dict, Any
 
 from .config import settings
 from .ollama_client import OllamaClient
 from .index_store import JsonlIndex
+from .security import require_api_key
+from .paperless_client import PaperlessClient
+from .utils import chunk_text
 
 
-app = FastAPI(title="Local AI Server", version="0.1.0")
+app = FastAPI(title="Local AI Server", version="0.2.1")
+
+# CORS
+import os
+cors_origins = os.getenv("CORS_ORIGINS", "*")
+origins = [o.strip() for o in cors_origins.split(",") if o.strip()] or ["*"]
+app.add_middleware(
+    CORSMiddleware,
+    allow_origins=origins,
+    allow_credentials=True,
+    allow_methods=["*"],
+    allow_headers=["*"],
+)
 ollama = OllamaClient(settings.ollama_host)
 index = JsonlIndex(settings.index_path)
 
@@ -137,8 +153,48 @@ class PaperlessHook(BaseModel):
 
 
 @app.post("/paperless/hook")
-def paperless_hook(hook: PaperlessHook) -> Dict[str, Any]:
-    # NOTE: 확장 지점 - paperless API를 조회하여 문서 텍스트/메타데이터를 받아
-    #       scripts/embed_ollama.py와 동일 로직으로 인덱스를 업데이트할 수 있습니다.
-    return {"status": "ack", "document_id": hook.document_id}
+def paperless_hook(hook: PaperlessHook, _: None = Depends(require_api_key)) -> Dict[str, Any]:
+    # Fetch text from Paperless and upsert into index
+    client = PaperlessClient(settings.paperless_base_url, settings.paperless_token)
+    text = client.get_document_text(hook.document_id)
+    parts = chunk_text(text)
+    model = settings.embedding_model
+    from .index_store import IndexRow
+    to_append = []
+    for i, t in enumerate(parts):
+        vec = ollama.embeddings(model, t)
+        to_append.append(IndexRow(id=f"paperless:{hook.document_id}:{i}", text=t, vector=vec, source="paperless"))
+    added = index.append(to_append)
+    return {"status": "indexed", "document_id": hook.document_id, "chunks": added}
+
+
+# OpenAI-compatible chat completions (minimal)
+class ChatCompletionsRequest(BaseModel):
+    model: str | None = None
+    messages: List[Dict[str, str]]
+    temperature: float | None = None
+    max_tokens: int | None = None
+
+
+@app.post("/v1/chat/completions")
+def chat_completions(req: ChatCompletionsRequest, _: None = Depends(require_api_key)) -> Dict[str, Any]:
+    chosen = req.model or settings.base_model
+    opts: Dict[str, Any] = {}
+    if req.temperature is not None:
+        opts["temperature"] = req.temperature
+    # Note: Ollama ignores max_tokens field; left here for interface similarity
+    resp = ollama.chat(chosen, req.messages, stream=False, options=opts)
+    # Minimal OpenAI-like response shape
+    return {
+        "id": "chatcmpl-local",
+        "object": "chat.completion",
+        "model": chosen,
+        "choices": [
+            {
+                "index": 0,
+                "message": resp.get("message", {"role": "assistant", "content": resp.get("response", "")}),
+                "finish_reason": resp.get("done_reason", "stop"),
+            }
+        ],
+    }