Update irpr/main.py

irpr/main.py

@@ -1,30 +1,25 @@
 from fastapi import FastAPI, HTTPException, Response, UploadFile, File, Request
-from fastapi.responses import HTMLResponse
+from fastapi.responses import HTMLResponse, JSONResponse
 from fastapi.staticfiles import StaticFiles
 from fastapi.middleware.cors import CORSMiddleware
 from fastapi.templating import Jinja2Templates
+import os, traceback
+
 from irpr.models import IngestRequest, GenerateRequest
-import os
 
-app = FastAPI(title="IR/PR Co-Pilot Pro", version="0.2.0")
+app = FastAPI(title="IR/PR Co-Pilot Pro", version="0.3.1")
 
-# --- CORS（同一オリジン運用想定だが念のため全許可） ---
 app.add_middleware(
-    CORSMiddleware,
-    allow_origins=["*"], allow_credentials=True, allow_methods=["*"], allow_headers=["*"],
+    CORSMiddleware, allow_origins=["*"], allow_credentials=True,
+    allow_methods=["*"], allow_headers=["*"],
 )
 
-# --- 静的 & 生成物の公開マウント ---
-os.makedirs("static", exist_ok=True)
-os.makedirs("templates", exist_ok=True)
-os.makedirs("data", exist_ok=True)
-os.makedirs("data/uploads", exist_ok=True)
-
+for d in ["static", "templates", "data", "data/uploads"]:
+    os.makedirs(d, exist_ok=True)
 app.mount("/static", StaticFiles(directory="static"), name="static")
 app.mount("/files", StaticFiles(directory="data"), name="files")
 templates = Jinja2Templates(directory="templates")
 
-# ---------------- UI ----------------
 @app.get("/", response_class=HTMLResponse)
 def ui(request: Request):
     return templates.TemplateResponse("index.html", {"request": request})
@@ -33,73 +28,70 @@ def ui(request: Request):
 def health():
     return {"ok": True, "service": "IR/PR Co-Pilot Pro"}
 
-# ------------- API: 取込 -------------
+# ---- ingest ----
 @app.post("/ingest/edinet")
 def ingest_edinet(req: IngestRequest):
     from rag.ingest import ingest_edinet_for_company
-    n = ingest_edinet_for_company(req.edinet_code, req.date)
-    return {"ingested_chunks": n}
+    try:
+        n = ingest_edinet_for_company(req.edinet_code, req.date)
+        return {"ok": True, "ingested_chunks": n}
+    except Exception as e:
+        return {"ok": False, "error": repr(e)}
 
 @app.post("/ingest/upload")
 async def ingest_upload(files: list[UploadFile] = File(...)):
     from rag.ingest import ingest_pdf_bytes
-    total = 0
-    saved = []
-    for f in files:
-        if not f.filename.lower().endswith(".pdf"):
-            continue
-        blob = await f.read()
-        # 保存してダウンロード可能に
-        save_path = os.path.join("data", "uploads", f.filename)
-        with open(save_path, "wb") as w:
-            w.write(blob)
-        source_url = f"/files/uploads/{f.filename}"
-        total += ingest_pdf_bytes(title=f.filename, source_url=source_url, pdf_bytes=blob)
-        saved.append(source_url)
-    return {"ingested_chunks": total, "saved": saved}
+    total = 0; saved = []
+    try:
+        for f in files:
+            if not f.filename.lower().endswith(".pdf"): continue
+            blob = await f.read()
+            save_path = os.path.join("data", "uploads", f.filename)
+            with open(save_path, "wb") as w: w.write(blob)
+            source_url = f"/files/uploads/{f.filename}"
+            total += ingest_pdf_bytes(title=f.filename, source_url=source_url, pdf_bytes=blob)
+            saved.append(source_url)
+        return {"ok": True, "ingested_chunks": total, "saved": saved}
+    except Exception as e:
+        return {"ok": False, "error": repr(e), "ingested_chunks": total, "saved": saved}
 
-# ------------- API: 生成 -------------
+# ---- generate ----
 @app.post("/generate/all")
 def generate_all(req: GenerateRequest):
-    from generators.summary import make_summary
-    from generators.qa import make_qa
-    from export.ppt import build_deck, save_pptx
-    from export.qa_csv import save_qa_csv
-
-    summary_text, links = make_summary(req.query)
-    sections = {
-        "highlights": _extract_section(summary_text, "業績ハイライト"),
-        "outlook":    _extract_section(summary_text, "見通し"),
-        "segments":   _extract_section(summary_text, "セグメント"),
-        "finance":    _extract_section(summary_text, "財務"),
-        "shareholder":_extract_section(summary_text, "株主還元"),
-        "esg":        _extract_section(summary_text, "ESG"),
-        "risks":      _extract_section(summary_text, "リスク"),
-    }
-    qa_list, links2 = make_qa(req.query, 30)
-    prs = build_deck(sections, links)
-    ppt_path = save_pptx(prs, "data/ir_summary.pptx")
-    csv_path = save_qa_csv(qa_list, links2, "data/qa_30.csv")
-    return {
-        "pptx": ppt_path.replace("data/", "/files/"),
-        "qa_csv": csv_path.replace("data/", "/files/"),
-        "links": links
-    }
+    try:
+        from generators.summary import make_summary
+        from generators.qa import make_qa
+        from export.ppt import build_deck, save_pptx
+        from export.qa_csv import save_qa_csv
 
-# ------------- API: EDINET PDF 直配 -------------
-@app.get("/proxy/edinet/{doc_id}")
-def proxy_edinet(doc_id: str, type: str = "pdf"):
-    if type != "pdf":
-        raise HTTPException(400, "unsupported type")
-    from rag.ingest import download_edinet_pdf
-    pdf = download_edinet_pdf(doc_id)
-    if not pdf:
-        raise HTTPException(404, "file not found")
-    return Response(content=pdf, media_type="application/pdf")
+        summary_text, links = make_summary(req.query)
+        sections = {
+            "highlights": _extract_section(summary_text, "業績ハイライト"),
+            "outlook":    _extract_section(summary_text, "見通し"),
+            "segments":   _extract_section(summary_text, "セグメント"),
+            "finance":    _extract_section(summary_text, "財務"),
+            "shareholder":_extract_section(summary_text, "株主還元"),
+            "esg":        _extract_section(summary_text, "ESG"),
+            "risks":      _extract_section(summary_text, "リスク"),
+        }
+        qa_list, links2 = make_qa(req.query, 30)
+        prs = build_deck(sections, links)
+        ppt_path = save_pptx(prs, "data/ir_summary.pptx")
+        csv_path = save_qa_csv(qa_list, links2, "data/qa_30.csv")
+        return {
+            "ok": True,
+            "pptx": ppt_path.replace("data/", "/files/"),
+            "qa_csv": csv_path.replace("data/", "/files/"),
+            "links": list(dict.fromkeys((links or []) + (links2 or [])))
+        }
+    except Exception as e:
+        # 500にせず詳細を返す（UIで表示）
+        traceback.print_exc()
+        return {"ok": False, "error": repr(e)}
 
-# ------------- helper -------------
 def _extract_section(text: str, head: str):
     import re
-    pat = rf"{head}[:：]\s*(.*?)(?:\n-\s*\S+[:：]|\Z)"
+    # 「見出し: 本文」形式を抽出。無ければ空。
+    pat = rf"{head}[:：]\s*(.*?)(?:\n[^\n]*[:：]|\Z)"
     m = re.search(pat, text, re.S)
     return (m.group(1).strip() if m else "").strip()