Update ui/ui_app.py

ui/ui_app.py

@@ -1,55 +1,234 @@
-# ui/ui_app.py の run_analyze を丸ごと置換
-def run_analyze(company: str, use_vision: bool, files: list[str], force_ocr: bool = False):
+# ui/ui_app.py
+from __future__ import annotations
+import os, io, json, base64, traceback
+import gradio as gr
+import pandas as pd
+import plotly.graph_objects as go
+
+from core.extract import parse_pdf, ExtractError
+from core.scoring import score_company
+from core.external_scoring import score_external_from_df
+from core.ai_judgement import make_ai_memo
+
+# ================= 共通ユーティリティ =================
+
+def _b64(img_bytes):
+    return base64.b64encode(img_bytes).decode("utf-8")
+
+def fin_to_df(fin):
+    rows = []
+    def add(cat, d):
+        for k, v in (d or {}).items():
+            rows.append({"category": cat, "item": k, "value": v})
+    add("balance_sheet", fin.get("balance_sheet"))
+    add("income_statement", fin.get("income_statement"))
+    add("cash_flows", fin.get("cash_flows"))
+    return pd.DataFrame(rows, columns=["category", "item", "value"])
+
+def df_to_fin(df):
+    out = {"balance_sheet": {}, "income_statement": {}, "cash_flows": {}}
+    for _, r in df.iterrows():
+        cat, item, val = str(r["category"]), str(r["item"]), r["value"]
+        try:
+            parsed = None if val in (None, "", "null") else float(str(val).replace(",",""))
+        except Exception:
+            parsed = None
+        if cat in out:
+            out[cat][item] = parsed
+    return out
+
+def radar(score):
+    labels = [d["metric"] for d in score["details"]]
+    values = [d["score"] for d in score["details"]]
+    fig = go.Figure()
+    fig.add_trace(go.Scatterpolar(r=values + values[:1], theta=labels + labels[:1], fill="toself"))
+    fig.update_layout(
+        polar=dict(radialaxis=dict(visible=True, range=[0, 100])),
+        showlegend=False,
+        margin=dict(l=20, r=20, t=30, b=20),
+        height=380,
+        title=f"総合スコア: {score['total_score']}（グレード: {score['grade']}）"
+    )
+    return fig
+
+# ================ OpenAI 抽出（Vision / Text） =================
+
+OPENAI_MODEL_VISION = os.environ.get("OPENAI_VISION_MODEL", "gpt-4o-mini")
+OPENAI_MODEL_TEXT   = os.environ.get("OPENAI_TEXT_MODEL",   "gpt-4o-mini")
+
+SYSTEM_JSON = """あなたは有能な財務アナリストです。
+与えられた決算書（画像またはテキスト）から、次の厳密な JSON 構造のみを日本語の単位なし・半角数値で返してください。分からない項目は null。
+{
+  "company": {"name": null},
+  "period": {"start_date": null, "end_date": null},
+  "balance_sheet": {
+    "total_assets": null, "total_liabilities": null, "total_equity": null,
+    "current_assets": null, "fixed_assets": null,
+    "current_liabilities": null, "long_term_liabilities": null
+  },
+  "income_statement": {
+    "sales": null, "cost_of_sales": null, "gross_profit": null,
+    "operating_expenses": null, "operating_income": null,
+    "ordinary_income": null, "net_income": null
+  },
+  "cash_flows": {
+    "operating_cash_flow": null, "investing_cash_flow": null, "financing_cash_flow": null
+  }
+}
+"""
+
+def _openai_client():
+    # openai==1.x の公式クライアント。proxies を渡さない（互換性エラー回避）。
+    from openai import OpenAI
+    key = os.environ.get("OPENAI_API_KEY")
+    if not key:
+        raise gr.Error("OPENAI_API_KEY が未設定です。Spaces → Settings → **Variables and secrets** に `OPENAI_API_KEY` を追加してください。")
+    return OpenAI(api_key=key, timeout=30)
+
+def extract_financials(images, text_blob, company_hint):
+    client = _openai_client()
+    if images:
+        content = [{"type": "text", "text": SYSTEM_JSON}]
+        if company_hint:
+            content.append({"type": "text", "text": f"会社名の候補: {company_hint}"})
+        for im in images:
+            content.append({"type": "input_image", "image_url": f"data:image/png;base64,{_b64(im)}"})
+        resp = client.chat.completions.create(
+            model=OPENAI_MODEL_VISION,
+            messages=[
+                {"role": "system", "content": "返答は必ず有効な JSON オブジェクトのみ。説明を含めない。"},
+                {"role": "user", "content": content},
+            ],
+            response_format={"type": "json_object"},
+            temperature=0.1,
+        )
+        return json.loads(resp.choices[0].message.content)
+    else:
+        prompt = f"{SYSTEM_JSON}\n\n以下は決算書のテキストです。上記の JSON だけを返してください。\n\n{text_blob or ''}"
+        resp = client.chat.completions.create(
+            model=OPENAI_MODEL_TEXT,
+            messages=[
+                {"role": "system", "content": "返答は必ず有効な JSON オブジェクトのみ。"},
+                {"role": "user", "content": prompt},
+            ],
+            response_format={"type": "json_object"},
+            temperature=0.1,
+        )
+        return json.loads(resp.choices[0].message.content)
+
+# ================== ハンドラ（型ヒントなしで安定化） ==================
+
+def run_analyze(company, use_vision, files, force_ocr):
     if not files:
         raise gr.Error("PDF をアップロードしてください。")
 
+    # 1) PDF抽出（テキスト→足りなければ画像化）
     try:
         images, raw_text, business_text, dbg = parse_pdf(files, force_ocr=force_ocr)
     except ExtractError as e:
-        # 環境起因など致命的なものはここでユーザーに明示
         raise gr.Error(f"PDF読み込みに失敗: {e}")
 
-    # --- Vision / Text フォールバック ---
+    # 2) Vision 優先 → 失敗ならテキスト
     try:
         if use_vision and images:
             fin = extract_financials(images, None, company or "")
         else:
             fin = extract_financials(None, raw_text, company or "")
     except Exception:
-        # Vision失敗→テキストへ
         try:
             fin = extract_financials(None, raw_text, company or "")
         except Exception as e:
             raise gr.Error(f"AI抽出に失敗: {e}")
 
     df = fin_to_df(fin)
-    score_internal = score_company(fin)                 # 既存社内ルール
-    fig = radar(score_internal)
+    score_int = score_company(fin)
+    fig = radar(score_int)
 
-    # 外部評価（定量化）: 既存の external_scoring を利用
+    # 3) 外部評価（定量化）
     try:
-        external_score = score_external_from_df(df)     # ← 名前はあなたの実装に合わせて
-    except Exception:
-        external_score = {"external_total": None, "items": []}
+        score_ext = score_external_from_df(df)
+    except Exception as e:
+        score_ext = {"name": "外部評価（失敗）", "external_total": None, "items": [], "notes": str(e)}
 
-    # AI 所見（中立・定量KPI＋市場/製品の補足）
+    # 4) AI 所見（中立）
     try:
-        insight = make_ai_memo(
+        memo = make_ai_memo(
             company=company or "",
             fin=fin,
-            score_internal=score_internal,
-            score_external=external_score,
+            score_internal=score_int,
+            score_external=score_ext,
             business_text=business_text
         )
     except Exception as e:
-        insight = f"AI所見の生成に失敗: {e}"
+        memo = f"AI所見の生成に失敗: {e}"
 
-    # 右ペインやタブへ返す
     return (
         json.dumps(fin, ensure_ascii=False, indent=2),
         df,
-        json.dumps(score_internal, ensure_ascii=False, indent=2),
+        json.dumps(score_int, ensure_ascii=False, indent=2),
         fig,
-        insight,
-        dbg  # ← 抽出ログ/デバッグ表示用（ui 側で Textbox/Code に出す）
+        memo,
+        json.dumps(score_ext, ensure_ascii=False, indent=2),
+        dbg
     )
+
+def run_recalc(df):
+    try:
+        fin = df_to_fin(df)
+        score_int = score_company(fin)
+        fig = radar(score_int)
+        return (
+            json.dumps(score_int, ensure_ascii=False, indent=2),
+            fig,
+            json.dumps(fin, ensure_ascii=False, indent=2)
+        )
+    except Exception as e:
+        tb = traceback.format_exc(limit=6)
+        raise gr.Error(f"再計算に失敗しました: {e}\n\n<pre style='white-space:pre-wrap'>{tb}</pre>")
+
+# ================== UI 組み立て ==================
+
+def build_ui():
+    with gr.Blocks(theme=gr.themes.Soft(primary_hue="indigo"), fill_height=True, analytics_enabled=False) as demo:
+        gr.Markdown("## 🧮 企業スコアリング（PDF解析 × OpenAI Vision）")
+
+        with gr.Row():
+            with gr.Column(scale=1):
+                company = gr.Textbox(label="企業名（任意）", placeholder="例：株式会社OO")
+                use_vision = gr.Checkbox(value=True, label="OpenAIでPDFをAI解析（Vision）")
+                force_ocr = gr.Checkbox(value=False, label="OCRを強制（スキャンPDF向け）")
+                files = gr.File(label="決算書PDF（複数可）", file_count="multiple", type="filepath")
+                run_btn = gr.Button("📄 PDFを解析してテンプレに反映", variant="primary")
+                recalc_btn = gr.Button("🔁 この表の値で再計算")
+                gr.Markdown("※ 画像化やVisionに失敗した場合はテキスト抽出に自動フォールバックします。")
+
+            with gr.Column(scale=1):
+                fin_json = gr.Code(label="抽出JSON（編集不可）", language="json", interactive=False)
+
+        with gr.Tabs():
+            with gr.Tab("抽出結果（表で編集可）"):
+                df_out = gr.Dataframe(headers=["category", "item", "value"], interactive=True, wrap=True)
+            with gr.Tab("スコアリング（内部ルール）"):
+                score_json = gr.Code(label="スコア（JSON）", language="json")
+                chart = gr.Plot(label="スコアレーダー")
+            with gr.Tab("AI診断（中立・日本語）"):
+                insight_md = gr.Markdown()
+            with gr.Tab("外部評価（定量化）"):
+                ext_json = gr.Code(label="外部評価JSON", language="json")
+            with gr.Tab("抽出ログ/デバッグ"):
+                debug_out = gr.Code(label="ログ", language="text")
+
+        run_btn.click(
+            run_analyze,
+            inputs=[company, use_vision, files, force_ocr],
+            outputs=[fin_json, df_out, score_json, chart, insight_md, ext_json, debug_out],
+            concurrency_limit=1
+        )
+        recalc_btn.click(
+            run_recalc,
+            inputs=[df_out],
+            outputs=[score_json, chart, fin_json],
+            concurrency_limit=1
+        )
+
+    return demo