Upload 13 files

app.py

@@ -1,673 +1,692 @@
-"""
-邮件智能分类与 Google Sheets 自动回填系统
-Gradio 前端界面（含配置面板）
-
-配置面板功能：
-- LLM API 密钥输入（OpenAI / DeepSeek 切换）
-- Google Sheets ID、工作表名称、Service Account JSON 配置
-- 列映射、模糊匹配参数高级选项
-- 连接测试（LLM / Sheets 分别测试）
-- 配置持久化保存至 ~/.wellfound/config.json
-- 启动时自动加载已保存配置
-"""
-import asyncio
-import json
-import logging
-import os
-import sys
-import threading
-from datetime import datetime
-from pathlib import Path
-from typing import Optional
-
-import gradio as gr
-
-sys.path.insert(0, str(Path(__file__).parent))
-
-from src import config as cfg
-from src.processor import (
-    EmailProcessor, EmailProcessResult,
-    STATUS_DONE, STATUS_AMBIGUOUS, STATUS_NOT_FOUND,
-    STATUS_PARSE_ERROR, STATUS_AI_ERROR, STATUS_WRITE_ERROR, STATUS_SKIPPED
-)
-from src.sheets_updater import get_sheets_updater, reset_sheets_updater
-
-logging.basicConfig(
-    level=logging.INFO,
-    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
-    datefmt="%H:%M:%S",
-)
-logger = logging.getLogger(__name__)
-
-# ---- 全局状态 ----
-_processing_lock = threading.Lock()
-_all_results: list = []
-_pending_confirmations: list = []
-_log_lines: list = []
-
-
-def _log(msg: str, level: str = "INFO"):
-    ts = datetime.now().strftime("%H:%M:%S")
-    line = f"[{ts}] [{level}] {msg}"
-    _log_lines.append(line)
-    if len(_log_lines) > 200:
-        _log_lines.pop(0)
-    (logger.error if level == "ERROR" else
-     logger.warning if level == "WARN" else logger.info)(msg)
-
-
-def _get_log_text() -> str:
-    return "\n".join(_log_lines[-50:]) if _log_lines else "（暂无日志）"
-
-
-def _results_to_table(results: list) -> list:
-    return [r.to_table_row for r in results]
-
-
-def _get_status_icon(status: str) -> str:
-    if STATUS_DONE in status:
-        return "✅"
-    if STATUS_AMBIGUOUS in status:
-        return "⚠️"
-    if "❌" in status:
-        return "❌"
-    if STATUS_SKIPPED in status:
-        return "⏭️"
-    return "🔄"
-
-
-# ========== 配置面板回调 ==========
-
-def _format_sa_json_for_display(raw) -> str:
-    if not raw:
-        return ""
-    if isinstance(raw, dict):
-        return json.dumps(raw, indent=2, ensure_ascii=False)
-    if isinstance(raw, str):
-        raw = raw.strip()
-        if not raw:
-            return ""
-        try:
-            return json.dumps(json.loads(raw), indent=2, ensure_ascii=False)
-        except json.JSONDecodeError:
-            return raw
-    return str(raw)
-
-
-def load_saved_config() -> list:
-    """启动时加载持久化配置，回填到界面组件。"""
-    d = cfg.get_config_dict()
-    _log("已加载持久化配置")
-    return [
-        d.get("llm_provider", "deepseek"),
-        d.get("openai_api_key", ""),
-        d.get("openai_model", "gpt-4o-mini"),
-        d.get("deepseek_api_key", ""),
-        d.get("deepseek_model", "deepseek-chat"),
-        d.get("google_sheet_id", ""),
-        d.get("google_sheet_name", "West"),
-        _format_sa_json_for_display(d.get("google_service_account_json", "")),
-        d.get("sheet_company_col", "C"),
-        d.get("sheet_email_type_col", "AA"),
-        d.get("sheet_email_body_col", "AB"),
-        d.get("sheet_email_date_col", "AC"),
-        d.get("sheet_timestamp_col", "AD"),
-        int(d.get("fuzzy_match_threshold", 85)),
-        float(d.get("fuzzy_ambiguous_gap", 3)),
-    ]
-
-
-def save_settings(
-    provider, openai_key, openai_model,
-    deepseek_key, deepseek_model,
-    sheet_id, sheet_name, sa_json_str,
-    col_company, col_type, col_body, col_date, col_ts,
-    fuzzy_thresh, fuzzy_gap,
-) -> str:
-    """保存配置到文件并同步到内存。"""
-    sa_parsed = _parse_sa_json(sa_json_str)
-
-    config_dict = {
-        "llm_provider": provider,
-        "openai_api_key": openai_key or "",
-        "openai_model": openai_model or "gpt-4o-mini",
-        "deepseek_api_key": deepseek_key or "",
-        "deepseek_model": deepseek_model or "deepseek-chat",
-        "google_sheet_id": sheet_id or "",
-        "google_sheet_name": sheet_name or "West",
-        "google_service_account_json": sa_parsed,
-        "sheet_company_col": col_company or "C",
-        "sheet_email_type_col": col_type or "AA",
-        "sheet_email_body_col": col_body or "AB",
-        "sheet_email_date_col": col_date or "AC",
-        "sheet_timestamp_col": col_ts or "AD",
-        "fuzzy_match_threshold": int(fuzzy_thresh or 85),
-        "fuzzy_ambiguous_gap": float(fuzzy_gap or 3),
-    }
-
-    success, msg = cfg.save_config(config_dict)
-    reset_sheets_updater()
-    _log(f"配置已保存：{msg}")
-    return msg
-
-
-def _parse_sa_json(s: str):
-    """解析 Service Account JSON 输入。"""
-    if not s or not s.strip():
-        return ""
-    s = s.strip()
-    try:
-        return json.loads(s)
-    except json.JSONDecodeError:
-        return s  # 原样返回，save_config 会报错
-
-
-def test_llm(provider, openai_key, openai_model,
-              deepseek_key, deepseek_model) -> str:
-    """测试 LLM API 连接。"""
-    import asyncio
-
-    if provider == "deepseek":
-        api_key = deepseek_key or ""
-        model = deepseek_model or "deepseek-chat"
-        base_url = "https://api.deepseek.com"
-        label = "DeepSeek"
-    else:
-        api_key = openai_key or ""
-        model = openai_model or "gpt-4o-mini"
-        base_url = None
-        label = "OpenAI"
-
-    if not api_key:
-        return f"❌ 请先填写 {label} API Key"
-
-    async def _run():
-        try:
-            from openai import AsyncOpenAI
-            kwargs = {"api_key": api_key}
-            if base_url:
-                kwargs["base_url"] = base_url
-            client = AsyncOpenAI(**kwargs)
-            resp = await client.chat.completions.create(
-                model=model,
-                messages=[{"role": "user", "content": "Reply with just: OK"}],
-                max_tokens=10,
-                timeout=15,
-            )
-            content = resp.choices[0].message.content
-            return f"✅ {label} API 连接成功！\n模型：{model}\n响应：{content}"
-        except Exception as e:
-            return f"❌ {label} API 连接失败：\n{str(e)}"
-
-    try:
-        return asyncio.run(_run())
-    except RuntimeError:
-        loop = asyncio.new_event_loop()
-        result = loop.run_until_complete(_run())
-        loop.close()
-        return result
-
-
-def test_sheets(sheet_id, sheet_name, sa_json_str) -> str:
-    """测试 Google Sheets 连接。"""
-    if not sheet_id:
-        return "❌ 请先填写 Google Sheet ID"
-
-    sa_info = _parse_sa_json(sa_json_str)
-    if not sa_info or (isinstance(sa_info, str) and not sa_info.startswith("{")):
-        return "❌ Service Account JSON 格式错误，请检查粘贴内容"
-
-    try:
-        import gspread
-        from google.oauth2.service_account import Credentials
-
-        SCOPES = [
-            "https://www.googleapis.com/auth/spreadsheets",
-            "https://www.googleapis.com/auth/drive.file",
-        ]
-        creds = Credentials.from_service_account_info(sa_info, scopes=SCOPES)
-        client = gspread.authorize(creds)
-        spreadsheet = client.open_by_key(sheet_id.strip())
-
-        try:
-            ws = spreadsheet.worksheet(sheet_name.strip())
-            col_idx = cfg._col_letter_to_index("C") + 1
-            col_values = ws.col_values(col_idx)
-            company_count = sum(1 for v in col_values[1:] if v.strip())
-
-            return (
-                f"✅ Google Sheets 连接成功！\n"
-                f"表格：{spreadsheet.title}\n"
-                f"工作表：{sheet_name}\n"
-                f"总行数：{ws.row_count}\n"
-                f"C列公司名数量：{company_count}"
-            )
-        except Exception:
-            names = [s.title for s in spreadsheet.worksheets()]
-            return (
-                f"✅ 表格连接成功，但工作表 '{sheet_name}' 不存在\n"
-                f"可用工作表：{names}"
-            )
-
-    except Exception as e:
-        return f"❌ Google Sheets 连接失败：\n{str(e)}"
-
-
-# ========== 邮件处理回调 ==========
-
-def process_emails(files, progress=gr.Progress()) -> tuple:
-    global _all_results, _pending_confirmations, _log_lines
-
-    if not files:
-        return [], [], _get_log_text(), "⚠️ 请先上传 .msg 文件", gr.update(visible=False)
-
-    if not _processing_lock.acquire(blocking=False):
-        return (_results_to_table(_all_results), [],
-                _get_log_text(), "⚠️ 处理中，请稍候...", gr.update(visible=False))
-
-    try:
-        _all_results = []
-        _pending_confirmations = []
-        _log_lines = []
-        file_paths = [f.name if hasattr(f, 'name') else f for f in files]
-        total = len(file_paths)
-        _log(f"开始处理 {total} 封邮件...")
-
-        try:
-            sheets_updater = get_sheets_updater(auto_init=True)
-            sheets_ok = sheets_updater.is_connected()
-        except Exception:
-            sheets_updater = None
-            sheets_ok = False
-
-        if not sheets_ok:
-            _log("⚠️ Google Sheets 未连接，将仅展示 AI 分析结果", "WARN")
-
-        processor = EmailProcessor(sheets_updater=sheets_updater if sheets_ok else None)
-
-        results_list: list = []
-        pending_list: list = []
-
-        async def _run():
-            sem = asyncio.Semaphore(cfg.get_max_concurrent())
-            done = [0]
-
-            async def _one(fp):
-                async with sem:
-                    r = await processor.process_single_email(fp)
-                    done[0] += 1
-                    results_list.append(r)
-                    if r.needs_confirmation:
-                        pending_list.append(r)
-                    icon = _get_status_icon(r.status)
-                    _log(f"[{done[0]}/{total}] {icon} {r.filename} -> {r.company_name} [{r.category}] {r.status}")
-                    progress(done[0] / total, desc=f"处理中 {done[0]}/{total}")
-
-            await asyncio.gather(*[_one(fp) for fp in file_paths])
-
-        try:
-            asyncio.run(_run())
-        except RuntimeError:
-            loop = asyncio.new_event_loop()
-            loop.run_until_complete(_run())
-            loop.close()
-
-        _all_results = results_list
-        _pending_confirmations = pending_list
-
-        done = sum(1 for r in results_list if STATUS_DONE in r.status)
-        skipped = sum(1 for r in results_list if STATUS_SKIPPED in r.status)
-        ambiguous = len(pending_list)
-        errors = sum(1 for r in results_list if "❌" in r.status)
-
-        summary = f"✅ 处理完成 | 成功：{done} | 跳过：{skipped} | 需确认：{ambiguous} | 失败：{errors} | 共 {total} 封"
-        _log(summary)
-
-        return (
-            _results_to_table(results_list),
-            _build_pending_table(pending_list),
-            _get_log_text(),
-            summary,
-            gr.update(visible=ambiguous > 0),
-        )
-
-    except Exception as e:
-        _log(f"处理异常：{e}", "ERROR")
-        import traceback
-        traceback.print_exc()
-        return (_results_to_table(_all_results), [],
-                _get_log_text(), f"❌ 处理失败：{str(e)}", gr.update(visible=False))
-    finally:
-        _processing_lock.release()
-
-
-def _build_pending_table(pending: list) -> list:
-    rows = []
-    for i, r in enumerate(pending):
-        cand_str = " | ".join(f"{c['name']}({c['score']:.0f}%)" for c in r.candidates[:3])
-        rows.append([i, r.filename, r.company_name, cand_str, r.category, r.date_str])
-    return rows
-
-
-def confirm_selection(pending_table_data, sel_idx, sel_company) -> tuple:
-    global _pending_confirmations, _all_results
-
-    if sel_idx < 0 or sel_idx >= len(_pending_confirmations):
-        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
-                _get_log_text(), "⚠️ 请选择有效的待确认项")
-
-    result = _pending_confirmations[sel_idx]
-    confirmed_row = -1
-    for c in result.candidates:
-        if c["name"] == sel_company:
-            confirmed_row = c["row"]
-            break
-
-    if confirmed_row < 0:
-        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
-                _get_log_text(), f"⚠️ 未找到所选公司 '{sel_company}' 的行号")
-
-    try:
-        updater = get_sheets_updater(auto_init=False)
-        if not updater.is_connected():
-            return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
-                    _get_log_text(), "❌ Google Sheets 未连接，无法写入")
-
-        processor = EmailProcessor(sheets_updater=updater)
-        wr = processor.confirm_and_write(result, confirmed_row, sel_company)
-
-        if wr.success:
-            _log(f"✅ 人工确认写入成功：{result.filename} -> {sel_company} (第{confirmed_row}行)")
-            _pending_confirmations.remove(result)
-        else:
-            _log(f"❌ 写入失败：{wr.error}", "ERROR")
-
-        return (
-            _results_to_table(_all_results),
-            _build_pending_table(_pending_confirmations),
-            _get_log_text(),
-            f"✅ 已确认写入：{result.company_name} -> {sel_company}" if wr.success
-            else f"❌ 写入失败：{wr.error}",
-        )
-
-    except Exception as e:
-        _log(f"确认写入异常：{e}", "ERROR")
-        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
-                _get_log_text(), f"❌ 写入异常：{str(e)}")
-
-
-def get_candidate_choices(sel_idx) -> gr.update:
-    idx = int(sel_idx) if sel_idx is not None else -1
-    if idx < 0 or idx >= len(_pending_confirmations):
-        return gr.update(choices=[], value=None)
-    result = _pending_confirmations[idx]
-    choices = [c["name"] for c in result.candidates]
-    return gr.update(choices=choices, value=choices[0] if choices else None)
-
-
-# ========== Gradio UI ==========
-
-def build_ui() -> gr.Blocks:
-    with gr.Blocks(
-        title="📧 邮件智能分类 & Sheets 回填",
-        theme=gr.themes.Soft(primary_hue="blue", secondary_hue="slate"),
-        css="""
-        .status-bar { background:#eff6ff; border-left:4px solid #3b82f6; padding:8px 16px; border-radius:6px; font-weight:500; margin-bottom:8px; }
-        .settings-panel { background:#f8fafc; border:1px solid #e2e8f0; border-radius:10px; padding:16px; margin-bottom:12px; }
-        .pending-section { border:2px solid #f59e0b; border-radius:10px; padding:16px; background:#fffbeb; margin-top:12px; }
-        .log-box { font-family:'Courier New',monospace; font-size:12px; background:#1e1e2e; color:#cdd6f4; }
-        """
-    ) as demo:
-
-        # ---- 标题 ----
-        gr.Markdown("""
-        # 📧 邮件智能分类 & Google Sheets 自动回填
-        > 上传招聘回复邮件（.msg），AI 自动分类、提取公司信息并回填到 Google Sheets
-        """)
-
-        # ========== 设置面板 ==========
-        with gr.Accordion("⚙️ 设置面板（点击展开配置 API 密钥和 Google Sheets）", open=False):
-            with gr.Group(elem_classes=["settings-panel"]):
-
-                # ---- LLM 配置 ----
-                gr.Markdown("### 🤖 LLM API 配置")
-
-                with gr.Row():
-                    llm_provider = gr.Dropdown(
-                        label="LLM 提供商", choices=["deepseek", "openai"],
-                        value="deepseek", info="选择使用的 LLM 服务"
-                    )
-                    llm_test_btn = gr.Button("🔌 测试 LLM 连接", size="sm")
-
-                llm_test_result = gr.Textbox(label="LLM 连接测试结果", interactive=False, lines=3, visible=False)
-
-                with gr.Row(visible=True) as openai_row:
-                    openai_api_key = gr.Textbox(label="OpenAI API Key", type="password", placeholder="sk-...")
-                    openai_model = gr.Textbox(label="OpenAI 模型", value="gpt-4o-mini", placeholder="gpt-4o-mini")
-
-                with gr.Row(visible=True) as deepseek_row:
-                    deepseek_api_key = gr.Textbox(label="DeepSeek API Key", type="password", placeholder="sk-...")
-                    deepseek_model = gr.Textbox(label="DeepSeek 模型", value="deepseek-chat", placeholder="deepseek-chat")
-
-                # JS：切换提供商时显示对应输入行
-                llm_provider.change(
-                    fn=None, inputs=[llm_provider],
-                    outputs=[openai_row, deepseek_row],
-                    js="(p)=> [p==='openai'?'visible':'hidden', p==='deepseek'?'visible':'hidden']"
-                )
-
-                # ---- Google Sheets 配置 ----
-                gr.Markdown("### 📊 Google Sheets 配置")
-
-                with gr.Row():
-                    sheet_id = gr.Textbox(label="Google Sheet ID",
-                        placeholder="从 URL 中复制 /d/ 和 /edit 之间的部分",
-                        info="https://docs.google.com/spreadsheets/d/[此处是ID]/edit")
-                    sheet_name = gr.Textbox(label="工作表名称", value="West",
-                        info="根据 Wellfound.xlsx 结构，数据在 West 工作表")
-
-                sa_json = gr.Textbox(
-                    label="Google Service Account JSON（粘贴完整内容）",
-                    lines=5,
-                    info="⚠️ 敏感信息仅保存在本地 ~/.wellfound/config.json"
-                )
-
-                with gr.Row():
-                    sheets_test_btn = gr.Button("📊 测试 Google Sheets 连接", size="sm")
-                    sheets_reload_btn = gr.Button("📥 重新加载公司列表", size="sm", variant="secondary")
-
-                sheets_test_result = gr.Textbox(label="Sheets 连接测试结果", interactive=False, lines=4, visible=False)
-
-                # ---- 高级选项 ----
-                with gr.Accordion("高级选项（可选）", open=False):
-                    gr.Markdown("#### 列映射（根据 Wellfound.xlsx 结构已预设）")
-                    with gr.Row():
-                        col_company = gr.Textbox(label="公司名列（C）", value="C", scale=1)
-                        col_type = gr.Textbox(label="邮件类型列（AA）", value="AA", scale=1)
-                        col_body = gr.Textbox(label="邮件正文列（AB）", value="AB", scale=1)
-                    with gr.Row():
-                        col_date = gr.Textbox(label="邮件日期列（AC）", value="AC", scale=1)
-                        col_ts = gr.Textbox(label="时间戳列（AD）", value="AD", scale=1)
-
-                    gr.Markdown("#### 模糊匹配参数")
-                    with gr.Row():
-                        fuzzy_thresh = gr.Slider(label="模糊匹配阈值（%）", minimum=50, maximum=100, value=85, step=1)
-                        fuzzy_gap = gr.Slider(label="歧义间隔（%）", minimum=0, maximum=10, value=3, step=0.5)
-
-                # ---- 保存 / 重置 ----
-                with gr.Row():
-                    save_btn = gr.Button("💾 保存配置", variant="primary", size="lg")
-                    reset_btn = gr.Button("🔄 重置为默认值", size="sm")
-
-                save_result = gr.Textbox(label="保存结果", interactive=False, visible=False)
-
-                # ---- 回调绑定：设置面板 ----
-                save_btn.click(
-                    fn=save_settings,
-                    inputs=[
-                        llm_provider, openai_api_key, openai_model,
-                        deepseek_api_key, deepseek_model,
-                        sheet_id, sheet_name, sa_json,
-                        col_company, col_type, col_body, col_date, col_ts,
-                        fuzzy_thresh, fuzzy_gap,
-                    ],
-                    outputs=[save_result],
-                ).then(
-                    fn=lambda s: gr.update(visible=bool(s)),
-                    inputs=[save_result],
-                    outputs=[save_result],
-                )
-
-                llm_test_btn.click(
-                    fn=test_llm,
-                    inputs=[llm_provider, openai_api_key, openai_model,
-                            deepseek_api_key, deepseek_model],
-                    outputs=[llm_test_result],
-                ).then(
-                    fn=lambda: gr.update(visible=True),
-                    outputs=[llm_test_result],
-                )
-
-                sheets_test_btn.click(
-                    fn=test_sheets,
-                    inputs=[sheet_id, sheet_name, sa_json],
-                    outputs=[sheets_test_result],
-                ).then(
-                    fn=lambda: gr.update(visible=True),
-                    outputs=[sheets_test_result],
-                )
-
-                sheets_reload_btn.click(
-                    fn=lambda: (reset_sheets_updater(), "✅ 已重置连接，下次处理时重新初始化")[1],
-                    outputs=[sheets_test_result],
-                ).then(
-                    fn=lambda: gr.update(visible=True),
-                    outputs=[sheets_test_result],
-                )
-
-                reset_btn.click(
-                    fn=lambda: (
-                        "deepseek", "", "gpt-4o-mini",
-                        "", "deepseek-chat",
-                        "", "West", "",
-                        "C", "AA", "AB", "AC", "AD",
-                        85, 3.0,
-                    ),
-                    outputs=[
-                        llm_provider, openai_api_key, openai_model,
-                        deepseek_api_key, deepseek_model,
-                        sheet_id, sheet_name, sa_json,
-                        col_company, col_type, col_body, col_date, col_ts,
-                        fuzzy_thresh, fuzzy_gap,
-                    ],
-                )
-
-        # ========== 主操作区 ==========
-        gr.Markdown("---")
-
-        with gr.Row():
-            with gr.Column(scale=3):
-                file_input = gr.File(
-                    label="📁 上传邮件文件（.msg）",
-                    file_types=[".msg"], file_count="multiple", height=120,
-                )
-                with gr.Row():
-                    process_btn = gr.Button("🚀 开始处理", variant="primary", size="lg")
-                    clear_btn = gr.Button("🗑️ 清空", variant="secondary", size="lg")
-
-            with gr.Column(scale=2):
-                gr.Markdown("""
-                **📋 操作说明**
-                1. 先在上方 **⚙️ 设置面板** 配置 API 密钥和 Sheets
-                2. 将 `.msg` 文件拖入上传区（支持多选）
-                3. 点击 **🚀 开始处理** 启动 AI 分析
-                4. 处理完成后在结果表格查看状态
-                5. 若有模糊匹配，在下方 **待确认区** 手动选择
-                """)
-
-        status_text = gr.Textbox(label="处理状态", value="⚠️ 请先在设置面板配置 API 密钥",
-                                  interactive=False, elem_classes=["status-bar"])
-
-        gr.Markdown("---")
-        gr.Markdown("### 📊 处理结果")
-
-        results_table = gr.Dataframe(
-            headers=["文件名", "发件人", "日期", "公司名", "分类", "邮件摘要", "状态", "备注"],
-            datatype=["str"] * 8, wrap=True, interactive=False,
-        )
-
-        gr.Markdown("---")
-        gr.Markdown("### 📝 处理日志")
-
-        log_display = gr.Textbox(label="日志输出", value="（暂无日志）",
-                                  lines=10, interactive=False, max_lines=15, elem_classes=["log-box"])
-
-        # ========== 待确认区 ==========
-        with gr.Group(elem_classes=["pending-section"]) as pending_section:
-            gr.Markdown("### ⚠️ 待人工确认项（模糊匹配）")
-            gr.Markdown("以下邮件的公司名存在多个相似匹配，请人工选择正确的公司后点击确认回填。")
-
-            pending_table = gr.Dataframe(
-                headers=["序号", "文件名", "AI提取公司", "候选公司列表", "分类", "日期"],
-                datatype=["number", "str", "str", "str", "str", "str"],
-                interactive=False,
-            )
-
-            with gr.Row():
-                with gr.Column(scale=1):
-                    pending_idx = gr.Number(label="选择待确认项序号（从0开始）", value=0, precision=0, minimum=0)
-                with gr.Column(scale=2):
-                    candidate_dropdown = gr.Dropdown(label="选择正确的公司", choices=[], value=None, interactive=True)
-                with gr.Column(scale=1):
-                    confirm_btn = gr.Button("✅ 确认回填", variant="primary")
-
-            confirm_status = gr.Textbox(label="确认结果", interactive=False)
-
-            pending_idx.change(fn=get_candidate_choices, inputs=[pending_idx], outputs=[candidate_dropdown])
-
-            confirm_btn.click(
-                fn=confirm_selection,
-                inputs=[pending_table, pending_idx, candidate_dropdown],
-                outputs=[results_table, pending_table, log_display, confirm_status],
-            )
-
-        # ---- 主区事件绑定 ----
-        process_btn.click(
-            fn=process_emails, inputs=[file_input],
-            outputs=[results_table, pending_table, log_display, status_text, pending_section],
-            show_progress=True,
-        )
-
-        def clear_all():
-            global _all_results, _pending_confirmations, _log_lines
-            _all_results = []
-            _pending_confirmations = []
-            _log_lines = []
-            return (None, [], [], "等待上传文件...", "（已清空）", gr.update(visible=False))
-
-        clear_btn.click(
-            fn=clear_all,
-            outputs=[file_input, results_table, pending_table,
-                     status_text, log_display, pending_section],
-        )
-
-        # ---- 启动时自动加载配置 ----
-        demo.load(fn=load_saved_config, outputs=[
-            llm_provider, openai_api_key, openai_model,
-            deepseek_api_key, deepseek_model,
-            sheet_id, sheet_name, sa_json,
-            col_company, col_type, col_body, col_date, col_ts,
-            fuzzy_thresh, fuzzy_gap,
-        ])
-
-    return demo
-
-
-if __name__ == "__main__":
-    app = build_ui()
-    app.launch(
-        server_name="0.0.0.0",
-        server_port=int(os.getenv("PORT", 7860)),
-        show_error=True,
-        debug=os.getenv("DEBUG", "false").lower() == "true",
-    )
+"""
+邮件智能处理系统
+Gradio 前端界面 — 双标签页架构
+
+标签页 1：📧 邮件智能分类 & Google Sheets 自动回填
+  - .msg 文件上传 → AI 分类 → 模糊匹配 → Sheets 写入
+  - LLM API 配置（OpenAI / DeepSeek 切换）
+  - Google Sheets 配置（Service Account + Sheet ID）
+  - 配置持久化保存至 ~/.wellfound/config.json
+  - 人工确认区（模糊匹配歧义处理）
+
+标签页 2：📦 PST 邮件管理
+  - PST/OST 文件加载与解析
+  - 邮件浏览、搜索、过滤
+  - EML / MBOX / MSG 格式导出
+  - 数据统计与详情预览
+"""
+import asyncio
+import json
+import logging
+import os
+import sys
+import threading
+from datetime import datetime
+from pathlib import Path
+from typing import Optional
+
+import gradio as gr
+
+sys.path.insert(0, str(Path(__file__).parent))
+
+from src import config as cfg
+from src.processor import (
+    EmailProcessor, EmailProcessResult,
+    STATUS_DONE, STATUS_AMBIGUOUS, STATUS_NOT_FOUND,
+    STATUS_PARSE_ERROR, STATUS_AI_ERROR, STATUS_WRITE_ERROR, STATUS_SKIPPED
+)
+from src.sheets_updater import get_sheets_updater, reset_sheets_updater
+
+# PST 邮件管理系统
+from pst_system.pst_app import build_pst_ui
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(name)s: %(message)s",
+    datefmt="%H:%M:%S",
+)
+logger = logging.getLogger(__name__)
+
+# ---- 全局状态 ----
+_processing_lock = threading.Lock()
+_all_results: list = []
+_pending_confirmations: list = []
+_log_lines: list = []
+
+
+def _log(msg: str, level: str = "INFO"):
+    ts = datetime.now().strftime("%H:%M:%S")
+    line = f"[{ts}] [{level}] {msg}"
+    _log_lines.append(line)
+    if len(_log_lines) > 200:
+        _log_lines.pop(0)
+    (logger.error if level == "ERROR" else
+     logger.warning if level == "WARN" else logger.info)(msg)
+
+
+def _get_log_text() -> str:
+    return "\n".join(_log_lines[-50:]) if _log_lines else "（暂无日志）"
+
+
+def _results_to_table(results: list) -> list:
+    return [r.to_table_row for r in results]
+
+
+def _get_status_icon(status: str) -> str:
+    if STATUS_DONE in status:
+        return "✅"
+    if STATUS_AMBIGUOUS in status:
+        return "⚠️"
+    if "❌" in status:
+        return "❌"
+    if STATUS_SKIPPED in status:
+        return "⏭️"
+    return "🔄"
+
+
+# ========== 配置面板回调 ==========
+
+def _format_sa_json_for_display(raw) -> str:
+    if not raw:
+        return ""
+    if isinstance(raw, dict):
+        return json.dumps(raw, indent=2, ensure_ascii=False)
+    if isinstance(raw, str):
+        raw = raw.strip()
+        if not raw:
+            return ""
+        try:
+            return json.dumps(json.loads(raw), indent=2, ensure_ascii=False)
+        except json.JSONDecodeError:
+            return raw
+    return str(raw)
+
+
+def load_saved_config() -> list:
+    """启动时加载持久化配置，回填到界面组件。"""
+    d = cfg.get_config_dict()
+    _log("已加载持久化配置")
+    return [
+        d.get("llm_provider", "deepseek"),
+        d.get("openai_api_key", ""),
+        d.get("openai_model", "gpt-4o-mini"),
+        d.get("deepseek_api_key", ""),
+        d.get("deepseek_model", "deepseek-chat"),
+        d.get("google_sheet_id", ""),
+        d.get("google_sheet_name", "West"),
+        _format_sa_json_for_display(d.get("google_service_account_json", "")),
+        d.get("sheet_company_col", "C"),
+        d.get("sheet_email_type_col", "AA"),
+        d.get("sheet_email_body_col", "AB"),
+        d.get("sheet_email_date_col", "AC"),
+        d.get("sheet_timestamp_col", "AD"),
+        int(d.get("fuzzy_match_threshold", 85)),
+        float(d.get("fuzzy_ambiguous_gap", 3)),
+    ]
+
+
+def save_settings(
+    provider, openai_key, openai_model,
+    deepseek_key, deepseek_model,
+    sheet_id, sheet_name, sa_json_str,
+    col_company, col_type, col_body, col_date, col_ts,
+    fuzzy_thresh, fuzzy_gap,
+) -> str:
+    """保存配置到文件并同步到内存。"""
+    sa_parsed = _parse_sa_json(sa_json_str)
+
+    config_dict = {
+        "llm_provider": provider,
+        "openai_api_key": openai_key or "",
+        "openai_model": openai_model or "gpt-4o-mini",
+        "deepseek_api_key": deepseek_key or "",
+        "deepseek_model": deepseek_model or "deepseek-chat",
+        "google_sheet_id": sheet_id or "",
+        "google_sheet_name": sheet_name or "West",
+        "google_service_account_json": sa_parsed,
+        "sheet_company_col": col_company or "C",
+        "sheet_email_type_col": col_type or "AA",
+        "sheet_email_body_col": col_body or "AB",
+        "sheet_email_date_col": col_date or "AC",
+        "sheet_timestamp_col": col_ts or "AD",
+        "fuzzy_match_threshold": int(fuzzy_thresh or 85),
+        "fuzzy_ambiguous_gap": float(fuzzy_gap or 3),
+    }
+
+    success, msg = cfg.save_config(config_dict)
+    reset_sheets_updater()
+    _log(f"配置已保存：{msg}")
+    return msg
+
+
+def _parse_sa_json(s: str):
+    """解析 Service Account JSON 输入。"""
+    if not s or not s.strip():
+        return ""
+    s = s.strip()
+    try:
+        return json.loads(s)
+    except json.JSONDecodeError:
+        return s
+
+
+def test_llm(provider, openai_key, openai_model,
+              deepseek_key, deepseek_model) -> str:
+    """测试 LLM API 连接。"""
+    import asyncio
+
+    if provider == "deepseek":
+        api_key = deepseek_key or ""
+        model = deepseek_model or "deepseek-chat"
+        base_url = "https://api.deepseek.com"
+        label = "DeepSeek"
+    else:
+        api_key = openai_key or ""
+        model = openai_model or "gpt-4o-mini"
+        base_url = None
+        label = "OpenAI"
+
+    if not api_key:
+        return f"❌ 请先填写 {label} API Key"
+
+    async def _run():
+        try:
+            from openai import AsyncOpenAI
+            kwargs = {"api_key": api_key}
+            if base_url:
+                kwargs["base_url"] = base_url
+            client = AsyncOpenAI(**kwargs)
+            resp = await client.chat.completions.create(
+                model=model,
+                messages=[{"role": "user", "content": "Reply with just: OK"}],
+                max_tokens=10,
+                timeout=15,
+            )
+            content = resp.choices[0].message.content
+            return f"✅ {label} API 连接成功！\n模型：{model}\n响应：{content}"
+        except Exception as e:
+            return f"❌ {label} API 连接失败：\n{str(e)}"
+
+    try:
+        return asyncio.run(_run())
+    except RuntimeError:
+        loop = asyncio.new_event_loop()
+        result = loop.run_until_complete(_run())
+        loop.close()
+        return result
+
+
+def test_sheets(sheet_id, sheet_name, sa_json_str) -> str:
+    """测试 Google Sheets 连接。"""
+    if not sheet_id:
+        return "❌ 请先填写 Google Sheet ID"
+
+    sa_info = _parse_sa_json(sa_json_str)
+    if not sa_info or (isinstance(sa_info, str) and not sa_info.startswith("{")):
+        return "❌ Service Account JSON 格式错误，请检查粘贴内容"
+
+    try:
+        import gspread
+        from google.oauth2.service_account import Credentials
+
+        SCOPES = [
+            "https://www.googleapis.com/auth/spreadsheets",
+            "https://www.googleapis.com/auth/drive.file",
+        ]
+        creds = Credentials.from_service_account_info(sa_info, scopes=SCOPES)
+        client = gspread.authorize(creds)
+        spreadsheet = client.open_by_key(sheet_id.strip())
+
+        try:
+            ws = spreadsheet.worksheet(sheet_name.strip())
+            col_idx = cfg._col_letter_to_index("C") + 1
+            col_values = ws.col_values(col_idx)
+            company_count = sum(1 for v in col_values[1:] if v.strip())
+
+            return (
+                f"✅ Google Sheets 连接成功！\n"
+                f"表格：{spreadsheet.title}\n"
+                f"工作表：{sheet_name}\n"
+                f"总行数：{ws.row_count}\n"
+                f"C列公司名数量：{company_count}"
+            )
+        except Exception:
+            names = [s.title for s in spreadsheet.worksheets()]
+            return (
+                f"✅ 表格连接成功，但工作表 '{sheet_name}' 不存在\n"
+                f"可用工作表：{names}"
+            )
+
+    except Exception as e:
+        return f"❌ Google Sheets 连接失败：\n{str(e)}"
+
+
+# ========== 邮件处理回调 ==========
+
+def process_emails(files, progress=gr.Progress()) -> tuple:
+    global _all_results, _pending_confirmations, _log_lines
+
+    if not files:
+        return [], [], _get_log_text(), "⚠️ 请先上传 .msg 文件", gr.update(visible=False)
+
+    if not _processing_lock.acquire(blocking=False):
+        return (_results_to_table(_all_results), [],
+                _get_log_text(), "⚠️ 处理中，请稍候...", gr.update(visible=False))
+
+    try:
+        _all_results = []
+        _pending_confirmations = []
+        _log_lines = []
+        file_paths = [f.name if hasattr(f, 'name') else f for f in files]
+        total = len(file_paths)
+        _log(f"开始处理 {total} 封邮件...")
+
+        try:
+            sheets_updater = get_sheets_updater(auto_init=True)
+            sheets_ok = sheets_updater.is_connected()
+        except Exception:
+            sheets_updater = None
+            sheets_ok = False
+
+        if not sheets_ok:
+            _log("⚠️ Google Sheets 未连接，将仅展示 AI 分析结果", "WARN")
+
+        processor = EmailProcessor(sheets_updater=sheets_updater if sheets_ok else None)
+
+        results_list: list = []
+        pending_list: list = []
+
+        async def _run():
+            sem = asyncio.Semaphore(cfg.get_max_concurrent())
+            done = [0]
+
+            async def _one(fp):
+                async with sem:
+                    r = await processor.process_single_email(fp)
+                    done[0] += 1
+                    results_list.append(r)
+                    if r.needs_confirmation:
+                        pending_list.append(r)
+                    icon = _get_status_icon(r.status)
+                    _log(f"[{done[0]}/{total}] {icon} {r.filename} -> {r.company_name} [{r.category}] {r.status}")
+                    progress(done[0] / total, desc=f"处理中 {done[0]}/{total}")
+
+            await asyncio.gather(*[_one(fp) for fp in file_paths])
+
+        try:
+            asyncio.run(_run())
+        except RuntimeError:
+            loop = asyncio.new_event_loop()
+            loop.run_until_complete(_run())
+            loop.close()
+
+        _all_results = results_list
+        _pending_confirmations = pending_list
+
+        done = sum(1 for r in results_list if STATUS_DONE in r.status)
+        skipped = sum(1 for r in results_list if STATUS_SKIPPED in r.status)
+        ambiguous = len(pending_list)
+        errors = sum(1 for r in results_list if "❌" in r.status)
+
+        summary = f"✅ 处理完成 | 成功：{done} | 跳过：{skipped} | 需确认：{ambiguous} | 失败：{errors} | 共 {total} 封"
+        _log(summary)
+
+        return (
+            _results_to_table(results_list),
+            _build_pending_table(pending_list),
+            _get_log_text(),
+            summary,
+            gr.update(visible=ambiguous > 0),
+        )
+
+    except Exception as e:
+        _log(f"处理异常：{e}", "ERROR")
+        import traceback
+        traceback.print_exc()
+        return (_results_to_table(_all_results), [],
+                _get_log_text(), f"❌ 处理失败：{str(e)}", gr.update(visible=False))
+    finally:
+        _processing_lock.release()
+
+
+def _build_pending_table(pending: list) -> list:
+    rows = []
+    for i, r in enumerate(pending):
+        cand_str = " | ".join(f"{c['name']}({c['score']:.0f}%)" for c in r.candidates[:3])
+        rows.append([i, r.filename, r.company_name, cand_str, r.category, r.date_str])
+    return rows
+
+
+def confirm_selection(pending_table_data, sel_idx, sel_company) -> tuple:
+    global _pending_confirmations, _all_results
+
+    if sel_idx < 0 or sel_idx >= len(_pending_confirmations):
+        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
+                _get_log_text(), "⚠️ 请选择有效的待确认项")
+
+    result = _pending_confirmations[sel_idx]
+    confirmed_row = -1
+    for c in result.candidates:
+        if c["name"] == sel_company:
+            confirmed_row = c["row"]
+            break
+
+    if confirmed_row < 0:
+        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
+                _get_log_text(), f"⚠️ 未找到所选公司 '{sel_company}' 的行号")
+
+    try:
+        updater = get_sheets_updater(auto_init=False)
+        if not updater.is_connected():
+            return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
+                    _get_log_text(), "❌ Google Sheets 未连接，无法写入")
+
+        processor = EmailProcessor(sheets_updater=updater)
+        wr = processor.confirm_and_write(result, confirmed_row, sel_company)
+
+        if wr.success:
+            _log(f"✅ 人工确认写入成功：{result.filename} -> {sel_company} (第{confirmed_row}行)")
+            _pending_confirmations.remove(result)
+        else:
+            _log(f"❌ 写入失败：{wr.error}", "ERROR")
+
+        return (
+            _results_to_table(_all_results),
+            _build_pending_table(_pending_confirmations),
+            _get_log_text(),
+            f"✅ 已确认写入：{result.company_name} -> {sel_company}" if wr.success
+            else f"❌ 写入失败：{wr.error}",
+        )
+
+    except Exception as e:
+        _log(f"确认写入异常：{e}", "ERROR")
+        return (_results_to_table(_all_results), _build_pending_table(_pending_confirmations),
+                _get_log_text(), f"❌ 写入异常：{str(e)}")
+
+
+def get_candidate_choices(sel_idx) -> gr.update:
+    idx = int(sel_idx) if sel_idx is not None else -1
+    if idx < 0 or idx >= len(_pending_confirmations):
+        return gr.update(choices=[], value=None)
+    result = _pending_confirmations[idx]
+    choices = [c["name"] for c in result.candidates]
+    return gr.update(choices=choices, value=choices[0] if choices else None)
+
+
+# ========== Gradio UI ==========
+
+def build_ui() -> gr.Blocks:
+    with gr.Blocks(
+        title="📧 邮件智能处理系统 — 分类回填 + PST 管理",
+        theme=gr.themes.Soft(primary_hue="blue", secondary_hue="slate"),
+        css="""
+        .status-bar { background:#eff6ff; border-left:4px solid #3b82f6; padding:8px 16px; border-radius:6px; font-weight:500; margin-bottom:8px; }
+        .settings-panel { background:#f8fafc; border:1px solid #e2e8f0; border-radius:10px; padding:16px; margin-bottom:12px; }
+        .pending-section { border:2px solid #f59e0b; border-radius:10px; padding:16px; background:#fffbeb; margin-top:12px; }
+        .log-box { font-family:'Courier New',monospace; font-size:12px; background:#1e1e2e; color:#cdd6f4; }
+        .tab-nav { font-size:15px; }
+        """
+    ) as demo:
+
+        # ========== 标签页容器 ==========
+        with gr.Tabs(elem_classes=["tab-nav"]):
+
+            # ===== Tab 1: 邮件智能分类 =====
+            with gr.TabItem("📧 邮件智能分类 & Sheets 回填", id="classifier"):
+
+                # ---- 标题 ----
+                gr.Markdown("""
+                # 📧 邮件智能分类 & Google Sheets 自动回填
+                > 上传招聘回复邮件（.msg），AI 自动分类、提取公司信息并回填到 Google Sheets
+                """)
+
+                # ===== 设置面板 =====
+                with gr.Accordion("⚙️ 设置面板（点击展开配置 API 密钥和 Google Sheets）", open=False):
+                    with gr.Group(elem_classes=["settings-panel"]):
+
+                        # ---- LLM 配置 ----
+                        gr.Markdown("### 🤖 LLM API 配置")
+
+                        with gr.Row():
+                            llm_provider = gr.Dropdown(
+                                label="LLM 提供商", choices=["deepseek", "openai"],
+                                value="deepseek", info="选择使用的 LLM 服务"
+                            )
+                            llm_test_btn = gr.Button("🔌 测试 LLM 连接", size="sm")
+
+                        llm_test_result = gr.Textbox(label="LLM 连接测试结果", interactive=False, lines=3, visible=False)
+
+                        with gr.Row(visible=True) as openai_row:
+                            openai_api_key = gr.Textbox(label="OpenAI API Key", type="password", placeholder="sk-...")
+                            openai_model = gr.Textbox(label="OpenAI 模型", value="gpt-4o-mini", placeholder="gpt-4o-mini")
+
+                        with gr.Row(visible=True) as deepseek_row:
+                            deepseek_api_key = gr.Textbox(label="DeepSeek API Key", type="password", placeholder="sk-...")
+                            deepseek_model = gr.Textbox(label="DeepSeek 模型", value="deepseek-chat", placeholder="deepseek-chat")
+
+                        llm_provider.change(
+                            fn=None, inputs=[llm_provider],
+                            outputs=[openai_row, deepseek_row],
+                            js="(p)=> [p==='openai'?'visible':'hidden', p==='deepseek'?'visible':'hidden']"
+                        )
+
+                        # ---- Google Sheets 配置 ----
+                        gr.Markdown("### 📊 Google Sheets 配置")
+
+                        with gr.Row():
+                            sheet_id = gr.Textbox(
+                                label="Google Sheet ID",
+                                placeholder="从 URL 中复制 /d/ 和 /edit 之间的部分",
+                                info="https://docs.google.com/spreadsheets/d/[此处是ID]/edit")
+                            sheet_name = gr.Textbox(label="工作表名称", value="West",
+                                info="根据 Wellfound.xlsx 结构，数据在 West 工作表")
+
+                        sa_json = gr.Textbox(
+                            label="Google Service Account JSON（粘贴完整内容）",
+                            lines=5,
+                            info="⚠️ 敏感信息仅保存在本地 ~/.wellfound/config.json"
+                        )
+
+                        with gr.Row():
+                            sheets_test_btn = gr.Button("📊 测试 Google Sheets 连接", size="sm")
+                            sheets_reload_btn = gr.Button("📥 重新加载公司列表", size="sm", variant="secondary")
+
+                        sheets_test_result = gr.Textbox(label="Sheets 连接测试结果", interactive=False, lines=4, visible=False)
+
+                        # ---- 高级选项 ----
+                        with gr.Accordion("高级选项（可选）", open=False):
+                            gr.Markdown("#### 列映射（根据 Wellfound.xlsx 结构已预设）")
+                            with gr.Row():
+                                col_company = gr.Textbox(label="公司名列（C）", value="C", scale=1)
+                                col_type = gr.Textbox(label="邮件类型列（AA）", value="AA", scale=1)
+                                col_body = gr.Textbox(label="邮件正文列（AB）", value="AB", scale=1)
+                            with gr.Row():
+                                col_date = gr.Textbox(label="邮件日期列（AC）", value="AC", scale=1)
+                                col_ts = gr.Textbox(label="时间戳列（AD）", value="AD", scale=1)
+
+                            gr.Markdown("#### 模糊匹配参数")
+                            with gr.Row():
+                                fuzzy_thresh = gr.Slider(label="模糊匹配阈值（%）", minimum=50, maximum=100, value=85, step=1)
+                                fuzzy_gap = gr.Slider(label="歧义间隔（%）", minimum=0, maximum=10, value=3, step=0.5)
+
+                        # ---- 保存 / 重置 ----
+                        with gr.Row():
+                            save_btn = gr.Button("💾 保存配置", variant="primary", size="lg")
+                            reset_btn = gr.Button("🔄 重置为默认值", size="sm")
+
+                        save_result = gr.Textbox(label="保存结果", interactive=False, visible=False)
+
+                        # 设置面板事件绑定
+                        save_btn.click(
+                            fn=save_settings,
+                            inputs=[
+                                llm_provider, openai_api_key, openai_model,
+                                deepseek_api_key, deepseek_model,
+                                sheet_id, sheet_name, sa_json,
+                                col_company, col_type, col_body, col_date, col_ts,
+                                fuzzy_thresh, fuzzy_gap,
+                            ],
+                            outputs=[save_result],
+                        ).then(
+                            fn=lambda s: gr.update(visible=bool(s)),
+                            inputs=[save_result],
+                            outputs=[save_result],
+                        )
+
+                        llm_test_btn.click(
+                            fn=test_llm,
+                            inputs=[llm_provider, openai_api_key, openai_model,
+                                    deepseek_api_key, deepseek_model],
+                            outputs=[llm_test_result],
+                        ).then(
+                            fn=lambda: gr.update(visible=True),
+                            outputs=[llm_test_result],
+                        )
+
+                        sheets_test_btn.click(
+                            fn=test_sheets,
+                            inputs=[sheet_id, sheet_name, sa_json],
+                            outputs=[sheets_test_result],
+                        ).then(
+                            fn=lambda: gr.update(visible=True),
+                            outputs=[sheets_test_result],
+                        )
+
+                        sheets_reload_btn.click(
+                            fn=lambda: (reset_sheets_updater(), "✅ 已重置连接，下次处理时重新初始化")[1],
+                            outputs=[sheets_test_result],
+                        ).then(
+                            fn=lambda: gr.update(visible=True),
+                            outputs=[sheets_test_result],
+                        )
+
+                        reset_btn.click(
+                            fn=lambda: (
+                                "deepseek", "", "gpt-4o-mini",
+                                "", "deepseek-chat",
+                                "", "West", "",
+                                "C", "AA", "AB", "AC", "AD",
+                                85, 3.0,
+                            ),
+                            outputs=[
+                                llm_provider, openai_api_key, openai_model,
+                                deepseek_api_key, deepseek_model,
+                                sheet_id, sheet_name, sa_json,
+                                col_company, col_type, col_body, col_date, col_ts,
+                                fuzzy_thresh, fuzzy_gap,
+                            ],
+                        )
+
+                # ===== 主操作区 =====
+                gr.Markdown("---")
+
+                with gr.Row():
+                    with gr.Column(scale=3):
+                        file_input = gr.File(
+                            label="📁 上传邮件文件（.msg）",
+                            file_types=[".msg"], file_count="multiple", height=120,
+                        )
+                        with gr.Row():
+                            process_btn = gr.Button("🚀 开始处理", variant="primary", size="lg")
+                            clear_btn = gr.Button("🗑️ 清空", variant="secondary", size="lg")
+
+                    with gr.Column(scale=2):
+                        gr.Markdown("""
+                        **📋 操作说明**
+                        1. 先在上方 **⚙️ 设置面板** 配置 API 密钥和 Sheets
+                        2. 将 `.msg` 文件拖入上传区（支持多选）
+                        3. 点击 **🚀 开始处理** 启动 AI 分析
+                        4. 处理完成后在结果表格查看状态
+                        5. 若有模糊匹配，在下方 **待确认区** 手动选择
+                        """)
+
+                status_text = gr.Textbox(label="处理状态", value="⚠️ 请先在设置面板配置 API 密钥",
+                                          interactive=False, elem_classes=["status-bar"])
+
+                gr.Markdown("---")
+                gr.Markdown("### 📊 处理结果")
+
+                results_table = gr.Dataframe(
+                    headers=["文件名", "发件人", "日期", "公司名", "分类", "邮件摘要", "状态", "备注"],
+                    datatype=["str"] * 8, wrap=True, interactive=False,
+                )
+
+                gr.Markdown("---")
+                gr.Markdown("### 📝 处理日志")
+
+                log_display = gr.Textbox(label="日志输出", value="（暂无日志）",
+                                          lines=10, interactive=False, max_lines=15, elem_classes=["log-box"])
+
+                # ===== 待确认区 =====
+                with gr.Group(elem_classes=["pending-section"]) as pending_section:
+                    gr.Markdown("### ⚠️ 待人工确认项（模糊匹配）")
+                    gr.Markdown("以下邮件的公司名存在多个相似匹配，请人工选择正确的公司后点击确认回填。")
+
+                    pending_table = gr.Dataframe(
+                        headers=["序号", "文件名", "AI提取公司", "候选公司列表", "分类", "日期"],
+                        datatype=["number", "str", "str", "str", "str", "str"],
+                        interactive=False,
+                    )
+
+                    with gr.Row():
+                        with gr.Column(scale=1):
+                            pending_idx = gr.Number(label="选择待确认项序号（从0开始）", value=0, precision=0, minimum=0)
+                        with gr.Column(scale=2):
+                            candidate_dropdown = gr.Dropdown(label="选择正确的公司", choices=[], value=None, interactive=True)
+                        with gr.Column(scale=1):
+                            confirm_btn = gr.Button("✅ 确认回填", variant="primary")
+
+                    confirm_status = gr.Textbox(label="确认结果", interactive=False)
+
+                    pending_idx.change(fn=get_candidate_choices, inputs=[pending_idx], outputs=[candidate_dropdown])
+
+                    confirm_btn.click(
+                        fn=confirm_selection,
+                        inputs=[pending_table, pending_idx, candidate_dropdown],
+                        outputs=[results_table, pending_table, log_display, confirm_status],
+                    )
+
+                # 主区事件绑定
+                process_btn.click(
+                    fn=process_emails, inputs=[file_input],
+                    outputs=[results_table, pending_table, log_display, status_text, pending_section],
+                    show_progress=True,
+                )
+
+                def clear_all():
+                    global _all_results, _pending_confirmations, _log_lines
+                    _all_results = []
+                    _pending_confirmations = []
+                    _log_lines = []
+                    return (None, [], [], "等待上传文件...", "（已清空）", gr.update(visible=False))
+
+                clear_btn.click(
+                    fn=clear_all,
+                    outputs=[file_input, results_table, pending_table,
+                             status_text, log_display, pending_section],
+                )
+
+            # ===== Tab 2: PST 邮件管理 =====
+            with gr.TabItem("📦 PST 邮件管理", id="pst_manager"):
+                build_pst_ui()
+
+        # ---- 启动时自动加载配置 ----
+        demo.load(fn=load_saved_config, outputs=[
+            llm_provider, openai_api_key, openai_model,
+            deepseek_api_key, deepseek_model,
+            sheet_id, sheet_name, sa_json,
+            col_company, col_type, col_body, col_date, col_ts,
+            fuzzy_thresh, fuzzy_gap,
+        ])
+
+    return demo
+
+
+if __name__ == "__main__":
+    app = build_ui()
+    app.launch(
+        server_name="0.0.0.0",
+        server_port=int(os.getenv("PORT", 7860)),
+        show_error=True,
+        debug=os.getenv("DEBUG", "false").lower() == "true",
+    )

requirements.txt

@@ -8,6 +8,9 @@ extract-msg==0.55.0
 beautifulsoup4==4.12.3
 html2text==2024.2.26
 
+# PST parsing (Outlook MAPI)
+pywin32
+
 # AI / LLM
 openai==1.51.0