feat: 重構 WIP 報表為即時 WIP 分布表

主要變更:
- 完全重寫 wip_report.html 為 Pivot 分布表格式
- 新增 3 個 WIP Distribution API (filter_options, pivot_columns, distribution)
- 橫向展開 Workcenter + Spec 組合，只顯示有資料的欄位
- 套用 WORKCENTER_GROUPS 合併邏輯 (切割→焊接→成型等12組)
- 新增 Lot Status 篩選 (Active/Hold，基於 HOLDREASONNAME)
- 排除 DUMMY 工單
- 摘要卡片顯示: 總LOT數、總數量、HOLD LOT數、HOLD數量

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

apps/portal.py

@@ -322,6 +322,13 @@ def get_current_wip_subquery(days_back=90):
     Uses ROW_NUMBER() analytic function for better performance.
     Only scans recent data (default 90 days) to reduce scan size.
     Filters out completed (8) and scrapped (128) status.
+    Excludes DUMMY orders (MFGORDERNAME = 'DUMMY').
+
+    Logic explanation:
+    - PARTITION BY CONTAINERNAME: Groups records by each LOT
+    - ORDER BY TXNDATE DESC: Orders by transaction time (newest first)
+    - rn = 1: Takes only the latest record for each LOT
+    - This gives us the current/latest status of each LOT
     """
     return f"""
         SELECT *
@@ -331,6 +338,7 @@ def get_current_wip_subquery(days_back=90):
             FROM DW_MES_WIP w
             WHERE w.TXNDATE >= SYSDATE - {days_back}
               AND w.STATUS NOT IN (8, 128)
+              AND (w.MFGORDERNAME IS NULL OR w.MFGORDERNAME <> 'DUMMY')
         )
         WHERE rn = 1
     """
@@ -1916,6 +1924,272 @@ def api_dashboard_utilization_heatmap():
     return jsonify({'success': False, 'error': '查詢失敗'}), 500
 
 
+# ============================================================
+# WIP Distribution Table API (即時 WIP 分布表)
+# ============================================================
+
+def query_wip_distribution_filter_options(days_back=90):
+    """取得 WIP 分布表的篩選選項: packages(PRODUCTLINENAME_LEF), types(PJ_TYPE), areas(PJ_PRODUCEREGION), lot_statuses"""
+    try:
+        base_sql = get_current_wip_subquery(days_back)
+        sql = f"""
+            SELECT
+                PRODUCTLINENAME_LEF,
+                PJ_TYPE,
+                PJ_PRODUCEREGION,
+                HOLDREASONNAME
+            FROM ({base_sql}) wip
+        """
+        df = read_sql_df(sql)
+
+        # 提取不重複值並排序
+        packages = sorted([x for x in df['PRODUCTLINENAME_LEF'].dropna().unique().tolist() if x])
+        types = sorted([x for x in df['PJ_TYPE'].dropna().unique().tolist() if x])
+        areas = sorted([x for x in df['PJ_PRODUCEREGION'].dropna().unique().tolist() if x])
+
+        # Lot 狀態: 根據 HOLDREASONNAME 判斷 - 有值=Hold, 無值=Active
+        lot_statuses = ['Active', 'Hold']
+
+        return {
+            'packages': packages,
+            'types': types,
+            'areas': areas,
+            'lot_statuses': lot_statuses
+        }
+    except Exception as exc:
+        print(f"WIP 篩選選項查詢失敗: {exc}")
+        import traceback
+        traceback.print_exc()
+        return None
+
+
+def query_wip_distribution_pivot_columns(filters=None, days_back=90):
+    """取得 WIP 分布表的 Pivot 欄位列表 (只回傳有資料的 Workcenter|Spec 組合)"""
+    try:
+        base_sql = get_current_wip_subquery(days_back)
+
+        # 建立篩選條件
+        where_conditions = []
+        if filters:
+            if filters.get('packages') and len(filters['packages']) > 0:
+                pkg_list = "', '".join(filters['packages'])
+                where_conditions.append(f"PRODUCTLINENAME_LEF IN ('{pkg_list}')")
+            if filters.get('types') and len(filters['types']) > 0:
+                type_list = "', '".join(filters['types'])
+                where_conditions.append(f"PJ_TYPE IN ('{type_list}')")
+            if filters.get('areas') and len(filters['areas']) > 0:
+                area_list = "', '".join(filters['areas'])
+                where_conditions.append(f"PJ_PRODUCEREGION IN ('{area_list}')")
+            # Lot 狀態篩選: Active = HOLDREASONNAME IS NULL, Hold = HOLDREASONNAME IS NOT NULL
+            if filters.get('lot_statuses') and len(filters['lot_statuses']) > 0:
+                status_conds = []
+                if 'Active' in filters['lot_statuses']:
+                    status_conds.append("HOLDREASONNAME IS NULL")
+                if 'Hold' in filters['lot_statuses']:
+                    status_conds.append("HOLDREASONNAME IS NOT NULL")
+                if status_conds:
+                    where_conditions.append(f"({' OR '.join(status_conds)})")
+
+        where_clause = " AND ".join(where_conditions) if where_conditions else "1=1"
+
+        sql = f"""
+            SELECT
+                WORKCENTERNAME,
+                SPECNAME as WC_SPEC,
+                COUNT(DISTINCT CONTAINERNAME) as LOT_COUNT
+            FROM ({base_sql}) wip
+            WHERE WORKCENTERNAME IS NOT NULL
+              AND {where_clause}
+            GROUP BY WORKCENTERNAME, SPECNAME
+            ORDER BY LOT_COUNT DESC
+        """
+        df = read_sql_df(sql)
+
+        # 轉換為 pivot 欄位列表，並套用 WORKCENTER_GROUPS 分組邏輯
+        pivot_columns = []
+        for _, row in df.iterrows():
+            wc = row['WORKCENTERNAME'] or ''
+            spec = row['WC_SPEC'] or ''
+            # 使用 get_workcenter_group 取得合併後的群組名稱和排序
+            group_name, order = get_workcenter_group(wc)
+            display_wc = group_name if group_name else wc  # 合併後的顯示名稱
+            key = f"{wc}|{spec}"  # key 仍使用原始 workcenter 以便匹配
+            pivot_columns.append({
+                'key': key,
+                'workcenter': wc,           # 原始 workcenter
+                'workcenter_group': display_wc,  # 合併後的群組名稱
+                'order': order,             # 排序順序
+                'spec': spec,
+                'count': int(row['LOT_COUNT'] or 0)
+            })
+
+        return pivot_columns
+    except Exception as exc:
+        print(f"WIP Pivot 欄位查詢失敗: {exc}")
+        import traceback
+        traceback.print_exc()
+        return None
+
+
+def query_wip_distribution(filters=None, limit=500, offset=0, days_back=90):
+    """查詢 WIP 分布表主數據，回傳每個 Lot 的基本資訊及其所在的 Workcenter|Spec"""
+    try:
+        base_sql = get_current_wip_subquery(days_back)
+
+        # 建立篩選條件
+        where_conditions = []
+        if filters:
+            if filters.get('packages') and len(filters['packages']) > 0:
+                pkg_list = "', '".join(filters['packages'])
+                where_conditions.append(f"PRODUCTLINENAME_LEF IN ('{pkg_list}')")
+            if filters.get('types') and len(filters['types']) > 0:
+                type_list = "', '".join(filters['types'])
+                where_conditions.append(f"PJ_TYPE IN ('{type_list}')")
+            if filters.get('areas') and len(filters['areas']) > 0:
+                area_list = "', '".join(filters['areas'])
+                where_conditions.append(f"PJ_PRODUCEREGION IN ('{area_list}')")
+            # Lot 狀態篩選: Active = HOLDREASONNAME IS NULL, Hold = HOLDREASONNAME IS NOT NULL
+            if filters.get('lot_statuses') and len(filters['lot_statuses']) > 0:
+                status_conds = []
+                if 'Active' in filters['lot_statuses']:
+                    status_conds.append("HOLDREASONNAME IS NULL")
+                if 'Hold' in filters['lot_statuses']:
+                    status_conds.append("HOLDREASONNAME IS NOT NULL")
+                if status_conds:
+                    where_conditions.append(f"({' OR '.join(status_conds)})")
+            if filters.get('search'):
+                search_term = filters['search'].replace("'", "''")
+                where_conditions.append(
+                    f"(UPPER(MFGORDERNAME) LIKE UPPER('%{search_term}%') "
+                    f"OR UPPER(CONTAINERNAME) LIKE UPPER('%{search_term}%'))"
+                )
+
+        where_clause = " AND ".join(where_conditions) if where_conditions else "1=1"
+
+        # 先查詢總筆數
+        count_sql = f"""
+            SELECT COUNT(DISTINCT CONTAINERNAME) as TOTAL_COUNT
+            FROM ({base_sql}) wip
+            WHERE {where_clause}
+        """
+        count_df = read_sql_df(count_sql)
+        total_count = int(count_df['TOTAL_COUNT'].iloc[0]) if len(count_df) > 0 else 0
+
+        # 分頁查詢主數據
+        start_row = offset + 1
+        end_row = offset + limit
+        sql = f"""
+            SELECT * FROM (
+                SELECT
+                    MFGORDERNAME,
+                    CONTAINERNAME,
+                    SPECNAME,
+                    PRODUCTLINENAME_LEF,
+                    WAFERLOT,
+                    PJ_TYPE,
+                    PJ_PRODUCEREGION,
+                    EQUIPMENTS,
+                    WORKCENTERNAME,
+                    STATUS,
+                    HOLDREASONNAME,
+                    QTY,
+                    QTY2,
+                    TXNDATE,
+                    ROW_NUMBER() OVER (ORDER BY TXNDATE DESC, MFGORDERNAME, CONTAINERNAME) as rn
+                FROM ({base_sql}) wip
+                WHERE {where_clause}
+            ) WHERE rn BETWEEN {start_row} AND {end_row}
+        """
+        df = read_sql_df(sql)
+
+        # 轉換為回傳格式
+        rows = []
+        for _, row in df.iterrows():
+            wc = row['WORKCENTERNAME'] or ''
+            spec = row['SPECNAME'] or ''
+            pivot_key = f"{wc}|{spec}"
+            # Lot 狀態判斷: HOLDREASONNAME 有值=Hold, 無值=Active
+            hold_reason = row['HOLDREASONNAME']
+            lot_status = 'Hold' if (pd.notna(hold_reason) and hold_reason) else 'Active'
+
+            rows.append({
+                'MFGORDERNAME': row['MFGORDERNAME'],
+                'CONTAINERNAME': row['CONTAINERNAME'],
+                'SPECNAME': row['SPECNAME'],
+                'PRODUCTLINENAME_LEF': row['PRODUCTLINENAME_LEF'],
+                'WAFERLOT': row['WAFERLOT'],
+                'PJ_TYPE': row['PJ_TYPE'],
+                'PJ_PRODUCEREGION': row['PJ_PRODUCEREGION'],
+                'EQUIPMENTS': row['EQUIPMENTS'],
+                'WORKCENTERNAME': row['WORKCENTERNAME'],
+                'LOT_STATUS': lot_status,
+                'HOLDREASONNAME': hold_reason if pd.notna(hold_reason) else None,
+                'QTY': int(row['QTY']) if pd.notna(row['QTY']) else 0,
+                'QTY2': int(row['QTY2']) if pd.notna(row['QTY2']) else 0,
+                'pivot_key': pivot_key
+            })
+
+        return {
+            'rows': rows,
+            'total_count': total_count,
+            'offset': offset,
+            'limit': limit
+        }
+    except Exception as exc:
+        print(f"WIP 分布表查詢失敗: {exc}")
+        import traceback
+        traceback.print_exc()
+        return None
+
+
+@app.route('/api/wip/distribution/filter_options')
+def api_wip_distribution_filter_options():
+    """API: 取得 WIP 分布表篩選選項"""
+    days_back = request.args.get('days_back', 90, type=int)
+    cache_key = make_cache_key("wip_dist_filter_options", days_back)
+    options = cache_get(cache_key)
+    if options is None:
+        options = query_wip_distribution_filter_options(days_back)
+        if options:
+            cache_set(cache_key, options, ttl=600)  # 10 分鐘快取
+    if options:
+        return jsonify({'success': True, 'data': options})
+    return jsonify({'success': False, 'error': '查詢失敗'}), 500
+
+
+@app.route('/api/wip/distribution/pivot_columns', methods=['POST'])
+def api_wip_distribution_pivot_columns():
+    """API: 取得 WIP 分布表 Pivot 欄位列表"""
+    data = request.get_json() or {}
+    filters = data.get('filters')
+    days_back = data.get('days_back', 90)
+
+    cache_key = make_cache_key("wip_dist_pivot_cols", days_back, filters)
+    columns = cache_get(cache_key)
+    if columns is None:
+        columns = query_wip_distribution_pivot_columns(filters, days_back)
+        if columns is not None:
+            cache_set(cache_key, columns, ttl=300)  # 5 分鐘快取
+    if columns is not None:
+        return jsonify({'success': True, 'data': columns})
+    return jsonify({'success': False, 'error': '查詢失敗'}), 500
+
+
+@app.route('/api/wip/distribution', methods=['POST'])
+def api_wip_distribution():
+    """API: 查詢 WIP 分布表主數據"""
+    data = request.get_json() or {}
+    filters = data.get('filters')
+    limit = min(data.get('limit', 500), 1000)  # 最大 1000 筆
+    offset = data.get('offset', 0)
+    days_back = data.get('days_back', 90)
+
+    result = query_wip_distribution(filters, limit, offset, days_back)
+    if result is not None:
+        return jsonify({'success': True, 'data': result})
+    return jsonify({'success': False, 'error': '查詢失敗'}), 500
+
+
 if __name__ == '__main__':
     print("正在測試數據庫連接...")
     conn = get_db_connection()