Upload app.py

app.py

@@ -33,6 +33,10 @@ load_dotenv()  # 加载本地 .env 文件
 
 # --- 全局配置和常量 ---
 TOKEN_FILE = 'token_data.json'
+ROOT_DIR = Path(__file__).resolve().parent
+CINEMA_CACHE_DIR = ROOT_DIR / "cinema_cache"
+NATIONAL_MARKET_MOVIE_FILE = CINEMA_CACHE_DIR / "national_market_movie_daily.csv"
+NATIONAL_MARKET_DAILY_FILE = CINEMA_CACHE_DIR / "national_market_daily_summary.csv"
 # --- 环境变量获取 (替代硬编码) ---
 # 使用 os.getenv 获取，如果获取不到默认为空字符串或特定默认值
 GAODE_API_KEY = os.getenv("GAODE_API_KEY", "")
@@ -821,12 +825,24 @@ def process_api_data(schedule_list, hall_seat_map, token=None, show_date=None):
     df.rename(columns={'movieName': '影片名称', 'showStartTime': '放映时间', 'soldBoxOffice': '总收入',
                        'soldTicketNum': '总人次', 'hallName': '影厅名称'}, inplace=True)
 
-    # 获取标准电影名列表并进行清洗
+    # 获取标准电影名列表（用于 movieNum 映射未命中时的回退）
     canonical_names = []
     if token and show_date:
         canonical_names = fetch_canonical_movie_names(token, show_date)
 
-    df['影片名称'] = df['影片名称'].apply(lambda x: clean_movie_title(x, canonical_names))
+    # 使用 movieNum 映射优先解析影片名（不含制式后缀，仅正式片名）
+    movie_num_col = df.get('movieNum')
+    if movie_num_col is not None:
+        df['影片名称'] = df.apply(
+            lambda row: resolve_movie_name_from_schedule_item(
+                row['影片名称'],
+                movie_num=row.get('movieNum'),
+                canonical_names=canonical_names
+            ),
+            axis=1
+        )
+    else:
+        df['影片名称'] = df['影片名称'].apply(lambda x: clean_movie_title(x, canonical_names))
 
     # 在 required_cols 中增加 '影厅名称'
     required_cols = ['影片名称', '放映时间', '座位数', '总收入', '总人次', '影厅名称']
@@ -1006,6 +1022,75 @@ def clean_movie_title(raw_title, canonical_names=None):
     return base_name
 
 
+# --- movieNum 影片名映射解析（与 history_sessions_monitor.py 策略一致） ---
+MOVIE_NUM_NAME_MAP_FILE = CINEMA_CACHE_DIR / "movie_num_name_map.json"
+
+
+def _normalize_movie_num_key(movie_num):
+    """将 movieNum 标准化为纯大写字母数字"""
+    return re.sub(r'[^A-Z0-9]', '', str(movie_num or '').strip().upper())
+
+
+@st.cache_data(show_spinner=False, ttl=120)
+def _load_movie_num_name_map():
+    """加载 movie_num_name_map.json 并返回 {normalized_key: official_name} 字典"""
+    if not MOVIE_NUM_NAME_MAP_FILE.exists():
+        return {}
+    try:
+        with open(MOVIE_NUM_NAME_MAP_FILE, 'r', encoding='utf-8') as f:
+            payload = json.load(f)
+    except Exception:
+        return {}
+    movie_num_map = payload.get('movie_num_map', {})
+    if not isinstance(movie_num_map, dict):
+        return {}
+    result = {}
+    for movie_num, entry in movie_num_map.items():
+        normalized_key = _normalize_movie_num_key(movie_num)
+        if not normalized_key:
+            continue
+        if isinstance(entry, str):
+            official_name = entry.strip()
+        elif isinstance(entry, dict):
+            official_name = str(entry.get('official_name') or '').strip()
+        else:
+            continue
+        if official_name:
+            result[normalized_key] = official_name
+    return result
+
+
+def resolve_movie_name_from_schedule_item(raw_movie_name, movie_num=None, canonical_names=None):
+    """
+    使用 movieNum 映射表解析影片正式名称（不含制式后缀，仅用于效率分析展示）。
+    优先级：movieNum 映射 > canonical_names 匹配 > 空格分割回退。
+    """
+    if not isinstance(raw_movie_name, str):
+        return raw_movie_name
+
+    # 1. 尝试 movieNum 映射
+    if movie_num:
+        num_key = _normalize_movie_num_key(movie_num)
+        if num_key:
+            name_map = _load_movie_num_name_map()
+            official_name = name_map.get(num_key)
+            if official_name:
+                return official_name
+
+    # 2. 回退到旧的 clean_movie_title 逻辑（不含制式后缀）
+    base_name = None
+    if canonical_names:
+        sorted_names = sorted(canonical_names, key=len, reverse=True)
+        for name in sorted_names:
+            if name in raw_movie_name:
+                base_name = name
+                break
+    if not base_name:
+        base_name = raw_movie_name.split(' ', 1)[0]
+
+    return base_name
+
+
 def get_available_movie_name_pool(show_date):
     token_data = load_token()
     token = token_data.get('token') if token_data else None
@@ -1031,7 +1116,11 @@ def get_available_movie_name_pool(show_date):
 
     canonical_names = fetch_canonical_movie_names(token, show_date) if token else []
     cleaned_movie_names = {
-        clean_movie_title(item.get('movieName'), canonical_names if canonical_names else None)
+        resolve_movie_name_from_schedule_item(
+            item.get('movieName'),
+            movie_num=item.get('movieNum'),
+            canonical_names=canonical_names if canonical_names else None
+        )
         for item in available_movies
         if item.get('movieName')
     }
@@ -1469,7 +1558,7 @@ def render_sales_summary_section(df, selected_date, key_prefix):
         return
 
     st.markdown("#### 销售总览 (套餐 Top 10 + 单品 Top 5)")
-    st.dataframe(final_summary, use_container_width=True, hide_index=True)
+    st.dataframe(final_summary, width="stretch", hide_index=True)
     st.markdown("##### 复制到 Excel")
     st.code(copy_text, language='text')
 
@@ -1775,7 +1864,18 @@ def build_daily_report_from_source_df(source_df, selected_date_str=None, token=N
     if token and display_date_str:
         canonical_names = fetch_canonical_movie_names(token, display_date_str)
 
-    df['影片'] = df['影片名称'].apply(lambda x: clean_movie_title(x, canonical_names))
+    # 使用 movieNum 映射优先解析影片名（不含制式后缀）
+    if 'movieNum' in df.columns:
+        df['影片'] = df.apply(
+            lambda row: resolve_movie_name_from_schedule_item(
+                row['影片名称'],
+                movie_num=row.get('movieNum'),
+                canonical_names=canonical_names
+            ),
+            axis=1
+        )
+    else:
+        df['影片'] = df['影片名称'].apply(lambda x: clean_movie_title(x, canonical_names))
     df['影厅'] = df['影厅名称']
     df['人数合计'] = df['总人次']
 
@@ -1804,6 +1904,9 @@ def process_and_filter_data_for_report(schedule_list, hall_seat_map, selected_da
         'soldTicketNum': '总人次'
     }, inplace=True)
     df['放映日期'] = selected_date_str
+    # 保留 movieNum 列供 build_daily_report_from_source_df 使用
+    if 'movieNum' not in df.columns:
+        df['movieNum'] = ''
 
     result_df, _ = build_daily_report_from_source_df(df, selected_date_str=selected_date_str, token=token)
     return result_df
@@ -2678,6 +2781,7 @@ def is_time_in_schedule_check_exempt_ranges(ts, ranges):
 
 def load_schedule_check_settings():
     """从 JSON 文件加载场次合理性检查设置"""
+    ensure_settings_file_synced(SCHEDULE_CHECK_SETTINGS_FILE)
     if os.path.exists(SCHEDULE_CHECK_SETTINGS_FILE):
         try:
             with open(SCHEDULE_CHECK_SETTINGS_FILE, 'r', encoding='utf-8') as f:
@@ -2768,6 +2872,7 @@ def save_schedule_check_settings(config=None, show_toast=True):
     try:
         with open(SCHEDULE_CHECK_SETTINGS_FILE, 'w', encoding='utf-8') as f:
             json.dump(settings_to_save, f, ensure_ascii=False, indent=4)
+        upload_settings_file_to_r2(SCHEDULE_CHECK_SETTINGS_FILE)
         if show_toast:
             st.toast("场次合理性检查设置已保存！", icon="💾")
         return True
@@ -2834,7 +2939,7 @@ def render_schedule_check_settings():
             rule1_enabled = st.checkbox("规则一：同影片最小开场间隔", key="schedule_check_rule1_enabled")
         with col2:
             st.number_input(
-                "规则一间隔(分钟)", min_value=0, step=5,
+                "规则一间隔(分钟)", min_value=30, step=5,
                 key="schedule_check_rule1_min_interval_minutes",
                 disabled=not rule1_enabled
             )
@@ -2844,13 +2949,13 @@ def render_schedule_check_settings():
             rule2_enabled = st.checkbox("规则二：X分钟内影片开场超过Y场", key="schedule_check_rule2_enabled")
         with col2:
             st.number_input(
-                "规则二阈值(场)", min_value=1, step=1,
+                "规则二阈值(场)", min_value=4, step=1,
                 key="schedule_check_rule2_threshold_sessions",
                 disabled=not rule2_enabled
             )
         with col3:
             st.number_input(
-                "规则二时间窗口(分钟)", min_value=5, step=5,
+                "规则二时间窗口(分钟)", min_value=20, step=5,
                 key="schedule_check_rule2_window_minutes",
                 disabled=not rule2_enabled
             )
@@ -2867,7 +2972,7 @@ def render_schedule_check_settings():
             rule3_enabled = st.checkbox("规则三：场次开场断档", key="schedule_check_rule3_enabled")
         with col2:
             st.number_input(
-                "规则三断档阈值(分钟)", min_value=0, step=5,
+                "规则三断档阈值(分钟)", min_value=30, step=5,
                 key="schedule_check_rule3_gap_threshold_minutes",
                 disabled=not rule3_enabled
             )
@@ -2896,7 +3001,7 @@ def render_schedule_check_settings():
             rule6_enabled = st.checkbox("规则六：影厅场次转换时间检查", key="schedule_check_rule6_enabled")
         with col2:
             st.number_input(
-                "场次转换基准(分钟)", min_value=0, step=1,
+                "场次转换基准(分钟)", min_value=10, step=1,
                 key="schedule_check_rule6_min_conversion_minutes",
                 disabled=not rule6_enabled
             )
@@ -2918,7 +3023,7 @@ def render_schedule_check_settings():
         col1, col2 = st.columns([1, 2])
         with col1:
             st.number_input(
-                "规则九热门TopN", min_value=1, step=1,
+                "规则九热门TopN", min_value=3, step=1,
                 key="schedule_check_rule9_hot_top_n",
                 disabled=not rule9_enabled
             )
@@ -2942,7 +3047,7 @@ def render_schedule_check_settings():
             rule12_enabled = st.checkbox("规则十二：次日票房前N影片必须有黄金场", key="schedule_check_rule12_enabled")
         with col2:
             st.number_input(
-                "规则十二票房TopN", min_value=1, step=1,
+                "规则十二票房TopN", min_value=5, step=1,
                 key="schedule_check_rule12_top_n",
                 disabled=not rule12_enabled
             )
@@ -3065,7 +3170,17 @@ def generate_schedule_check_logs(schedule_list, date_str, today_analysis_df=None
 
     df_original['simpleHallName'] = df_original['Hall'].apply(simplify_hall)
     df_check = df_original.sort_values(by='startTime').reset_index(drop=True)
-    df_check['clean_filmName'] = df_check['filmName'].apply(clean_movie_title)
+    # 使用 movieNum 映射优先解析影片名（不含制式后缀）
+    if 'movieNum' in df_check.columns:
+        df_check['clean_filmName'] = df_check.apply(
+            lambda row: resolve_movie_name_from_schedule_item(
+                row['filmName'],
+                movie_num=row.get('movieNum'),
+            ),
+            axis=1
+        )
+    else:
+        df_check['clean_filmName'] = df_check['filmName'].apply(clean_movie_title)
 
     next_day_bo_data = fetch_realtime_box_office(date_str)
     movie_box_office = {}
@@ -3855,7 +3970,7 @@ def build_api_gantt_rows(schedule_list, date_str, token, hall_seat_map=None):
         if movie_length <= 0:
             movie_length = int((end_dt - start_dt).total_seconds() / 60)
 
-        film_name = clean_movie_title(raw_movie_name, canonical_names if canonical_names else None)
+        film_name = resolve_movie_name_from_schedule_item(raw_movie_name, movie_num=item.get('movieNum'), canonical_names=canonical_names if canonical_names else None)
         film_language = item.get('movieLanguage', '') or ''
         imagery = item.get('movieMediaType', '') or ''
         movie_num = item.get('movieNum', '') or ''
@@ -4094,7 +4209,103 @@ def render_nextday_pool_history_table(base_date_str):
 
     st.dataframe(
         summary_df.style.format({'总票房': '{:,.2f}'}),
-        use_container_width=True,
+        width="stretch",
+        hide_index=True,
+    )
+
+
+
+def load_nextday_pool_cache_for_app():
+    """加载次日电影池统计缓存 CSV 和 JSON 元数据"""
+    nextday_pool_csv = CINEMA_CACHE_DIR / "history_sessions_monitor_nextday_pool_stats.csv"
+    nextday_pool_json = CINEMA_CACHE_DIR / "history_sessions_monitor_nextday_pool_stats.json"
+    if not nextday_pool_csv.exists():
+        return pd.DataFrame(), {}
+    try:
+        df = pd.read_csv(nextday_pool_csv)
+    except Exception:
+        return pd.DataFrame(), {}
+    meta = {}
+    if nextday_pool_json.exists():
+        try:
+            with open(nextday_pool_json, 'r', encoding='utf-8') as f:
+                meta = json.load(f)
+        except Exception:
+            pass
+    return df, meta
+
+
+def render_nextday_pool_cache_table():
+    """渲染次日电影池统计缓存表"""
+    pool_df, meta = load_nextday_pool_cache_for_app()
+    if pool_df.empty:
+        return
+
+    nextday_date = meta.get('nextday_date', '')
+    reference_date = meta.get('reference_date', '')
+    generated_at = meta.get('generated_at', '')
+    status = meta.get('status', '')
+
+    title_parts = ["### 📊 次日电影池统计数据"]
+    if nextday_date:
+        title_parts[0] += f"（目标日期：{nextday_date}）"
+    st.markdown(title_parts[0])
+
+    caption_parts = []
+    if reference_date:
+        caption_parts.append(f"统计截止：{reference_date}")
+    if generated_at:
+        caption_parts.append(f"生成时间：{generated_at}")
+    if status:
+        caption_parts.append(status)
+    if caption_parts:
+        st.caption(" | ".join(caption_parts))
+
+    # 定义次日电影池的列顺序
+    nextday_pool_windows = [
+        ("最近一天", 1),
+        ("最近三天", 3),
+        ("最近一周", 7),
+        ("最近一个月", 30),
+    ]
+    display_columns = ["影片", "历史票房", "历史场次", "历史效率"]
+    for label, _days in nextday_pool_windows:
+        display_columns.extend([f"{label}票房", f"{label}场次", f"{label}效率"])
+
+    available_columns = [col for col in display_columns if col in pool_df.columns]
+    remaining_columns = [col for col in pool_df.columns if col not in available_columns]
+    display_df = pool_df[available_columns + remaining_columns].copy()
+
+    def _format_nextday_metric(value, pattern):
+        if value is None:
+            return 'n/a'
+        if isinstance(value, str):
+            if value.strip().lower() in {'', 'n/a', 'na', 'nan', '<na>'}:
+                return 'n/a'
+        try:
+            if pd.isna(value):
+                return 'n/a'
+        except Exception:
+            pass
+        try:
+            return pattern.format(float(value))
+        except Exception:
+            return str(value)
+
+    format_config = {
+        '历史票房': lambda v: _format_nextday_metric(v, '{:,.2f}'),
+        '历史场次': lambda v: _format_nextday_metric(v, '{:,.0f}'),
+        '历史效率': lambda v: _format_nextday_metric(v, '{:.2f}'),
+    }
+    for label, _days in nextday_pool_windows:
+        format_config[f'{label}票房'] = lambda v: _format_nextday_metric(v, '{:,.2f}')
+        format_config[f'{label}场次'] = lambda v: _format_nextday_metric(v, '{:,.0f}')
+        format_config[f'{label}效率'] = lambda v: _format_nextday_metric(v, '{:.2f}')
+
+    format_config = {k: v for k, v in format_config.items() if k in display_df.columns}
+    st.dataframe(
+        display_df.style.format(format_config),
+        width="stretch",
         hide_index=True,
     )
 
@@ -4142,7 +4353,18 @@ def display_analysis_results(df_raw, data_source_name, date_for_display, query_t
     date_str = f"{date_for_display} " if date_for_display else ""
     total_revenue, total_attendance, total_sessions = df_raw['总收入'].sum(), df_raw['总人次'].sum(), len(df_raw)
     st.markdown(
-        f"> {date_str}数据总览：总票房 **¥{total_revenue:,.2f}** | 总人次 **{total_attendance:,.0f}** | 总场次 **{total_sessions:,.0f}**")
+        f"""
+        <div style="margin: 12px 0 18px; padding: 14px 18px; border-left: 6px solid #D83B01; background: #FFF4ED;">
+            <span style="font-size: 18px; font-weight: 700; color: #5C1F00;">{date_str}总票房：</span>
+            <span style="font-size: 26px; font-weight: 800; color: #D83B01;">¥{total_revenue:,.2f}</span>
+            <span style="font-size: 18px; font-weight: 700; color: #5C1F00;"> | 总人次：</span>
+            <span style="font-size: 26px; font-weight: 800; color: #D83B01;">{total_attendance:,.0f}</span>
+            <span style="font-size: 18px; font-weight: 700; color: #5C1F00;"> | 总场次：</span>
+            <span style="font-size: 26px; font-weight: 800; color: #D83B01;">{total_sessions:,.0f}</span>
+        </div>
+        """,
+        unsafe_allow_html=True,
+    )
 
     format_config = {'座位数': '{:,.0f}', '场次': '{:,.0f}', '人次': '{:,.0f}', '票房': '{:,.2f}', '均价': '{:.2f}',
                      '座次比': '{:.2%}', '场次比': '{:.2%}', '票房比': '{:.2%}', '座次效率': '{:.2f}',
@@ -4169,10 +4391,10 @@ def display_analysis_results(df_raw, data_source_name, date_for_display, query_t
 
     st.markdown("#### 全天排片效率分析");
     st.dataframe(full_day_analysis.style.format(format_config).apply(style_efficiency, axis=1),
-                 use_container_width=True, hide_index=True)
+                 width="stretch", hide_index=True)
     st.markdown("#### 黄金时段排片效率分析 (14:00-21:00)");
     st.dataframe(prime_time_analysis.style.format(format_config).apply(style_efficiency, axis=1),
-                 use_container_width=True, hide_index=True)
+                 width="stretch", hide_index=True)
     if not full_day_analysis.empty:
         st.markdown("### 排片效率汇总")
         full_day_summary = full_day_analysis.rename(
@@ -4189,10 +4411,10 @@ def display_analysis_results(df_raw, data_source_name, date_for_display, query_t
                                  '全部座次效率': '{:.2f}', '全部场次效率': '{:.2f}', '黄金时段座次效率': '{:.2f}',
                                  '黄金时段场次效率': '{:.2f}'}
         st.dataframe(summary_df.style.format(summary_format_config).apply(style_summary_efficiency, axis=1),
-                     use_container_width=True, hide_index=True)
+                     width="stretch", hide_index=True)
 
 
-        st.markdown("## 📈 全国大盘实时票房查询")
+        st.markdown("### 📈 全国大盘实时票房查询")
         
         # 增加手续费复选框，默认不包含
         include_fee = st.checkbox("包含手续费", value=False)
@@ -4244,7 +4466,7 @@ def display_analysis_results(df_raw, data_source_name, date_for_display, query_t
                         '总场次': '{:,.0f}',
                         '总人次': '{:,.0f}'
                     }
-                    st.dataframe(today_df.style.format(format_dict), use_container_width=True, hide_index=True)
+                    st.dataframe(today_df.style.format(format_dict), width="stretch", hide_index=True)
                 else:
                     st.info(f"未能获取到 {today_str} 的全国大盘数据。")
                     
@@ -4282,11 +4504,13 @@ def display_analysis_results(df_raw, data_source_name, date_for_display, query_t
                         '总场次': '{:,.0f}',
                         '总人次': '{:,.0f}'
                     }
-                    st.dataframe(tomorrow_df.style.format(format_dict), use_container_width=True, hide_index=True)
+                    st.dataframe(tomorrow_df.style.format(format_dict), width="stretch", hide_index=True)
                 else:
                     st.info(f"未能获取到 {tomorrow_str} 的全国大盘预售数据。")
 
-        st.markdown(f"## 🎬 {today_str} 排片甘特图")
+        render_nextday_pool_cache_table()
+
+        st.markdown(f"### 🎬 {today_str} 排片甘特图")
         render_today_api_gantt_chart(today_str)
 
 
@@ -4313,7 +4537,18 @@ def fetch_and_process_daily_sessions(date_str, quiet=False):
     if token:
         canonical_names = fetch_canonical_movie_names(token, date_str)
 
-    df['影片名称_清理后'] = df['movieName'].apply(lambda x: clean_movie_title(x, canonical_names))
+    # 使用 movieNum 映射优先解析影片名（不含制式后缀）
+    if 'movieNum' in df.columns:
+        df['影片名称_清理后'] = df.apply(
+            lambda row: resolve_movie_name_from_schedule_item(
+                row.get('movieName', ''),
+                movie_num=row.get('movieNum'),
+                canonical_names=canonical_names
+            ),
+            axis=1
+        )
+    else:
+        df['影片名称_清理后'] = df['movieName'].apply(lambda x: clean_movie_title(x, canonical_names))
 
     sessions_map = df.groupby('影片名称_清理后').size().to_dict()
     # 返回 (映射表, 总场次, 原始排片列表)
@@ -4634,7 +4869,7 @@ def main():
                                  '座次效率': '{:.2f}', '场次效率': '{:.2f}', '次日场数': '{:,.0f}'}
                 display_df['均价'] = pd.to_numeric(display_df['均价'], errors='coerce').replace([np.inf, -np.inf],
                                                                                                 np.nan)
-                st.dataframe(display_df.style.format(report_format, na_rep="#DIV/0!"), use_container_width=True,
+                st.dataframe(display_df.style.format(report_format, na_rep="#DIV/0!"), width="stretch",
                              hide_index=True)
 
                 n_diff = st.session_state.today_movie_count - st.session_state.previous_day_movie_count
@@ -4824,7 +5059,7 @@ def main():
                 st.session_state.daily_report_df.style.format({
                     '人数合计': '{:,.0f}', '座位数': '{:,.0f}', '上座率%': '{:.2f}%'
                 }),
-                use_container_width=True, hide_index=True
+                width="stretch", hide_index=True
             )
             total_attendance = pd.to_numeric(
                 st.session_state.daily_report_df.get('人数合计', 0),
@@ -5020,7 +5255,7 @@ def main():
                     with tab_views[0]:
                         st.markdown(display_pdf(led_output['pdf']), unsafe_allow_html=True)
                     if 'png' in led_output:
-                        with tab_views[1]: st.image(led_output['png'], use_container_width=True)
+                        with tab_views[1]: st.image(led_output['png'], width="stretch")
             else:
                 st.error("未能成功生成 '修改 LED 屏排片表'。请检查数据源。")
 
@@ -5087,7 +5322,7 @@ def main():
                                    '时长(分钟)': format_play_time(item['play_time']),
                                    '文件名': format_content_name_with_explanation(item['content_name'])}
                                   for item in movie_list_sorted]
-                    st.dataframe(pd.DataFrame(view2_data), hide_index=True, use_container_width=True)
+                    st.dataframe(pd.DataFrame(view2_data), hide_index=True, width="stretch")
                     st.markdown("#### 按影厅查看影片内容")
                     hall_tabs = st.tabs(list(halls_data.keys()))
                     for tab, hall_name in zip(hall_tabs, halls_data.keys()):
@@ -5097,7 +5332,7 @@ def main():
                                            '时长(分钟)': format_play_time(item['details']['play_time']),
                                            '文件名': format_content_name_with_explanation(item['content_name'])} for item in
                                           halls_data[hall_name]]
-                            st.dataframe(pd.DataFrame(view1_data), hide_index=True, use_container_width=True)
+                            st.dataframe(pd.DataFrame(view1_data), hide_index=True, width="stretch")
                 except Exception as e:
                     st.error(f"查询TMS服务器时出错: {e}")