Update app.py

app.py

@@ -7,224 +7,240 @@ import base64
 import matplotlib.gridspec as gridspec
 import math
 from matplotlib.backends.backend_pdf import PdfPages
-from matplotlib.patches import Rectangle # Replaced FancyBboxPatch
 
-# --- Constants ---
+# 常量定义
 SPLIT_TIME = "17:30"
 BUSINESS_START = "09:30"
 BUSINESS_END = "01:30"
-BORDER_COLOR = '#A9A9A9'
+BORDER_COLOR = 'grey'  # 边框颜色更新为灰色
 DATE_COLOR = '#A9A9A9'
-SEQ_COLOR = '#A9A9A9' # Color for the new serial number
+A5_WIDTH_IN = 5.83  # A5 宽度 (英寸)
+A5_HEIGHT_IN = 8.27 # A5 高度 (英寸)
 
 def process_schedule(file):
     """处理上传的 Excel 文件，生成排序和分组后的打印内容"""
     try:
         # 读取 Excel，跳过前 8 行
         df = pd.read_excel(file, skiprows=8)
-
+        
         # 提取所需列 (G9, H9, J9)
         df = df.iloc[:, [6, 7, 9]]  # G, H, J 列
         df.columns = ['Hall', 'StartTime', 'EndTime']
-
+        
         # 清理数据
         df = df.dropna(subset=['Hall', 'StartTime', 'EndTime'])
-
+        
         # 转换影厅格式为 "#号" 格式
-        df['Hall'] = df['Hall'].str.extract(r'(\d+)号').astype(str) + ' '
-
+        df['Hall'] = df['Hall'].str.extract(r'(\d+)').astype(str) + '号厅 '
+        
         # 保存原始时间字符串用于诊断
         df['original_end'] = df['EndTime']
-
+        
         # 转换时间为 datetime 对象
         base_date = datetime.today().date()
-        # Using errors='coerce' will turn unparseable times into NaT (Not a Time)
-        df['StartTime'] = pd.to_datetime(df['StartTime'], errors='coerce')
-        df['EndTime'] = pd.to_datetime(df['EndTime'], errors='coerce')
-        df = df.dropna(subset=['StartTime', 'EndTime']) # Drop rows where time conversion failed
-
+        df['StartTime'] = pd.to_datetime(df['StartTime'])
+        df['EndTime'] = pd.to_datetime(df['EndTime'])
+        
         # 设置基准时间
-        business_start_time = datetime.strptime(BUSINESS_START, "%H:%M").time()
-        business_end_time = datetime.strptime(BUSINESS_END, "%H:%M").time()
-
-        # 处理跨天情况：结束时间小于开始时间，则结束时间加一天
-        # This logic handles cases like 9:30 AM to 1:30 AM (next day)
-        df['EndTime_adjusted'] = df.apply(
-            lambda row: row['EndTime'] + timedelta(days=1) if row['EndTime'].time() < row['StartTime'].time() else row['EndTime'],
-            axis=1
-        )
+        business_start = datetime.strptime(f"{base_date} {BUSINESS_START}", "%Y-%m-%d %H:%M")
+        business_end = datetime.strptime(f"{base_date} {BUSINESS_END}", "%Y-%m-%d %H:%M")
         
-        # 按散场时间排序 (using the adjusted time)
-        df = df.sort_values('EndTime_adjusted')
+        # 处理跨天情况
+        if business_end < business_start:
+            business_end += timedelta(days=1)
+        
+        # 标准化所有时间到同一天
+        for idx, row in df.iterrows():
+            end_time = row['EndTime']
+            # 如果结束时间在凌晨0点到9点之间，则认为是第二天的场次
+            if end_time.hour < 9:
+                df.at[idx, 'EndTime'] = end_time + timedelta(days=1)
+            # 如果开始时间在21点后，结束时间在9点前，也认为是跨天场次
+            elif row['StartTime'].hour >= 21 and end_time.hour < 9:
+                df.at[idx, 'EndTime'] = end_time + timedelta(days=1)
+
+        # 为了正确筛选和排序，创建一个统一比较的时间列
+        # 该列将所有时间都映射到正确的日期上
+        df['time_for_comparison'] = df['EndTime'].apply(
+            lambda x: x if x.hour >= 9 else x + timedelta(days=1) if x.date() == base_date else x
+        )
 
+        # 筛选营业时间内的场次
+        valid_times = (df['time_for_comparison'] >= business_start) & (df['time_for_comparison'] <= business_end)
+        df = df[valid_times]
+        
+        # 按散场时间排序
+        df = df.sort_values('time_for_comparison')
+        
         # 分割数据
-        split_dt = datetime.strptime(SPLIT_TIME, "%H:%M").time()
+        split_time = datetime.strptime(f"{base_date} {SPLIT_TIME}", "%Y-%m-%d %H:%M")
         
-        part1 = df[df['EndTime_adjusted'].dt.time <= split_dt].copy()
-        part2 = df[df['EndTime_adjusted'].dt.time > split_dt].copy()
-
-        # 格式化时间显示 (use original EndTime for display)
+        part1 = df[df['time_for_comparison'] <= split_time].copy()
+        part2 = df[df['time_for_comparison'] > split_time].copy()
+        
+        # 格式化时间显示
         for part in [part1, part2]:
-            part['EndTime_formatted'] = part['EndTime'].dt.strftime('%-I:%M')
-
-        # 读取日期单元格 C6
-        date_df = pd.read_excel(file, skiprows=5, nrows=1, usecols=[2], header=None)
+            part['EndTime'] = part['EndTime'].dt.strftime('%-H:%M')
+        
+        # 精确读取C6单元格的日期
+        date_df = pd.read_excel(
+            file,
+            skiprows=5,
+            nrows=1,
+            usecols=[2],
+            header=None
+        )
         date_cell = date_df.iloc[0, 0]
-
+        
         try:
             if isinstance(date_cell, str):
-                # Assuming format like '2023-10-27'
                 date_str = datetime.strptime(date_cell, '%Y-%m-%d').strftime('%Y-%m-%d')
             else:
-                # Assuming it's a datetime object
                 date_str = pd.to_datetime(date_cell).strftime('%Y-%m-%d')
         except:
             date_str = datetime.today().strftime('%Y-%m-%d')
-
-        return part1[['Hall', 'EndTime_formatted']], part2[['Hall', 'EndTime_formatted']], date_str
-
+        
+        return part1[['Hall', 'EndTime']], part2[['Hall', 'EndTime']], date_str
+    
     except Exception as e:
         st.error(f"处理文件时出错: {str(e)}")
         return None, None, None
 
-
 def create_print_layout(data, title, date_str):
-    """
-    创建符合新要求的打印布局 (PNG 和 PDF)。
-    1. 动态计算边距。
-    2. 使用灰色虚线圆点作为单元格边框。
-    3. 单元格内容区域为单元格的90%。
-    4. 在左上角添加灰色序号。
-    """
+    """创建符合新布局要求的打印视图 (PNG 和 PDF)"""
     if data.empty:
         return None
 
-    # --- Constants ---
-    A5_WIDTH_IN = 5.83
-    A5_HEIGHT_IN = 8.27
-    DPI = 300
-    NUM_COLS = 3
-
-    # --- Setup Figure ---
-    fig = plt.figure(figsize=(A5_WIDTH_IN, A5_HEIGHT_IN), dpi=DPI)
-    
-    # --- Font Setup ---
-    plt.rcParams['font.family'] = 'sans-serif'
-    plt.rcParams['font.sans-serif'] = ['Arial Unicode MS', 'Heiti TC', 'sans-serif']
-
-
-    # --- Data Preparation ---
+    # --- 布局计算 (要求1) ---
     total_items = len(data)
-    # Augment data with an original index for numbering
-    data_values_with_index = [(i, row) for i, row in enumerate(data.values.tolist())]
-    
-    # Pad data to be a multiple of NUM_COLS
-    padded_total = math.ceil(total_items / NUM_COLS) * NUM_COLS
-    while len(data_values_with_index) < padded_total:
-        data_values_with_index.append((None, ['', '']))
-
-    num_rows = padded_total // NUM_COLS
-
-    # --- Layout Calculation (Request 1) ---
-    if num_rows > 0:
-        # "A5 paper height / num_rows / 4 is the padding for all sides"
-        padding_in = (A5_HEIGHT_IN / num_rows / 4)
-        # Cap padding to prevent it from being excessively large
-        padding_in = min(padding_in, 0.5) 
-    else:
-        padding_in = 0.25 # Default padding if no rows
-
-    # Convert padding to relative figure coordinates for subplots_adjust
-    left_margin = padding_in / A5_WIDTH_IN
-    right_margin = 1 - left_margin
-    bottom_margin = padding_in / A5_HEIGHT_IN
-    top_margin = 1 - bottom_margin
-    
-    # Adjust overall figure margins
-    fig.subplots_adjust(left=left_margin, right=right_margin, top=top_margin, bottom=bottom_margin, hspace=0.4, wspace=0.4)
-
-    # --- Grid & Font Size ---
-    gs = gridspec.GridSpec(num_rows + 1, NUM_COLS, height_ratios=[0.2] + [1] * num_rows, figure=fig)
+    num_cols = 3
+    # 计算行数，小数位进一
+    num_rows = math.ceil(total_items / num_cols)
+    if num_rows == 0: # 避免除以零
+        return None
 
-    if num_rows > 0:
-        content_area_height_in = A5_HEIGHT_IN * (top_margin - bottom_margin)
-        cell_height_in = content_area_height_in / num_rows * (1 - fig.subplotpars.hspace)
-        base_fontsize = min(40, max(10, cell_height_in * 72 * 0.4)) # 72 pt/inch, 40% of cell height
-    else:
-        base_fontsize = 20
+    # 计算留空距离
+    # A5高度除以行数，再除以4，得到上下左右留空
+    margin_y = (A5_HEIGHT_IN / num_rows) / 4
+    margin_x = margin_y # 保持长宽比
+
+    # 计算绘图区域在figure中的相对位置
+    left_margin = margin_x / A5_WIDTH_IN
+    right_margin = 1 - (margin_x / A5_WIDTH_IN)
+    bottom_margin = margin_y / A5_HEIGHT_IN
+    top_margin = 1 - (margin_y / A5_HEIGHT_IN)
+
+    # --- 创建图形 ---
+    # 使用相同的 figure 和函数来确保 PNG 和 PDF 的一致性
+    fig = plt.figure(figsize=(A5_WIDTH_IN, A5_HEIGHT_IN), dpi=300)
+
+    # --- 内部绘图函数 ---
+    def process_figure(figure):
+        plt.rcParams['font.family'] = 'sans-serif'
+        plt.rcParams['font.sans-serif'] = ['Arial Unicode MS', 'Heiti TC', 'SimHei'] # 添加备用中文字体
+
+        # 创建网格布局
+        gs = gridspec.GridSpec(
+            num_rows + 1, num_cols,
+            figure=figure,
+            left=left_margin, right=right_margin,
+            bottom=bottom_margin, top=top_margin,
+            hspace=0, wspace=0, # 单元格之间无间距，边框会自己形成网格
+            height_ratios=[0.2] + [1] * num_rows # 日期行高度较小
+        )
 
-    # --- Z-Sort (Column-major) Data for Layout ---
-    rows_per_col_layout = num_rows
-    sorted_data = [(None, ['',''])] * padded_total
-    for i, item_tuple in enumerate(data_values_with_index):
-        if item_tuple[0] is not None:
-            original_data_index = i # Index from the time-sorted list
-            row_in_col = original_data_index % rows_per_col_layout
-            col_idx = original_data_index // rows_per_col_layout
-            new_grid_index = row_in_col * NUM_COLS + col_idx
-            if new_grid_index < len(sorted_data):
-                sorted_data[new_grid_index] = item_tuple
+        # 动态计算字体大小
+        cell_height_in = (A5_HEIGHT_IN - 2 * margin_y) / num_rows
+        # 字体大小与单元格高度关联，0.35为经验系数，可以微调
+        base_fontsize = cell_height_in * 72 * 0.35
 
-    # --- Drawing Logic ---
-    for grid_idx, item_tuple in enumerate(sorted_data):
-        original_index, (hall, end_time) = item_tuple
+        data_values = data.values.tolist()
         
-        if original_index is not None:
-            row_grid = grid_idx // NUM_COLS + 1 # +1 because date is in row 0
-            col_grid = grid_idx % NUM_COLS
-            
-            ax = fig.add_subplot(gs[row_grid, col_grid])
-            ax.set_axis_off()
-
-            # --- Cell Border (Request 2) & Content Area (Request 3) ---
-            # Draw a dotted rectangle. Content will be placed inside this.
-            # Making the rect slightly smaller creates a visual 90% area.
-            cell_border = Rectangle((0.05, 0.05), 0.9, 0.9,
-                                    edgecolor=BORDER_COLOR,
-                                    facecolor='none',
-                                    linestyle=(0, (1, 1.5)), # Dotted line with round caps
-                                    linewidth=1,
-                                    transform=ax.transAxes,
-                                    clip_on=False)
-            ax.add_patch(cell_border)
+        # 补全空位，以对齐网格
+        while len(data_values) % num_cols != 0:
+            data_values.append(['', ''])
+
+        rows_per_col_layout = math.ceil(len(data_values) / num_cols)
+
+        # 按列优先（Z字形）排序数据
+        sorted_data = [['', '']] * len(data_values)
+        item_count = 0
+        for i, item in enumerate(data_values):
+            if item[0] and item[1]:
+                row_in_col = i % rows_per_col_layout
+                col_idx = i // rows_per_col_layout
+                new_index = row_in_col * num_cols + col_idx
+                if new_index < len(sorted_data):
+                    sorted_data[new_index] = item
+                    item_count += 1
+        
+        # 重新整理数据，移除空位，方便添加序号
+        final_data = [item for item in sorted_data if item[0] and item[1]]
 
-            # --- Cell Content ---
+        # 绘制数据单元格
+        for idx, (hall, end_time) in enumerate(final_data):
+            row_grid = idx // num_cols + 1
+            col_grid = idx % num_cols
+            
+            ax = figure.add_subplot(gs[row_grid, col_grid])
+
+            # --- 修改开始：绘制单元格 (要求2) ---
+            # 设置所有边框为灰色点状虚线
+            for spine in ax.spines.values():
+                spine.set_visible(True)
+                spine.set_linestyle((0, (1, 2))) # 点状线样式
+                spine.set_edgecolor(BORDER_COLOR)
+                spine.set_linewidth(1)
+            
+            # --- 内容与序号 (要求3 & 4) ---
+            # 主内容
             display_text = f"{hall}{end_time}"
+            # 使用 text 函数并设置 bbox 来控制内容宽度为90%
             ax.text(0.5, 0.5, display_text,
                     fontsize=base_fontsize,
                     fontweight='bold',
-                    ha='center', va='center',
-                    transform=ax.transAxes)
-
-            # --- Cell Numbering (Request 4) ---
-            # Serial number is original_index + 1
-            ax.text(0.12, 0.82, str(original_index + 1),
-                    fontsize=base_fontsize * 0.5,
-                    color=SEQ_COLOR,
+                    ha='center',
+                    va='center',
+                    transform=ax.transAxes,
+                    bbox=dict(boxstyle='square,pad=0', fc='none', ec='none',
+                              width=0.9, height=0.9) # 隐形 bbox 控制宽度
+                   )
+
+            # 左上角序号
+            ax.text(0.08, 0.92, str(idx + 1),
+                    fontsize=base_fontsize * 0.4, # 序号字体较小
+                    color=DATE_COLOR,
                     fontweight='normal',
-                    ha='center', va='center',
+                    ha='left',
+                    va='top',
                     transform=ax.transAxes)
-
-    # --- Date Header ---
-    ax_date = fig.add_subplot(gs[0, :])
-    ax_date.set_axis_off()
-    ax_date.text(0, 0.5, f"{date_str} {title}",
-                 fontsize=base_fontsize * 0.6,
-                 color=DATE_COLOR,
-                 fontweight='bold',
-                 ha='left', va='center',
-                 transform=ax_date.transAxes)
-
-    # --- Save to Buffers ---
-    # Save PNG
+            
+            ax.set_xticks([])
+            ax.set_yticks([])
+
+        # 添加日期和标��
+        ax_date = figure.add_subplot(gs[0, :])
+        ax_date.text(0.01, 0.5, f"{date_str} {title}",
+                     fontsize=base_fontsize * 0.6,
+                     color=DATE_COLOR,
+                     fontweight='bold',
+                     ha='left',
+                     va='center',
+                     transform=ax_date.transAxes)
+        ax_date.set_axis_off()
+
+    # --- 处理和保存图形 ---
+    process_figure(fig)
+
+    # 保存为 PNG
     png_buffer = io.BytesIO()
-    fig.savefig(png_buffer, format='png', bbox_inches='tight', pad_inches=0.02)
+    fig.savefig(png_buffer, format='png', bbox_inches='tight', pad_inches=0)
     png_buffer.seek(0)
     png_base64 = base64.b64encode(png_buffer.getvalue()).decode()
 
-    # Save PDF
+    # 保存为 PDF
     pdf_buffer = io.BytesIO()
-    fig.savefig(pdf_buffer, format='pdf', bbox_inches='tight', pad_inches=0.02)
+    fig.savefig(pdf_buffer, format='pdf', bbox_inches='tight', pad_inches=0)
     pdf_buffer.seek(0)
     pdf_base64 = base64.b64encode(pdf_buffer.getvalue()).decode()
     
@@ -237,45 +253,43 @@ def create_print_layout(data, title, date_str):
 
 def display_pdf(base64_pdf):
     """在Streamlit中嵌入显示PDF"""
-    pdf_display = f'<iframe src="{base64_pdf}" width="100%" height="800" type="application/pdf"></iframe>'
+    pdf_display = f'<iframe src="data:application/pdf;base64,{base64_pdf}" width="100%" height="800" type="application/pdf"></iframe>'
     return pdf_display
 
-# --- Streamlit UI ---
+# --- Streamlit 界面 ---
 st.set_page_config(page_title="散厅时间快捷打印", layout="wide")
 st.title("散厅时间快捷打印")
 
 uploaded_file = st.file_uploader("上传【放映场次核对表.xls】文件", type=["xls", "xlsx"])
 
 if uploaded_file:
-    # Use new column name 'EndTime_formatted' for display
     part1, part2, date_str = process_schedule(uploaded_file)
+
     if part1 is not None and part2 is not None:
-        part1_data_for_layout = part1[['Hall', 'EndTime_formatted']]
-        part2_data_for_layout = part2[['Hall', 'EndTime_formatted']]
-        
-        part1_output = create_print_layout(part1_data_for_layout, "A", date_str)
-        part2_output = create_print_layout(part2_data_for_layout, "C", date_str)
+        with st.spinner('正在生成预览图...'):
+            part1_output = create_print_layout(part1, "白班", date_str)
+            part2_output = create_print_layout(part2, "晚班", date_str)
 
         col1, col2 = st.columns(2)
 
         with col1:
-            st.subheader("白班散场预览（散场时间 ≤ 17:30）")
+            st.subheader(f"白班散场预览 (≤{SPLIT_TIME})")
             if part1_output:
                 tab1_1, tab1_2 = st.tabs(["PDF 预览", "PNG 预览"])
                 with tab1_1:
                     st.markdown(display_pdf(part1_output['pdf']), unsafe_allow_html=True)
                 with tab1_2:
-                    st.image(part1_output['png'])
+                    st.image(f"data:image/png;base64,{part1_output['png']}")
             else:
                 st.info("白班部分没有数据")
 
         with col2:
-            st.subheader("夜班散场预览（散场时间 > 17:30）")
+            st.subheader(f"晚班散场预览 (>{SPLIT_TIME})")
             if part2_output:
                 tab2_1, tab2_2 = st.tabs(["PDF 预览", "PNG 预览"])
                 with tab2_1:
                     st.markdown(display_pdf(part2_output['pdf']), unsafe_allow_html=True)
                 with tab2_2:
-                    st.image(part2_output['png'])
+                    st.image(f"data:image/png;base64,{part2_output['png']}")
             else:
-                st.info("夜班部分没有数据")
+                st.info("晚班部分没有数据")