Update app.py

app.py

@@ -6,101 +6,64 @@ import io
 import base64
 import math
 from matplotlib.backends.backend_pdf import PdfPages
-from matplotlib.patches import Rectangle # Replaced FancyBboxPatch
 
 # --- Constants ---
 SPLIT_TIME = "17:30"
 BUSINESS_START = "09:30"
 BUSINESS_END = "01:30"
-BORDER_COLOR = '#CCCCCC' # Light gray for dotted border
+BORDER_COLOR = 'grey' # Changed to grey for the new design
 DATE_COLOR = '#A9A9A9'
-A5_WIDTH_INCH = 5.83
-A5_HEIGHT_INCH = 8.27
-
 
 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"""
+    """
+    Processes the uploaded Excel file to extract, clean, and sort screening times.
+    """
     try:
-        # 读取 Excel，跳过前 8 行
+        # Read Excel, skipping header rows
         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
+        # Extract required columns (G, H, J)
+        df = df.iloc[:, [6, 7, 9]]
         df.columns = ['Hall', 'StartTime', 'EndTime']
 
-        # 清理数据
+        # Clean data: drop rows with missing values
         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) + ' '
-
-        # 保存原始时间字符串用于诊断
-        df['original_end'] = df['EndTime']
+        # Format Hall to "# " format
+        df['Hall'] = df['Hall'].str.extract(r'(\d+)').astype(str) + ' '
 
-        # 转换时间为 datetime 对象
+        # Convert time columns to datetime objects
         base_date = datetime.today().date()
         df['StartTime'] = pd.to_datetime(df['StartTime'])
         df['EndTime'] = pd.to_datetime(df['EndTime'])
 
-        # 设置基准时间
-        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")
-
-        # 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
+        # --- Handle overnight screenings ---
+        # If a show ends after midnight (e.g., 1:30 AM), it belongs to the previous day's schedule.
+        # We handle this by adding a day to its datetime object.
         for idx, row in df.iterrows():
-            end_time = row['EndTime']
-            if end_time.hour < 9:
-                df.at[idx, 'EndTime'] = end_time + timedelta(days=1)
-
-            if row['StartTime'].hour >= 21 and end_time.hour < 9:
-                df.at[idx, 'EndTime'] = end_time + timedelta(days=1)
-
-        # 筛选营业时间内的场次
-        df['time_for_comparison'] = df['EndTime'].apply(
-            lambda x: datetime.combine(base_date, x.time())
-        )
-
-        df.loc[df['time_for_comparison'].dt.hour < 9, 'time_for_comparison'] += timedelta(days=1)
-
-        valid_times = (
-            ((df['time_for_comparison'] >= datetime.combine(base_date, business_start.time())) &
-             (df['time_for_comparison'] <= datetime.combine(base_date + timedelta(days=1), business_end.time())))
-        )
+            if row['EndTime'].hour < 9: # Assuming any end time before 9 AM is part of the previous night
+                df.at[idx, 'EndTime'] = row['EndTime'] + timedelta(days=1)
 
-        df = df[valid_times]
+        # Create a comparable time column that correctly handles the business day logic
+        df['time_for_comparison'] = df['EndTime']
 
-        # 按散场时间排序
+        # Sort screenings by their end time
         df = df.sort_values('EndTime')
 
-        # 分割数据
-        split_time = datetime.strptime(f"{base_date} {SPLIT_TIME}", "%Y-%m-%d %H:%M")
-        split_time_for_comparison = df['time_for_comparison'].apply(
-            lambda x: datetime.combine(base_date, split_time.time())
-        )
+        # Split data into day and night shifts
+        split_datetime = datetime.combine(base_date, datetime.strptime(SPLIT_TIME, "%H:%M").time())
+        part1 = df[df['time_for_comparison'] <= split_datetime].copy()
+        part2 = df[df['time_for_comparison'] > split_datetime].copy()
 
-        part1 = df[df['time_for_comparison'] <= split_time_for_comparison].copy()
-        part2 = df[df['time_for_comparison'] > split_time_for_comparison].copy()
-
-        # 格式化时间显示
+        # Format the time display string (e.g., "5:30")
         for part in [part1, part2]:
             part['EndTime'] = part['EndTime'].dt.strftime('%-I:%M')
 
-        # 精确读取C6单元格
-        date_df = pd.read_excel(
-            file,
-            skiprows=5,    # 跳过前5行（0-4）
-            nrows=1,       # 只读1行
-            usecols=[2],   # 第三列（C列）
-            header=None    # 无表头
-        )
+        # Precisely read the date from cell C6
+        date_df = pd.read_excel(file, skiprows=5, nrows=1, usecols=[2], header=None)
         date_cell = date_df.iloc[0, 0]
 
         try:
-            # ��理不同日期格式
             if isinstance(date_cell, str):
                 date_str = datetime.strptime(date_cell, '%Y-%m-%d').strftime('%Y-%m-%d')
             else:
@@ -111,47 +74,44 @@ def process_schedule(file):
         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)}")
+        st.error(f"Error processing file: {str(e)}")
         return None, None, None
 
-
-def create_print_layout(data, title, date_str):
-    """创建符合新设计要求的打印布局 (PNG 和 PDF)"""
-    if data.empty:
-        return None
-
-    # --- 1. 初始化 Figure ---
-    fig = plt.figure(figsize=(A5_WIDTH_INCH, A5_HEIGHT_INCH), dpi=300)
-    # 使用整个figure区域，手动控制所有边距
-    fig.subplots_adjust(left=0, right=1, top=1, bottom=0)
-
+def _draw_grid_on_figure(fig, data, title, date_str):
+    """
+    Internal helper function to draw the new grid layout onto a Matplotlib figure.
+    """
     plt.rcParams['font.family'] = 'sans-serif'
-    plt.rcParams['font.sans-serif'] = ['Arial Unicode MS']
+    plt.rcParams['font.sans-serif'] = ['Arial Unicode MS'] # Font that supports Chinese characters
 
-    # --- 2. 数据准备和布局计算 ---
     total_items = len(data)
+    if total_items == 0:
+        return
+
     num_cols = 3
-    data_rows = math.ceil(total_items / num_cols)
-    if data_rows == 0: data_rows = 1 # 避免除以零
-
-    # 为header分配固定高度 (占总高度的8%), 剩余为数据区
-    header_height_norm = 0.08
-    data_area_height_norm = 1.0 - header_height_norm
-
-    # 每个数据单元格的尺寸 (标准化坐标, 0-1)
-    cell_height_norm = data_area_height_norm / data_rows
-    cell_width_norm = 1.0 / num_cols
-    
-    # 根据单元格高度计算留白距离 (高度的 1/4 作为边距)
-    # 这将被用作单元格内部的padding
-    spacing_norm_y = cell_height_norm / 4
-    spacing_norm_x = cell_width_norm / 4 # 对称的边距
-
-    # 准备数据 (Z字形排序)
+    num_rows = math.ceil(total_items / num_cols)
+
+    A5_WIDTH_IN = 5.83
+    A5_HEIGHT_IN = 8.27
+
+    # 1. Redesign layout based on precise grid calculations
+    margin_y = (A5_HEIGHT_IN / num_rows) / 4
+    margin_x = margin_y # Use symmetric margins for a cleaner look
+
+    # Prevent margins from becoming too large on pages with few items
+    if A5_WIDTH_IN < 2 * margin_x or A5_HEIGHT_IN < 2 * margin_y:
+        margin_x = A5_WIDTH_IN / 10
+        margin_y = A5_HEIGHT_IN / 10
+
+    printable_width = A5_WIDTH_IN - 2 * margin_x
+    printable_height = A5_HEIGHT_IN - 2 * margin_y
+    cell_width = printable_width / num_cols
+    cell_height = printable_height / num_rows
+
+    # Prepare data: Sort into column-first (Z-style) order for layout
     data_values = data.values.tolist()
     while len(data_values) % num_cols != 0:
-        data_values.append(['', ''])
-
+        data_values.append(['', '']) # Pad data for a full grid
     rows_per_col_layout = math.ceil(len(data_values) / num_cols)
     sorted_data = [['', '']] * len(data_values)
     for i, item in enumerate(data_values):
@@ -162,80 +122,102 @@ def create_print_layout(data, title, date_str):
             if new_index < len(sorted_data):
                 sorted_data[new_index] = item
 
-    # --- 3. 绘制元素 ---
+    # --- Draw each cell onto the figure ---
+    item_counter = 0
+    for idx, (hall, end_time) in enumerate(sorted_data):
+        if not (hall and end_time):
+            continue
+
+        item_counter += 1
+        row_grid = idx // num_cols
+        col_grid = idx % num_cols
+
+        # Calculate position for each cell's axes in Figure coordinates [left, bottom, width, height]
+        ax_left_in = margin_x + col_grid * cell_width
+        ax_bottom_in = margin_y + (num_rows - 1 - row_grid) * cell_height # Y-axis from bottom
+        ax_pos = [
+            ax_left_in / A5_WIDTH_IN,
+            ax_bottom_in / A5_HEIGHT_IN,
+            cell_width / A5_WIDTH_IN,
+            cell_height / A5_HEIGHT_IN,
+        ]
+        ax = fig.add_axes(ax_pos)
+
+        # 2. Change Cell Border to a gray, dotted line
+        for spine in ax.spines.values():
+            spine.set_visible(True)
+            spine.set_linestyle(':') # Dotted line style
+            spine.set_edgecolor(BORDER_COLOR)
+            spine.set_linewidth(1)
+
+        # 4. Add Cell Index Number
+        ax.text(0.07, 0.93, str(item_counter),
+                transform=ax.transAxes,
+                fontsize=9,
+                color='grey',
+                ha='left',
+                va='top')
+
+        # 3. Adjust Cell Content
+        display_text = f"{hall}{end_time}"
+
+        # Dynamically estimate font size to fill 90% of the cell width
+        # This is a heuristic that provides a good balance of size and spacing.
+        font_scale_factor = 1.7
+        estimated_fontsize = (cell_width * 72 / len(display_text)) * font_scale_factor * 0.9
+        max_fontsize = cell_height * 72 * 0.6 # Cap font size to 60% of cell height
+        final_fontsize = min(estimated_fontsize, max_fontsize)
+
+        ax.text(0.5, 0.5, display_text,
+                fontsize=final_fontsize,
+                fontweight='bold',
+                ha='center',
+                va='center',
+                transform=ax.transAxes)
+
+        ax.set_xticks([])
+        ax.set_yticks([])
+
+    # Add date and title to the top margin of the figure
+    title_fontsize = margin_y * 72 * 0.3 # Scale title font size with margin
+    fig.text(margin_x / A5_WIDTH_IN, 1 - (margin_y * 0.5 / A5_HEIGHT_IN),
+             f"{date_str} {title}",
+             fontsize=title_fontsize,
+             color=DATE_COLOR,
+             fontweight='bold',
+             ha='left',
+             va='center')
+
+
+def create_print_layout(data, title, date_str):
+    """
+    Creates the final print-ready output in both PNG and PDF formats using the new grid layout.
+    """
+    if data.empty:
+        return None
 
-    # 绘制页眉 (日期和标题)
-    ax_date = fig.add_axes([0, data_area_height_norm, 1, header_height_norm])
-    ax_date.text(0.02, 0.5, f"{date_str} {title}",
-                 fontsize=12, color=DATE_COLOR, fontweight='bold',
-                 ha='left', va='center', transform=ax_date.transAxes)
-    ax_date.set_axis_off()
+    # --- Create separate figures for PNG and PDF to ensure no cross-contamination ---
+    png_fig = plt.figure(figsize=(5.83, 8.27), dpi=300)
+    pdf_fig = plt.figure(figsize=(5.83, 8.27), dpi=300)
 
-    # 动态计算主字体大小，使其适应单元格高度
-    # 1英寸 = 72点。字体大小占单元格高度的40%左右
-    base_fontsize = (cell_height_norm * A5_HEIGHT_INCH) * 72 * 0.4
+    # --- Draw the layout on both figures ---
+    _draw_grid_on_figure(png_fig, data, title, date_str)
+    _draw_grid_on_figure(pdf_fig, data, title, date_str)
 
-    # 绘制数据单元格
-    for idx, (hall, end_time) in enumerate(sorted_data):
-        if hall and end_time:
-            row_grid = idx // num_cols
-            col_grid = idx % num_cols
-
-            # 计算单元格(Axes)的位置和尺寸
-            ax_left = col_grid * cell_width_norm
-            # Y坐标从下往上，所以要从数据区顶部开始往下减
-            ax_bottom = data_area_height_norm - (row_grid + 1) * cell_height_norm
-            
-            ax = fig.add_axes([ax_left, ax_bottom, cell_width_norm, cell_height_norm])
-            ax.set_axis_off()
-
-            # 要求 2 & 3: 创建点状虚线边框和呼吸感
-            # 通过在Axes内部绘制一个稍小的矩形来实现
-            border_padding_x = spacing_norm_x / 2
-            border_padding_y = spacing_norm_y / 2
-            
-            border_rect = Rectangle(
-                (border_padding_x, border_padding_y),             # 左下角坐标 (x,y)
-                1 - 2 * border_padding_x,                         # 宽度
-                1 - 2 * border_padding_y,                         # 高度
-                transform=ax.transAxes,
-                edgecolor=BORDER_COLOR,
-                facecolor='none',
-                linestyle=':',  # 点状虚线
-                linewidth=2     # 让点更明显
-            )
-            ax.add_patch(border_rect)
-
-            # 要求 4: 在左上角添加序号
-            ax.text(border_padding_x + 0.02, 1 - border_padding_y - 0.02, str(idx + 1),
-                    fontsize=base_fontsize * 0.3, # 序号字体较小
-                    color='grey',
-                    ha='left', va='top',
-                    transform=ax.transAxes)
-
-            # 要求 3: 居中显示主要文本
-            display_text = f"{hall}{end_time}"
-            ax.text(0.5, 0.5, display_text,
-                    fontsize=base_fontsize,
-                    fontweight='bold',
-                    ha='center', va='center',
-                    transform=ax.transAxes)
-
-    # --- 4. 保存到内存 ---
-    # 保存为 PNG
+    # --- Save PNG to a memory buffer ---
     png_buffer = io.BytesIO()
-    fig.savefig(png_buffer, format='png', bbox_inches='tight', pad_inches=0.02)
+    png_fig.savefig(png_buffer, format='png', bbox_inches='tight', pad_inches=0.02)
     png_buffer.seek(0)
     png_base64 = base64.b64encode(png_buffer.getvalue()).decode()
+    plt.close(png_fig)
 
-    # 保存为 PDF
+    # --- Save PDF to a memory buffer ---
     pdf_buffer = io.BytesIO()
     with PdfPages(pdf_buffer) as pdf:
-        pdf.savefig(fig, bbox_inches='tight', pad_inches=0.02)
+        pdf.savefig(pdf_fig, bbox_inches='tight', pad_inches=0.02)
     pdf_buffer.seek(0)
     pdf_base64 = base64.b64encode(pdf_buffer.getvalue()).decode()
-
-    plt.close(fig)
+    plt.close(pdf_fig)
 
     return {
         'png': f'data:image/png;base64,{png_base64}',
@@ -243,29 +225,32 @@ 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>'
+    """
+    Generates the HTML to embed and display a PDF in Streamlit.
+    """
+    pdf_display = f'<iframe src="data:application/pdf;base64,{base64_pdf}" width="100%" height="800" type="application/pdf"></iframe>'
     return pdf_display
 
-# --- Streamlit 界面 (无变化) ---
+# --- Streamlit User Interface ---
 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"])
+uploaded_file = st.file_uploader("上传【放映场次核对表.xls】文件", type=["xls", "xlsx"])
 
 if uploaded_file:
-    part1, part2, date_str = process_schedule(uploaded_file)
+    part1_data, part2_data, date_string = process_schedule(uploaded_file)
 
-    if part1 is not None and part2 is not None:
-        # 生成包含 PNG 和 PDF 的字典
-        part1_output = create_print_layout(part1, "A", date_str)
-        part2_output = create_print_layout(part2, "C", date_str)
+    if part1_data is not None and part2_data is not None:
+        # Generate layouts for both day and night shifts
+        part1_output = create_print_layout(part1_data, "A", date_string)
+        part2_output = create_print_layout(part2_data, "C", date_string)
 
         col1, col2 = st.columns(2)
 
         with col1:
-            st.subheader("白班散场预览（时间 ≤ 17:30）")
+            st.subheader("白班散场预览 (时间 ≤ 17:30)")
             if part1_output:
+                # Use tabs for PDF and PNG previews
                 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)
@@ -275,8 +260,9 @@ if uploaded_file:
                 st.info("白班部分没有数据")
 
         with col2:
-            st.subheader("夜班散场预览（时间 > 17:30）")
+            st.subheader("夜班散场预览 (时间 > 17:30)")
             if part2_output:
+                # Use tabs for PDF and PNG previews
                 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)