app.py

import pandas as pd
import streamlit as st
import matplotlib.pyplot as plt
import matplotlib.font_manager as font_manager
import io
import base64
import os
from datetime import datetime, timedelta
import math
from pypinyin import lazy_pinyin, Style
from matplotlib.backends.backend_pdf import PdfPages

def get_font(size=14):
    """Loads a font file, searching for common names."""
    font_path = "simHei.ttc"
    if not os.path.exists(font_path):
        font_path = "SimHei.ttf"
    return font_manager.FontProperties(fname=font_path, size=size)

def get_pinyin_abbr(text):
    """Gets the first letter of the Pinyin for the first two Chinese characters of a text."""
    if not text:
        return ""
    # Extract the first two Chinese characters
    chars = [c for c in text if '\u4e00' <= c <= '\u9fff']
    if len(chars) < 2:
        chars = chars + [''] * (2 - len(chars))
    else:
        chars = chars[:2]
    # Get the first letter of the Pinyin
    pinyin_list = lazy_pinyin(chars, style=Style.FIRST_LETTER)
    return ''.join(pinyin_list).upper()

def process_schedule(file):
    """Processes the uploaded Excel file to extract and clean schedule data."""
    try:
        # Try to read the date from a specific cell
        date_df = pd.read_excel(file, header=None, skiprows=7, nrows=1, usecols=[3])
        date_str = pd.to_datetime(date_df.iloc[0, 0]).strftime('%Y-%m-%d')
        base_date = pd.to_datetime(date_str).date()
    except:
        # Fallback to the current date if reading fails
        date_str = datetime.today().strftime('%Y-%m-%d')
        base_date = datetime.today().date()
    
    try:
        # Read the main schedule data
        df = pd.read_excel(file, header=9, usecols=[1, 2, 4, 5])
        df.columns = ['Hall', 'StartTime', 'EndTime', 'Movie']
        df['Hall'] = df['Hall'].ffill()
        df.dropna(subset=['StartTime', 'EndTime', 'Movie'], inplace=True)
        df['Hall'] = df['Hall'].astype(str).str.extract(r'(\d+号)')
        
        # Convert times to datetime objects
        df['StartTime_dt'] = pd.to_datetime(df['StartTime'], format='%H:%M', errors='coerce').apply(
            lambda t: t.replace(year=base_date.year, month=base_date.month, day=base_date.day) if pd.notnull(t) else t
        )
        df['EndTime_dt'] = pd.to_datetime(df['EndTime'], format='%H:%M', errors='coerce').apply(
            lambda t: t.replace(year=base_date.year, month=base_date.month, day=base_date.day) if pd.notnull(t) else t
        )
        # Handle overnight screenings
        df.loc[df['EndTime_dt'] < df['StartTime_dt'], 'EndTime_dt'] += timedelta(days=1)
        df = df.sort_values(['Hall', 'StartTime_dt'])
        
        # Merge consecutive screenings of the same movie
        merged_rows = []
        for hall, group in df.groupby('Hall'):
            group = group.sort_values('StartTime_dt')
            current = None
            for _, row in group.iterrows():
                if current is None:
                    current = row.copy()
                else:
                    if row['Movie'] == current['Movie']:
                        current['EndTime_dt'] = row['EndTime_dt']
                    else:
                        merged_rows.append(current)
                        current = row.copy()
            if current is not None:
                merged_rows.append(current)
        
        merged_df = pd.DataFrame(merged_rows)
        
        # Adjust start and end times
        merged_df['StartTime_dt'] = merged_df['StartTime_dt'] - timedelta(minutes=10)
        merged_df['EndTime_dt'] = merged_df['EndTime_dt'] - timedelta(minutes=5)
        
        merged_df['StartTime_str'] = merged_df['StartTime_dt'].dt.strftime('%H:%M')
        merged_df['EndTime_str'] = merged_df['EndTime_dt'].dt.strftime('%H:%M')
        
        return merged_df[['Hall', 'Movie', 'StartTime_str', 'EndTime_str']], date_str
    except:
        return None, date_str

def create_print_layout(data, date_str):
    """Creates the PNG and PDF print layouts from the schedule data."""
    if data is None or data.empty:
        return None
    
    # Create PNG image for preview
    png_fig = plt.figure(figsize=(8.27, 11.69), dpi=300)
    png_ax = png_fig.add_subplot(111)
    png_ax.set_axis_off()
    png_fig.subplots_adjust(left=0.02, right=0.98, top=0.98, bottom=0.02)
    
    # Create PDF image for download
    pdf_fig = plt.figure(figsize=(8.27, 11.69), dpi=300)
    pdf_ax = pdf_fig.add_subplot(111)
    pdf_ax.set_axis_off()
    pdf_fig.subplots_adjust(left=0.02, right=0.98, top=0.98, bottom=0.02)
    
    def process_figure(fig, ax, is_pdf=False):
        """Shared function to draw the schedule on a matplotlib figure."""
        date_font = get_font(12)
        
        # Calculate row height to fill the page
        total_movies = len(data)
        # Add 2 to the total to create top and bottom margins
        total_divisions = total_movies + 2 
        
        available_height_ax = 0.98 - 0.02  # Use 96% of the page height
        row_height_ax = available_height_ax / total_divisions

        # Convert relative height to point size for fonts
        fig_height_pts = fig.get_figheight() * 72
        row_height_pts = row_height_ax * fig_height_pts
        
        # Set font size to 90% of the calculated row height
        font_size = row_height_pts * 0.9
        hall_font_size = font_size * 0.8
        
        hall_font = get_font(hall_font_size)
        movie_font = get_font(font_size)
        
        # Add date stamp to the top left
        ax.text(0.00, 1.00, date_str, fontsize=12, color='#A9A9A9',
                ha='left', va='top', fontproperties=date_font, transform=ax.transAxes, zorder=2)
        
        halls = sorted(data['Hall'].unique(), key=lambda h: int(h.replace('号','')) if h else 0)
        
        # Start drawing from the top, leaving one row_height as a margin
        y_position = 0.98 - row_height_ax
        
        for i, hall in enumerate(halls):
            hall_data = data[data['Hall'] == hall]
            hall_num = hall.replace("号", "")
            hall_text = f"${hall_num}^{{\\#}}$"
            movie_count = 1
            
            for _, row in hall_data.iterrows():
                # Hall Number (left-aligned)
                if movie_count == 1:
                    ax.text(0.03, y_position, hall_text,
                            fontsize=hall_font_size, fontweight='bold',
                            ha='left', va='top', fontproperties=hall_font,
                            transform=ax.transAxes, zorder=2)
                
                pinyin_abbr = get_pinyin_abbr(row['Movie'])
                
                # New content order and alignment
                # 1. Movie Name (right-aligned)
                ax.text(0.55, y_position, row['Movie'],
                        fontsize=font_size, ha='right', va='top', fontproperties=movie_font,
                        transform=ax.transAxes, zorder=2)
                
                # 2. Sequence Number (left-aligned)
                ax.text(0.57, y_position, f"{movie_count}.",
                        fontsize=font_size, ha='left', va='top', fontproperties=movie_font,
                        transform=ax.transAxes, zorder=2)
                
                # 3. Pinyin Abbreviation (left-aligned)
                ax.text(0.63, y_position, pinyin_abbr,
                        fontsize=font_size, ha='left', va='top', fontproperties=movie_font,
                        transform=ax.transAxes, zorder=2)
                        
                # 4. Time (left-aligned)
                ax.text(0.72, y_position, f"{row['StartTime_str']} - {row['EndTime_str']}",
                        fontsize=font_size, ha='left', va='top', fontproperties=movie_font,
                        transform=ax.transAxes, zorder=2)
                
                y_position -= row_height_ax
                movie_count += 1
            
            # Draw a black line to separate halls
            if i < len(halls) - 1:
                line_y = y_position + (row_height_ax / 2)
                ax.plot([0.03, 0.97], [line_y, line_y], color='black', linewidth=1, transform=ax.transAxes, zorder=1)

    # Process both the PNG and PDF figures
    process_figure(png_fig, png_ax)
    process_figure(pdf_fig, pdf_ax, is_pdf=True)
    
    # Save PNG to buffer
    png_buffer = io.BytesIO()
    png_fig.savefig(png_buffer, format='png', bbox_inches='tight', pad_inches=0.05)
    png_buffer.seek(0)
    image_base64 = base64.b64encode(png_buffer.getvalue()).decode()
    plt.close(png_fig)
    
    # Save PDF to buffer
    pdf_buffer = io.BytesIO()
    with PdfPages(pdf_buffer) as pdf:
        pdf.savefig(pdf_fig, bbox_inches='tight', pad_inches=0.05)
    pdf_buffer.seek(0)
    pdf_base64 = base64.b64encode(pdf_buffer.getvalue()).decode()
    plt.close(pdf_fig)
    
    return {
        'png': f"data:image/png;base64,{image_base64}",
        'pdf': f"data:application/pdf;base64,{pdf_base64}"
    }

def display_pdf(base64_pdf):
    """Generates the HTML to embed a PDF in Streamlit."""
    pdf_display = f"""
    <iframe src="{base64_pdf}" width="100%" height="800" type="application/pdf"></iframe>
    """
    return pdf_display

# --- Streamlit App ---
st.set_page_config(page_title="LED 屏幕时间表打印", layout="wide")
st.title("LED 屏幕时间表打印")

uploaded_file = st.file_uploader("选择打开【放映时间核对表.xls】文件", accept_multiple_files=False, type=["xls"])

if uploaded_file:
    with st.spinner("文件正在处理中，请稍候..."):
        schedule, date_str = process_schedule(uploaded_file)
        if schedule is not None:
            output = create_print_layout(schedule, date_str)
            
            # Create tabs for PDF and PNG previews
            tab1, tab2 = st.tabs(["PDF 预览", "PNG 预览"])
            
            with tab1:
                st.markdown(display_pdf(output['pdf']), unsafe_allow_html=True)
            
            with tab2:
                st.image(output['png'], use_container_width=True)
        else:
            st.error("无法处理文件，请检查文件格式或内容是否正确。")