Dockerfile

# 轻量级LaTeX处理器，适用于Hugging Face Spaces
FROM ubuntu:22.04

# 设置环境变量
ENV DEBIAN_FRONTEND=noninteractive
ENV TZ=Asia/Shanghai
ENV PYTHONUNBUFFERED=1

# 安装系统依赖
RUN apt-get update && apt-get install -y \
    # 基础工具
    curl \
    wget \
    git \
    unzip \
    # Python环境
    python3 \
    python3-pip \
    python3-venv \
    # LaTeX环境
    texlive-full \
    texlive-lang-chinese \
    texlive-xetex \
    texlive-luatex \
    # 字体支持
    fonts-noto-cjk \
    fonts-noto-cjk-extra \
    fonts-wqy-microhei \
    fonts-wqy-zenhei \
    # PDF工具
    poppler-utils \
    # 其他工具
    pandoc \
    imagemagick \
    && rm -rf /var/lib/apt/lists/*

# 安装中文字体（基于教程中提到的字体）
RUN mkdir -p /usr/share/fonts/chinese
COPY fonts/ /usr/share/fonts/chinese/ 2>/dev/null || true

# 如果没有本地字体文件，下载一些常用中文字体
RUN cd /usr/share/fonts/chinese && \
    # 下载思源字体
    wget -q https://github.com/adobe-fonts/source-han-sans/releases/download/2.004R/SourceHanSansCN.zip && \
    unzip -q SourceHanSansCN.zip && \
    rm SourceHanSansCN.zip && \
    # 更新字体缓存
    fc-cache -fv

# 设置工作目录
WORKDIR /app

# 安装Python依赖
RUN pip3 install --no-cache-dir \
    gradio \
    streamlit \
    fastapi \
    uvicorn \
    python-multipart \
    aiofiles \
    jinja2 \
    PyPDF2 \
    reportlab \
    matplotlib \
    pillow \
    numpy \
    pandas

# 创建应用目录结构
RUN mkdir -p /app/templates \
    && mkdir -p /app/uploads \
    && mkdir -p /app/outputs \
    && mkdir -p /app/temp

# 如果有本地requirements.txt，复制并安装
COPY requirements.txt /app/ 2>/dev/null || echo "gradio" > /app/requirements.txt
RUN if [ -f /app/requirements.txt ]; then pip3 install --no-cache-dir -r /app/requirements.txt; fi

# 创建Python应用文件
RUN python3 -c "
import os
app_code = '''import gradio as gr
import subprocess
import tempfile
import os
import shutil
from pathlib import Path

def compile_latex(latex_code, compiler=\"xelatex\"):
    \"\"\"编译LaTeX代码并返回PDF\"\"\"
    with tempfile.TemporaryDirectory() as temp_dir:
        # 写入LaTeX文件
        tex_file = os.path.join(temp_dir, \"document.tex\")
        with open(tex_file, \"w\", encoding=\"utf-8\") as f:
            f.write(latex_code)
        
        try:
            # 编译LaTeX
            result = subprocess.run(
                [compiler, \"-interaction=nonstopmode\", \"-output-directory\", temp_dir, tex_file],
                capture_output=True,
                text=True,
                timeout=30
            )
            
            pdf_file = os.path.join(temp_dir, \"document.pdf\")
            log_file = os.path.join(temp_dir, \"document.log\")
            
            # 读取日志
            log_content = \"\"
            if os.path.exists(log_file):
                with open(log_file, \"r\", encoding=\"utf-8\", errors=\"ignore\") as f:
                    log_content = f.read()
            
            if os.path.exists(pdf_file):
                # 复制PDF到输出目录
                output_pdf = \"/app/outputs/output.pdf\"
                shutil.copy2(pdf_file, output_pdf)
                return output_pdf, f\"编译成功!\\n\\n编译日志:\\n{log_content}\"
            else:
                return None, f\"编译失败!\\n\\n错误信息:\\n{result.stderr}\\n\\n日志:\\n{log_content}\"
                
        except subprocess.TimeoutExpired:
            return None, \"编译超时（30秒）\"
        except Exception as e:
            return None, f\"编译出错: {str(e)}\"

# 创建Gradio界面
def create_interface():
    with gr.Blocks(title=\"LaTeX编译器 - 支持中文\") as demo:
        gr.Markdown(\"# LaTeX编译器（支持中文字体）\")
        gr.Markdown(\"基于教程中的完整texlive环境，支持XeLaTeX编译中文文档\")
        
        with gr.Row():
            with gr.Column():
                latex_input = gr.Textbox(
                    label=\"LaTeX代码\",
                    placeholder=\"请输入LaTeX代码...\",
                    lines=20,
                    value=\"\"\"\\\\documentclass{article}
\\\\usepackage{xeCJK}
\\\\setCJKmainfont{SimSun}

\\\\title{中文LaTeX示例}
\\\\author{作者姓名}
\\\\date{\\\\today}

\\\\begin{document}
\\\\maketitle

\\\\section{介绍}
这是一个支持中文的LaTeX文档示例。

\\\\section{数学公式}
爱因斯坦的质能方程：$E = mc^2$

\\\\end{document}\"\"\"
                )
                
                compiler = gr.Radio(
                    choices=[\"xelatex\", \"pdflatex\", \"lualatex\"],
                    value=\"xelatex\",
                    label=\"编译器选择\"
                )
                
                compile_btn = gr.Button(\"编译PDF\", variant=\"primary\")
            
            with gr.Column():
                output_file = gr.File(label=\"生成的PDF\")
                log_output = gr.Textbox(label=\"编译日志\", lines=15, interactive=False)
        
        compile_btn.click(
            fn=compile_latex,
            inputs=[latex_input, compiler],
            outputs=[output_file, log_output]
        )
    
    return demo

if __name__ == \"__main__\":
    # 确保输出目录存在
    os.makedirs(\"/app/outputs\", exist_ok=True)
    
    # 启动应用
    demo = create_interface()
    demo.launch(
        server_name=\"0.0.0.0\",
        server_port=7860,
        share=False
    )
'''

with open('/app/app.py', 'w', encoding='utf-8') as f:
    f.write(app_code)
"

# 设置权限
RUN chmod +x /app/app.py

# 暴露端口
EXPOSE 7860

# 健康检查
HEALTHCHECK --interval=30s --timeout=10s --start-period=5s --retries=3 \
  CMD curl -f http://localhost:7860/health || exit 1

# 启动命令
CMD ["python3", "/app/app.py"]