docker环境利用centos7镜像 + miniconda + python3.9 + wkhtmltopdf 构建html转图片服务

1、目录结构

html2image
——Dockerfile
——main.py
——requirements.txt

2、Dockerfile

FROM centos:7

WORKDIR /app
COPY . /app/

RUN curl -O https://github.com/wkhtmltopdf/packaging/releases/download/0.12.6-1/wkhtmltox-0.12.6-1.centos7.x86_64.rpm \
    && curl -O https://mirrors.tuna.tsinghua.edu.cn/anaconda/miniconda/Miniconda3-latest-Linux-x86_64.sh \
    && curl -o /etc/yum.repos.d/CentOS-Base.repo https://repo.huaweicloud.com/repository/conf/CentOS-7-reg.repo \
    && yum install -y fontconfig freetype libX11 libXext libXrender libjpeg libpng openssl xorg-x11-fonts-75dpi xorg-x11-fonts-Type1 \
    && rpm -ivh wkhtmltox-0.12.6-1.centos7.x86_64.rpm\
    && rm -rf /usr/share/fonts/* && cp simsun.ttc /usr/share/fonts/ \
    && fc-cache -fv \
    && bash Miniconda3-latest-Linux-x86_64.sh -b -p /opt/conda \
    && rm -f Miniconda3-latest-Linux-x86_64.sh \
    && ln -s /opt/conda/etc/profile.d/conda.sh /etc/profile.d/conda.sh \
    && echo ". /opt/conda/etc/profile.d/conda.sh" >> ~/.bashrc \
    && source /root/.bashrc \
    && conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/ \
    && conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/r/ \
    && conda config --add channels https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/msys2 \
    && conda config --set custom_channels.auto https://mirrors.tuna.tsinghua.edu.cn/anaconda/cloud/ \
    && conda config --set show_channel_urls yes \
    && conda create -n py39 python=3.9 -y \
    && conda activate py39 \
    && pip install -r requirements.txt -i https://mirrors.tuna.tsinghua.edu.cn/pypi/web/simple 

EXPOSE 54248
CMD ["bash", "-c", "source /opt/conda/etc/profile.d/conda.sh && conda activate py39 && uvicorn main:app --host 0.0.0.0 --port 54248"]

• 由于我用的是centos7镜像，所以最好提前下载好wkhtmltox的安装包
• Miniconda3的安装脚本Miniconda3-latest-Linux-x86_64.sh
• 华为镜像源（用来替换掉centos7里面的镜像源，原来的太慢了，我是centos7所以下载的Centos-7-reg.repo）
• 字体文件（进入c:/windows/fonts/，里面的字体随便复制一个支持中文的到当前目录下就行）我这里是使用的宋体字simsun.ttc

3、main.py

• 我这里是用的fastapi搭建的web服务，主要避免多客户端访问时候造成服务器卡顿

import imgkit
from fastapi import FastAPI, HTTPException, BackgroundTasks
from pydantic import BaseModel
from io import BytesIO
from fastapi.responses import StreamingResponse

app = FastAPI()
cfg = imgkit.config(wkhtmltoimage="/usr/local/bin/wkhtmltoimage")

class ImageRequest(BaseModel):
    url: str = None
    text: str = None

async def generate_image(url: str = None, text: str = None) -> BytesIO:
    """生成图像并返回二进制数据"""
    if url:
        img = imgkit.from_url(url, False, config=cfg)
    elif text:
        img = imgkit.from_string(text, False, config=cfg)
    else:
        raise ValueError("Either 'url' or 'text' must be provided")
    
    image_bytes = BytesIO(img)
    image_bytes.seek(0)
    return image_bytes

@app.post("/convert/")
async def convert(request: ImageRequest, background_tasks: BackgroundTasks):
    """处理图像生成请求"""
    if not request.url and not request.text:
        raise HTTPException(status_code=400, detail="Must provide either 'url' or 'text'.")

    # 这里不再使用后台任务，而是直接返回图像
    image_bytes = await generate_image(request.url, request.text)
    
    # 返回图像内容
    return StreamingResponse(image_bytes, media_type="image/png")

@app.get("/status/")
async def status():
    """返回服务器的状态"""
    return {"status": "Server is running"}

制作镜像

• 进入html2image目录

 docker build -t html2image .

• 运行镜像

docker run -d -p:54248:54248 html2image

使用postman校验接口，发现可以实现根据url生成图片，或者根据文本生成图片，效果如下图所示