Python 把 HTML 中的 img 元素的 src 转换成data URL | Convert the src link of img tag of HTML into data URL

html_img_src2datauri.py

#!/usr/bin/env python3
# coding: utf-8

__author__ = "ChenyangGao <https://chenyanggao.github.io/>"
__version__ = (0, 1)
__all__ = ["process"]

if __name__ == "__main__":
    from argparse import ArgumentParser

    parser = ArgumentParser(description="把 html 文件中 <img> 元素的 src 属性转换成 datauri。注意：所有 html 文件会被原地替换，并不生成新文件")
    parser.add_argument("pathlist", nargs="+", metavar="path_to_file_or_dir", help="html 所在的文件或文件夹")
    parser.add_argument("-e", "--encoding", help="html 文件的编码")

    args = parser.parse_args()


from base64 import b64encode
from contextlib import contextmanager
from mimetypes import guess_type
from os import chdir, getcwd, path
from re import compile as re_compile
from urllib.parse import urlsplit
from urllib.request import urlopen


# 正则表达式，匹配有 src 属性的 <img> 元素
cre_src = re_compile(r'(<img\b[^>]+?\bsrc=")(?P<uri>[^"]+)')


@contextmanager
def tempcwd(dir_):
    "上下文管理器，切换当前工作目录"
    if dir_ in ("", "."):
        yield
        return
    cwdold = getcwd()
    chdir(dir_)
    try:
        yield
    finally:
        chdir(cwdold)


# TODO: 以后将会支持 ftp(s)://, dav(s)://, sftp:// 等协议
def _read(path):
    "读取文件的二进制数据"
    if path.startswith(("http://", "https://")):
        return urlopen(path).read()
    return open(path, "rb").read()


def _repl(m):
    "正则表达式替换时用到的函数，用于把搜索到的 uri 替换成 datauri"
    uri = m["uri"]
    # 如果本身就是 datauri，则原样返回
    if uri.startswith("data:"):
        return uri[0]
    mimetype = guess_type(urlsplit(uri).path)[0]
    # 如果判断不了 mimetype，则原样返回
    if mimetype is None:
        return uri[0]
    # 如果读取失败，则原样返回
    try:
        content = _read(uri)
    except:
        return uri[0]
    # datauri 使用 base64 编码二进制数据
    return "%sdata:%s;base64,%s" % (m[1], mimetype, b64encode(content).decode('ascii'))


def process(path_file, path_save=None, encoding=None):
    """读取 html 文件，把其中可读取的图片路径转化并替换为 datauri，并保存修改后的 html

    :param path_file: 原始 html 文件的路径
    :param path_save: 保存更改后的 html 文件的路径，如果为 None，那么等于参数 `path_file`
    :param encoding: 文件的编码
    """
    if path_save is None:
        path_save = path_file
    dir_, name = path.split(path_file)
    with tempcwd(dir_):
        text = open(name, encoding=encoding).read()
        text_new = cre_src.sub(_repl, text)
    if path_file != path_save or text != text_new:
        open(path_save, "w", encoding=encoding).write(text_new)


if __name__ == "__main__":
    from glob import glob

    encoding = args.encoding

    def process_one(path_):
        print("-" * 20)
        print("[PROCESSING]", path_)
        try:
            process(path_, encoding=encoding)
        except Exception as exc:
            print("😭 FAILED:", exc)
        else:
            print("😄 SUCCESS")

    for path_ in args.pathlist:
        if path.isdir(path_):
            for p in glob(path.join(path_, "**", "*.htm"), recursive=True):
                process_one(p)
            for p in glob(path.join(path_, "**", "*.html"), recursive=True):
                process_one(p)
        else:
            process_one(path_)