ghmath2codecogs.py

#!/usr/bin/env python3

import re

URL = "https://latex.codecogs.com/svg.image?%s"

def main(input, output):
    lines = []
    with open(input) as f:
        for line in f:
            lines.append(line)

    with open(output, 'w') as f:
        math = None
        ws = None
        for line in lines:
            if math is None:
                m = re.match('(\s*)```\s*math', line)
                if m:
                    # start collecting math lines
                    math = []
                    ws = m.group(1)

                else:
                    f.write(line)

            else:
                m = re.match('(\s*)```', line)
                if m:
                    # merge all math, removing newlines
                    math = ''.join(m[:-1]+' ' for m in math)
                    # deduplicate a whitespace, but whitespace is
                    # important so keep one space
                    math = re.sub('\s+', ' ', math)
                    # and strip
                    math = math.strip()

                    # create the codecogs link
                    f.write('%s<p align="center">\n' % ws)
                    f.write('%s<img\n' % ws)
                    f.write('%s    alt="%s"\n' % (ws, math))
                    f.write('%s    src="%s"\n' % (ws, URL % ''.join(
                        # for codecogs we need to url-encode our latex
                        c if re.match('[a-zA-Z_]', c)
                            else '%%%02x' % ord(c)
                        for c in math)))
                    f.write('%s>\n' % ws)
                    f.write('%s</p>\n' % ws)

                    # reset
                    math = None
                    ws = None
        
                else:
                    math.append(line)


if __name__ == "__main__":
    import sys
    import argparse
    parser = argparse.ArgumentParser(
        description="Rewrite a markdown file so GitHub math blocks "
            "are replaced with embedded images generated by codecogs.",
        allow_abbrev=False)
    parser.add_argument(
        'input',
        help="Input markdown file.")
    parser.add_argument(
        '-o', '--output',
        required=True,
        help="Output markdown file.")
    sys.exit(main(**{k: v
        for k, v in vars(parser.parse_args()).items()
        if v is not None}))