Convert inline links to reference links in markdown

md_inline_to_ref.py

import os
import re
import sys


RE_MD_INLINE_LINK = re.compile(' \[(?P<name>[\w -_]*)\]\((?P<link>[\w \-/\.:]*)\)')


if __name__ == "__main__":
    src_path = sys.argv[1]
    assert os.path.exists(src_path)

    links = {}
    with open(src_path, 'r') as src:
        dst_path = src_path + ".new"
        with open(dst_path, 'w') as dst:
            link_id = 1
            for line in src:
                r = RE_MD_INLINE_LINK.search(line)
                if r is None:
                    dst.write(line)
                    continue

                output_link_id = None
                if r.groupdict()['link'] not in links:
                    links[r.groupdict()['link']] = link_id
                    output_link_id = link_id
                    link_id = link_id + 1
                else:
                    output_link_id = links[r.groupdict()['link']]

                line = re.sub(RE_MD_INLINE_LINK,
                              r' [\g<name>][{}]'.format(output_link_id),
                              line)
                dst.write(line)

            # Now output the links at the bottom of the page
            for link, id_ in sorted(links.iteritems(),
                                    key=lambda (k, v): (v, k)):
                dst.write('[{}]: {}\n'.format(id_, link))
                print "[{}]: {}".format(id_, link)