Function html2groff

cppman/formatter/cppreference.py:198–317

Convert HTML text from cppreference.com to Groff-formatted text.

(data, name)

def html2groff(data, name):
    """Convert HTML text from cppreference.com to Groff-formatted text."""
    # Remove header and footer
    try:
        data = data[data.index('<div id="cpp-content-base">'):]
        data = data[:data.index('<div class="printfooter">') + 25]
    except ValueError:
        pass

    # Remove non-printable characters
    data = ''.join([x for x in data if x in string.printable])

    for table in re.findall(
            r'<table class="(?:wikitable|dsctable)"[^>]*>.*?</table>',
            data, re.S):
        tbl = parse_table(table)
        # Escape column with '.' as prefix
        tbl = re.compile(r'T{\n(\..*?)\nT}', re.S).sub(r'T{\n\\E \1\nT}', tbl)
        data = data.replace(table, tbl)

    # Pre replace all
    for rp in rps:
        data = re.compile(rp[0], rp[2]).sub(rp[1], data)

    # Remove non-printable characters
    data = ''.join([x for x in data if x in string.printable])

    # Upper case all section headers
    for st in re.findall(r'.SH .*\n', data):
        data = data.replace(st, st.upper())

    # Add tags to member/inherited member functions
    # e.g. insert -> vector::insert
    #
    # .SE is a pseudo macro I created which means 'SECTION END'
    # The reason I use it is because I need a marker to know where section
    # ends.
    # re.findall find patterns which does not overlap, which means if I do
    # this: secs = re.findall(r'\n\.SH "(.+?)"(.+?)\.SH', data, re.S)
    # re.findall will skip the later .SH tag and thus skip the later section.
    # To fix this, '.SE' is used to mark the end of the section so the next
    # '.SH' can be find by re.findall

    try:
        idx = data.index('.IEND')
    except ValueError:
        idx = None

    def add_header_multi(prefix, g):
        if ',' in g.group(1):
            res = ', '.join(['%s::%s' % (prefix, x.strip())
                             for x in g.group(1).split(',')])
        else:
            res = '%s::%s' % (prefix, g.group(1))

        return '\n.IP "%s"' % res

    if idx: