html_aware_wrap

main.py

def html_aware_wrap(string, width=72):
    """Wraps some HTML to the given width.

    This ignores the HTML for the purpose of determining the width of each
    line.

    Warning: This takes a pretty trivial approach to matching HTML tags, so
    buyer beware.
    """
    lines = [""]
    pos = 0
    current_line_length = 0

    TAG_RE = re.compile(ur"(<.+?>)", flags=re.MULTILINE|re.UNICODE)
    TEXT_RE = re.compile(ur"(.+?)(\s|<.+?>|$)", flags=re.MULTILINE|re.UNICODE)

    while True:
        if pos < len(string) and string[pos] == "\n":
            pos += 1
            current_line_length = 0
            lines.append("")
            continue

        for regex in (TAG_RE, TEXT_RE):
            match = regex.match(string, pos)
            if match:
                if regex is not TAG_RE:
                    if current_line_length + len(match.group(1)) > width:
                        lines.append("")
                        current_line_length = 0

                    current_line_length += len(match.group(1))

                pos += len(match.group(1))
                lines[-1] += match.group(1)

                break
        else:
            break

    if pos < len(string):
        raise RuntimeError()

    return [i.strip() for i in lines]