Convert between XMP and PRE sections using pygments

pyg.py

#!/usr/bin/env python

import sys
import os
import re
import subprocess
import shutil

r_tag = re.compile(r"(?s)<[^>]+>")
r_pre = re.compile(r"(?s)<pre([^>]+)>(.*?)</pre>")
r_xmp = re.compile(r"(?s)<xmp([^>]+)>(.*?)</xmp>")
r_pyg = re.compile(r"\bpyg-([a-z]+)\b")

prefix = len("<div class=\"highlight\"><pre>")
suffix = len("</pre></div>\n")

def xmp_to_pre(match):
    attributes, content = match.groups()
    syntax = r_pyg.search(attributes)
    if not syntax:
        return match.group(0)
    syntax = syntax.group(1)

    command = ["pygmentize", "-l",  syntax, "-f", "html"]
    p = subprocess.Popen(command,
        stdin=subprocess.PIPE, stdout=subprocess.PIPE)
    stdout, stderr = p.communicate(content)

    result = stdout[prefix:-suffix]
    result = result.replace("\t", "<span class=\"tab\">    </span>")
    return "<pre%s>\n%s</pre>" % (attributes, result)

def pre_to_xmp(match):
    attributes, content = match.groups()
    syntax = r_pyg.search(attributes)
    if not syntax:
        return match.group(0)

    result = content.replace("<span class=\"tab\">    </span>", "\t")
    result = r_tag.sub("", result)
    result = result.replace("&lt;", "<")
    result = result.replace("&gt;", ">")
    result = result.replace("&quot;", "\"")
    result = result.replace("&amp;", "&")
    return "<xmp%s>%s</xmp>" % (attributes, result)

def pre(filename):
    with open(filename) as f:
        text = f.read()

    text = r_xmp.sub(xmp_to_pre, text)

    with open(filename, "w") as o:
        o.write(text)

def xmp(filename):
    with open(filename) as f:
        text = f.read()

    text = r_pre.sub(pre_to_xmp, text)

    with open(filename, "w") as o:
        o.write(text)

def main():
    command = sys.argv[1]
    filename = sys.argv[2]

    backup = os.path.expanduser("~/.pyg-backup")
    shutil.copy2(filename, backup)

    if command == "pre":
        pre(filename)
    elif command == "xmp":
        xmp(filename)
    else: print "?"

if __name__ == "__main__":
    main()

What's xmp? :D

Ancient long-forgotten magics that let you write markup embedded in HTML without having to use all that silly escaped entity nonsense. Don't tell anybody, they might try to ban it.