#!/usr/bin/env python
# encoding: utf-8

import urllib
import re
from xml.etree import ElementTree
from xml.parsers.expat import ExpatError
import htmlentitydefs
import os
import glob


_UNESCAPE = re.compile(ur'&\w+?;', re.UNICODE)


def unescape(s):
    if s is None:
        return ''

    def fixup(m):
        ent = m.group(0)[1:-1]
        return unichr(htmlentitydefs.name2codepoint[ent])
    try:
        return _UNESCAPE.sub(fixup, s)
    except:
        print 'unescape failed: %s' % repr(s)
        raise


class UnknownVariable(Exception):
    pass


class UnsupportedVariableExpression(Exception):
    pass


def replace_vars(m):
    """Replace vars in 'content' portion.

    :m: match object
    :returns: string

    """
    var = m.group(1)
    default = m.group(2)

    if not re.match(r'\w+$', var):
        raise UnsupportedVariableExpression(var)

    translate_vars = {
        'TM_PHP_OPEN_TAG_WITH_ECHO': 'g:UltiSnipsOpenTagWithEcho',
        'TM_PHP_OPEN_TAG': 'g:UltiSnipsOpenTag',
        'PHPDOC_AUTHOR': 'g:snips_author',
    }
    # TODO: TM_SELECTED_TEXT/([\t ]*).*/$1/m

    if var in translate_vars:
        newvar = translate_vars[var]
    else:
        # TODO: this could be autogenerated
        raise UnknownVariable(var)

    return "`!v exists('%s') ? %s : '%s'`" % (newvar, newvar, default)


def parse_content(c):
    try:
        data = ElementTree.fromstring(c)[0]

        rv = {}
        for k, v in zip(data[::2], data[1::2]):
            rv[k.text] = unescape(v.text)

        if re.search(r'\$\{\D', rv['content']):
            rv['content'] = re.sub(
                r'\$\{([^\d}][^}:]*)(?::([^}]*))?\}',
                replace_vars,
                rv['content'])

        return rv
    except (ExpatError, ElementTree.ParseError) as detail:
        print '   Syntax Error: %s' % (detail,)
        print c
        return None
    except UnknownVariable as detail:
        print '   Unknown variable: %s' % (detail,)
        return None
    except UnsupportedVariableExpression as detail:
        print '   Unsupported variable expression: %s' % (detail,)
        return None


def fetch_snippets_from_svn(name):
    base_url = 'http://svn.textmate.org/trunk/Bundles/' + name + '.tmbundle/'
    snippet_idx = base_url + 'Snippets/'

    idx_list = urllib.urlopen(snippet_idx).read()

    rv = []
    for link in re.findall('<li>(.*?)</li>', idx_list):
        m = re.match(r'<a\s*href="(.*)"\s*>(.*)</a>', link)
        link, name = m.groups()
        if name == '..':
            continue

        name = unescape(name.rsplit('.', 1)[0])  # remove Extension
        print "Fetching data for Snippet '%s'" % name
        content = urllib.urlopen(snippet_idx + link).read()

        cont = parse_content(content)
        if cont:
            rv.append((name, cont))

    return rv


def fetch_snippets_from_dir(path):
    """Fetch snippets from a given path."""

    rv = []
    for filename in glob.glob(os.path.join(path, '*.tmSnippet')):
        print 'Reading file %s' % filename
        f = open(filename)
        content = f.read()

        cont = parse_content(content)
        if cont:
            name = os.path.splitext(os.path.basename(filename))[0]
            rv.append((name, cont))
    return rv


def write_snippets(snip_descr, f):

    for name, d in snip_descr:
        if 'tabTrigger' not in d:
            continue

        if 'content' not in d or d['content'] is None:
            print 'SKIP: %s (no content)' % (d,)
            continue

        f.write('snippet %s "%s"\n' % (d['tabTrigger'], name))
        f.write(d['content'].encode('utf-8') + '\n')
        f.write('endsnippet\n\n')


if __name__ == '__main__':
    import sys

    bundle = sys.argv[1]

    if os.path.isdir(bundle):
        name = sys.argv[2]
        rv = fetch_snippets_from_dir(bundle)
    else:
        rv = fetch_snippets_from_svn(bundle)
        name = bundle.lower()

    write_snippets(rv, open('tm_' + name + '.snippets', 'w'))