source: products/quintagroup.plonegooglesitemaps/branches/two_request/utils.py

import re

from Missing import MV as Missing_Value
from Products.CMFCore.utils import getToolByName

from quintagroup.canonicalpath.interfaces import ICanonicalLink

ADD_ZOPE = re.compile('^/')
ADD_PLONE = re.compile('^[^http://|https://|\\\]')
OPERATIONS_PARSE = re.compile(r"(.?[^\\])/(.*[^\\]|)/(.*[^\\]|)/")
_marker = []

    
def searchAndReplace(string, what, withs):
    """Emulate sed command s/"""
    #TODO: Better solution would be to compile
    #'what' and 'withs'
    res = re.sub(what, withs, string)
    return res
OPERATORS = {'s': searchAndReplace, }

def urlFilter(url, operations):
    """ Method dedicated to filter url according with operations """
    for operator, what, withs in operations:
        url = OPERATORS[operator](url, what, withs.replace("\\", ""))
    return url 

def getUrlsObjects(objects):
    """Method dedicated to get url for each object"""
    results = set()
    for ob in objects:
        url = _marker
        if hasattr(ob, 'canonical_link'):
            url = ob.canonical_link
        if url in [Missing_Value, _marker]:
            url = ICanonicalLink(ob.getObject()).canonical_link
        #TODO: Remove or replace following condition
        #it is senseless in the case we need intelligent
        #result set. Better condition would be to place
        #freshest brain into result
        if url in results:
            continue
        #TODO: replace brain with only data necessary to
        #generate sitemap
        results.add(url)
        yield url, ob

def applyOperations(objects, operations):
    """Parse Operations"""
    operations = [OPERATIONS_PARSE.match(op).groups() for op in operations]
    result = {}
    for url, ob in getUrlsObjects(objects):
        url = urlFilter(url, operations)
        result[url] = ob
    return result

def additionalURLs(context):
    """Add URLs to sitemap that arn't objects"""
    res = []

    plone_home = getToolByName(context, 'portal_url')()
    root = context.getPhysicalRoot().absolute_url()
    URLs = context.getUrls()

    for url in URLs:
        if ADD_ZOPE.match(url):
            res.append(root + url)
        elif ADD_PLONE.match(url):
            res.append(plone_home + '/' + url)
        else:
            res.append(url)
    return res