view MoinMoin/action/ @ 5910:7e7e1cbb9d3f

security: fix remote code execution vulnerability in twikidraw/anywikidraw actions We have wikiutil.taintfilename() to make user supplied filenames safe, so that they can't contain any "special" characters like path separators, etc. It is used at many places in moin, but wasn't used here. :|
author Thomas Waldmann <tw AT waldmann-edv DOT de>
date Sat, 29 Dec 2012 15:05:29 +0100
parents bf0b4b96dcb4
children e9ef58bdad15
line wrap: on
line source
# -*- coding: iso-8859-1 -*-
    MoinMoin - LikePages action

    This action generates a list of pages that either start or end
    with the same word as the current pagename. If only one matching
    page is found, that page is displayed directly.

    @copyright: 2001 Richard Jones <>,
                2001 Juergen Hermann <>
    @license: GNU GPL, see COPYING for details.

import re

from MoinMoin import config, wikiutil
from import difflib
from MoinMoin.Page import Page

def execute(pagename, request):
    _ = request.getText
    start, end, matches = findMatches(pagename, request)

    # Error?
    if isinstance(matches, (str, unicode)):
        request.theme.add_msg(wikiutil.escape(matches), "info")
        Page(request, pagename).send_page()

    # No matches
    if not matches:
        request.theme.add_msg(_('No pages like "%s"!') % (wikiutil.escape(pagename), ), "error")
        Page(request, pagename).send_page()

    # One match - display it
    if len(matches) == 1:
        request.theme.add_msg(_('Exactly one page like "%s" found, redirecting to page.') % (wikiutil.escape(pagename), ), "info")
        Page(request, matches.keys()[0]).send_page()

    # more than one match, list 'em
    # This action generate data using the user language

    request.theme.send_title(_('Pages like "%s"') % (pagename), pagename=pagename)

    # Start content - IMPORTANT - without content div, there is no
    # direction support!

    showMatches(pagename, request, start, end, matches)

    # End content and send footer

def findMatches(pagename, request, s_re=None, e_re=None):
    """ Find like pages

    @param pagename: name to match
    @param request: current reqeust
    @param s_re: start re for wiki matching
    @param e_re: end re for wiki matching
    @rtype: tuple
    @return: start word, end word, matches dict
    # Get full list of pages, with no filtering - very fast. We will
    # first search for like pages, then filter the results.
    pages = request.rootpage.getPageList(user='', exists='')

    # Remove current page
    except ValueError:

    # Get matches using wiki way, start and end of word
    start, end, matches = wikiMatches(pagename, pages, start_re=s_re,

    # Get the best 10 close matches
    close_matches = {}
    found = 0
    for name in closeMatches(pagename, pages):
        # Skip names already in matches
        if name in matches:

        # Filter deleted pages or pages the user can't read
        page = Page(request, name)
        if page.exists() and
            close_matches[name] = 8
            found += 1
            # Stop after 10 matches
            if found == 10:

    # Filter deleted pages or pages the user can't read from
    # matches. Order is important!
    for name in matches.keys(): # we need .keys() because we modify the dict
        page = Page(request, name)
        if not (page.exists() and
            del matches[name]

    # Finally, merge both dicts

    return start, end, matches

def wikiMatches(pagename, pages, start_re=None, end_re=None):
    Get pages that starts or ends with same word as this page

    Matches are ranked like this:
        4 - page is subpage of pagename
        3 - match both start and end
        2 - match end
        1 - match start

    @param pagename: page name to match
    @param pages: list of page names
    @param start_re: start word re (compile regex)
    @param end_re: end word re (compile regex)
    @rtype: tuple
    @return: start, end, matches dict
    if start_re is None:
        start_re = re.compile('([%s][%s]+)' % (config.chars_upper,
    if end_re is None:
        end_re = re.compile('([%s][%s]+)$' % (config.chars_upper,

    # If we don't get results with wiki words matching, fall back to
    # simple first word and last word, using spaces.
    words = pagename.split()
    match = start_re.match(pagename)
    if match:
        start =
        start = words[0]

    match =
    if match:
        end =
        end = words[-1]

    matches = {}
    subpage = pagename + '/'

    # Find any matching pages and rank by type of match
    for name in pages:
        if name.startswith(subpage):
            matches[name] = 4
            if name.startswith(start):
                matches[name] = 1
            if name.endswith(end):
                matches[name] = matches.get(name, 0) + 2

    return start, end, matches

def closeMatches(pagename, pages):
    """ Get close matches.

    Return all matching pages with rank above cutoff value.

    @param pagename: page name to match
    @param pages: list of page names
    @rtype: list
    @return: list of matching pages, sorted by rank
    # Match using case insensitive matching
    # Make mapping from lowerpages to pages - pages might have same name
    # with different case (although its stupid).
    lower = {}
    for name in pages:
        key = name.lower()
        if key in lower:
            lower[key] = [name]

    # Get all close matches
    all_matches = difflib.get_close_matches(pagename.lower(), lower.keys(),
                                            len(lower), cutoff=0.6)

    # Replace lower names with original names
    matches = []
    for name in all_matches:

    return matches

def showMatches(pagename, request, start, end, matches, show_count=True):
    keys = matches.keys()
    _showMatchGroup(request, matches, keys, 8, pagename, show_count)
    _showMatchGroup(request, matches, keys, 4, "%s/..." % pagename, show_count)
    _showMatchGroup(request, matches, keys, 3, "%s...%s" % (start, end), show_count)
    _showMatchGroup(request, matches, keys, 1, "%s..." % (start, ), show_count)
    _showMatchGroup(request, matches, keys, 2, "...%s" % (end, ), show_count)

def _showMatchGroup(request, matches, keys, match, title, show_count=True):
    _ = request.getText
    matchcount = matches.values().count(match)

    if matchcount:
        if show_count:
            # Render title line
                _('%(matchcount)d %(matches)s for "%(title)s"') % {
                    'matchcount': matchcount,
                    'matches': ' ' + (_('match'), _('matches'))[matchcount != 1],
                    'title': title}))

        # Render links
        for key in keys:
            if matches[key] == match:
                request.write(request.formatter.pagelink(1, key, generated=True))
                request.write(request.formatter.pagelink(0, key, generated=True))