#! /usr/local/bin/python
# -*- coding: utf-8 -*-

# Overview
# --------
#
# this script will display the users not fulfilling their meeting quotas
#
# Run it as the web user, or the user that has access to the wiki data files.
#
# Highly Koumbit-specific.
# 
# Still lacking:
#
# 3. finding *missing* people (ie. people that do not appear in any meeting, we need an external list, parse MembresDuComiteDeTravail?)
# 4. finding people with a valid excuse (parse the "Absences" field)
# 5. split report between meeting de reflexion and other involvements
# 6. simply output which users are problematic
# 
# Done:
# 1. alias recognition (ie. Omar vs OmarBickell and Antoine vs TheAnarcat)
# 2. typo correction (same problem as above)
#
# (C) Copyleft 2007, The Anarcat <anarcat@koumbit.org>
# Licensed under the GPLv2 or any later version

import sys, operator, os

os.chdir('/export/wiki/config')
sys.path.insert(0, '/export/wiki/config')

from MoinMoin import wikiutil
from MoinMoin.Page import Page
from MoinMoin.user import User
from farmconfig import wikis
import re
import time
import farmconfig

from MoinMoin.request import RequestCLI

# the pattern of meeting pages. the filter_pages will further reduce the list based on dates extracted from this pattern
meeting_pages = re.compile('^.*/(....)-(..)-(..)$')

# the pattern to detect the assistance listing, the main loop uses this to extract the actual user listings from meeting pages
presences = re.compile('Pr.*sence.*$\s*((\s*\*.*\n)*)',re.MULTILINE)

# alias -> realname mapping, used by sanitize_user_line to fix typos and match users with their aliases
aliases  = { 'antoine' : 'theanarcat', 'omar' : 'omarbickell' , 'marco' : 'marcangles', 'patrickhetu' : 'patrickhétu', 'jennigalewski' : 'jenny', 'jennygalewski' : 'jenny', 'antoinebeaupré' : 'theanarcat', 'alexis' : 'alexisboulva', 'marcel-olivierbrodeur' : 'marcelolivierbrodeur' }

# take a user formatted as a single line in a bullet list and output the user only
def sanitize_user_line(line):
    # strip starting bullet
    line = re.sub('\s*\*\s*','',line)
    # just take the first word, lowercased
    user = line.split(' ')[0].lower().encode('utf-8')
    # alias expansion
    if user in aliases:
        user = aliases[user]
    return user

# just find the meeting pages, in the last 2 months
#
# XXX: this algorithm is stupid, for dates calculation, i'm tired and can't think straight tonight
def filter_pages(name):
    matches = meeting_pages.match(name)
    if matches:
        now_year = time.localtime()[0]
        now_month = time.localtime()[1]
        meeting_year = int(matches.group(1))
        meeting_month = int(matches.group(2))
        lower_bound = now_month - 2
        if lower_bound < 1:
            now_year = now_year - 1
            # simple: if in feb, lower is 0, then 12, if jan, lower is -1, then 11
            lower_bound = 12 + lower_bound
        #print "now: %d %d, meeting: %d %d, lower: %d" % ( now_year, now_month, meeting_year, meeting_month, lower_bound)
        if now_year == meeting_year and meeting_month >= lower_bound:
            return name 
        #else:
        #   print "not within timerange: " + name.encode('utf-8')
    return None
    
# main script
def main():
    request = RequestCLI('wiki.koumbit.net/')
    # the user we masquerade as to have all the pages
    user = User(request,name='TheAnarcat')
    pagelist = request.rootpage.getPageList(user=user,filter=filter_pages)
    
    # the dictionnary mapping user -> meetings participations
    participation = {}
    # main loop: analyse each valid page for assistance
    for name in pagelist:
        body = Page(request,name).get_raw_body()
        #r = re.compile('Pr.*sence.*$(.*)^---',re.MULTILINE | re.DOTALL)
        #r = re.compile("''Pr.*sence.*''*$\s*\(\(^\s\s*\*\s*\S*$\)*\)",re.MULTILINE)
        matches = presences.search(body)
        if matches:
            #print "page %s conforms to pattern" % name.encode('utf-8')
            users = matches.group(1)
            if not users:
                print "no users found??"
            users = map(sanitize_user_line, users.split("\n"))
            #print "USERS: " + ", ".join(users).encode('utf-8')
            for u in users:
                # add the meeting to this users' list
                if not u in participation:
                    participation[u] = ""
                participation[u] += ", " + name
    # final display of the user assistance
    for u, meeting in participation.iteritems():
        print " %s:: %s" % ( u, meeting.encode('utf-8'))

if __name__ == "__main__":
    main()