import re
import json
import sys
from wikitools import *
import userpass
import mwparserfromhell

site = wiki.Wiki()
site.login(userpass.username, userpass.password)
etitles = ['Template:Infobox journal', 'Template:Infobox magazine']
journaltitles = ['Template:Infobox journal']
journalprojects = ['Template:WikiProject Academic Journals']
magazineprojects = ['Template:WikiProject Magazines']
bannershells = ['Template:WikiProject banner shell']

def updatelists():
    global bannershells
    global magazineprojects
    global journalprojects
    global journaltitles

    params = {'action':'query',
              'list':'backlinks',
              'bltitle':magazineprojects[0],
              'bllimit':'5000',
              'blfilterredir':'redirects',
              'blnamespace':'10'
              }
    for pageelement in api.APIRequest(site, params).query(False)['query']['backlinks']:
        magazineprojects.append(pageelement['title'])

    params = {'action':'query',
              'list':'backlinks',
              'bltitle':journalprojects[0],
              'bllimit':'5000',
              'blfilterredir':'redirects',
              'blnamespace':'10'
              }
    for pageelement in api.APIRequest(site, params).query(False)['query']['backlinks']:
        journalprojects.append(pageelement['title'])

    params = {'action':'query',
              'list':'backlinks',
              'bltitle':journaltitles[0],
              'bllimit':'5000',
              'blfilterredir':'redirects',
              'blnamespace':'10'
              }
    for pageelement in api.APIRequest(site, params).query(False)['query']['backlinks']:
        journaltitles.append(pageelement['title'])

    params = {'action':'query',
              'list':'backlinks',
              'bltitle':bannershells[0],
              'bllimit':'5000',
              'blfilterredir':'redirects',
              'blnamespace':'10'
              }
    for pageelement in api.APIRequest(site, params).query(False)['query']['backlinks']:
       bannershells.append(pageelement['title'])

    print(magazineprojects)
    print(journalprojects)
    print(bannershells)
    print(journaltitles)

epages1 = []
epages2 = []

def startAllowed():
    if page.Page(site, 'User:DatBot/run/task9') != 'true':
        return False
    else:
        return True

def getembeds():
    global epages
    for pagetitle in etitles:
        params = {'action':'query',
                  'list':'embeddedin',
                  'eititle':pagetitle,
                  'eilimit':'5000',
                  'eifilterredir':'nonredirects',
                  'einamespace':'0'
                  }
        req = api.APIRequest(site, params)
        res = req.query(False)
        length = len(res['query']['embeddedin'])
        while length >= 5000:
            for pagename in res['query']['embeddedin']:
                if pagetitle in journaltitles:
                    epages1.append(pagename['title'])
                else:
                    epages2.append(pagename['title'])
            length = len(res['query']['embeddedin'])
            try:
                params['eicontinue'] = res['continue']['eicontinue']
            except KeyError:
                pass
            req = api.APIRequest(site, params)
            res = req.query(False)

def changepage(pagetitle, isjournal, originalpage):
    global bannershells

    if page.Page(site, pagetitle).exists:
        filetext = page.Page(site, pagetitle).getWikiText()
    else:
        filetext = ''

    if not allow_bots(filetext):
        return

    if isjournal:
        stringadd = "{{WikiProject Academic Journals|class=File}}"
        transcludepage = etitles[0]
    else:
        transcludepage = etitles[1]
        stringadd = "{{WikiProject Magazines|class=File}}"

    if any(bannershell in filetext for bannershell[9:] in bannershells):
        re.sub(r'(%s.*\|1=)' % bannershell, '\g<1>\n%s' % stringadd, filetext, re.IGNORECASE)
    else:
        filetext = '%s\n' %  stringadd + filetext

    page.Page(site, pagetitle).edit(text = filetext,
                                    bot = True,
                                    summary = "Adding %s because [[%s]] transcludes %s ([[Wikipedia:Bots/Requests for approval/DatBot 9|BOT]])" % (stringadd, originalpage, transcludepage))


def main():
    global epages1
    global epages2
    global journalprojects
    global magazineprojects

    if not startAllowed():
        return
    updatelists()
    getembeds()
    filechecked = open('pageschecked.txt', 'r')
    pageschecked = filechecked.read()
    pageschecked = pageschecked.decode('utf-8')
    filechecked.close()
    openfile = open('pageschecked.txt', 'a')
    for epage in epages1 + list(set(epages2) - set(epages1)):
        if epage in pageschecked:
            continue
        else:
            try:
                # https://regex101.com/r/BXBwby/1
                filename = re.findall(r'\|\s*(?:image|cover|image_file|logo)\s*=\s*(?:\[\[)?(?:(?:File|Image)\s*:\s*)?(\w[^\|<\]]*)',
                          page.Page(site, epage).getWikiText(), re.IGNORECASE | re.M)[0]

                filetalk = "File talk:%s" % filename
                if not page.Page(site, filetalk).exists:
                    filetext = ''
                else:
                    filetext = page.Page(site, filetalk).getWikiText()

                if epage in epages1:
                    if any(pageelement in filetext.lower() for pageelement[9:] in journalprojects):
                        continue

                    changepage(filetalk, True, epage)
                if epage in epages2:
                    if any(pageelement in filetext.lower() for pageelement[9:] in magazineprojects):
                        continue

                    changepage(filetalk, False, epage)

            except IndexError:
                continue
            openfile.write("%s\n" % epage.encode('utf-8')) # all refreshed monthly
    openfile.close()

def allow_bots(text):
    user = 'datbot'
    text = mwparserfromhell.parse(text)
    for tl in text.filter_templates():
        if tl.name.matches(['bots', 'nobots']):
            break
    else:
        return True
    for param in tl.params:
        bots = [x.lower().strip() for x in param.value.split(",")]
        if param.name == 'allow':
            if ''.join(bots) == 'none': return False
            for bot in bots:
                if bot in (user, 'all'):
                    return True
        elif param.name == 'deny':
            if ''.join(bots) == 'none': return True
            for bot in bots:
                if bot in (user, 'all'):
                    return False
    if (tl.name.matches('nobots') and len(tl.params) == 0):
        return False
    return True

if __name__ == "__main__":
    main()