# -*- coding: utf-8 -*-

from django.conf import settings

from django.utils.http import urlquote

from haystack.constants import DJANGO_ID

from haystack.query import SearchQuerySet

from hdabo.models import Tag, TaggedSheet

from wikitools import api, wiki

def normalize_tag(tag):

    if len(tag) == 0:

        return tag

    tag = tag.strip()

    tag = tag.replace("_", " ")

    tag = " ".join(tag.split())

    tag = tag[0].upper() + tag[1:]

    return tag

def urlize_for_wikipedia(label):

    return urlquote(label.replace(" ", "_"))

def __is_homonymie(page_dict):

    for cat in page_dict.get(u"categories", []):

        if u'Catégorie:Homonymie' in cat.get(u"title", u"") or u'Category:Disambiguation pages' in cat.get(u"title", u""):

            return True

    return False

def query_wikipedia_title(site, label=None, pageid=None):

    params = {'action':'query', 'prop':'info|categories|langlinks', 'inprop':'url', 'lllimit':'500', 'cllimit':'500', 'rvprop':'ids'}

    if label:

        params['titles'] = label

    else:

        params['pageids'] = pageid

    wpquery = api.APIRequest(site, params) #@UndefinedVariable

    response = wpquery.query()

    original_response = response

    def return_null_result():

        return { 'new_label': None, 'status': Tag.TAG_URL_STATUS_DICT["null_result"], 'wikipedia_url': None, 'pageid': None, 'dbpedia_uri': None, 'revision_id': None, 'response': response }

    query_dict = response['query']

    # get page if multiple pages or none -> return Tag.null_result

    pages = query_dict.get("pages", {})

    if len(pages) > 1 or len(pages) == 0:

        return return_null_result()

    page = pages.values()[0]

    if u"invalid" in page or u"missing" in page:

        return return_null_result()

    url = page.get(u'fullurl', None)

    pageid = page.get(u'pageid', None)

    new_label = page[u'title']

    if __is_homonymie(page):

        status = Tag.TAG_URL_STATUS_DICT["homonyme"]

    elif u"redirect" in page:

        status = Tag.TAG_URL_STATUS_DICT["redirection"]

    else:

        status = Tag.TAG_URL_STATUS_DICT["match"]

    if status == Tag.TAG_URL_STATUS_DICT["redirection"]:

        params['redirects'] = True

        wpquery = api.APIRequest(site, params) #@UndefinedVariable    

        response = wpquery.query()

        query_dict = response['query']

        pages = query_dict.get("pages", {})

        #we know that we have at least one answer        

        if len(pages) > 1 or len(pages) == 0:

            return return_null_result()

        page = pages.values()[0]

    revision_id = page.get('lastrevid', None)

    #process language to extract the english label

    english_label = None

    if status == Tag.TAG_URL_STATUS_DICT['match'] or status == Tag.TAG_URL_STATUS_DICT['redirection']:

        lang_links = page.get('langlinks', [])

        for lang_info_dict in lang_links:

            if lang_info_dict['lang'] == "en":

                english_label = lang_info_dict["*"]

                break

    if english_label and "#" not in english_label:

        dbpedia_uri = settings.DBPEDIA_URI_TEMPLATE % (urlize_for_wikipedia(english_label))

    else:

        dbpedia_uri = None

def get_or_create_tag(tag_label):

    tag_label_normalized = normalize_tag(tag_label)

    # We get the wikipedia references for the tag_label

    # We get or create the tag object

    tag = None

    for t in Tag.objects.filter(label__iexact=tag_label_normalized):

        if tag is None or t.url_status != Tag.TAG_URL_STATUS_DICT['null_result']:

            tag = t

            if tag.url_status != Tag.TAG_URL_STATUS_DICT['null_result']:

                break

    if tag is None:

        tag = Tag(label=tag_label_normalized, original_label=tag_label)

        created = True

    else:

        created = False

    site = wiki.Wiki(settings.WIKIPEDIA_API_URL) #@UndefinedVariable

    if created:

        wp_res = query_wikipedia_title(site, label=tag_label_normalized)

        # We save the datas

        if new_label is not None:

            tag.label = new_label

        if status is not None:

            tag.url_status = status

        tag.wikipedia_url = url            

        tag.wikipedia_pageid = pageid

        tag.dbpedia_uri = dbpedia_uri 

        tag.save()

    elif tag.wikipedia_pageid:

        wp_res = query_wikipedia_title(site, pageid=tag.wikipedia_pageid)

        wikipedia_revision_id = wp_res['revision_id']

    else:

        wikipedia_revision_id = None

    return tag, wikipedia_revision_id, created

def process_tag(site, tag, verbosity=0):

    wp_res = query_wikipedia_title(site, label=tag.label)

    new_label, status, url, pageid, response, dbpedia_uri, revision_id = wp_res['new_label'], wp_res['status'], wp_res['wikipedia_url'], wp_res['pageid'], wp_res['response'], wp_res["dbpedia_uri"], wp_res["revision_id"]

    if verbosity >= 2 :

        print "response from query to %s with parameters %s :" % (site.apibase, repr(new_label))

        print repr(response)

    prev_wikipedia_pageid = tag.wikipedia_pageid

    if new_label is not None:

        tag.label = new_label

    if status is not None:

        tag.url_status = status

    tag.wikipedia_url = url

    tag.wikipedia_pageid = pageid

    tag.dbpedia_uri = dbpedia_uri

    tag.save()

    if prev_wikipedia_pageid != pageid:

        TaggedSheet.objects.filter(tag=tag).update(wikipedia_revision_id=revision_id)

def reorder_datasheet_tags(ds):

    ts_list = []

    for ts in ds.taggedsheet_set.all():

        ts.index_note = 0

        kwargs = {DJANGO_ID + "__exact": unicode(ds.pk)}

        results = SearchQuerySet().filter(title=ts.tag.label).filter_or(description=ts.tag.label).filter(**kwargs)

        if len(results) > 0:

            ts.index_note += results[0].score

            ts.save()

        ts_list.append(ts)

    ts_list.sort(key=lambda t: (-t.index_note, t.order))

    for k, ts in enumerate(ts_list):

        ts.order = k + 1

        ts.save()

    if ds.manual_order:

        ds.manual_order = False

        ds.save()