X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/ca7bffa0f660afefb98d37ac002b9652e1c858e5..4c0b1c26b2e4789b194f34c907c822c55f04af3c:/apps/search/views.py?ds=inline diff --git a/apps/search/views.py b/apps/search/views.py index 98184299a..f7aa77cd9 100644 --- a/apps/search/views.py +++ b/apps/search/views.py @@ -1,21 +1,19 @@ # -*- coding: utf-8 -*- - +# This file is part of Wolnelektury, licensed under GNU Affero GPLv3 or later. +# Copyright © Fundacja Nowoczesna Polska. See NOTICE for more information. +# from django.conf import settings from django.shortcuts import render_to_response, get_object_or_404 from django.template import RequestContext -from django.contrib.auth.decorators import login_required from django.views.decorators import cache -from django.http import HttpResponse, HttpResponseRedirect, Http404, HttpResponsePermanentRedirect +from django.http import HttpResponse, HttpResponseRedirect, Http404, HttpResponsePermanentRedirect, JsonResponse from django.utils.translation import ugettext as _ from catalogue.utils import split_tags from catalogue.models import Book, Tag, Fragment from pdcounter.models import Author as PDCounterAuthor, BookStub as PDCounterBook -from catalogue.views import JSONResponse -from search import Search, SearchResult -from lucene import StringReader +from search.index import Search, SearchResult from suggest.forms import PublishingSuggestForm -from time import sleep import re #import enchant import json @@ -28,6 +26,13 @@ def match_word_re(word): return "[[:<:]]%s[[:>:]]" % word +query_syntax_chars = re.compile(r"[\\/*:(){}]") + + +def remove_query_syntax_chars(query, replace=' '): + return query_syntax_chars.sub(' ', query) + + def did_you_mean(query, tokens): return query # change = {} @@ -54,10 +59,13 @@ def did_you_mean(query, tokens): # return query +@cache.never_cache def hint(request): prefix = request.GET.get('term', '') if len(prefix) < 2: - return JSONResponse([]) + return JsonResponse([], safe=False) + + prefix = remove_query_syntax_chars(prefix) search = Search() # tagi beda ograniczac tutaj @@ -65,9 +73,6 @@ def hint(request): # jezeli tagi dot tylko ksiazki, to wazne zeby te nowe byly w tej samej ksiazce # jesli zas dotycza themes, to wazne, zeby byly w tym samym fragmencie. - tags = search.hint_tags(prefix, pdcounter=True) - books = search.hint_books(prefix) - def is_dupe(tag): if isinstance(tag, PDCounterAuthor): if filter(lambda t: t.slug == tag.slug and t != tag, tags): @@ -77,31 +82,55 @@ def hint(request): return True return False - tags = filter(lambda t: not is_dupe(t), tags) - def category_name(c): if c.startswith('pd_'): c = c[len('pd_'):] return _(c) - callback = request.GET.get('callback', None) - data = [{'label': t.name, - 'category': category_name(t.category), - 'id': t.id, - 'url': t.get_absolute_url()} - for t in tags] + \ - [{'label': b.title, + try: + limit = int(request.GET.get('max', '')) + except ValueError: + limit = -1 + else: + if limit < 1: + limit = -1 + + data = [] + + tags = search.hint_tags(prefix, pdcounter=True) + tags = filter(lambda t: not is_dupe(t), tags) + for t in tags: + if not limit: + break + limit -= 1 + data.append({ + 'label': t.name, + 'category': category_name(t.category), + 'id': t.id, + 'url': t.get_absolute_url() + }) + if limit: + books = search.hint_books(prefix) + for b in books: + if not limit: + break + limit -= 1 + data.append({ + 'label': b.title, 'category': _('book'), 'id': b.id, - 'url': b.get_absolute_url()} - for b in books] + 'url': b.get_absolute_url() + }) + + callback = request.GET.get('callback', None) if callback: return HttpResponse("%s(%s);" % (callback, json.dumps(data)), content_type="application/json; charset=utf-8") else: - return JSONResponse(data) + return JsonResponse(data, safe=False) +@cache.never_cache def main(request): results = {} @@ -114,6 +143,9 @@ def main(request): return render_to_response('catalogue/search_too_short.html', {'prefix': query}, context_instance=RequestContext(request)) + + query = remove_query_syntax_chars(query) + search = Search() theme_terms = search.index.analyze(text=query, field="themes_pl") \ @@ -124,15 +156,17 @@ def main(request): tags = split_tags(tags) author_results = search.search_phrase(query, 'authors', book=True) + translator_results = search.search_phrase(query, 'translators', book=True) + title_results = search.search_phrase(query, 'title', book=True) # Boost main author/title results with mixed search, and save some of its results for end of list. # boost author, title results - author_title_mixed = search.search_some(query, ['authors', 'title', 'tags'], query_terms=theme_terms) + author_title_mixed = search.search_some(query, ['authors', 'translators', 'title', 'tags'], query_terms=theme_terms) author_title_rest = [] for b in author_title_mixed: - also_in_mixed = filter(lambda ba: ba.book_id == b.book_id, author_results + title_results) + also_in_mixed = filter(lambda ba: ba.book_id == b.book_id, author_results + translator_results + title_results) for b2 in also_in_mixed: b2.boost *= 1.1 if also_in_mixed is []: @@ -155,30 +189,24 @@ def main(request): return True return False return f - f = already_found(author_results + title_results + text_phrase) + f = already_found(author_results + translator_results + title_results + text_phrase) everywhere = filter(lambda x: not f(x), everywhere) author_results = SearchResult.aggregate(author_results) + translator_results = SearchResult.aggregate(translator_results) title_results = SearchResult.aggregate(title_results) everywhere = SearchResult.aggregate(everywhere, author_title_rest) for field, res in [('authors', author_results), + ('translators', translator_results), ('title', title_results), ('text', text_phrase), ('text', everywhere)]: res.sort(reverse=True) - print "get snips %s, res size %d" % (field, len(res)) for r in res: - print "Get snippets for %s" % r search.get_snippets(r, query, field, 3) - # for r in res: - # for h in r.hits: - # h['snippets'] = map(lambda s: - # re.subn(r"(^[ \t\n]+|[ \t\n]+$)", u"", - # re.subn(r"[ \t\n]*\n[ \t\n]*", u"\n", s)[0])[0], h['snippets']) - # suggestion = did_you_mean(query, search.get_tokens(toks, field="SIMPLE")) suggestion = u'' def ensure_exists(r): @@ -188,22 +216,25 @@ def main(request): return False author_results = filter(ensure_exists, author_results) + translator_results = filter(ensure_exists, translator_results) title_results = filter(ensure_exists, title_results) text_phrase = filter(ensure_exists, text_phrase) everywhere = filter(ensure_exists, everywhere) - results = author_results + title_results + text_phrase + everywhere + results = author_results + translator_results + title_results + text_phrase + everywhere # ensure books do exists & sort them - results.sort(reverse=True) - - if len(results) == 1: - fragment_hits = filter(lambda h: 'fragment' in h, results[0].hits) - if len(fragment_hits) == 1: - #anchor = fragment_hits[0]['fragment'] - #frag = Fragment.objects.get(anchor=anchor) - return HttpResponseRedirect(fragment_hits[0]['fragment'].get_absolute_url()) - return HttpResponseRedirect(results[0].book.get_absolute_url()) - elif len(results) == 0: + for res in (author_results, translator_results, title_results, text_phrase, everywhere): + res.sort(reverse=True) + + # We don't want to redirect to book text, but rather display result page even with one result. + # if len(results) == 1: + # fragment_hits = filter(lambda h: 'fragment' in h, results[0].hits) + # if len(fragment_hits) == 1: + # #anchor = fragment_hits[0]['fragment'] + # #frag = Fragment.objects.get(anchor=anchor) + # return HttpResponseRedirect(fragment_hits[0]['fragment'].get_absolute_url()) + # return HttpResponseRedirect(results[0].book.get_absolute_url()) + if len(results) == 0: form = PublishingSuggestForm(initial={"books": query + ", "}) return render_to_response('catalogue/search_no_hits.html', {'tags': tags, @@ -216,6 +247,7 @@ def main(request): {'tags': tags, 'prefix': query, 'results': {'author': author_results, + 'translator': translator_results, 'title': title_results, 'content': text_phrase, 'other': everywhere},