X-Git-Url: https://git.mdrn.pl/wolnelektury.git/blobdiff_plain/1801c65004af14a8ed50d60e47d75a415753bbf7..33af401eb4abe15a37d0147807d06a9b19eaa802:/src/search/views.py diff --git a/src/search/views.py b/src/search/views.py index 64382e24b..b6e290b87 100644 --- a/src/search/views.py +++ b/src/search/views.py @@ -1,16 +1,16 @@ -# -*- 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 -from django.template import RequestContext +from django.http.response import HttpResponseRedirect +from django.shortcuts import render from django.views.decorators import cache from django.http import HttpResponse, JsonResponse -from catalogue.utils import split_tags from catalogue.models import Book, Tag -from search.index import Search, SearchResult +from pdcounter.models import Author +from picture.models import Picture +from search.index import Search, SearchResult, PictureResult from suggest.forms import PublishingSuggestForm import re import json @@ -25,7 +25,7 @@ def match_word_re(word): return "[[:<:]]%s[[:>:]]" % word -query_syntax_chars = re.compile(r"[\\/*:(){}]") +query_syntax_chars = re.compile(r"[\\/*:(){}?.[\]+]") def remove_query_syntax_chars(query, replace=' '): @@ -59,8 +59,8 @@ def did_you_mean(query, tokens): @cache.never_cache -def hint(request): - prefix = request.GET.get('term', '') +def hint(request, mozhint=False, param='term'): + prefix = request.GET.get(param, '') if len(prefix) < 2: return JsonResponse([], safe=False) @@ -75,7 +75,7 @@ def hint(request): limit = 20 authors = Tag.objects.filter( - category='author', name__iregex='\m' + prefix).only('name', 'id', 'slug', 'category') + category='author', name_pl__iregex='\m' + prefix).only('name', 'id', 'slug', 'category') data = [ { 'label': author.name, @@ -87,12 +87,23 @@ def hint(request): if len(data) < limit: data += [ { - 'label': '%s, %s' % (b.title, b.author_unicode()), + 'label': b.title, + 'author': b.author_unicode(), 'id': b.id, 'url': b.get_absolute_url() } - for b in Book.objects.filter(title__iregex='\m' + prefix)[:limit-len(data)] + for b in Book.objects.filter(findable=True, title__iregex='\m' + prefix)[:limit-len(data)] ] + + if mozhint: + data = [ + prefix, + [ + item['label'] + for item in data + ] + ] + callback = request.GET.get('callback', None) if callback: return HttpResponse("%s(%s);" % (callback, json.dumps(data)), @@ -104,43 +115,100 @@ def hint(request): @cache.never_cache def main(request): query = request.GET.get('q', '') - query = ' '.join(query.split()) - # filter out private use characters - import unicodedata - query = ''.join(ch for ch in query if unicodedata.category(ch) != 'Co') + + format = request.GET.get('format') + lang = request.GET.get('lang') + epoch = request.GET.get('epoch') + kind = request.GET.get('kind') + genre = request.GET.get('genre') if len(query) < 2: - return render_to_response( - 'catalogue/search_too_short.html', {'prefix': query}, - context_instance=RequestContext(request)) + return render( + request, 'catalogue/search_too_short.html', + {'prefix': query}) elif len(query) > 256: - return render_to_response( - 'catalogue/search_too_long.html', {'prefix': query}, context_instance=RequestContext(request)) + return render( + request, 'catalogue/search_too_long.html', + {'prefix': query}) - query = remove_query_syntax_chars(query) + query = prepare_query(query) + if not (format or lang or epoch or kind or genre): + pd_authors = search_pd_authors(query) + else: + pd_authors = [] + if not format or format != 'obraz': + books = search_books( + query, + lang=lang, + only_audio=format=='audio', + only_synchro=format=='synchro', + epoch=epoch, + kind=kind, + genre=genre + ) + else: + books = [] + if (not format or format == 'obraz') and not lang: + pictures = search_pictures( + query, + epoch=epoch, + kind=kind, + genre=genre + ) + else: + pictures = [] + + suggestion = '' - words = query.split() - if len(words) > 10: - query = ' '.join(words[:10]) + if not (books or pictures or pd_authors): + form = PublishingSuggestForm(initial={"books": query + ", "}) + return render( + request, + 'catalogue/search_no_hits.html', + { + 'form': form, + 'did_you_mean': suggestion + }) - search = Search() + if not (books or pictures) and len(pd_authors) == 1: + return HttpResponseRedirect(pd_authors[0].get_absolute_url()) - tags = search.hint_tags(query, pdcounter=True, prefix=False) - tags = split_tags(tags) + return render( + request, + 'catalogue/search_multiple_hits.html', + { + 'pd_authors': pd_authors, + 'books': books, + 'pictures': pictures, + 'did_you_mean': suggestion, + 'set': { + 'lang': lang, + 'format': format, + 'epoch': epoch, + 'kind': kind, + 'genre': genre, + }, + 'tags': { + 'epoch': Tag.objects.filter(category='epoch', for_books=True), + 'genre': Tag.objects.filter(category='genre', for_books=True), + 'kind': Tag.objects.filter(category='kind', for_books=True), + }, + }) +def search_books(query, lang=None, only_audio=False, only_synchro=False, epoch=None, kind=None, genre=None): + search = Search() results_parts = [] - search_fields = [] + words = query.split() fieldsets = ( - (['authors'], True), - (['title'], True), - (['metadata'], True), - (['text', 'themes_pl'], False), + (['authors', 'authors_nonstem'], True), + (['title', 'title_nonstem'], True), + (['metadata', 'metadata_nonstem'], True), + (['text', 'text_nonstem', 'themes_pl', 'themes_pl_nonstem'], False), ) - for fieldset, is_book in fieldsets: - search_fields += fieldset - results_parts.append(search.search_words(words, search_fields, book=is_book)) - + for fields, is_book in fieldsets: + search_fields += fields + results_parts.append(search.search_words(words, search_fields, required=fields, book=is_book)) results = [] ids_results = {} for results_part in results_parts: @@ -151,38 +219,100 @@ def main(request): else: results.append(result) ids_results[book_id] = result - + descendant_ids = set( + Book.objects.filter(id__in=ids_results, ancestor__in=ids_results).values_list('id', flat=True)) + results = [result for result in results if result.book_id not in descendant_ids] for result in results: search.get_snippets(result, query, num=3) - suggestion = u'' - def ensure_exists(r): try: - return r.book + if not r.book: + return False except Book.DoesNotExist: return False - results = filter(ensure_exists, results) + if lang and r.book.language != lang: + return False + if only_audio and not r.book.has_mp3_file(): + return False + if only_synchro and not r.book.has_daisy_file(): + return False + if epoch and not r.book.tags.filter(category='epoch', slug=epoch).exists(): + return False + if kind and not r.book.tags.filter(category='kind', slug=kind).exists(): + return False + if genre and not r.book.tags.filter(category='genre', slug=genre).exists(): + return False - if not results: - form = PublishingSuggestForm(initial={"books": query + ", "}) - return render_to_response( - 'catalogue/search_no_hits.html', - { - 'tags': tags, - 'prefix': query, - 'form': form, - 'did_you_mean': suggestion - }, - context_instance=RequestContext(request)) + return True - return render_to_response( - 'catalogue/search_multiple_hits.html', - { - 'tags': tags['author'] + tags['kind'] + tags['genre'] + tags['epoch'] + tags['theme'], - 'prefix': query, - 'results': results, - 'did_you_mean': suggestion - }, - context_instance=RequestContext(request)) + results = [r for r in results if ensure_exists(r)] + return results + + +def search_pictures(query, epoch=None, kind=None, genre=None): + search = Search() + results_parts = [] + search_fields = [] + words = query.split() + fieldsets = ( + (['authors', 'authors_nonstem'], True), + (['title', 'title_nonstem'], True), + (['metadata', 'metadata_nonstem'], True), + (['themes_pl', 'themes_pl_nonstem'], False), + ) + for fields, is_book in fieldsets: + search_fields += fields + results_parts.append(search.search_words(words, search_fields, required=fields, book=is_book, picture=True)) + results = [] + ids_results = {} + for results_part in results_parts: + for result in sorted(PictureResult.aggregate(results_part), reverse=True): + picture_id = result.picture_id + if picture_id in ids_results: + ids_results[picture_id].merge(result) + else: + results.append(result) + ids_results[picture_id] = result + + def ensure_exists(r): + try: + if not r.picture: + return False + except Picture.DoesNotExist: + return False + + if epoch and not r.picture.tags.filter(category='epoch', slug=epoch).exists(): + return False + if kind and not r.picture.tags.filter(category='kind', slug=kind).exists(): + return False + if genre and not r.picture.tags.filter(category='genre', slug=genre).exists(): + return False + + return True + + results = [r for r in results if ensure_exists(r)] + return results + + +def search_pd_authors(query): + pd_authors = Author.objects.filter(name__icontains=query) + existing_slugs = Tag.objects.filter( + category='author', slug__in=list(pd_authors.values_list('slug', flat=True))) \ + .values_list('slug', flat=True) + pd_authors = pd_authors.exclude(slug__in=existing_slugs) + return pd_authors + + +def prepare_query(query): + query = ' '.join(query.split()) + # filter out private use characters + import unicodedata + query = ''.join(ch for ch in query if unicodedata.category(ch) != 'Co') + query = remove_query_syntax_chars(query) + + words = query.split() + if len(words) > 10: + query = ' '.join(words[:10]) + return query