pipeline bug

[wolnelektury.git] / apps / search / views.py

diff --git a/apps/search/views.py b/apps/search/views.py
index 2109a73..f7aa77c 100644 (file)
--- 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") \
@@ -191,7 +223,8 @@ def main(request):
 
     results = author_results + translator_results + title_results + text_phrase + everywhere
     # ensure books do exists & sort them
-    results.sort(reverse=True)
+    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: