escape user-provided strings used in regular expressions

diff --git a/src/api/handlers.py b/src/api/handlers.py
index eb56e20..8810626 100644 (file)
--- a/src/api/handlers.py
+++ b/src/api/handlers.py
@@ -19,6 +19,7 @@ from picture.models import Picture
 from picture.forms import PictureImportForm
 
 from stats.utils import piwik_track
+from wolnelektury.utils import re_escape
 
 from . import emitters  # Register our emitters
 
@@ -334,6 +335,7 @@ class FilterBooksHandler(AnonymousBooksHandler):
         if (search_string is not None) and len(search_string) < 3:
             search_string = None
         if search_string:
+            search_string = re_escape(search_string)
             books_author = books.filter(cached_author__iregex='\m' + search_string)
             books_title = books.filter(title__iregex='\m' + search_string)
             books_title = books_title.exclude(id__in=list(books_author.values_list('id', flat=True)))

diff --git a/src/search/views.py b/src/search/views.py
index 70a216e..b0f0641 100644 (file)
--- a/src/search/views.py
+++ b/src/search/views.py
@@ -17,6 +17,8 @@ from suggest.forms import PublishingSuggestForm
 import re
 import json
 
+from wolnelektury.utils import re_escape
+
 
 def match_word_re(word):
     if 'sqlite' in settings.DATABASES['default']['ENGINE']:
@@ -29,7 +31,7 @@ query_syntax_chars = re.compile(r"[\\/*:(){}]")
 
 
 def remove_query_syntax_chars(query, replace=' '):
-    return query_syntax_chars.sub(' ', query)
+    return query_syntax_chars.sub(replace, query)
 
 
 def did_you_mean(query, tokens):
@@ -64,7 +66,7 @@ def hint(request):
     if len(prefix) < 2:
         return JsonResponse([], safe=False)
 
-    prefix = remove_query_syntax_chars(prefix)
+    prefix = re_escape(' '.join(remove_query_syntax_chars(prefix).split()))
 
     try:
         limit = int(request.GET.get('max', ''))
@@ -81,7 +83,7 @@ def hint(request):
             'id': author.id,
             'url': author.get_absolute_url(),
         }
-        for author in Tag.objects.filter(category='author', name__iregex='\m' + prefix)[:10]
+        for author in Tag.objects.filter(category='author', name__iregex=u'\m' + prefix)[:10]
     ]
     if len(data) < limit:
         data += [

diff --git a/src/wolnelektury/utils.py b/src/wolnelektury/utils.py
index 8c5ead6..d20039c 100644 (file)
--- a/src/wolnelektury/utils.py
+++ b/src/wolnelektury/utils.py
@@ -12,6 +12,7 @@ from functools import wraps
 import pytz
 from inspect import getargspec
 
+import re
 from django.core.mail import send_mail
 from django.http import HttpResponse
 from django.template import RequestContext
@@ -149,3 +150,8 @@ class UnicodeCSVWriter(object):
     def writerows(self, rows):
         for row in rows:
             self.writerow(row)
+
+
+# the original re.escape messes with unicode
+def re_escape(s):
+    return re.sub(r"[(){}\[\].*?|^$\\+-]", r"\\\g<0>", s)