search: ascii everywhere and unformalized whitespace

core/search_indexes.py

@@ -45,6 +45,9 @@ class UserIndex(indexes.SearchIndex, indexes.Indexable):
     def get_updated_field(self):
         return "last_update"
 
+    def prepare_auto(self, obj):
+        return self.prepared_data["auto"].strip()
+
 
 class IndexSignalProcessor(signals.BaseSignalProcessor):
     def setup(self):

core/templates/search/indexes/core/user_auto.txt

@@ -1,3 +1,13 @@
-{{ object.first_name }}
-{{ object.last_name }}
-{{ object.nick_name }}
+{% load replace %}
+
+{% with first=object.first_name|safe|slugify last=object.last_name|safe|slugify nick=object.nick_name|default_if_none:""|safe|slugify %}
+
+{{ first|replace:"|-| " }}
+{{ last|replace:"|-| " }}
+{{ nick|replace:"|-| " }}
+{{ first|cut:"-" }}
+{{ last|cut:"-" }}
+{{ nick|cut:"-" }}
+{{ first|cut:"-" }}{{ last|cut:"-" }}
+
+{% endwith %}

core/templatetags/replace.py

@@ -0,0 +1,22 @@
+from django.template.exceptions import TemplateSyntaxError
+from django import template
+from django.template.defaultfilters import stringfilter
+
+register = template.Library()
+
+
+# arg should be of the form "|foo|bar" where the first character is the
+# separator between old and new in value.replace(old, new)
+@register.filter
+@stringfilter
+def replace(value, arg):
+    # s.replace('', '') == s so len(arg) == 2 is fine
+    if len(arg) < 2:
+        raise TemplateSyntaxError("badly formatted argument")
+
+    arg = arg.split(arg[0])
+
+    if len(arg) != 3:
+        raise TemplateSyntaxError("badly formatted argument")
+
+    return value.replace(arg[1], arg[2])

core/views/site.py

@@ -30,6 +30,7 @@ from django.contrib.auth.decorators import login_required
 from django.utils import html
 from django.views.generic import ListView, TemplateView
 from django.conf import settings
+from django.utils.text import slugify
 
 import json
 
@@ -73,7 +74,13 @@ def notification(request, notif_id):
 
 def search_user(query, as_json=False):
     try:
-        res = SearchQuerySet().models(User).autocomplete(auto=html.escape(query))[:20]
+        # slugify turns everything into ascii and every whitespace into -
+        # it ends by removing duplicate - (so '- - ' will turn into '-')
+        # replace('-', ' ') because search is whitespace based
+        query = slugify(query).replace("-", " ")
+        # is this necessary? it's not done when indexing users
+        query = html.escape(query)
+        res = SearchQuerySet().models(User).autocomplete(auto=query)[:20]
         return [r.object for r in res]
     except TypeError:
         return []