Support search on Asian languages

nitely · Jan 28, 2022 · 13f338c · 13f338c
1 parent 01e5799
commit 13f338c
Show file tree

Hide file tree

Showing 3 changed files with 16 additions and 1 deletion.
diff --git a/spirit/core/conf/defaults.py b/spirit/core/conf/defaults.py
@@ -202,6 +202,12 @@
 #: Latin, Greek, and Cyrillic charsets
 ST_EXTENDED_FONT = False
 
+#: This enables search on Asian languages,
+#: and across word boundaries. It'll increase the index size.
+#: Changing this setting requires re-building the
+#: search index by running `python manage.py rebuild_index`
+ST_NGRAM_SEARCH = False
+
 # Tests helper
 ST_TESTS_RATELIMIT_NEVER_EXPIRE = False
 

diff --git a/spirit/extra/project_template/project_name/settings/prod.py b/spirit/extra/project_template/project_name/settings/prod.py
@@ -125,3 +125,7 @@
 
 # https://spirit.readthedocs.io/en/latest/settings.html#spirit.core.conf.defaults.ST_SITE_URL
 ST_SITE_URL = 'https://example.com/'
+
+# If using elasticsearch, this fixes:
+# https://github.com/django-haystack/django-haystack/issues/1057
+ELASTICSEARCH_DEFAULT_NGRAM_SEARCH_ANALYZER = 'standard'
diff --git a/spirit/search/search_indexes.py b/spirit/search/search_indexes.py
@@ -27,9 +27,14 @@ def convert(self, value):
  return bool(value)
 
 
+TEXT_FIELD = indexes.CharField
+if settings.ST_NGRAM_SEARCH:
+ TEXT_FIELD = indexes.NgramField
+
+
 class TopicIndex(indexes.SearchIndex, indexes.Indexable):
 
- text = indexes.CharField(document=True, use_template=True, stored=False)
+ text = TEXT_FIELD(document=True, use_template=True, stored=False)
  category_id = indexes.IntegerField(model_attr='category_id', stored=False)
  is_removed = BooleanField(stored=False)