add uax_url_email analyzer

artsy · Jul 25, 2019 · 4fc1a4f · 4fc1a4f
1 parent 040903a
commit 4fc1a4f
Showing 1 changed file with 6 additions and 1 deletion.
diff --git a/lib/estella/analysis.rb b/lib/estella/analysis.rb
@@ -18,6 +18,9 @@ module Analysis
     NGRAM_ANALYZER =
       { type: 'custom', tokenizer: 'standard_tokenizer', filter: %w[lowercase asciifolding front_ngram_filter] }
 
+    EMAIL_ANALYZER =
+      { type: 'custom', tokenizer: 'uax_url_email', filter: %w[lowercase stop] }
+
     DEFAULT_ANALYSIS = {
       tokenizer: {
         standard_tokenizer: { type: 'standard' }
@@ -30,6 +33,7 @@ module Analysis
         snowball_analyzer: SNOWBALL_ANALYZER,
         shingle_analyzer: SHINGLE_ANALYZER,
         ngram_analyzer: NGRAM_ANALYZER,
+        email_analyzer: EMAIL_ANALYZER,
         search_analyzer: DEFAULT_ANALYZER
       }
     }
@@ -38,7 +42,8 @@ module Analysis
       default: { type: 'text', analyzer: 'default_analyzer' },
       snowball: { type: 'text', analyzer: 'snowball_analyzer' },
       shingle: { type: 'text', analyzer: 'shingle_analyzer' },
-      ngram: { type: 'text', analyzer: 'ngram_analyzer', search_analyzer: 'search_analyzer' }
+      ngram: { type: 'text', analyzer: 'ngram_analyzer', search_analyzer: 'search_analyzer' },
+      email: { type: 'text', analyzer: 'email_analyzer'}
     }
 
     DEFAULT_FIELD_FACTORS = {