Merge branch '224459-more-performant-elasticsearch-file-path-regex' into 'master'

Speed up Advanced global search regex for file path segments

See merge request gitlab-org/gitlab!35292

doc/development/elasticsearch.md

@@ -111,11 +111,7 @@ Patterns:
 - `'"((?:\\"|[^"]|\\")*)"'`: captures terms inside quotes, removing the quotes
 - `"'((?:\\'|[^']|\\')*)'"`: same as above, for single-quotes
 - `'\.([^.]+)(?=\.|\s|\Z)'`: separate terms with periods in-between
-- `'\/?([^\/]+)(?=\/|\b)'`: separate path terms `like/this/one`
-
-#### `edgeNGram_filter`
-
-Uses an [Edge NGram token filter](https://www.elastic.co/guide/en/elasticsearch/reference/5.5/analysis-edgengram-tokenfilter.html) to allow inputs with only parts of a token to find the token. For example it would turn `glasses` into permutations starting with `gl` and ending with `glasses`, which would allow a search for "`glass`" to find the original token `glasses`
+- `'([\p{L}_.-]+)'` : some common chars in file names to keep the whole filename intact (eg. `my_file-ñame.txt`)
 
 ## Gotchas
 

ee/changelogs/unreleased/224459-more-performant-elasticsearch-file-path-regex.yml

+---
+title: Speed up Advanced global search regex for file path segments
+merge_request: 35292
+author:
+type: performance

ee/lib/elastic/latest/config.rb

@@ -38,7 +38,7 @@ module Elastic
               code_analyzer: {
                 type: 'custom',
                 tokenizer: 'whitespace',
-                filter: %w(code lowercase asciifolding)
+                filter: %w(code lowercase asciifolding remove_duplicates)
               },
               code_search_analyzer: {
                 type: 'custom',
@@ -61,7 +61,7 @@ module Elastic
                   '"((?:\\"|[^"]|\\")*)"', # capture terms inside quotes, removing the quotes
                   "'((?:\\'|[^']|\\')*)'", # same as above, for single quotes
                   '\.([^.]+)(?=\.|\s|\Z)', # separate terms on periods
-                  '\/?([^\/]+)(?=\/|\b)' # separate path terms (like/this/one)
+                  '([\p{L}_.-]+)' # some common chars in file names to keep the whole filename intact (eg. my_file-name.txt)
                 ]
               }
             },

ee/spec/lib/gitlab/elastic/search_results_spec.rb

@@ -622,9 +622,21 @@ RSpec.describe Gitlab::Elastic::SearchResults, :elastic, :sidekiq_might_not_need
           Foo.bar(x)
 
           include "bikes-3.4"
+          /a/longer/file-path/absolute_with_specials.txt
+          another/file-path/relative-with-specials.txt
+          /file-path/components-within-slashes/
+          another/file-path/differeñt-lønguage.txt
 
           us-east-2
           bye
+
+          MyJavaClass::javaLangStaticMethodCall
+          $my_perl_object->perlMethodCall
+          LanguageWithSingleColon:someSingleColonMethodCall
+          WouldHappenInManyLanguages,tokenAfterCommaWithNoSpace
+          ParenthesesBetweenTokens)tokenAfterParentheses
+          a.b.c=missing_token_around_equals
+
         FILE
       end
       let(:file_name) { 'elastic_specialchars_test.md' }
@@ -651,6 +663,46 @@ RSpec.describe Gitlab::Elastic::SearchResults, :elastic, :sidekiq_might_not_need
         expect(search_for('"and;colons:too$"')).to include(file_name)
         expect(search_for('bar\(x\)')).to include(file_name)
       end
+
+      it 'finds absolute file paths with slashes and other special chars' do
+        expect(search_for('"absolute_with_specials.txt"')).to include(file_name)
+      end
+
+      it 'finds relative file paths with slashes and other special chars' do
+        expect(search_for('"relative-with-specials.txt"')).to include(file_name)
+      end
+
+      it 'finds file path components within slashes for directories' do
+        expect(search_for('"components-within-slashes"')).to include(file_name)
+      end
+
+      it 'finds file paths for various languages' do
+        expect(search_for('"differeñt-lønguage.txt"')).to include(file_name)
+      end
+
+      it 'finds java style static method call after ::' do
+        expect(search_for('javaLangStaticMethodCall')).to include(file_name)
+      end
+
+      it 'finds perl object method call' do
+        expect(search_for('perlMethodCall')).to include(file_name)
+      end
+
+      it 'finds tokens after a colon' do
+        expect(search_for('someSingleColonMethodCall')).to include(file_name)
+      end
+
+      it 'finds tokens after a comma with no space' do
+        expect(search_for('tokenAfterCommaWithNoSpace')).to include(file_name)
+      end
+
+      it 'finds a token directly after parentheses' do
+        expect(search_for('tokenAfterParentheses')).to include(file_name)
+      end
+
+      it 'finds a token after = without a space' do
+        expect(search_for('missing_token_around_equals')).to include(file_name)
+      end
     end
   end