Remove special-casing of Synonym filters in AnalysisRegistry

modules/analysis-common/src/main/java/org/elasticsearch/analysis/common/CommonAnalysisPlugin.java

@@ -276,6 +276,8 @@ public Map<String, AnalysisProvider<TokenFilterFactory>> getTokenFilters() {
         filters.put("sorani_normalization", SoraniNormalizationFilterFactory::new);
         filters.put("stemmer_override", requiresAnalysisSettings(StemmerOverrideTokenFilterFactory::new));
         filters.put("stemmer", StemmerTokenFilterFactory::new);
+        filters.put("synonym", requiresAnalysisSettings(SynonymTokenFilterFactory::new));
+        filters.put("synonym_graph", requiresAnalysisSettings(SynonymGraphTokenFilterFactory::new));
         filters.put("trim", TrimTokenFilterFactory::new);
         filters.put("truncate", requiresAnalysisSettings(TruncateTokenFilterFactory::new));
         filters.put("unique", UniqueTokenFilterFactory::new);

server/src/main/java/org/elasticsearch/index/analysis/ESSolrSynonymParser.java → modules/analysis-common/src/main/java/org/elasticsearch/analysis/common/ESSolrSynonymParser.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.logging.log4j.Logger;
 import org.apache.logging.log4j.LogManager;

server/src/main/java/org/elasticsearch/index/analysis/ESWordnetSynonymParser.java → modules/analysis-common/src/main/java/org/elasticsearch/analysis/common/ESWordnetSynonymParser.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.logging.log4j.Logger;
 import org.apache.logging.log4j.LogManager;

server/src/main/java/org/elasticsearch/index/analysis/SynonymGraphTokenFilterFactory.java → modules/analysis-common/src/main/java/org/elasticsearch/analysis/common/SynonymGraphTokenFilterFactory.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.lucene.analysis.Analyzer;
 import org.apache.lucene.analysis.TokenStream;
@@ -26,16 +26,18 @@
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.env.Environment;
 import org.elasticsearch.index.IndexSettings;
+import org.elasticsearch.index.analysis.CharFilterFactory;
+import org.elasticsearch.index.analysis.TokenFilterFactory;
+import org.elasticsearch.index.analysis.TokenizerFactory;
 
-import java.io.IOException;
 import java.util.List;
 import java.util.function.Function;
 
 public class SynonymGraphTokenFilterFactory extends SynonymTokenFilterFactory {
 
-    public SynonymGraphTokenFilterFactory(IndexSettings indexSettings, Environment env, AnalysisRegistry analysisRegistry,
-                                     String name, Settings settings) throws IOException {
-        super(indexSettings, env, analysisRegistry, name, settings);
+    public SynonymGraphTokenFilterFactory(IndexSettings indexSettings, Environment env,
+                                     String name, Settings settings) {
+        super(indexSettings, env, name, settings);
     }
 
     @Override

server/src/main/java/org/elasticsearch/index/analysis/SynonymTokenFilterFactory.java → modules/analysis-common/src/main/java/org/elasticsearch/analysis/common/SynonymTokenFilterFactory.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.lucene.analysis.Analyzer;
 import org.apache.lucene.analysis.TokenStream;
@@ -26,8 +26,13 @@
 import org.elasticsearch.common.settings.Settings;
 import org.elasticsearch.env.Environment;
 import org.elasticsearch.index.IndexSettings;
+import org.elasticsearch.index.analysis.AbstractTokenFilterFactory;
+import org.elasticsearch.index.analysis.Analysis;
+import org.elasticsearch.index.analysis.CharFilterFactory;
+import org.elasticsearch.index.analysis.CustomAnalyzer;
+import org.elasticsearch.index.analysis.TokenFilterFactory;
+import org.elasticsearch.index.analysis.TokenizerFactory;
 
-import java.io.IOException;
 import java.io.Reader;
 import java.io.StringReader;
 import java.util.List;
@@ -41,8 +46,8 @@ public class SynonymTokenFilterFactory extends AbstractTokenFilterFactory {
     protected final Settings settings;
     protected final Environment environment;
 
-    public SynonymTokenFilterFactory(IndexSettings indexSettings, Environment env, AnalysisRegistry analysisRegistry,
-                                      String name, Settings settings) throws IOException {
+    public SynonymTokenFilterFactory(IndexSettings indexSettings, Environment env,
+                                      String name, Settings settings) {
         super(indexSettings, name, settings);
         this.settings = settings;
 

modules/analysis-common/src/test/java/org/elasticsearch/analysis/common/CommonAnalysisFactoryTests.java

@@ -25,7 +25,6 @@
 import org.apache.lucene.analysis.reverse.ReverseStringFilterFactory;
 import org.apache.lucene.analysis.snowball.SnowballPorterFilterFactory;
 import org.elasticsearch.index.analysis.SoraniNormalizationFilterFactory;
-import org.elasticsearch.index.analysis.SynonymTokenFilterFactory;
 import org.elasticsearch.indices.analysis.AnalysisFactoryTestCase;
 
 import java.util.List;

server/src/test/java/org/elasticsearch/index/analysis/ESSolrSynonymParserTests.java → modules/analysis-common/src/test/java/org/elasticsearch/analysis/common/ESSolrSynonymParserTests.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.lucene.analysis.CharArraySet;
 import org.apache.lucene.analysis.StopFilter;

server/src/test/java/org/elasticsearch/index/analysis/ESWordnetSynonymParserTests.java → modules/analysis-common/src/test/java/org/elasticsearch/analysis/common/ESWordnetSynonymParserTests.java

@@ -17,7 +17,7 @@
  * under the License.
  */
 
-package org.elasticsearch.index.analysis;
+package org.elasticsearch.analysis.common;
 
 import org.apache.lucene.analysis.CharArraySet;
 import org.apache.lucene.analysis.StopFilter;

modules/analysis-common/src/test/resources/rest-api-spec/test/analysis-common/40_token_filters.yml

@@ -223,6 +223,68 @@
     - match:  { tokens.0.token: Foo }
     - match:  { tokens.1.token: Bar! }
 
+---
+"synonym":
+    - do:
+        indices.create:
+          index: test
+          body:
+            settings:
+              analysis:
+                filter:
+                  my_synonym:
+                    type: synonym
+                    synonyms: ["car,auto"]
+
+    - do:
+        indices.analyze:
+          index: test
+          body:
+            text: what car magazine
+            tokenizer: whitespace
+            filter: [ my_synonym ]
+    - length: { tokens: 4 }
+    - match:  { tokens.0.token: what }
+    - match:  { tokens.0.position: 0 }
+    - match:  { tokens.1.token: car }
+    - match:  { tokens.1.position: 1 }
+    - match:  { tokens.2.token: auto }
+    - match:  { tokens.2.position: 1 }
+    - match:  { tokens.3.token: magazine }
+    - match:  { tokens.3.position: 2 }
+
+---
+"synonym_graph":
+    - do:
+        indices.create:
+          index: test
+          body:
+            settings:
+              analysis:
+                filter:
+                  my_graph_synonym:
+                    type: synonym_graph
+                    synonyms: [ "guinea pig,cavy" ]
+
+    - do:
+        indices.analyze:
+          index: test
+          body:
+            text: my guinea pig snores
+            tokenizer: whitespace
+            filter: [ my_graph_synonym ]
+    - length: { tokens: 5 }
+    - match:  { tokens.0.token: my }
+    - match:  { tokens.1.token: cavy }
+    - match:  { tokens.1.position: 1 }
+    - match:  { tokens.1.positionLength: 2 }
+    - match:  { tokens.2.token: guinea }
+    - match:  { tokens.2.position: 1 }
+    - match:  { tokens.3.token: pig }
+    - match:  { tokens.3.position: 2 }
+    - match:  { tokens.4.token: snores }
+    - match:  { tokens.4.position: 3 }
+
 ---
 "synonym_graph and flatten_graph":
     - do:

server/src/main/java/org/elasticsearch/index/analysis/AnalysisRegistry.java

@@ -158,16 +158,8 @@ public IndexAnalyzers build(IndexSettings indexSettings) throws IOException {
 
     public Map<String, TokenFilterFactory> buildTokenFilterFactories(IndexSettings indexSettings) throws IOException {
         final Map<String, Settings> tokenFiltersSettings = indexSettings.getSettings().getGroups(INDEX_ANALYSIS_FILTER);
-        Map<String, AnalysisModule.AnalysisProvider<TokenFilterFactory>> tokenFilters = new HashMap<>(this.tokenFilters);
-        /*
-         * synonym and synonym_graph are different than everything else since they need access to the tokenizer factories for the index.
-         * instead of building the infrastructure for plugins we rather make it a real exception to not pollute the general interface and
-         * hide internal data-structures as much as possible.
-         */
-        tokenFilters.put("synonym", requiresAnalysisSettings((is, env, name, settings) -> new SynonymTokenFilterFactory(is, env, this, name, settings)));
-        tokenFilters.put("synonym_graph", requiresAnalysisSettings((is, env, name, settings) -> new SynonymGraphTokenFilterFactory(is, env, this, name, settings)));
-
-        return buildMapping(Component.FILTER, indexSettings, tokenFiltersSettings, Collections.unmodifiableMap(tokenFilters), prebuiltAnalysis.preConfiguredTokenFilters);
+        return buildMapping(Component.FILTER, indexSettings, tokenFiltersSettings,
+            Collections.unmodifiableMap(this.tokenFilters), prebuiltAnalysis.preConfiguredTokenFilters);
     }
 
     public Map<String, TokenizerFactory> buildTokenizerFactories(IndexSettings indexSettings) throws IOException {
@@ -222,18 +214,7 @@ public AnalysisProvider<TokenFilterFactory> getTokenFilterProvider(String tokenF
         if (tokenFilterSettings.containsKey(tokenFilter)) {
             Settings currentSettings = tokenFilterSettings.get(tokenFilter);
             String typeName = currentSettings.get("type");
-            /*
-             * synonym and synonym_graph are different than everything else since they need access to the tokenizer factories for the index.
-             * instead of building the infrastructure for plugins we rather make it a real exception to not pollute the general interface and
-             * hide internal data-structures as much as possible.
-             */
-            if ("synonym".equals(typeName)) {
-                return requiresAnalysisSettings((is, env, name, settings) -> new SynonymTokenFilterFactory(is, env, this, name, settings));
-            } else if ("synonym_graph".equals(typeName)) {
-                return requiresAnalysisSettings((is, env, name, settings) -> new SynonymGraphTokenFilterFactory(is, env, this, name, settings));
-            } else {
-                return getAnalysisProvider(Component.FILTER, tokenFilters, tokenFilter, typeName);
-            }
+            return getAnalysisProvider(Component.FILTER, tokenFilters, tokenFilter, typeName);
         } else {
             return getTokenFilterProvider(tokenFilter);
         }
@@ -257,19 +238,6 @@ public AnalysisProvider<CharFilterFactory> getCharFilterProvider(String charFilt
         }
     }
 
-    private static <T> AnalysisModule.AnalysisProvider<T> requiresAnalysisSettings(AnalysisModule.AnalysisProvider<T> provider) {
-        return new AnalysisModule.AnalysisProvider<T>() {
-            @Override
-            public T get(IndexSettings indexSettings, Environment environment, String name, Settings settings) throws IOException {
-                return provider.get(indexSettings, environment, name, settings);
-            }
-            @Override
-            public boolean requiresAnalysisSettings() {
-                return true;
-            }
-        };
-    }
-
     enum Component {
         ANALYZER {
             @Override

server/src/test/java/org/elasticsearch/index/mapper/TextFieldMapperTests.java

@@ -19,6 +19,7 @@
 
 package org.elasticsearch.index.mapper;
 
+import org.apache.lucene.analysis.MockSynonymAnalyzer;
 import org.apache.lucene.analysis.TokenStream;
 import org.apache.lucene.analysis.tokenattributes.CharTermAttribute;
 import org.apache.lucene.document.FieldType;
@@ -55,6 +56,7 @@
 import org.elasticsearch.index.mapper.TextFieldMapper.TextFieldType;
 import org.elasticsearch.index.query.MatchPhraseQueryBuilder;
 import org.elasticsearch.index.query.QueryShardContext;
+import org.elasticsearch.index.search.MatchQuery;
 import org.elasticsearch.index.shard.IndexShard;
 import org.elasticsearch.plugins.Plugin;
 import org.elasticsearch.test.ESSingleNodeTestCase;
@@ -82,10 +84,6 @@ public class TextFieldMapperTests extends ESSingleNodeTestCase {
     @Before
     public void setup() {
         Settings settings = Settings.builder()
-            .put("index.analysis.filter.mySynonyms.type", "synonym")
-            .putList("index.analysis.filter.mySynonyms.synonyms", Collections.singletonList("car, auto"))
-            .put("index.analysis.analyzer.synonym.tokenizer", "standard")
-            .put("index.analysis.analyzer.synonym.filter", "mySynonyms")
             // Stop filter remains in server as it is part of lucene-core
             .put("index.analysis.analyzer.my_stop_analyzer.tokenizer", "standard")
             .put("index.analysis.analyzer.my_stop_analyzer.filter", "stop")
@@ -739,7 +737,7 @@ public void testFastPhraseMapping() throws IOException {
             .endObject()
             .startObject("synfield")
             .field("type", "text")
-            .field("analyzer", "synonym")
+            .field("analyzer", "standard")  // will be replaced with MockSynonymAnalyzer
             .field("index_phrases", true)
             .endObject()
             .endObject()
@@ -766,11 +764,13 @@ public void testFastPhraseMapping() throws IOException {
         assertThat(q5,
             is(new PhraseQuery.Builder().add(new Term("field", "sparkle")).add(new Term("field", "stopword"), 2).build()));
 
-        Query q6 = new MatchPhraseQueryBuilder("synfield", "motor car").toQuery(queryShardContext);
+        MatchQuery matchQuery = new MatchQuery(queryShardContext);
+        matchQuery.setAnalyzer(new MockSynonymAnalyzer());
+        Query q6 = matchQuery.parse(MatchQuery.Type.PHRASE, "synfield", "motor dogs");
         assertThat(q6, is(new MultiPhraseQuery.Builder()
             .add(new Term[]{
-                new Term("synfield._index_phrase", "motor car"),
-                new Term("synfield._index_phrase", "motor auto")})
+                new Term("synfield._index_phrase", "motor dogs"),
+                new Term("synfield._index_phrase", "motor dog")})
             .build()));
 
         ParsedDocument doc = mapper.parse(SourceToParse.source("test", "type", "1", BytesReference

server/src/test/java/org/elasticsearch/index/search/MultiMatchQueryTests.java

@@ -73,11 +73,7 @@ protected Collection<Class<? extends Plugin>> getPlugins() {
 
     @Before
     public void setup() throws IOException {
-        Settings settings = Settings.builder()
-            .put("index.analysis.filter.syns.type","synonym")
-            .putList("index.analysis.filter.syns.synonyms","quick,fast")
-            .put("index.analysis.analyzer.syns.tokenizer","standard")
-            .put("index.analysis.analyzer.syns.filter","syns").build();
+        Settings settings = Settings.builder().build();
         IndexService indexService = createIndex("test", settings);
         MapperService mapperService = indexService.mapperService();
         String mapping = "{\n" +
@@ -87,11 +83,11 @@ public void setup() throws IOException {
                 "                  \"properties\":{\n" +
                 "                        \"first\": {\n" +
                 "                            \"type\":\"text\",\n" +
-                "                            \"analyzer\":\"syns\"\n" +
+                "                            \"analyzer\":\"standard\"\n" +
                 "                        }," +
                 "                        \"last\": {\n" +
                 "                            \"type\":\"text\",\n" +
-                "                            \"analyzer\":\"syns\"\n" +
+                "                            \"analyzer\":\"standard\"\n" +
                 "                        }" +
                 "                   }" +
                 "            }\n" +
@@ -221,25 +217,27 @@ public void testMultiMatchCrossFieldsWithSynonyms() throws IOException {
         QueryShardContext queryShardContext = indexService.newQueryShardContext(
             randomInt(20), null, () -> { throw new UnsupportedOperationException(); }, null);
 
+        MultiMatchQuery parser = new MultiMatchQuery(queryShardContext);
+        parser.setAnalyzer(new MockSynonymAnalyzer());
+        Map<String, Float> fieldNames = new HashMap<>();
+        fieldNames.put("name.first", 1.0f);
+
         // check that synonym query is used for a single field
-        Query parsedQuery =
-            multiMatchQuery("quick").field("name.first")
-                .type(MultiMatchQueryBuilder.Type.CROSS_FIELDS).toQuery(queryShardContext);
+        Query parsedQuery = parser.parse(MultiMatchQueryBuilder.Type.CROSS_FIELDS, fieldNames, "dogs", null);
         Term[] terms = new Term[2];
-        terms[0] = new Term("name.first", "quick");
-        terms[1] = new Term("name.first", "fast");
+        terms[0] = new Term("name.first", "dog");
+        terms[1] = new Term("name.first", "dogs");
         Query expectedQuery = new SynonymQuery(terms);
         assertThat(parsedQuery, equalTo(expectedQuery));
 
         // check that blended term query is used for multiple fields
-        parsedQuery =
-            multiMatchQuery("quick").field("name.first").field("name.last")
-                .type(MultiMatchQueryBuilder.Type.CROSS_FIELDS).toQuery(queryShardContext);
+        fieldNames.put("name.last", 1.0f);
+        parsedQuery = parser.parse(MultiMatchQueryBuilder.Type.CROSS_FIELDS, fieldNames, "dogs", null);
         terms = new Term[4];
-        terms[0] = new Term("name.first", "quick");
-        terms[1] = new Term("name.first", "fast");
-        terms[2] = new Term("name.last", "quick");
-        terms[3] = new Term("name.last", "fast");
+        terms[0] = new Term("name.first", "dog");
+        terms[1] = new Term("name.first", "dogs");
+        terms[2] = new Term("name.last", "dog");
+        terms[3] = new Term("name.last", "dogs");
         float[] boosts = new float[4];
         Arrays.fill(boosts, 1.0f);
         expectedQuery = BlendedTermQuery.dismaxBlendedQuery(terms, boosts, 1.0f);

test/framework/src/main/java/org/elasticsearch/indices/analysis/AnalysisFactoryTestCase.java

@@ -31,8 +31,6 @@
 import org.elasticsearch.index.analysis.ShingleTokenFilterFactory;
 import org.elasticsearch.index.analysis.StandardTokenizerFactory;
 import org.elasticsearch.index.analysis.StopTokenFilterFactory;
-import org.elasticsearch.index.analysis.SynonymGraphTokenFilterFactory;
-import org.elasticsearch.index.analysis.SynonymTokenFilterFactory;
 import org.elasticsearch.plugins.AnalysisPlugin;
 import org.elasticsearch.test.ESTestCase;
 
@@ -169,8 +167,8 @@ private static String toCamelCase(String s) {
         .put("stemmeroverride",           MovedToAnalysisCommon.class)
         .put("stop",                      StopTokenFilterFactory.class)
         .put("swedishlightstem",          MovedToAnalysisCommon.class)
-        .put("synonym",                   SynonymTokenFilterFactory.class)
-        .put("synonymgraph",              SynonymGraphTokenFilterFactory.class)
+        .put("synonym",                   MovedToAnalysisCommon.class)
+        .put("synonymgraph",              MovedToAnalysisCommon.class)
         .put("trim",                      MovedToAnalysisCommon.class)
         .put("truncate",                  MovedToAnalysisCommon.class)
         .put("turkishlowercase",          MovedToAnalysisCommon.class)