Revert "Reapply "Bratseth/special tokens""

4 files changed, 89 insertions, 61 deletions

diff --git a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParseTestCase.java b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParseTestCase.java
index cef8ae1751c..6afea895f3a 100644
--- a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParseTestCase.java
+++ b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParseTestCase.java
@@ -18,14 +18,16 @@ import com.yahoo.prelude.query.PhraseSegmentItem;
 import com.yahoo.prelude.query.PrefixItem;
 import com.yahoo.prelude.query.RankItem;
 import com.yahoo.prelude.query.SubstringItem;
+import com.yahoo.prelude.query.SubstringItem;
 import com.yahoo.prelude.query.SuffixItem;
 import com.yahoo.prelude.query.TaggableItem;
 import com.yahoo.prelude.query.WordItem;
-import com.yahoo.language.process.SpecialTokens;
+import com.yahoo.prelude.query.parser.SpecialTokens;
 import com.yahoo.prelude.query.parser.TestLinguistics;
 import com.yahoo.search.Query;
 import org.junit.Test;
 
+import java.util.Collections;
 import java.util.Iterator;
 
 import static org.junit.Assert.assertEquals;
@@ -1637,7 +1639,7 @@ public class ParseTestCase {
 
     @Test
     public void testNonSpecialTokenParsing() {
-        ParsingTester customTester = new ParsingTester(SpecialTokens.empty());
+        ParsingTester customTester = new ParsingTester(new SpecialTokens("default"));
         customTester.assertParsed("OR c or c with (AND tcp ip)", "c# or c++ with tcp/ip", Query.Type.ANY);
     }
 
diff --git a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParsingTester.java b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParsingTester.java
index fd7e4cbe0e6..17155fff5de 100644
--- a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParsingTester.java
+++ b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/ParsingTester.java
@@ -11,8 +11,8 @@ import com.yahoo.prelude.IndexFacts;
 import com.yahoo.prelude.IndexModel;
 import com.yahoo.prelude.query.Item;
 import com.yahoo.prelude.query.NullItem;
-import com.yahoo.language.process.SpecialTokenRegistry;
-import com.yahoo.language.process.SpecialTokens;
+import com.yahoo.prelude.query.parser.SpecialTokenRegistry;
+import com.yahoo.prelude.query.parser.SpecialTokens;
 import com.yahoo.search.Query;
 import com.yahoo.search.config.IndexInfoConfig;
 import com.yahoo.search.query.parser.Parsable;
@@ -20,9 +20,6 @@ import com.yahoo.search.query.parser.Parser;
 import com.yahoo.search.query.parser.ParserEnvironment;
 import com.yahoo.search.query.parser.ParserFactory;
 
-import java.util.ArrayList;
-import java.util.List;
-
 import static org.junit.Assert.assertEquals;
 import static org.junit.Assert.assertNotNull;
 import static org.junit.Assert.assertTrue;
@@ -35,7 +32,7 @@ import static org.junit.Assert.assertTrue;
 public class ParsingTester {
 
     private static final Linguistics linguistics = new SimpleLinguistics();
-    private final IndexFacts indexFacts;
+    private IndexFacts indexFacts;
     private SpecialTokenRegistry tokenRegistry;
 
     public ParsingTester() {
@@ -52,10 +49,11 @@ public class ParsingTester {
 
     public ParsingTester(IndexFacts indexFacts, SpecialTokens specialTokens) {
         indexFacts.freeze();
+        specialTokens.freeze();
 
         this.indexFacts = indexFacts;
         tokenRegistry = new SpecialTokenRegistry();
-        tokenRegistry = new SpecialTokenRegistry(List.of(specialTokens));
+        tokenRegistry.addSpecialTokens(specialTokens);
     }
 
     /**
@@ -74,13 +72,13 @@ public class ParsingTester {
      * This can be used to add new tokens and passing the resulting special tokens to the constructor of this.
      */
     public static SpecialTokens createSpecialTokens() {
-        List<SpecialTokens.Token> tokens = new ArrayList<>();
-        tokens.add(new SpecialTokens.Token("c++"));
-        tokens.add(new SpecialTokens.Token(".net", "dotnet"));
-        tokens.add(new SpecialTokens.Token("tcp/ip"));
-        tokens.add(new SpecialTokens.Token("c#"));
-        tokens.add(new SpecialTokens.Token("special-token-fs","firstsecond"));
-        return new SpecialTokens("default", tokens);
+        SpecialTokens tokens = new SpecialTokens("default");
+        tokens.addSpecialToken("c++", null);
+        tokens.addSpecialToken(".net", "dotnet");
+        tokens.addSpecialToken("tcp/ip", null);
+        tokens.addSpecialToken("c#", null);
+        tokens.addSpecialToken("special-token-fs","firstsecond");
+        return tokens;
     }
 
     /**
diff --git a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/TokenizerTestCase.java b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/TokenizerTestCase.java
index e10fbd71c72..aa2e9dbcf75 100644
--- a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/TokenizerTestCase.java
+++ b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/TokenizerTestCase.java
@@ -6,13 +6,12 @@ import com.yahoo.prelude.Index;
 import com.yahoo.prelude.IndexFacts;
 import com.yahoo.prelude.IndexModel;
 import com.yahoo.prelude.SearchDefinition;
-import com.yahoo.language.process.SpecialTokenRegistry;
-import com.yahoo.language.process.SpecialTokens;
+import com.yahoo.prelude.query.parser.SpecialTokenRegistry;
+import com.yahoo.prelude.query.parser.SpecialTokens;
 import com.yahoo.prelude.query.parser.Token;
 import com.yahoo.prelude.query.parser.Tokenizer;
 import org.junit.Test;
 
-import java.util.ArrayList;
 import java.util.Collections;
 import java.util.List;
 
@@ -40,11 +39,13 @@ import static org.junit.Assert.assertTrue;
  */
 public class TokenizerTestCase {
 
+    private SpecialTokenRegistry defaultRegistry = new SpecialTokenRegistry("file:src/test/java/com/yahoo/prelude/query/parser/test/replacingtokens.cfg");
+
     @Test
     public void testPlainTokenization() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("drive (to hwy88, 88) +or language:en ugcapi_1 & &a");
 
         assertEquals(new Token(WORD, "drive"), tokens.get(0));
@@ -86,7 +87,7 @@ public class TokenizerTestCase {
     public void testOneSpecialToken() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("c++ lovers, please apply");
 
         assertEquals(new Token(WORD, "c++"), tokens.get(0));
@@ -96,7 +97,7 @@ public class TokenizerTestCase {
     public void testSpecialTokenCombination() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("c#, c++ or .net know, not tcp/ip");
 
         assertEquals(new Token(WORD, "c#"), tokens.get(0));
@@ -122,9 +123,10 @@ public class TokenizerTestCase {
      */
     @Test
     public void testSpecialTokenCJK() {
+        assertEquals("Special tokens configured", 6, defaultRegistry.getSpecialTokens("default").size());
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
         tokenizer.setSubstringSpecialTokens(true);
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("replacing"));
+        tokenizer.setSpecialTokens(defaultRegistry.getSpecialTokens("default"));
 
         List<?> tokens = tokenizer.tokenize("fooc#bar,c++with spacebarknowknowknow,knowknownot know");
         assertEquals(new Token(WORD, "foo"), tokens.get(0));
@@ -149,7 +151,7 @@ public class TokenizerTestCase {
     public void testSpecialTokenCaseInsensitive() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("The AS/400 is great");
 
         assertEquals(new Token(WORD, "The"), tokens.get(0));
@@ -165,7 +167,7 @@ public class TokenizerTestCase {
     public void testSpecialTokenNonMatch() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("c++ c+ aS/400 i/o .net i/ooo ap.net");
 
         assertEquals(new Token(WORD, "c++"), tokens.get(0));
@@ -188,9 +190,18 @@ public class TokenizerTestCase {
 
     @Test
     public void testSpecialTokenConfigurationDefault() {
+        String tokenFile = "file:src/test/java/com/yahoo/prelude/query/parser/test/specialtokens.cfg";
+
+        SpecialTokenRegistry r = new SpecialTokenRegistry(tokenFile);
+        assertEquals("Special tokens configured", 6,
+                r.getSpecialTokens("default").size());
+        assertEquals("Special tokens configured", 4,
+                r.getSpecialTokens("other").size());
+
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(
+                r.getSpecialTokens("default"));
         List<?> tokens = tokenizer.tokenize(
                 "with space, c++ or .... know, not b.s.d.");
 
@@ -213,9 +224,18 @@ public class TokenizerTestCase {
 
     @Test
     public void testSpecialTokenConfigurationOther() {
+        String tokenFile = "file:src/test/java/com/yahoo/prelude/query/parser/test/specialtokens.cfg";
+
+        SpecialTokenRegistry r = new SpecialTokenRegistry(tokenFile);
+        assertEquals("Special tokens configured", 6,
+                r.getSpecialTokens("default").size());
+        assertEquals("Special tokens configured", 4,
+                r.getSpecialTokens("other").size());
+
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("other"));
+        tokenizer.setSpecialTokens(
+                r.getSpecialTokens("other"));
         List<?> tokens = tokenizer.tokenize(
                 "with space,!!!*** [huh] or ------ " + "know, &&&%%% b.s.d.");
 
@@ -247,9 +267,26 @@ public class TokenizerTestCase {
     }
 
     @Test
+    public void testSpecialTokenConfigurationMissing() {
+        String tokenFile = "file:source/bogus/specialtokens.cfg";
+
+        SpecialTokenRegistry r = new SpecialTokenRegistry(tokenFile);
+
+        Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
+
+        tokenizer.setSpecialTokens(r.getSpecialTokens("other"));
+        List<?> tokens = tokenizer.tokenize("c++");
+
+        assertEquals(new Token(WORD, "c"), tokens.get(0));
+        assertEquals(new Token(PLUS, "+"), tokens.get(1));
+        assertEquals(new Token(PLUS, "+"), tokens.get(2));
+    }
+
+    @Test
     public void testTokenReplacing() {
+        assertEquals("Special tokens configured", 6, defaultRegistry.getSpecialTokens("default").size());
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("replacing"));
+        tokenizer.setSpecialTokens(defaultRegistry.getSpecialTokens("default"));
 
         List<?> tokens = tokenizer.tokenize("with space, c++ or .... know, not b.s.d.");
         assertEquals(new Token(WORD, "with-space"), tokens.get(0));
@@ -708,7 +745,7 @@ public class TokenizerTestCase {
     public void testSingleQuoteAsWordCharacter() {
         Tokenizer tokenizer = new Tokenizer(new SimpleLinguistics());
 
-        tokenizer.setSpecialTokens(createSpecialTokens().getSpecialTokens("default"));
+        tokenizer.setSpecialTokens(createSpecialTokens());
         List<?> tokens = tokenizer.tokenize("drive (to hwy88, 88) +or language:en nalle:a'a ugcapi_1 'a' 'a a'");
 
         assertEquals(new Token(WORD, "drive"), tokens.get(0));
@@ -744,38 +781,17 @@ public class TokenizerTestCase {
         assertEquals(new Token(WORD, "a'"), tokens.get(30));
     }
 
-    private SpecialTokenRegistry createSpecialTokens() {
-        List<SpecialTokens.Token> tokens = new ArrayList<>();
-        tokens.add(new SpecialTokens.Token("c+"));
-        tokens.add(new SpecialTokens.Token("c++"));
-        tokens.add(new SpecialTokens.Token(".net"));
-        tokens.add(new SpecialTokens.Token("tcp/ip"));
-        tokens.add(new SpecialTokens.Token("i/o"));
-        tokens.add(new SpecialTokens.Token("c#"));
-        tokens.add(new SpecialTokens.Token("AS/400"));
-        tokens.add(new SpecialTokens.Token("...."));
-        tokens.add(new SpecialTokens.Token("b.s.d."));
-        tokens.add(new SpecialTokens.Token("with space"));
-        tokens.add(new SpecialTokens.Token("dvd\\xB1r"));
-        SpecialTokens defaultTokens = new SpecialTokens("default", tokens);
-
-        tokens = new ArrayList<>();
-        tokens.add(new SpecialTokens.Token("[huh]"));
-        tokens.add(new SpecialTokens.Token("&&&%%%"));
-        tokens.add(new SpecialTokens.Token("------"));
-        tokens.add(new SpecialTokens.Token("!!!***"));
-        SpecialTokens otherTokens = new SpecialTokens("other", tokens);
-
-        tokens = new ArrayList<>();
-        tokens.add(new SpecialTokens.Token("...."));
-        tokens.add(new SpecialTokens.Token("c++", "cpp"));
-        tokens.add(new SpecialTokens.Token("b.s.d."));
-        tokens.add(new SpecialTokens.Token("with space", "with-space"));
-        tokens.add(new SpecialTokens.Token("c#"));
-        tokens.add(new SpecialTokens.Token("know", "knuwww"));
-        SpecialTokens replacingTokens = new SpecialTokens("replacing", tokens);
-
-        return new SpecialTokenRegistry(List.of(defaultTokens, otherTokens, replacingTokens));
+    private SpecialTokens createSpecialTokens() {
+        SpecialTokens tokens = new SpecialTokens("default");
+
+        tokens.addSpecialToken("c+", null);
+        tokens.addSpecialToken("c++", null);
+        tokens.addSpecialToken(".net", null);
+        tokens.addSpecialToken("tcp/ip", null);
+        tokens.addSpecialToken("i/o", null);
+        tokens.addSpecialToken("c#", null);
+        tokens.addSpecialToken("AS/400", null);
+        return tokens;
     }
 
 }
diff --git a/container-search/src/test/java/com/yahoo/prelude/query/parser/test/replacingtokens.cfg b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/replacingtokens.cfg
new file mode 100644
index 00000000000..6a189de0164
--- /dev/null
+++ b/container-search/src/test/java/com/yahoo/prelude/query/parser/test/replacingtokens.cfg
@@ -0,0 +1,12 @@
+tokenlist[1]
+tokenlist[0].name default
+tokenlist[0].tokens[6]
+tokenlist[0].tokens[0].token ....
+tokenlist[0].tokens[1].token c++
+tokenlist[0].tokens[1].replace cpp
+tokenlist[0].tokens[2].token b.s.d.
+tokenlist[0].tokens[3].token with space
+tokenlist[0].tokens[3].replace with-space
+tokenlist[0].tokens[4].token c#
+tokenlist[0].tokens[5].token know
+tokenlist[0].tokens[5].replace knuwww