From 0c3afd47b0e2f43bc9aa9e4e60d33c104c37b81b Mon Sep 17 00:00:00 2001
From: Jon Bratseth <bratseth@gmail.com>
Date: Tue, 15 Jun 2021 19:43:59 +0200
Subject: Require replacements to be applied during tokenization

---
 .../linguistics/LinguisticsAnnotator.java          | 14 ++---------
 .../linguistics/LinguisticsAnnotatorTestCase.java  | 27 +++++++++-------------
 2 files changed, 13 insertions(+), 28 deletions(-)

(limited to 'indexinglanguage')

diff --git a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotator.java b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotator.java
index 8b6ef83f05e..81a5305a778 100644
--- a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotator.java
+++ b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotator.java
@@ -115,9 +115,7 @@ public class LinguisticsAnnotator {
                 }
                 return;
             }
-            if ( ! token.isIndexable()) {
-                return;
-            }
+            if ( ! token.isIndexable()) return;
         }
         String orig = token.getOrig();
         int pos = (int)token.getOffset();
@@ -137,9 +135,6 @@ public class LinguisticsAnnotator {
 
             String lowercasedTerm = lowercasedOrig;
             String term = token.getTokenString();
-            if (term != null) {
-                term = tokenizer.getReplacementTerm(term);
-            }
             if (term != null) {
                 lowercasedTerm = toLowerCase(term);
             }
@@ -155,12 +150,7 @@ public class LinguisticsAnnotator {
             }
         } else {
             String term = token.getTokenString();
-            if (term != null) {
-                term = tokenizer.getReplacementTerm(term);
-            }
-            if (term == null || term.trim().isEmpty()) {
-                return;
-            }
+            if (term == null || term.trim().isEmpty()) return;
             if (termOccurrences.termCountBelowLimit(term))  {
                 parent.span(pos, len).annotate(lowerCaseTermAnnotation(term, token.getOrig()));
             }
diff --git a/indexinglanguage/src/test/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotatorTestCase.java b/indexinglanguage/src/test/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotatorTestCase.java
index afbcf597a46..5f436720990 100644
--- a/indexinglanguage/src/test/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotatorTestCase.java
+++ b/indexinglanguage/src/test/java/com/yahoo/vespa/indexinglanguage/linguistics/LinguisticsAnnotatorTestCase.java
@@ -19,6 +19,7 @@ import org.junit.Test;
 import org.mockito.Mockito;
 
 import java.util.*;
+import java.util.stream.Collectors;
 
 import static org.junit.Assert.assertEquals;
 import static org.junit.Assert.assertTrue;
@@ -30,12 +31,6 @@ public class LinguisticsAnnotatorTestCase {
 
     private static final AnnotatorConfig CONFIG = new AnnotatorConfig();
 
-    // --------------------------------------------------------------------------------
-    //
-    // Tests
-    //
-    // --------------------------------------------------------------------------------
-
     @Test
     public void requireThatAnnotateFailsWithZeroTokens() {
         assertAnnotations(null, "foo");
@@ -145,7 +140,7 @@ public class LinguisticsAnnotatorTestCase {
                     continue;
                 }
                 assertAnnotations(expected, "foo",
-                                  newLinguistics(Arrays.asList(newToken("foo", "foo", type, specialToken)),
+                                  newLinguistics(List.of(newToken("foo", "foo", type, specialToken)),
                                                  Collections.singletonMap("foo", "bar")));
             }
         }
@@ -159,7 +154,7 @@ public class LinguisticsAnnotatorTestCase {
         StringFieldValue val = new StringFieldValue("foo");
         val.setSpanTree(spanTree);
 
-        Linguistics linguistics = newLinguistics(Arrays.asList(newToken("foo", "bar", TokenType.ALPHABETIC, false)),
+        Linguistics linguistics = newLinguistics(List.of(newToken("foo", "bar", TokenType.ALPHABETIC, false)),
                                                  Collections.<String, String>emptyMap());
         new LinguisticsAnnotator(linguistics, CONFIG).annotate(val);
 
@@ -253,11 +248,15 @@ public class LinguisticsAnnotatorTestCase {
     private static class MyTokenizer implements Tokenizer {
 
         final List<Token> tokens;
-        final Map<String, String> replacementTerms;
 
         public MyTokenizer(List<? extends Token> tokens, Map<String, String> replacementTerms) {
-            this.tokens = new ArrayList<>(tokens);
-            this.replacementTerms = replacementTerms;
+            this.tokens = tokens.stream().map(token -> replace(token, replacementTerms)).collect(Collectors.toList());
+        }
+
+        private Token replace(Token token, Map<String, String> replacementTerms) {
+            var simpleToken = (SimpleToken)token;
+            simpleToken.setTokenString(replacementTerms.getOrDefault(token.getTokenString(), token.getTokenString()));
+            return simpleToken;
         }
 
         @Override
@@ -265,10 +264,6 @@ public class LinguisticsAnnotatorTestCase {
             return tokens;
         }
 
-        @Override
-        public String getReplacementTerm(String term) {
-            String replacement = replacementTerms.get(term);
-            return replacement != null ? replacement : term;
-        }
     }
+
 }
-- 
cgit v1.2.3