diff options
author | Jon Bratseth <bratseth@vespa.ai> | 2023-11-10 21:38:23 +0100 |
---|---|---|
committer | Jon Bratseth <bratseth@vespa.ai> | 2023-11-10 21:38:23 +0100 |
commit | 90965807bd8a6134fe92f7058b3b0a3287050c2a (patch) | |
tree | ada0de730852649e4e648c0c4093e9288988ea77 | |
parent | d74701fe719494819eeb7f5c1af4b59a5c652df6 (diff) |
Prefer first stem to original if non equal
13 files changed, 136 insertions, 111 deletions
diff --git a/document/src/main/java/com/yahoo/document/annotation/Annotation.java b/document/src/main/java/com/yahoo/document/annotation/Annotation.java index 3d9300550ff..012e21ecb85 100644 --- a/document/src/main/java/com/yahoo/document/annotation/Annotation.java +++ b/document/src/main/java/com/yahoo/document/annotation/Annotation.java @@ -129,7 +129,7 @@ public class Annotation implements Comparable<Annotation> { } /** - * WARNING! Should only be used by deserializers! Sets the span node that this annotation points to. + * WARNING! Should only be used by deserializers! Sets the span node that this annotation points to. * * @param spanNode the span node that this annotation shall point to. */ @@ -221,10 +221,9 @@ public class Annotation implements Comparable<Annotation> { @Override public String toString() { - String retval = "annotation of type " + type; - retval += ((value == null) ? " (no value)" : " (with value)"); - retval += ((spanNode == null) ? " (no span)" : (" with span "+spanNode)); - return retval; + return type.getName() + " annotation, " + + (value == null ? "no value" : "value '" + value + "'") + ", " + + (spanNode == null ? "no span" : "span " + spanNode); } diff --git a/document/src/main/java/com/yahoo/document/annotation/AnnotationContainer.java b/document/src/main/java/com/yahoo/document/annotation/AnnotationContainer.java index 6e2b986a478..ac2e6aefa1b 100644 --- a/document/src/main/java/com/yahoo/document/annotation/AnnotationContainer.java +++ b/document/src/main/java/com/yahoo/document/annotation/AnnotationContainer.java @@ -23,11 +23,7 @@ abstract class AnnotationContainer { */ abstract void annotate(Annotation annotation); - /** - * Returns a mutable collection of annotations. - * - * @return a mutable collection of annotations. - */ + /** Returns a mutable collection of the annotations in this. */ abstract Collection<Annotation> annotations(); /** diff --git a/document/src/main/java/com/yahoo/document/annotation/AnnotationType2AnnotationContainer.java b/document/src/main/java/com/yahoo/document/annotation/AnnotationType2AnnotationContainer.java index d8709baa3a1..01b8f990bb4 100644 --- a/document/src/main/java/com/yahoo/document/annotation/AnnotationType2AnnotationContainer.java +++ b/document/src/main/java/com/yahoo/document/annotation/AnnotationType2AnnotationContainer.java @@ -16,6 +16,7 @@ import java.util.NoSuchElementException; */ // TODO: Should this be removed? public class AnnotationType2AnnotationContainer extends IteratingAnnotationContainer { + private final Multimap<AnnotationType, Annotation> annotationType2Annotation = Multimaps.newMultimap(new IdentityHashMap<>(), ArrayList::new); @Override @@ -31,7 +32,6 @@ public class AnnotationType2AnnotationContainer extends IteratingAnnotationConta } @Override - @SuppressWarnings("unchecked") Collection<Annotation> annotations() { return annotationType2Annotation.values(); } @@ -56,12 +56,12 @@ public class AnnotationType2AnnotationContainer extends IteratingAnnotationConta } private class NonRecursiveIterator implements Iterator<Annotation> { + private final IdentityHashMap<SpanNode, SpanNode> nodes; private final Iterator<Annotation> annotationIt; private Annotation next = null; private boolean nextCalled; - @SuppressWarnings("unchecked") public NonRecursiveIterator(IdentityHashMap<SpanNode, SpanNode> nodes) { this.nodes = nodes; this.annotationIt = annotationType2Annotation.values().iterator(); @@ -106,4 +106,5 @@ public class AnnotationType2AnnotationContainer extends IteratingAnnotationConta nextCalled = false; } } + } diff --git a/document/src/main/java/com/yahoo/document/annotation/IteratingAnnotationContainer.java b/document/src/main/java/com/yahoo/document/annotation/IteratingAnnotationContainer.java index 54af80e6e43..c7c3545fc09 100644 --- a/document/src/main/java/com/yahoo/document/annotation/IteratingAnnotationContainer.java +++ b/document/src/main/java/com/yahoo/document/annotation/IteratingAnnotationContainer.java @@ -5,30 +5,29 @@ import java.util.IdentityHashMap; import java.util.Iterator; /** - * @author <a href="mailto:einarmr@yahoo-inc.com">Einar M R Rosenvinge</a> + * @author Einar M R Rosenvinge */ abstract class IteratingAnnotationContainer extends AnnotationContainer { @Override Iterator<Annotation> iterator(SpanNode node) { - IdentityHashMap<SpanNode, SpanNode> nodes = new IdentityHashMap<SpanNode, SpanNode>(); + IdentityHashMap<SpanNode, SpanNode> nodes = new IdentityHashMap<>(); nodes.put(node, node); return iterator(nodes); } @Override Iterator<Annotation> iteratorRecursive(SpanNode node) { - IdentityHashMap<SpanNode, SpanNode> nodes = new IdentityHashMap<SpanNode, SpanNode>(); + IdentityHashMap<SpanNode, SpanNode> nodes = new IdentityHashMap<>(); nodes.put(node, node); - { - Iterator<SpanNode> childrenIt = node.childIteratorRecursive(); - while (childrenIt.hasNext()) { - SpanNode child = childrenIt.next(); - nodes.put(child, child); - } + Iterator<SpanNode> childrenIt = node.childIteratorRecursive(); + while (childrenIt.hasNext()) { + SpanNode child = childrenIt.next(); + nodes.put(child, child); } return iterator(nodes); } abstract Iterator<Annotation> iterator(IdentityHashMap<SpanNode, SpanNode> nodes); + } diff --git a/document/src/main/java/com/yahoo/document/annotation/ListAnnotationContainer.java b/document/src/main/java/com/yahoo/document/annotation/ListAnnotationContainer.java index c2c22558a32..b8d8cd692d8 100644 --- a/document/src/main/java/com/yahoo/document/annotation/ListAnnotationContainer.java +++ b/document/src/main/java/com/yahoo/document/annotation/ListAnnotationContainer.java @@ -13,6 +13,7 @@ import java.util.NoSuchElementException; * @author Einar M R Rosenvinge */ public class ListAnnotationContainer extends IteratingAnnotationContainer { + private final List<Annotation> annotations = new LinkedList<>(); @Override @@ -38,9 +39,8 @@ public class ListAnnotationContainer extends IteratingAnnotationContainer { @Override public boolean equals(Object o) { if (this == o) return true; - if (!(o instanceof ListAnnotationContainer)) return false; - ListAnnotationContainer that = (ListAnnotationContainer) o; - if (!annotations.equals(that.annotations)) return false; + if (!(o instanceof ListAnnotationContainer other)) return false; + if (!annotations.equals(other.annotations)) return false; return true; } @@ -50,8 +50,9 @@ public class ListAnnotationContainer extends IteratingAnnotationContainer { } private class AnnotationIterator implements Iterator<Annotation> { - private IdentityHashMap<SpanNode, SpanNode> nodes; - private PeekableListIterator<Annotation> base; + + private final IdentityHashMap<SpanNode, SpanNode> nodes; + private final PeekableListIterator<Annotation> base; private boolean nextCalled = false; AnnotationIterator(ListIterator<Annotation> baseIt, IdentityHashMap<SpanNode, SpanNode> nodes) { @@ -91,4 +92,5 @@ public class ListAnnotationContainer extends IteratingAnnotationContainer { nextCalled = false; } } + } diff --git a/document/src/main/java/com/yahoo/document/annotation/SpanList.java b/document/src/main/java/com/yahoo/document/annotation/SpanList.java index 9530689c3f1..04bdf5c892e 100644 --- a/document/src/main/java/com/yahoo/document/annotation/SpanList.java +++ b/document/src/main/java/com/yahoo/document/annotation/SpanList.java @@ -11,13 +11,14 @@ import java.util.ListIterator; /** * A node in a Span tree that can have child nodes. * - * @author <a href="mailto:einarmr@yahoo-inc.com">Einar M R Rosenvinge</a> + * @author Einar M R Rosenvinge */ public class SpanList extends SpanNode { + public static final byte ID = 2; private final List<SpanNode> children; - private int cachedFrom = Integer.MIN_VALUE; //triggers calculateFrom() - private int cachedTo = Integer.MIN_VALUE; //triggers calculateTo() + private int cachedFrom = Integer.MIN_VALUE; // triggers calculateFrom() + private int cachedTo = Integer.MIN_VALUE; // triggers calculateTo() /** Creates a new SpanList. */ public SpanList() { @@ -39,7 +40,7 @@ public class SpanList extends SpanNode { * @param other the SpanList to copy. */ public SpanList(SpanList other) { - this.children = new LinkedList<SpanNode>(); + this.children = new LinkedList<>(); for (SpanNode otherNode : other.children) { if (otherNode instanceof Span) { children.add(new Span((Span) otherNode)); @@ -86,15 +87,15 @@ public class SpanList extends SpanNode { /** Create a span, add it to this list and return it */ public Span span(int from, int length) { - Span span=new Span(from,length); + Span span = new Span(from, length); add(span); return span; } void setInvalid() { - //invalidate ourselves: + // invalidate ourselves: super.setInvalid(); - //invalidate all our children: + // invalidate all our children: for (SpanNode node : children()) { node.setInvalid(); } @@ -213,20 +214,12 @@ public class SpanList extends SpanNode { return this; } - /** - * Returns a <strong>modifiable</strong> list of the immediate children of this SpanList. - * - * @return a <strong>modifiable</strong> list of the immediate children of this SpanList. - */ + /** Returns a modifiable list of the immediate children of this SpanList. */ protected List<SpanNode> children() { return children; } - /** - * Returns the number of children this SpanList holds. - * - * @return the number of children this SpanList holds. - */ + /** Returns the number of children this SpanList holds. */ public int numChildren() { return children().size(); } @@ -394,11 +387,9 @@ public class SpanList extends SpanNode { @Override public boolean equals(Object o) { if (this == o) return true; - if (!(o instanceof SpanList)) return false; + if (!(o instanceof SpanList spanList)) return false; if (!super.equals(o)) return false; - SpanList spanList = (SpanList) o; - if (children() != null ? !children().equals(spanList.children()) : spanList.children() != null) return false; return true; @@ -415,4 +406,5 @@ public class SpanList extends SpanNode { public String toString() { return "SpanList with " + children().size() + " children"; } + } diff --git a/document/src/main/java/com/yahoo/document/annotation/SpanNode2AnnotationContainer.java b/document/src/main/java/com/yahoo/document/annotation/SpanNode2AnnotationContainer.java index a66639b3bfd..f563fd0d6cd 100644 --- a/document/src/main/java/com/yahoo/document/annotation/SpanNode2AnnotationContainer.java +++ b/document/src/main/java/com/yahoo/document/annotation/SpanNode2AnnotationContainer.java @@ -18,6 +18,7 @@ import java.util.List; * @author Einar M R Rosenvinge */ class SpanNode2AnnotationContainer extends AnnotationContainer { + private final Multimap<SpanNode, Annotation> spanNode2Annotation = Multimaps.newMultimap(new IdentityHashMap<>(), ArrayList::new); @Override diff --git a/document/src/main/java/com/yahoo/document/annotation/SpanTree.java b/document/src/main/java/com/yahoo/document/annotation/SpanTree.java index f785cf3b3ec..2faf17ce428 100644 --- a/document/src/main/java/com/yahoo/document/annotation/SpanTree.java +++ b/document/src/main/java/com/yahoo/document/annotation/SpanTree.java @@ -19,6 +19,7 @@ import java.util.Iterator; import java.util.LinkedList; import java.util.List; import java.util.Map; +import java.util.stream.Collectors; /** * A SpanTree holds a root node of a tree of SpanNodes, and a List of Annotations pointing to these nodes @@ -431,7 +432,7 @@ public class SpanTree implements Iterable<Annotation>, SpanNodeParent, Comparabl } /** - * Adds an Annotation to the internal list of annotations for this SpanTree. Use this when + * Adds an Annotation to the internal list of annotations for this SpanTree. Use this when * adding an Annotation that shall annotate a SpanNode. Upon return, Annotation.getSpanNode() * returns the given node. * @@ -446,7 +447,7 @@ public class SpanTree implements Iterable<Annotation>, SpanNodeParent, Comparabl } /** - * Adds an Annotation to the internal list of annotations for this SpanTree. Use this when + * Adds an Annotation to the internal list of annotations for this SpanTree. Use this when * adding an Annotation that shall annotate a SpanNode. Upon return, Annotation.getSpanNode() * returns the given node. This one is unchecked and assumes that the SpanNode is valid and has * already been attached to the Annotation. @@ -538,10 +539,7 @@ public class SpanTree implements Iterable<Annotation>, SpanNodeParent, Comparabl } } - /** - * Returns an Iterator over all annotations in this tree. Note that the iteration order is non-deterministic. - * @return an Iterator over all annotations in this tree. - */ + /** Returns an Iterator over all annotations in this tree. Note that the iteration order is non-deterministic. */ public Iterator<Annotation> iterator() { return annotations.annotations().iterator(); } @@ -641,7 +639,9 @@ public class SpanTree implements Iterable<Annotation>, SpanNodeParent, Comparabl @Override public String toString() { - return "SpanTree '" + name + "'"; + return "SpanTree '" + name + "' with root: " + root + + ( annotations.annotations().size() > 5 ? "" : + ", annotations: " + annotations.annotations().stream().map(Annotation::toString).collect(Collectors.joining(", "))); } @Override diff --git a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/AnnotatorConfig.java b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/AnnotatorConfig.java index 684bae3bf97..5c1bf0813c4 100644 --- a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/AnnotatorConfig.java +++ b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/linguistics/AnnotatorConfig.java @@ -13,7 +13,7 @@ public class AnnotatorConfig implements Cloneable { private Language language; private StemMode stemMode; private boolean removeAccents; - private int maxTermOccurences; + private int maxTermOccurrences; private int maxTokenizeLength; public static final int DEFAULT_MAX_TERM_OCCURRENCES; @@ -29,7 +29,7 @@ public class AnnotatorConfig implements Cloneable { language = Language.ENGLISH; stemMode = StemMode.NONE; removeAccents = false; - maxTermOccurences = DEFAULT_MAX_TERM_OCCURRENCES; + maxTermOccurrences = DEFAULT_MAX_TERM_OCCURRENCES; maxTokenizeLength = DEFAULT_MAX_TOKENIZE_LENGTH; } @@ -37,7 +37,7 @@ public class AnnotatorConfig implements Cloneable { language = rhs.language; stemMode = rhs.stemMode; removeAccents = rhs.removeAccents; - maxTermOccurences = rhs.maxTermOccurences; + maxTermOccurrences = rhs.maxTermOccurrences; maxTokenizeLength = rhs.maxTokenizeLength; } @@ -74,11 +74,11 @@ public class AnnotatorConfig implements Cloneable { } public int getMaxTermOccurrences() { - return maxTermOccurences; + return maxTermOccurrences; } public AnnotatorConfig setMaxTermOccurrences(int maxTermCount) { - this.maxTermOccurences = maxTermCount; + this.maxTermOccurrences = maxTermCount; return this; } @@ -109,7 +109,7 @@ public class AnnotatorConfig implements Cloneable { if (removeAccents != rhs.removeAccents) { return false; } - if (maxTermOccurences != rhs.maxTermOccurences) { + if (maxTermOccurrences != rhs.maxTermOccurrences) { return false; } if (maxTokenizeLength != rhs.maxTokenizeLength) { @@ -121,6 +121,7 @@ public class AnnotatorConfig implements Cloneable { @Override public int hashCode() { return getClass().hashCode() + language.hashCode() + stemMode.hashCode() + - Boolean.valueOf(removeAccents).hashCode() + maxTermOccurences + maxTokenizeLength; + Boolean.valueOf(removeAccents).hashCode() + maxTermOccurrences + maxTokenizeLength; } + } 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 04019800d59..74afd30d7ef 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 @@ -34,8 +34,8 @@ public class LinguisticsAnnotator { final Map<String, Integer> termOccurrences = new HashMap<>(); final int maxOccurrences; - public TermOccurrences(int maxOccurences) { - this.maxOccurrences = maxOccurences; + public TermOccurrences(int maxOccurrences) { + this.maxOccurrences = maxOccurrences; } boolean termCountBelowLimit(String term) { @@ -126,16 +126,15 @@ public class LinguisticsAnnotator { } if (mode == StemMode.ALL) { Span where = parent.span((int)token.getOffset(), token.getOrig().length()); - addAnnotation(where, token.getOrig(), token.getOrig(), termOccurrences); String lowercasedOrig = toLowerCase(token.getOrig()); String termOrIfNullOrig = lowercasedOrig; String term = token.getTokenString(); if (term != null) { - termOrIfNullOrig = term; - } - if (! lowercasedOrig.equals(termOrIfNullOrig)) { addAnnotation(where, term, token.getOrig(), termOccurrences); + termOrIfNullOrig = term; + if ( ! term.equals(lowercasedOrig)) + addAnnotation(where, token.getOrig(), token.getOrig(), termOccurrences); } for (int i = 0; i < token.getNumStems(); i++) { String stem = token.getStem(i); 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 8baa4851f5d..c52b877ba3b 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 @@ -19,7 +19,6 @@ 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; @@ -29,8 +28,6 @@ import static org.junit.Assert.assertTrue; */ public class LinguisticsAnnotatorTestCase { - private static final AnnotatorConfig CONFIG = new AnnotatorConfig(); - @Test public void requireThatAnnotateFailsWithZeroTokens() { assertAnnotations(null, "foo"); @@ -42,7 +39,7 @@ public class LinguisticsAnnotatorTestCase { if (type.isIndexable()) { continue; } - assertAnnotations(null, "foo", newToken("foo", "bar", type)); + assertAnnotations(null, "foo", token("foo", "bar", type)); } } @@ -54,7 +51,27 @@ public class LinguisticsAnnotatorTestCase { if (!type.isIndexable()) { continue; } - assertAnnotations(expected, "foo", newToken("foo", "bar", type)); + assertAnnotations(expected, "foo", token("foo", "bar", type)); + } + } + + @Test + public void requireThatIndexableTokenStringsAreAnnotatedWithModeALL() { + SpanTree expected = new SpanTree(SpanTrees.LINGUISTICS); + expected.spanList().span(0, 5).annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("tesla"))); + var span2 = expected.spanList().span(0, 4); + span2.annotate(new Annotation(AnnotationTypes.TERM)); + span2.annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("car"))); + var span3 = expected.spanList().span(0, 8); + span3.annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("modelxes"))); + span3.annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("modelx"))); + span3.annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("mex"))); + for (TokenType type : TokenType.values()) { + if (!type.isIndexable()) continue; + assertAnnotations(expected, "Tesla cars", new AnnotatorConfig().setStemMode("ALL"), + token("Tesla", "tesla", type), + token("cars", "car", type), + SimpleToken.fromStems("ModelXes", List.of("modelxes", "modelx", "mex"))); } } @@ -63,7 +80,7 @@ public class LinguisticsAnnotatorTestCase { SpanTree expected = new SpanTree(SpanTrees.LINGUISTICS); expected.spanList().span(0, 3).annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("bar"))); for (TokenType type : TokenType.values()) { - assertAnnotations(expected, "foo", newToken("foo", "bar", type, true)); + assertAnnotations(expected, "foo", token("foo", "bar", type, true)); } } @@ -76,7 +93,7 @@ public class LinguisticsAnnotatorTestCase { if (!specialToken && !type.isIndexable()) { continue; } - assertAnnotations(expected, "foo", newToken("foo", "foo", type, specialToken)); + assertAnnotations(expected, "foo", token("foo", "foo", type, specialToken)); } } } @@ -90,7 +107,7 @@ public class LinguisticsAnnotatorTestCase { if (!specialToken && !type.isIndexable()) { continue; } - assertAnnotations(expected, "foo", newToken("foo", "BaR", type, specialToken)); + assertAnnotations(expected, "foo", token("foo", "BaR", type, specialToken)); } } } @@ -102,11 +119,11 @@ public class LinguisticsAnnotatorTestCase { expected.spanList().span(3, 3).annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("bar"))); expected.spanList().span(6, 3).annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("baz"))); - SimpleToken token = newToken("FOOBARBAZ", "foobarbaz", TokenType.ALPHABETIC) - .addComponent(newToken("FOO", "foo", TokenType.ALPHABETIC).setOffset(0)) - .addComponent(newToken("BARBAZ", "barbaz", TokenType.ALPHABETIC).setOffset(3) - .addComponent(newToken("BAR", "bar", TokenType.ALPHABETIC).setOffset(3)) - .addComponent(newToken("BAZ", "baz", TokenType.ALPHABETIC).setOffset(6))); + SimpleToken token = token("FOOBARBAZ", "foobarbaz", TokenType.ALPHABETIC) + .addComponent(token("FOO", "foo", TokenType.ALPHABETIC).setOffset(0)) + .addComponent(token("BARBAZ", "barbaz", TokenType.ALPHABETIC).setOffset(3) + .addComponent(token("BAR", "bar", TokenType.ALPHABETIC).setOffset(3)) + .addComponent(token("BAZ", "baz", TokenType.ALPHABETIC).setOffset(6))); assertAnnotations(expected, "foobarbaz", token); } @@ -116,11 +133,11 @@ public class LinguisticsAnnotatorTestCase { expected.spanList().span(0, 9).annotate(new Annotation(AnnotationTypes.TERM, new StringFieldValue("foobarbaz"))); - SimpleToken token = newToken("FOOBARBAZ", "foobarbaz", TokenType.ALPHABETIC).setSpecialToken(true) - .addComponent(newToken("FOO", "foo", TokenType.ALPHABETIC).setOffset(0)) - .addComponent(newToken("BARBAZ", "barbaz", TokenType.ALPHABETIC).setOffset(3) - .addComponent(newToken("BAR", "bar", TokenType.ALPHABETIC).setOffset(3)) - .addComponent(newToken("BAZ", "baz", TokenType.ALPHABETIC).setOffset(6))); + SimpleToken token = token("FOOBARBAZ", "foobarbaz", TokenType.ALPHABETIC).setSpecialToken(true) + .addComponent(token("FOO", "foo", TokenType.ALPHABETIC).setOffset(0)) + .addComponent(token("BARBAZ", "barbaz", TokenType.ALPHABETIC).setOffset(3) + .addComponent(token("BAR", "bar", TokenType.ALPHABETIC).setOffset(3)) + .addComponent(token("BAZ", "baz", TokenType.ALPHABETIC).setOffset(6))); assertAnnotations(expected, "foobarbaz", token); } @@ -140,7 +157,8 @@ public class LinguisticsAnnotatorTestCase { continue; } assertAnnotations(expected, "foo", - newLinguistics(List.of(newToken("foo", "foo", type, specialToken)), + new AnnotatorConfig(), + newLinguistics(List.of(token("foo", "foo", type, specialToken)), Collections.singletonMap("foo", "bar"))); } } @@ -154,11 +172,9 @@ public class LinguisticsAnnotatorTestCase { StringFieldValue val = new StringFieldValue("foo"); val.setSpanTree(spanTree); - Linguistics linguistics = newLinguistics(List.of(newToken("foo", "bar", TokenType.ALPHABETIC, false)), + Linguistics linguistics = newLinguistics(List.of(token("foo", "bar", TokenType.ALPHABETIC, false)), Collections.<String, String>emptyMap()); - new LinguisticsAnnotator(linguistics, CONFIG).annotate(val); - - assertTrue(new LinguisticsAnnotator(linguistics, CONFIG).annotate(val)); + assertTrue(new LinguisticsAnnotator(linguistics, new AnnotatorConfig()).annotate(val)); assertEquals(spanTree, val.getSpanTree(SpanTrees.LINGUISTICS)); } @@ -186,7 +202,7 @@ public class LinguisticsAnnotatorTestCase { } @Test - public void requireThatMaxTermOccurencesIsHonored() { + public void requireThatMaxTermOccurrencesIsHonored() { final String inputTerm = "foo"; final String stemmedInputTerm = "bar"; // completely different from // inputTerm for safer test @@ -204,7 +220,7 @@ public class LinguisticsAnnotatorTestCase { StringBuilder input = new StringBuilder(); Token[] tokens = new Token[inputTermOccurence]; for (int i = 0; i < inputTermOccurence; ++i) { - SimpleToken t = newToken(inputTerm, stemmedInputTerm, type); + SimpleToken t = token(inputTerm, stemmedInputTerm, type); t.setOffset(i * paddedInputTerm.length()); tokens[i] = t; input.append(paddedInputTerm); @@ -214,28 +230,29 @@ public class LinguisticsAnnotatorTestCase { } // -------------------------------------------------------------------------------- - // // Utilities - // - // -------------------------------------------------------------------------------- - private static SimpleToken newToken(String orig, String stem, TokenType type) { - return newToken(orig, stem, type, false); + private static SimpleToken token(String orig, String stem, TokenType type) { + return token(orig, stem, type, false); } - private static SimpleToken newToken(String orig, String stem, TokenType type, boolean specialToken) { + private static SimpleToken token(String orig, String stem, TokenType type, boolean specialToken) { return new SimpleToken(orig).setTokenString(stem) .setType(type) .setSpecialToken(specialToken); } private static void assertAnnotations(SpanTree expected, String value, Token... tokens) { - assertAnnotations(expected, value, newLinguistics(Arrays.asList(tokens), Collections.<String, String>emptyMap())); + assertAnnotations(expected, value, new AnnotatorConfig(), newLinguistics(Arrays.asList(tokens), Collections.emptyMap())); + } + + private static void assertAnnotations(SpanTree expected, String value, AnnotatorConfig config, Token... tokens) { + assertAnnotations(expected, value, config, newLinguistics(Arrays.asList(tokens), Collections.emptyMap())); } - private static void assertAnnotations(SpanTree expected, String str, Linguistics linguistics) { + private static void assertAnnotations(SpanTree expected, String str, AnnotatorConfig config, Linguistics linguistics) { StringFieldValue val = new StringFieldValue(str); - assertEquals(expected != null, new LinguisticsAnnotator(linguistics, CONFIG).annotate(val)); + assertEquals(expected != null, new LinguisticsAnnotator(linguistics, config).annotate(val)); assertEquals(expected, val.getSpanTree(SpanTrees.LINGUISTICS)); } @@ -255,6 +272,7 @@ public class LinguisticsAnnotatorTestCase { private Token replace(Token token, Map<String, String> replacementTerms) { var simpleToken = (SimpleToken)token; + System.out.println("Token: " + token + ", getTokenString: " + token.getTokenString()); simpleToken.setTokenString(replacementTerms.getOrDefault(token.getTokenString(), token.getTokenString())); return simpleToken; } diff --git a/linguistics/src/main/java/com/yahoo/language/simple/SimpleToken.java b/linguistics/src/main/java/com/yahoo/language/simple/SimpleToken.java index 6cc68c7ac14..809e9b8d133 100644 --- a/linguistics/src/main/java/com/yahoo/language/simple/SimpleToken.java +++ b/linguistics/src/main/java/com/yahoo/language/simple/SimpleToken.java @@ -15,35 +15,48 @@ import java.util.Objects; public class SimpleToken implements Token { private final List<Token> components = new ArrayList<>(); - private final String orig; + private final String original; private TokenType type = TokenType.UNKNOWN; private TokenScript script = TokenScript.UNKNOWN; private String tokenString; + private List<String> stems = null; // Any additional stems after tokenString private boolean specialToken = false; private long offset = 0; - public SimpleToken(String orig) { - this(orig, null); + public SimpleToken(String original) { + this(original, (String)null); } - public SimpleToken(String orig, String tokenString) { - this.orig = orig; + public SimpleToken(String original, String tokenString) { + this.original = original; this.tokenString = tokenString; } + /** Exposed as fromStems */ + private SimpleToken(String original, List<String> stems) { + this.type = TokenType.ALPHABETIC; // Only type which may have stems + this.original = original; + this.tokenString = stems.get(0); + this.stems = List.copyOf(stems.subList(1, stems.size())); + } + @Override public String getOrig() { - return orig; + return original; } @Override public int getNumStems() { - return tokenString != null ? 1 : 0; + return (tokenString != null ? 1 : 0) + (stems != null ? stems.size() : 0); } @Override public String getStem(int i) { - return tokenString; + if (i == 0) + return tokenString; + if (stems != null && i-1 < stems.size()) + return stems.get(i-1); + return tokenString; // TODO Vespa 9: throw new IllegalArgumentException() instead } @Override @@ -131,12 +144,12 @@ public class SimpleToken implements Token { @Override public int hashCode() { - return orig.hashCode(); + return original.hashCode(); } @Override public String toString() { - return "token '" + orig + "'"; + return "token '" + original + "'"; } public String toDetailString() { @@ -171,4 +184,8 @@ public class SimpleToken implements Token { return getType().isIndexable() && (getOrig().length() > 0); } + public static SimpleToken fromStems(String original, List<String> stems) { + return new SimpleToken(original, stems); + } + } diff --git a/linguistics/src/main/java/com/yahoo/language/simple/SimpleTokenizer.java b/linguistics/src/main/java/com/yahoo/language/simple/SimpleTokenizer.java index 98a84a48095..b72d2bd6d37 100644 --- a/linguistics/src/main/java/com/yahoo/language/simple/SimpleTokenizer.java +++ b/linguistics/src/main/java/com/yahoo/language/simple/SimpleTokenizer.java @@ -106,7 +106,7 @@ public class SimpleTokenizer implements Tokenizer { String oldToken = token; token = stemmer.stem(token); String newToken = token; - log.log(Level.FINEST, () -> "stem '" + oldToken+"' to '" + newToken+"'"); + log.log(Level.FINEST, () -> "stem '" + oldToken + "' to '" + newToken + "'"); } String result = token; log.log(Level.FINEST, () -> "processed token is: " + result); |