blob: e93b56d4828e74750aed0a814101a8e3304f975a [file] [log] [blame]
diff --git a/NOTICE.txt b/NOTICE.txt
index c8396cb01e8..1626d8e1dea 100644
--- a/NOTICE.txt
+++ b/NOTICE.txt
@@ -202,11 +202,11 @@ Nori Korean Morphological Analyzer - Apache Lucene Integration
This software includes a binary and/or source version of data from
- mecab-ko-dic-2.0.3-20170922
+ mecab-ko-dic-2.1.1-20180720
which can be obtained from
- https://bitbucket.org/eunjeon/mecab-ko-dic/downloads/mecab-ko-dic-2.0.3-20170922.tar.gz
+ https://bitbucket.org/eunjeon/mecab-ko-dic/downloads/mecab-ko-dic-2.1.1-20180720.tar.gz
The floating point precision conversion in NumericUtils.Float16Converter is derived from work by
Jeroen van der Zijp, granted for use under the Apache license.
diff --git a/gradle/generation/nori.gradle b/gradle/generation/nori.gradle
index 4fadb11d6f7..fc10f0f561e 100644
--- a/gradle/generation/nori.gradle
+++ b/gradle/generation/nori.gradle
@@ -54,7 +54,7 @@ configure(project(":lucene:analysis:nori")) {
dependsOn deleteDictionaryData
dependsOn sourceSets.main.runtimeClasspath
- def dictionaryName = "mecab-ko-dic-2.0.3-20170922"
+ def dictionaryName = "mecab-ko-dic-2.1.1-20180720"
def dictionarySource = "https://bitbucket.org/eunjeon/mecab-ko-dic/downloads/${dictionaryName}.tar.gz"
def dictionaryFile = file("${buildDir}/generate/${dictionaryName}.tar.gz")
def unpackedDir = file("${buildDir}/generate/${dictionaryName}")
diff --git a/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/util/UnknownDictionaryBuilder.java b/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/util/UnknownDictionaryBuilder.java
index cbf3119719a..634c2a599ee 100644
--- a/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/util/UnknownDictionaryBuilder.java
+++ b/lucene/analysis/nori/src/java/org/apache/lucene/analysis/ko/util/UnknownDictionaryBuilder.java
@@ -28,7 +28,7 @@ import java.util.List;
import org.apache.lucene.analysis.ko.dict.CharacterDefinition;
class UnknownDictionaryBuilder {
- private static final String NGRAM_DICTIONARY_ENTRY = "NGRAM,1798,3559,3677,SY,*,*,*,*,*,*,*";
+ private static final String NGRAM_DICTIONARY_ENTRY = "NGRAM,1801,3559,3677,SY,*,*,*,*,*,*,*";
private String encoding;
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/ConnectionCosts.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/ConnectionCosts.dat
index 7fad91e727d..3b92dcff709 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/ConnectionCosts.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/ConnectionCosts.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$buffer.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$buffer.dat
index d7cc866ca34..5d951670592 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$buffer.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$buffer.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$fst.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$fst.dat
index 91668d24a78..53082e33a37 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$fst.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$fst.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$posDict.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$posDict.dat
index c4db79866c5..851e9d5df56 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$posDict.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$posDict.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$targetMap.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$targetMap.dat
index 4661bf8f68e..85cd99141ce 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$targetMap.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/TokenInfoDictionary$targetMap.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$buffer.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$buffer.dat
index c9a18eef34c..27b2d343151 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$buffer.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$buffer.dat differ
diff --git a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$posDict.dat b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$posDict.dat
index f1ceebd734f..75479935289 100644
Binary files a/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$posDict.dat and b/lucene/analysis/nori/src/resources/org/apache/lucene/analysis/ko/dict/UnknownDictionary$posDict.dat differ