Fix: raise error on malformed CSV.

Signed-off-by: dblock <[email protected]>
opensearch-project · Nov 8, 2022 · 578b7c3 · 578b7c3
1 parent 88757b4
commit 578b7c3
Show file tree

Hide file tree

Showing 2 changed files with 12 additions and 0 deletions.
diff --git a/...alysis-kuromoji/src/main/java/org/opensearch/index/analysis/KuromojiTokenizerFactory.java b/...alysis-kuromoji/src/main/java/org/opensearch/index/analysis/KuromojiTokenizerFactory.java
@@ -79,6 +79,9 @@ public KuromojiTokenizerFactory(IndexSettings indexSettings, Environment env, St
 
  private static String parse(String rule, Set<String> dup) {
  String[] values = CSVUtil.parse(rule);
+ if (values.length == 0) {
+ throw new IllegalArgumentException("Malformed csv in user dictionary.");
+ }
  if (dup.add(values[0]) == false) {
  throw new IllegalArgumentException("Found duplicate term [" + values[0] + "] in user dictionary.");
  }

diff --git a/.../analysis-kuromoji/src/test/java/org/opensearch/index/analysis/KuromojiAnalysisTests.java b/.../analysis-kuromoji/src/test/java/org/opensearch/index/analysis/KuromojiAnalysisTests.java
@@ -379,6 +379,15 @@ public void testKuromojiAnalyzerInvalidUserDictOption() throws Exception {
  );
  }
 
+ public void testKuromojiAnalyzerEmptyDictRule() throws Exception {
+ Settings settings = Settings.builder()
+ .put("index.analysis.analyzer.my_analyzer.type", "kuromoji")
+ .putList("index.analysis.analyzer.my_analyzer.user_dictionary_rules", "\"")
+ .build();
+ RuntimeException exc = expectThrows(RuntimeException.class, () -> createTestAnalysis(settings));
+ assertThat(exc.getMessage(), equalTo("Line [1]: Malformed csv in user dictionary."));
+ }
+
  public void testKuromojiAnalyzerDuplicateUserDictRule() throws Exception {
  Settings settings = Settings.builder()
  .put("index.analysis.analyzer.my_analyzer.type", "kuromoji")