SOLR-17960 Remove deprecated TikaLanguageIdentifierUpdateProcessor (#3787)

Author: janhoy

solr/CHANGES.txt

@@ -263,6 +263,9 @@ Dependency Upgrades
 Other Changes
 ---------------------
 
+* SOLR-17960: Removed TikaLanguageIdentifierUpdateProcessor. Use LangDetectLanguageIdentifierUpdateProcessor
+  or OpenNLPLangDetectUpdateProcessor instead. (janhoy)
+
 * SOLR-15730: SolrJ modules like SolrJ-Zookeeper are now opt-in from a Maven POM perspective.
   Previously, the modules would come transitively.
   (David Smiley)

solr/modules/langid/README.md

@@ -35,6 +35,5 @@ for more information.
 
 Dependencies
 ------------
-The Tika detector depends on Tika Core (which is part of the extraction module)
 The Langdetect detector depends on LangDetect library
 The OpenNLP detector depends on OpenNLP tools and requires a previously trained user-supplied model

solr/modules/langid/build.gradle

@@ -23,7 +23,6 @@ dependencies {
   implementation project(':solr:core')
   implementation project(':solr:solrj')
 
-  implementation(libs.apache.tika.core) { transitive = false }
   implementation libs.commonsio.commonsio
   implementation libs.cybozulabs.langdetect
   implementation libs.apache.opennlp.tools

solr/modules/langid/gradle.lockfile

@@ -116,7 +116,6 @@ org.apache.lucene:lucene-spatial3d:10.3.1=jarValidation,runtimeClasspath,runtime
 org.apache.lucene:lucene-suggest:10.3.1=jarValidation,runtimeClasspath,runtimeLibs,solrPlatformLibs,testRuntimeClasspath
 org.apache.lucene:lucene-test-framework:10.3.1=jarValidation,testCompileClasspath,testRuntimeClasspath
 org.apache.opennlp:opennlp-tools:2.5.6=compileClasspath,jarValidation,runtimeClasspath,runtimeLibs,testCompileClasspath,testRuntimeClasspath
-org.apache.tika:tika-core:1.28.5=compileClasspath,jarValidation,runtimeClasspath,runtimeLibs,testCompileClasspath,testRuntimeClasspath
 org.apache.zookeeper:zookeeper-jute:3.9.4=compileClasspath,jarValidation,runtimeClasspath,runtimeLibs,solrPlatformLibs,testCompileClasspath,testRuntimeClasspath
 org.apache.zookeeper:zookeeper:3.9.4=compileClasspath,jarValidation,runtimeClasspath,runtimeLibs,solrPlatformLibs,testCompileClasspath,testRuntimeClasspath
 org.apiguardian:apiguardian-api:1.1.2=jarValidation,testRuntimeClasspath

solr/modules/langid/src/java/org/apache/solr/update/processor/TikaLanguageIdentifierUpdateProcessor.java

@@ -56,27 +56,10 @@
 
   <requestHandler name="/update" class="solr.UpdateRequestHandler" >
     <lst name="defaults">
-      <str name="update.chain">lang_id_tika</str>
+      <str name="update.chain">lang_id_lang_detect</str>
     </lst>
   </requestHandler>
 
-  <updateRequestProcessorChain name="lang_id_tika">
-    <processor class="org.apache.solr.update.processor.TikaLanguageIdentifierUpdateProcessorFactory">
-      <!-- Can take defaults, invariants and appends just like req handlers-->
-      <lst name="defaults">
-        <bool name="langid">true</bool>
-        <str name="langid.fl">name,subject</str>
-        <bool name="langid.map">true</bool>
-        <str name="langid.langField">language_s</str>
-        <str name="langid.langsField">language_sm</str>
-        <str name="langid.map.lcmap">th:thai</str>
-        <float name="threshold">0.5</float>
-        <str name="langid.fallback">fallback</str>
-      </lst>
-    </processor>
-    <processor class="solr.RunUpdateProcessorFactory" />
-  </updateRequestProcessorChain>
-
   <updateRequestProcessorChain name="lang_id_lang_detect">
     <processor class="org.apache.solr.update.processor.LangDetectLanguageIdentifierUpdateProcessorFactory">
       <!-- Can take defaults, invariants and appends just like req handlers-->

solr/modules/langid/src/java/org/apache/solr/update/processor/TikaLanguageIdentifierUpdateProcessorFactory.java

@@ -39,9 +39,7 @@ public abstract class LanguageIdentifierUpdateProcessorFactoryTestCase extends S
   public static void beforeClass() throws Exception {
     initCore("solrconfig-languageidentifier.xml", "schema.xml", getFile("langid/solr"));
     SolrCore core = h.getCore();
-    UpdateRequestProcessorChain chained = core.getUpdateProcessingChain("lang_id_tika");
-    assertNotNull(chained);
-    chained = core.getUpdateProcessingChain("lang_id_lang_detect");
+    UpdateRequestProcessorChain chained = core.getUpdateProcessingChain("lang_id_lang_detect");
     assertNotNull(chained);
     chained = core.getUpdateProcessingChain("lang_id_opennlp");
     assertNotNull(chained);