Increase highlighter fragment size for snippets

Author: kderusso

x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/rank/textsimilarity/TextSimilarityRankBuilder.java

@@ -52,15 +52,15 @@ public class TextSimilarityRankBuilder extends RankBuilder {
     public static final String NAME = "text_similarity_reranker";
 
     /**
-     * The default token size limit of the Elastic reranker.
+     * The default token size limit of the Elastic reranker is 512.
      */
     private static final int RERANK_TOKEN_SIZE_LIMIT = 512;
 
     /**
-     * A safe default token size limit for other reranker models.
+     * 4096 is a safe default token size limit for other reranker models.
      * Reranker models with smaller token limits will be truncated.
      */
-    private static final int DEFAULT_TOKEN_SIZE_LIMIT = 4096;
+    private static final int DEFAULT_TOKEN_SIZE_LIMIT = 4_096;
 
     public static final LicensedFeature.Momentary TEXT_SIMILARITY_RERANKER_FEATURE = LicensedFeature.momentary(
         null,

x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/rank/textsimilarity/TextSimilarityRerankingRankFeaturePhaseRankShardContext.java

@@ -34,6 +34,10 @@ public class TextSimilarityRerankingRankFeaturePhaseRankShardContext extends Rer
 
     private final SnippetRankInput snippetRankInput;
 
+    // Rough approximation of token size vs. characters in highlight fragments.
+    // TODO highlighter should be able to set fragment size by token not length
+    private static final int TOKEN_SIZE_LIMIT_MULTIPLIER = 5;
+
     public TextSimilarityRerankingRankFeaturePhaseRankShardContext(String field, @Nullable SnippetRankInput snippetRankInput) {
         super(field);
         this.snippetRankInput = snippetRankInput;
@@ -81,8 +85,9 @@ public void prepareForFetch(SearchContext context) {
                     : DEFAULT_NUM_SNIPPETS;
                 highlightBuilder.numOfFragments(numSnippets);
                 // Rely on the model to determine the fragment size
-                // TODO highlighter should be able to set fragment size by token not length
-                highlightBuilder.fragmentSize(snippetRankInput.tokenSizeLimit());
+                int tokenSizeLimit = snippetRankInput.tokenSizeLimit();
+                int fragmentSize = tokenSizeLimit * TOKEN_SIZE_LIMIT_MULTIPLIER;
+                highlightBuilder.fragmentSize(fragmentSize);
                 SearchHighlightContext searchHighlightContext = highlightBuilder.build(context.getSearchExecutionContext());
                 context.highlight(searchHighlightContext);
             } catch (IOException e) {