Add possible suggestion to the empty results response

src/main/java/io/quarkus/search/app/SearchService.java

@@ -1,7 +1,9 @@
 package io.quarkus.search.app;
 
+import java.lang.reflect.Field;
 import java.util.List;
 
+import com.google.gson.JsonArray;
 import jakarta.enterprise.context.ApplicationScoped;
 import jakarta.enterprise.event.Observes;
 import jakarta.inject.Inject;
@@ -22,14 +24,20 @@
 
 import io.quarkus.runtime.LaunchMode;
 
+import org.hibernate.search.backend.elasticsearch.ElasticsearchExtension;
+import org.hibernate.search.backend.elasticsearch.search.query.ElasticsearchSearchResult;
+import org.hibernate.search.backend.elasticsearch.search.query.impl.ElasticsearchSearchQueryImpl;
 import org.hibernate.search.engine.search.common.BooleanOperator;
 import org.hibernate.search.engine.search.common.ValueModel;
 import org.hibernate.search.engine.search.predicate.dsl.SimpleQueryFlag;
+import org.hibernate.search.engine.search.query.SearchQuery;
 import org.hibernate.search.mapper.pojo.standalone.mapping.SearchMapping;
 
 import org.eclipse.microprofile.openapi.annotations.Operation;
 import org.jboss.resteasy.reactive.RestQuery;
 
+import com.google.gson.JsonObject;
+
 import io.vertx.ext.web.Router;
 
 @ApplicationScoped
@@ -68,7 +76,7 @@ public SearchResult<GuideSearchHit> search(@RestQuery @DefaultValue(QuarkusVersi
             @RestQuery @DefaultValue("1") @Min(0) @Max(value = 10, message = MAX_FOR_PERF_MESSAGE) int contentSnippets,
             @RestQuery @DefaultValue("100") @Min(0) @Max(value = 200, message = MAX_FOR_PERF_MESSAGE) int contentSnippetsLength) {
         try (var session = searchMapping.createSession()) {
-            var result = session.search(Guide.class)
+            SearchQuery<GuideSearchHit> query = session.search(Guide.class)
                     .select(f -> f.composite().from(
                             f.id(),
                             f.field("type"),
@@ -111,28 +119,74 @@ public SearchResult<GuideSearchHit> search(@RestQuery @DefaultValue(QuarkusVersi
                     .highlighter(f -> f.fastVector()
                             // Highlighters are going to use spans-with-classes so that we will have more control over styling the visual on the search results screen.
                             .tag("<span class=\"" + highlightCssClass + "\">", "</span>"))
-                    .highlighter("highlighter_title_or_summary", f -> f.fastVector()
-                            // We want the whole text of the field, regardless of whether it has a match or not.
-                            .noMatchSize(TITLE_OR_SUMMARY_MAX_SIZE)
-                            .fragmentSize(TITLE_OR_SUMMARY_MAX_SIZE)
-                            // We want the whole text as a single fragment
-                            .numberOfFragments(1))
-                    .highlighter("highlighter_content", f -> f.fastVector()
-                            // If there's no match in the full content we don't want to return anything.
-                            .noMatchSize(0)
-                            // Content is really huge, so we want to only get small parts of the sentences.
-                            // We give control to the caller on the content snippet length and the number of these fragments
-                            .numberOfFragments(contentSnippets)
-                            .fragmentSize(contentSnippetsLength)
-                            // The rest of fragment configuration is static
-                            .orderByScore(true)
-                            // We don't use sentence boundaries because those can result in huge fragments
-                            .boundaryScanner().chars().boundaryMaxScan(10).end())
+                    .highlighter(
+                            "highlighter_title_or_summary", f -> f.fastVector()
+                                    // We want the whole text of the field, regardless of whether it has a match or not.
+                                    .noMatchSize(TITLE_OR_SUMMARY_MAX_SIZE)
+                                    .fragmentSize(TITLE_OR_SUMMARY_MAX_SIZE)
+                                    // We want the whole text as a single fragment
+                                    .numberOfFragments(1))
+                    .highlighter(
+                            "highlighter_content", f -> f.fastVector()
+                                    // If there's no match in the full content we don't want to return anything.
+                                    .noMatchSize(0)
+                                    // Content is really huge, so we want to only get small parts of the sentences.
+                                    // We give control to the caller on the content snippet length and the number of these fragments
+                                    .numberOfFragments(contentSnippets)
+                                    .fragmentSize(contentSnippetsLength)
+                                    // The rest of fragment configuration is static
+                                    .orderByScore(true)
+                                    // We don't use sentence boundaries because those can result in huge fragments
+                                    .boundaryScanner().chars().boundaryMaxScan(10).end())
                     .sort(f -> f.score().then().field(language.addSuffix("title_sort")))
                     .routing(QuarkusVersionAndLanguageRoutingBinder.searchKeys(version, language))
                     .totalHitCountThreshold(TOTAL_HIT_COUNT_THRESHOLD + (page + 1) * PAGE_SIZE)
+                    .toQuery();
+
+            hackQuery(query, q, language, highlightCssClass);
+
+            var result = query
+                    .extension(ElasticsearchExtension.get())
+
                     .fetch(page * PAGE_SIZE, PAGE_SIZE);
-            return new SearchResult<>(result);
+
+            return new SearchResult<>(result, extractSuggestion(result));
+        }
+    }
+
+    private String extractSuggestion(ElasticsearchSearchResult<GuideSearchHit> result) {
+        JsonArray options = result.responseBody().getAsJsonObject("suggest")
+                .getAsJsonArray("didYouMean")
+                .get(0).getAsJsonObject()
+                .getAsJsonArray("options");
+        if (options != null && !options.isEmpty()) {
+            return options.get(0).getAsJsonObject().get("highlighted").getAsString();
+        } else {
+            return null;
+        }
+    }
+
+    private void hackQuery(SearchQuery<GuideSearchHit> query, String q, Language language, String highlightCssClass) {
+        try {
+            Field payloadField = ElasticsearchSearchQueryImpl.class.getDeclaredField("payload");
+            payloadField.setAccessible(true);
+            JsonObject payload = (JsonObject) payloadField.get(query);
+            JsonObject suggest = new JsonObject();
+            payload.add("suggest", suggest);
+            suggest.addProperty("text", q);
+            JsonObject suggestion = new JsonObject();
+            suggest.add("didYouMean", suggestion);
+            JsonObject phrase = new JsonObject();
+            suggestion.add("phrase", phrase);
+            phrase.addProperty("field", language.addSuffix("fullContent_suggestion"));
+            phrase.addProperty("size", 1);
+            phrase.addProperty("gram_size", 1);
+            JsonObject highlight = new JsonObject();
+            phrase.add("highlight", highlight);
+            highlight.addProperty("pre_tag", "<span class=\"" + highlightCssClass + "\">");
+            highlight.addProperty("post_tag", "</span>");
+        } catch (NoSuchFieldException | IllegalAccessException e) {
+            throw new RuntimeException(e);
         }
     }
 

src/main/java/io/quarkus/search/app/dto/SearchResult.java

@@ -2,12 +2,12 @@
 
 import java.util.List;
 
-public record SearchResult<T>(Total total, List<T> hits) {
+public record SearchResult<T>(Total total, List<T> hits, String suggestion) {
 
-    public SearchResult(org.hibernate.search.engine.search.query.SearchResult<T> result) {
+    public SearchResult(org.hibernate.search.engine.search.query.SearchResult<T> result, String suggestion) {
         this(new Total(result.total().isHitCountExact() ? result.total().hitCount() : null,
                 result.total().hitCountLowerBound()),
-                result.hits());
+                result.hits(), suggestion);
     }
 
     public record Total(Long exact, Long lowerBound) {

src/main/java/io/quarkus/search/app/entity/Guide.java

@@ -48,11 +48,13 @@ public class Guide {
 
     @I18nFullTextField(highlightable = Highlightable.FAST_VECTOR, termVector = TermVector.WITH_POSITIONS_OFFSETS, analyzerPrefix = AnalysisConfigurer.DEFAULT, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
     @I18nFullTextField(name = "title_autocomplete", analyzerPrefix = AnalysisConfigurer.AUTOCOMPLETE, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
+    @I18nFullTextField(name = "title_suggestion", analyzerPrefix = AnalysisConfigurer.SUGGESTION, searchAnalyzerPrefix = AnalysisConfigurer.SUGGESTION)
     @I18nKeywordField(name = "title_sort", normalizerPrefix = AnalysisConfigurer.SORT, searchable = Searchable.NO, sortable = Sortable.YES)
     public I18nData<String> title = new I18nData<>();
 
     @I18nFullTextField(highlightable = Highlightable.FAST_VECTOR, termVector = TermVector.WITH_POSITIONS_OFFSETS, analyzerPrefix = AnalysisConfigurer.DEFAULT, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
     @I18nFullTextField(name = "summary_autocomplete", analyzerPrefix = AnalysisConfigurer.AUTOCOMPLETE, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
+    @I18nFullTextField(name = "summary_suggestion", analyzerPrefix = AnalysisConfigurer.SUGGESTION, searchAnalyzerPrefix = AnalysisConfigurer.SUGGESTION)
     public I18nData<String> summary = new I18nData<>();
 
     @I18nFullTextField(analyzerPrefix = AnalysisConfigurer.DEFAULT, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
@@ -61,6 +63,7 @@ public class Guide {
 
     @I18nFullTextField(name = "fullContent", valueBridge = @ValueBridgeRef(type = InputProviderHtmlBodyTextBridge.class), highlightable = Highlightable.FAST_VECTOR, termVector = TermVector.WITH_POSITIONS_OFFSETS, analyzerPrefix = AnalysisConfigurer.DEFAULT, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
     @I18nFullTextField(name = "fullContent_autocomplete", valueBridge = @ValueBridgeRef(type = InputProviderHtmlBodyTextBridge.class), analyzerPrefix = AnalysisConfigurer.AUTOCOMPLETE, searchAnalyzerPrefix = AnalysisConfigurer.DEFAULT_SEARCH)
+    @I18nFullTextField(name = "fullContent_suggestion", valueBridge = @ValueBridgeRef(type = InputProviderHtmlBodyTextBridge.class), analyzerPrefix = AnalysisConfigurer.SUGGESTION, searchAnalyzerPrefix = AnalysisConfigurer.SUGGESTION)
     @IndexingDependency(reindexOnUpdate = ReindexOnUpdate.NO)
     public I18nData<InputProvider> htmlFullContentProvider = new I18nData<>();
 

src/main/java/io/quarkus/search/app/hibernate/AnalysisConfigurer.java

@@ -25,6 +25,7 @@ public class AnalysisConfigurer implements ElasticsearchAnalysisConfigurer {
 
     public static final String DEFAULT = "basic_analyzer";
     public static final String DEFAULT_SEARCH = DEFAULT + "_search";
+    public static final String SUGGESTION = "suggestion";
     public static final String AUTOCOMPLETE = "autocomplete";
     public static final String SORT = "sort";
     // This is simplified by assuming no default package, lowercase package names and capitalized class name,
@@ -36,6 +37,10 @@ public static String defaultAnalyzer(Language language) {
         return language.addSuffix(DEFAULT);
     }
 
+    public static String suggestionAnalyzer(Language language) {
+        return language.addSuffix(SUGGESTION);
+    }
+
     public static String defaultSearchAnalyzer(Language language) {
         return language.addSuffix(DEFAULT_SEARCH);
     }
@@ -101,6 +106,16 @@ void configureEnglishLikeLanguage(ElasticsearchAnalysisConfigurationContext cont
                         "asciifolding")
                 .charFilters("html_strip");
 
+        context.analyzer(suggestionAnalyzer(language)).custom()
+                .tokenizer("standard")
+                .tokenFilters(
+                        // To make all words in lowercase.
+                        "lowercase",
+                        // To convert characters into ascii ones, e.g. à to a or ę to e etc.
+                        "asciifolding",
+                        "shingle")
+                .charFilters("html_strip");
+
         // The analyzer to be applied to the user-input text.
         context.analyzer(defaultSearchAnalyzer(language)).custom()
                 .tokenizer("standard")
@@ -158,6 +173,16 @@ void configureJapanese(ElasticsearchAnalysisConfigurationContext context) {
                         "icu_normalizer",
                         "html_strip");
 
+        context.analyzer(suggestionAnalyzer(language)).custom()
+                .tokenizer("kuromoji_tokenizer")
+                .tokenFilters(
+                        // To make all words in lowercase.
+                        "lowercase",
+                        // To convert characters into ascii ones, e.g. à to a or ę to e etc.
+                        "asciifolding",
+                        "shingle")
+                .charFilters("html_strip");
+
         context.analyzer(defaultSearchAnalyzer(language)).custom()
                 .tokenizer("kuromoji_tokenizer")
                 .tokenFilters(
@@ -211,6 +236,16 @@ void configureChinese(ElasticsearchAnalysisConfigurationContext context) {
                         "asciifolding")
                 .charFilters("html_strip");
 
+        context.analyzer(suggestionAnalyzer(language)).custom()
+                .tokenizer("smartcn_tokenizer")
+                .tokenFilters(
+                        // To make all words in lowercase.
+                        "lowercase",
+                        // To convert characters into ascii ones, e.g. à to a or ę to e etc.
+                        "asciifolding",
+                        "shingle")
+                .charFilters("html_strip");
+
         // The analyzer to be applied to the user-input text.
         context.analyzer(defaultSearchAnalyzer(language)).custom()
                 .tokenizer("smartcn_tokenizer")

src/main/resources/web/app/qs-form.ts

@@ -10,6 +10,7 @@ export const QS_NEXT_PAGE_EVENT = 'qs-next-page';
 export interface QsResult {
   hits: QsHit[];
   hasMoreHits: boolean;
+  suggestion: string;
 }
 
 export interface QsHit {

src/main/resources/web/app/qs-target.ts

@@ -4,6 +4,7 @@ import './qs-guide'
 import {QS_NEXT_PAGE_EVENT, QS_RESULT_EVENT, QS_START_EVENT, QsResult} from "./qs-form";
 import debounce from 'lodash/debounce';
 import icons from "./assets/icons";
+import { unsafeHTML } from 'lit/directives/unsafe-html.js';
 
 
 /**
@@ -92,7 +93,8 @@ export class QsTarget extends LitElement {
       if (this._result.hits.length === 0) {
         return html`
           <div id="qs-target" class="no-hits">
-            <p>Sorry, no ${this.type}s matched your search. Please try again.</p>
+            <p>Sorry, no ${this.type}s matched your search.
+              ${unsafeHTML(this._result.suggestion ? `Did you mean ${this._result.suggestion}` : 'Please try again.')}</p>
           </div>
         `;
       }