diff --git a/kahuna/app/controllers/KahunaController.scala b/kahuna/app/controllers/KahunaController.scala index 7674168c42a..d77e6313653 100644 --- a/kahuna/app/controllers/KahunaController.scala +++ b/kahuna/app/controllers/KahunaController.scala @@ -3,7 +3,7 @@ package controllers import com.gu.mediaservice.lib.argo.ArgoHelpers import com.gu.mediaservice.lib.auth.Authentication.Principal import com.gu.mediaservice.lib.auth.{Authentication, Authorisation, BaseControllerWithLoginRedirects} -import lib.{EnableAISearch, ExampleSwitch, FeatureSwitches, KahunaConfig, UseCqlChips} +import lib.{EnableAISearch, EnableAISearchManualMode, ExampleSwitch, FeatureSwitches, KahunaConfig, UseCqlChips} import play.api.mvc.ControllerComponents import play.api.libs.json._ @@ -38,7 +38,7 @@ class KahunaController( val isIFramed = request.headers.get("Sec-Fetch-Dest").contains("iframe") val featureSwitches = new FeatureSwitches( - List(ExampleSwitch, UseCqlChips, EnableAISearch) + List(ExampleSwitch, UseCqlChips, EnableAISearch, EnableAISearchManualMode) ) val featureSwitchesWithClientValues = featureSwitches.getClientSwitchValues(featureSwitches.getFeatureSwitchCookies(request.cookies.get)) val featureSwitchesJson = Json.stringify(Json.toJson(featureSwitches.getFeatureSwitchesToStringify(featureSwitchesWithClientValues))) diff --git a/kahuna/app/lib/FeatureSwitch.scala b/kahuna/app/lib/FeatureSwitch.scala index 67e78114b86..a5c52148047 100644 --- a/kahuna/app/lib/FeatureSwitch.scala +++ b/kahuna/app/lib/FeatureSwitch.scala @@ -22,6 +22,12 @@ object EnableAISearch extends FeatureSwitch( default = false ) +object EnableAISearchManualMode extends FeatureSwitch( + key = "enable-ai-search-manual-mode", + title = "Expose the AI search manual/hybrid toggle (off = always hybrid)", + default = false +) + class FeatureSwitches(featureSwitches: List[FeatureSwitch]){ // Feature switches are defined here, but updated by setting a cookie following the pattern e.g. "feature-switch-my-key" // for a switch called "my-key". diff --git a/kahuna/public/js/components/gr-chips/gr-chips.js b/kahuna/public/js/components/gr-chips/gr-chips.js index 5911eee453e..6deb99e4b36 100644 --- a/kahuna/public/js/components/gr-chips/gr-chips.js +++ b/kahuna/public/js/components/gr-chips/gr-chips.js @@ -27,7 +27,6 @@ grChips.controller('grChipsCtrl', ['$scope', function($scope) { $grChipsCtrl.configureNgModel = function(ngModelCtrl, onChangeExpr, autoCompleteExpr, validKeysExpr, autofocus, placeholder) { $grChipsCtrl.defaultPlaceholder = placeholder || 'Search for images... (type + for advanced search)'; - $grChipsCtrl.aiSearchPlaceholder = 'Search for conceptual images using AI search... (no filters available)'; $grChipsCtrl.onChange = () => onChangeExpr($scope, {$chips: $grChipsCtrl.items}); @@ -63,9 +62,7 @@ grChips.controller('grChipsCtrl', ['$scope', function($scope) { }); $grChipsCtrl.getPlaceholder = function() { - return $grChipsCtrl.useAiSearch ? - $grChipsCtrl.aiSearchPlaceholder : - $grChipsCtrl.defaultPlaceholder; + return $grChipsCtrl.defaultPlaceholder; }; $grChipsCtrl.isEmpty = function() { @@ -179,9 +176,6 @@ grChips.directive('grChips', ['$parse', function($parse) { template: template, controller: 'grChipsCtrl', controllerAs: '$grChipsCtrl', - bindToController: { - useAiSearch: ' diff --git a/kahuna/public/js/search/index.js b/kahuna/public/js/search/index.js index 8897e6006ee..722e85cb930 100644 --- a/kahuna/public/js/search/index.js +++ b/kahuna/public/js/search/index.js @@ -179,6 +179,7 @@ search.config(['$stateProvider', '$urlMatcherFactoryProvider', 'until', 'orderBy', 'useAISearch', + 'aiQuery', 'vecWeight', 'dateField', 'takenSince', diff --git a/kahuna/public/js/search/query.html b/kahuna/public/js/search/query.html index 8cfde8b1157..2824af8c7e8 100644 --- a/kahuna/public/js/search/query.html +++ b/kahuna/public/js/search/query.html @@ -1,12 +1,13 @@
- + search + + + auto_awesome + + + + + + + + + + + + + + + lexical + + vector + {{ searchQuery.vecWeight | number:2 }} + + + + -
+
-
+
diff --git a/kahuna/public/js/search/query.js b/kahuna/public/js/search/query.js index db7facc3a08..d47da3c91f5 100644 --- a/kahuna/public/js/search/query.js +++ b/kahuna/public/js/search/query.js @@ -80,13 +80,30 @@ query.controller('SearchQueryCtrl', [ ctrl.initialShowPaidEvent = ($stateParams.nonFree === undefined && ctrl.usePermissionsFilter) ? false : true; ctrl.shouldDisplayAISearchOption = getFeatureSwitchActive("enable-ai-search"); + ctrl.shouldDisplayAIManualModeOption = getFeatureSwitchActive("enable-ai-search-manual-mode"); if (!ctrl.shouldDisplayAISearchOption) { ctrl.useAISearch = false; ctrl.vecWeight = undefined; + ctrl.aiMode = 'hybrid'; } else { ctrl.useAISearch = ($stateParams.useAISearch === 'true' || $stateParams.useAISearch === true) ? true : false; - ctrl.vecWeight = $stateParams.vecWeight; + // vecWeight defaults to 1.0 (fully vector) when in hybrid mode. + const parsedWeight = parseFloat($stateParams.vecWeight); + ctrl.vecWeight = Number.isFinite(parsedWeight) ? parsedWeight : 1.0; + // Manual mode is implied by an existing aiQuery in the URL; otherwise default to hybrid. + // When the manual-mode feature switch is off, force hybrid regardless of URL params. + ctrl.aiMode = (ctrl.shouldDisplayAIManualModeOption && $stateParams.aiQuery && $stateParams.aiQuery.trim()) ? 'manual' : 'hybrid'; } + ctrl.showWeightSlider = false; + + // Bridge for the range input so two-way binding can debounce numerically. + ctrl.vecWeightModel = function (newValue) { + if (arguments.length) { + const n = parseFloat(newValue); + ctrl.vecWeight = Number.isFinite(n) ? n : 1.0; + } + return ctrl.vecWeight; + }; //--react - angular interop events-- function raisePayableImagesEvent(showPaid) { @@ -288,7 +305,7 @@ query.controller('SearchQueryCtrl', [ } ctrl.filter.nonFree = nonFreeCheck; - sendTelemetryForQuery(ctrl.filter.query, nonFreeCheck, uploadedByMe, ctrl.useAISearch); + sendTelemetryForQuery(ctrl.filter.query, nonFreeCheck, uploadedByMe, !!ctrl.filter.aiQuery); if (ctrl.collectionSearch && !curCollectionSearch) { storage.setJs("orderBy", CollectionSortOption.value); ctrl.ordering["orderBy"] = CollectionSortOption.value; @@ -389,8 +406,6 @@ query.controller('SearchQueryCtrl', [ Object.keys($stateParams). // Exclude date-related filters, managed separately in dateFilter filter(key => dateFilterParams.indexOf(key) === -1). - // Exclude useAISearch, managed separately by its own dedicated watcher - filter(key => key !== 'useAISearch'). forEach(setAndWatchParam); // URL parameters are not decoded when taken out of the params. @@ -451,20 +466,46 @@ query.controller('SearchQueryCtrl', [ $scope.$watch(() => ctrl.ordering.orderBy, onValChange(newVal => { $state.go('search.results', {...ctrl.filter, orderBy: newVal}); })); + $scope.$watch(() => ctrl.useAISearch, () => { // Note: $watch expressions execute at least once during initialization, so this is executed on page refresh. // This is the behaviour we want so that the URL is updated based on the AI search toggle if (ctrl.useAISearch) { + // When entering AI mode, include the current mode/weight in the URL but DON'T inject an + // aiQuery unless we're already in manual mode with one set. $state.go('search.results', { ...ctrl.filter, useAISearch: true, - vecWeight: ctrl.vecWeight + vecWeight: ctrl.aiMode === 'hybrid' ? ctrl.vecWeight : null, + aiQuery: ctrl.aiMode === 'manual' ? ctrl.filter.aiQuery : null }); } else { - $state.go('search.results', {...ctrl.filter, useAISearch: null}); + // When leaving AI mode, clear all AI-specific params so we don't carry stale state. + $state.go('search.results', {...ctrl.filter, useAISearch: null, vecWeight: null, aiQuery: null}); } }); + // Toggling between hybrid and manual AI modes adjusts which URL params are live. + $scope.$watch(() => ctrl.aiMode, onValChange((newMode) => { + if (!ctrl.useAISearch) { return; } + if (newMode === 'hybrid') { + // Hybrid: clear the manual aiQuery so the backend takes the hybrid path. + ctrl.filter.aiQuery = ''; + $state.go('search.results', {...ctrl.filter, aiQuery: null, vecWeight: ctrl.vecWeight}); + } else { + // Manual: stop sending vecWeight (it's a hybrid-only knob). + $state.go('search.results', {...ctrl.filter, vecWeight: null}); + } + })); + + // Hybrid mode slider for the vector/lexical weight (only meaningful while useAISearch is on + // and the manual aiQuery dual-box is not in use). + $scope.$watch(() => ctrl.vecWeight, onValChange((newVal) => { + if (ctrl.useAISearch && ctrl.aiMode === 'hybrid') { + $state.go('search.results', {...ctrl.filter, vecWeight: newVal}); + } + })); + $scope.$watchCollection(() => ctrl.dateFilter, onValChange(({field, since, until}) => { // Translate dateFilter to actual state and query params $state.go('search.results', {...ctrl.filter, ...{ @@ -529,7 +570,7 @@ query.controller('SearchQueryCtrl', [ const { nonFree, uploadedByMe } = ctrl.filter; - sendTelemetryForQuery(ctrl.filter.query, nonFree, uploadedByMe, ctrl.useAISearch); + sendTelemetryForQuery(ctrl.filter.query, nonFree, uploadedByMe, !!ctrl.filter.aiQuery); }]); query.directive('searchQuery', [function() { diff --git a/kahuna/public/js/search/results.js b/kahuna/public/js/search/results.js index 1268fd9a9e2..bc44510d2a9 100644 --- a/kahuna/public/js/search/results.js +++ b/kahuna/public/js/search/results.js @@ -204,7 +204,7 @@ results.controller('SearchResultsCtrl', [ ctrl.newImagesCount = 0; ctrl.newImagesLastCheckedMoment = moment(); - ctrl.needsQuery = $stateParams.useAISearch && (!$stateParams.query || !$stateParams.query.trim()); + ctrl.needsQuery = false; // Map to track image->position and help remove duplicates let imagesPositions; @@ -303,7 +303,8 @@ results.controller('SearchResultsCtrl', [ // (i.e. the uploadTime of the newest result in the set) // TODO: avoid this initial search (two API calls to init!) - const isAiSearch = !!$stateParams.useAISearch; + // AI/KNN search activates when the useAISearch flag is set. + const isAiSearch = !!$stateParams.useAISearch && $stateParams.useAISearch !== 'false'; const initialSearchParams = isAiSearch ? {offset: 0, length: $window._clientConfig.aiSearchResultLimit} : {length: 1, orderBy: 'newest'}; @@ -318,7 +319,7 @@ results.controller('SearchResultsCtrl', [ }); ctrl.loadRange = function(start, end) { - if ($stateParams.useAISearch) { + if (!!$stateParams.useAISearch && $stateParams.useAISearch !== 'false') { return; } @@ -424,7 +425,7 @@ results.controller('SearchResultsCtrl', [ function checkForNewImages() { // Polling for new images is meaningless for AI search — results are // ranked by vector similarity, not upload time. - if ($stateParams.useAISearch) { return; } + if (!!$stateParams.useAISearch && $stateParams.useAISearch !== 'false') { return; } $timeout(() => { // Use explicit `until`, or blank it to find new images @@ -578,6 +579,7 @@ results.controller('SearchResultsCtrl', [ length: length, orderBy: orderBy, useAISearch: $stateParams.useAISearch, + aiQuery: $stateParams.aiQuery, vecWeight: $stateParams.vecWeight, hasRightsAcquired: $stateParams.hasRightsAcquired, hasCrops: $stateParams.hasCrops, diff --git a/kahuna/public/js/search/structured-query/structured-query.js b/kahuna/public/js/search/structured-query/structured-query.js index 4fa365f942a..6583d4defc3 100644 --- a/kahuna/public/js/search/structured-query/structured-query.js +++ b/kahuna/public/js/search/structured-query/structured-query.js @@ -73,8 +73,7 @@ grStructuredQuery.directive("grStructuredQuery", [ ng-model="ctrl.structuredQuery" gr:valid-keys="ctrl.filterFields" gr:on-change="ctrl.structuredQueryChanged($chips)" - gr:autocomplete="ctrl.getSuggestions($chip)" - use-ai-search="searchQuery.useAISearch"> + gr:autocomplete="ctrl.getSuggestions($chip)"> ; match the structured-query styling rather than .text-input */ + background-color: transparent; + color: inherit; + border: none; + outline: none; + padding: 0; + height: 100%; + width: 100%; + min-width: 0; + font: inherit; +} + +.search__ai-controls { + display: inline-flex; + align-items: center; + margin-left: 8px; + gap: 8px; + flex-shrink: 0; + color: #ccc; +} + +.search__ai-mode-toggle { + display: inline-flex; + align-items: center; + gap: 6px; +} + +.search__ai-mode-toggle label { + display: inline-flex; + align-items: center; + gap: 3px; + cursor: pointer; +} + +.search__ai-weight { + display: inline-flex; + align-items: center; + gap: 6px; +} + +.search__ai-weight-toggle { + background: none; + border: 1px solid #666; + color: #ccc; + padding: 2px 4px; + cursor: pointer; + border-radius: 3px; + display: inline-flex; + align-items: center; +} + +.search__ai-weight-toggle:hover { + border-color: #999; +} + +.search__ai-weight-slider { + display: inline-flex; + align-items: center; + gap: 4px; + font-size: 11px; +} + +.search__ai-weight-slider input[type="range"] { + width: 100px; +} + +.search__ai-weight-value { + min-width: 2.6em; + text-align: right; + font-variant-numeric: tabular-nums; } .search { diff --git a/media-api/app/controllers/MediaApi.scala b/media-api/app/controllers/MediaApi.scala index 5b32930b4a6..6698b576cd2 100644 --- a/media-api/app/controllers/MediaApi.scala +++ b/media-api/app/controllers/MediaApi.scala @@ -89,6 +89,7 @@ class MediaApi( "countAll", "persisted", "useAISearch", + "aiQuery", "vecWeight" ).mkString(",") @@ -601,7 +602,7 @@ class MediaApi( ) } - def semanticSearchByImage(imageId: String, k: Int): Future[SearchResults] = { + def semanticSearchByImage(imageId: String, k: Int, filter: Option[com.sksamuel.elastic4s.requests.searches.queries.Query]): Future[SearchResults] = { for { maybeImage <- elasticSearch.getImageById(imageId) maybeEmbedding = maybeImage @@ -612,19 +613,36 @@ class MediaApi( searchResults <- maybeEmbedding match { // If we have an embedding, perform the KNN search. If not, return an empty result set. case Some(embedding) => - elasticSearch.knnSearch(embedding, k = k, numCandidates = Math.max(k * 2, 100)) + elasticSearch.knnSearch(embedding, k = k, numCandidates = Math.max(k * 2, 100), filter = filter) case None => Future.successful(SearchResults(Nil, total = 0, extraCounts = None)) } } yield searchResults } - def semanticSearchByText(query: String, k: Int, vecWeight: Option[Double]): Future[SearchResults] = { - // Normalise key so that "Dogs" and "dogs " share a cache entry. + // Manual mode: user supplies an explicit `aiQuery` (KNN) and an optional Elasticsearch + // filter built from the regular SearchParams (the structured/metadata query, dates, etc.). + // KNN ranks; filter constrains the candidate set. + def semanticSearchByTextManual(query: String, k: Int, filter: Option[com.sksamuel.elastic4s.requests.searches.queries.Query]): Future[SearchResults] = { val cacheKey = query.trim.toLowerCase - val markers = logMarker ++ Map("query" -> query) + if (embeddingCache.getIfPresent(cacheKey).isDefined) { + logger.info(markers, s"AI search embedding cache hit query=$query") + } else { + logger.info(markers, s"AI search embedding cache miss query=$query") + } + val embeddingFuture = embeddingCache.get(cacheKey) + for { + embedding <- embeddingFuture + searchResults <- elasticSearch.knnSearch(embedding, k = k, numCandidates = Math.max(k * 2, 100), filter = filter) + } yield searchResults + } + // Hybrid mode (default): a single text query, combined KNN + BM25 ranking with + // a configurable vector weight (defaults to fully-vector when omitted). + def semanticSearchByTextHybrid(query: String, k: Int, vecWeight: Option[Double]): Future[SearchResults] = { + val cacheKey = query.trim.toLowerCase + val markers = logMarker ++ Map("query" -> query) if (embeddingCache.getIfPresent(cacheKey).isDefined) { logger.info(markers, s"AI search embedding cache hit query=$query") } else { @@ -632,9 +650,6 @@ class MediaApi( } val weight = vecWeight.getOrElse(1.0) - - // cache.get(key) is atomic: if two requests race on the same key, only one - // load fires and both callers receive the same Future. val embeddingFuture = embeddingCache.get(cacheKey) logger.info(markers, s"vecWeight for query '$query' is $weight") @@ -650,28 +665,52 @@ class MediaApi( } yield searchResults } - def performAiSearchAndRespond(query: String, vecWeight: Option[Double]): Future[Result] = { + def performManualAiSearchAndRespond(aiQuery: String, filter: Option[com.sksamuel.elastic4s.requests.searches.queries.Query]): Future[Result] = { val k = config.aiSearchResultLimit - val searchResultsFuture = parseAiSearchMode(query) match { - case SimilarSearch(imageId) => semanticSearchByImage(imageId, k) - case TextSearch(textQuery) => semanticSearchByText(textQuery, k, vecWeight) + val searchResultsFuture = parseAiSearchMode(aiQuery) match { + case SimilarSearch(imageId) => semanticSearchByImage(imageId, k, filter) + case TextSearch(textQuery) => semanticSearchByTextManual(textQuery, k, filter) } + searchResultsFuture.map(aiSearchResponseFromResults) + } + def performHybridAiSearchAndRespond(query: String, vecWeight: Option[Double]): Future[Result] = { + val k = config.aiSearchResultLimit + val searchResultsFuture = parseAiSearchMode(query) match { + case SimilarSearch(imageId) => semanticSearchByImage(imageId, k, filter = None) + case TextSearch(textQuery) => semanticSearchByTextHybrid(textQuery, k, vecWeight) + } searchResultsFuture.map(aiSearchResponseFromResults) } + // AI search is gated behind the `useAISearch` flag. Inside AI mode we route between + // two ranking strategies: + // - manual dual-box: when `aiQuery` is set, treat the structured `query` as a filter + // and rank purely by KNN against the user's separate AI-text input. + // - hybrid (default): single text query combined KNN+BM25, with optional `vecWeight`. if (_searchParams.useAISearch.contains(true)) { - _searchParams.query match { - // Short-circuit polling requests (length=0) and empty queries to avoid unnecessary Bedrock/KNN calls - case _ if _searchParams.length == 0 => - emptyAiSearchResponse - case Some(q) if !q.isBlank => - performAiSearchAndRespond(q, _searchParams.vecWeight) - // Empty queries do not make sense for AI search as we can - // only rank results once we have a meaningful vector to compare with. - // So return 0 results if the query was empty. - case _ => - emptyAiSearchResponse + if (_searchParams.length == 0) { + emptyAiSearchResponse + } else { + val effectiveParams = if (canViewDeletedImages) { + _searchParams.copy(uploadedBy = Some(Authentication.getIdentity(request.user))) + } else { + _searchParams + } + _searchParams.aiQuery match { + case Some(aiQuery) if !aiQuery.isBlank => + val filter = Some(elasticSearch.buildSearchQuery(effectiveParams)) + performManualAiSearchAndRespond(aiQuery, filter) + case _ => + _searchParams.query match { + case Some(q) if !q.isBlank => + performHybridAiSearchAndRespond(q, _searchParams.vecWeight) + // Empty queries do not make sense for AI search as we can + // only rank results once we have a meaningful vector to compare with. + case _ => + emptyAiSearchResponse + } + } } } else { val searchParams = if (canViewDeletedImages) { diff --git a/media-api/app/lib/elasticsearch/ElasticSearch.scala b/media-api/app/lib/elasticsearch/ElasticSearch.scala index 175fe3d63c6..525fd1e05ce 100644 --- a/media-api/app/lib/elasticsearch/ElasticSearch.scala +++ b/media-api/app/lib/elasticsearch/ElasticSearch.scala @@ -178,17 +178,22 @@ class ElasticSearch( } } - def knnSearch(queryEmbedding: List[Float], k: Int, numCandidates: Int) + def knnSearch(queryEmbedding: List[Float], k: Int, numCandidates: Int, filter: Option[Query] = None) (implicit ex: ExecutionContext, logMarker: LogMarker): Future[SearchResults] = { - val knn = Knn("embedding.cohereEmbedV4.image") + val baseKnn = Knn("embedding.cohereEmbedV4.image") .queryVector(queryEmbedding.map(_.toDouble)) .k(k) .numCandidates(numCandidates) + val knn = filter.fold(baseKnn)(baseKnn.filter) + val searchRequest = ElasticDsl.search(imagesCurrentAlias) .knn(knn) .size(k) + // Print the exact ES request body to stdout so it's visible in the sbt console. + println(s"[ES knn search] POST ${imagesCurrentAlias}/_search\n${prettyEsRequest(searchRequest)}") + executeAndLog(withSearchQueryTimeout(searchRequest), "knn search").map { r => val imageHits = r.result.hits.hits.map(resolveHit).toSeq.flatten.map(i => (i.instance.id, i)) SearchResults(hits = imageHits, total = imageHits.length, extraCounts = None) @@ -278,7 +283,13 @@ class ElasticSearch( } } - def search(params: SearchParams)(implicit ex: ExecutionContext, request: AuthenticatedRequest[AnyContent, Principal], logMarker: LogMarker = MarkerMap()): Future[SearchResults] = { + /** + * Build the combined Elasticsearch query (structured/BM25 query + metadata filters) + * for the given SearchParams. Used by `search` for full ranking, and exposed so that + * `knnSearch` callers can pass it as a KNN filter (e.g. for hybrid "filter by metadata, + * rank by vector" searches). + */ + def buildSearchQuery(params: SearchParams): Query = { val query: Query = queryBuilder.makeQuery(params.structuredQuery) val uploadTimeFilter = filters.date("uploadTime", params.since, params.until) @@ -360,9 +371,14 @@ class ElasticSearch( ++ printUsageFilter ).toNel.map(filter => filter.list.toList.reduceLeft(filters.and(_, _))).toOption - val withFilter = filterOpt.map { f => + filterOpt.map { f => boolQuery() must (query) filter f }.getOrElse(query) + } + + def search(params: SearchParams)(implicit ex: ExecutionContext, request: AuthenticatedRequest[AnyContent, Principal], logMarker: LogMarker = MarkerMap()): Future[SearchResults] = { + val withFilter: Query = buildSearchQuery(params) + val sort = params.orderBy match { case Some("dateAddedToCollection") => sorts.dateAddedToCollectionDescending @@ -407,6 +423,9 @@ class ElasticSearch( .size(params.length) .sortBy(sort) + // Print the exact ES request body to stdout so it's visible in the sbt console. + println(s"[ES image search] POST ${imagesCurrentAlias}/_search\n${prettyEsRequest(searchRequest)}") + executeAndLog(searchRequest, "image search"). toMetric(Some(mediaApiMetrics.searchQueries), List(mediaApiMetrics.searchTypeDimension("results")))(_.result.took).map { r => logSearchQueryIfTimedOut(searchRequest, r.result) @@ -558,4 +577,19 @@ class ElasticSearch( private def logSearchQueryIfTimedOut(req: SearchRequest, res: SearchResponse) = if (res.isTimedOut) logger.info(s"SearchQuery was TimedOut after $SearchQueryTimeout \nquery: ${req.show}") + /** Pretty-print an ES request body, eliding any `query_vector` arrays so the output stays readable. + * `req.show` produces something like `POST /index/_search\n{...body...}`; we keep the request line + * but re-indent the JSON body underneath it. + */ + private def prettyEsRequest(req: SearchRequest): String = { + val raw = req.show + val firstBrace = raw.indexOf('{') + val (prefix, body) = + if (firstBrace >= 0) (raw.substring(0, firstBrace), raw.substring(firstBrace)) + else ("", raw) + val elided = body.replaceAll("\"query_vector\"\\s*:\\s*\\[[^\\]]*\\]", "\"query_vector\":\"\"") + val prettyBody = scala.util.Try(Json.prettyPrint(Json.parse(elided))).getOrElse(elided) + prefix + prettyBody + } + } diff --git a/media-api/app/lib/elasticsearch/ElasticSearchModel.scala b/media-api/app/lib/elasticsearch/ElasticSearchModel.scala index 4bac2fb3f44..443ac608042 100644 --- a/media-api/app/lib/elasticsearch/ElasticSearchModel.scala +++ b/media-api/app/lib/elasticsearch/ElasticSearchModel.scala @@ -86,6 +86,7 @@ case class SearchParams( printUsageFilters: Option[PrintUsageFilters] = None, shouldFlagGraphicImages: Boolean = false, useAISearch: Option[Boolean] = None, + aiQuery: Option[String] = None, vecWeight: Option[Double] = None ) @@ -178,6 +179,7 @@ object SearchParams { printUsageFilters, shouldFlagGraphicImages = false, request.getQueryString("useAISearch") flatMap parseBooleanFromQuery, + request.getQueryString("aiQuery").filter(_.trim.nonEmpty), request.getQueryString("vecWeight") flatMap parseBoundedDoubleFromQuery, ) } @@ -207,7 +209,10 @@ object SearchParams { "hasMetadata" -> listToCommas(searchParams.hasMetadata), "persisted" -> searchParams.persisted.map(_.toString), "usageStatus" -> listToCommas(searchParams.usageStatus.map(_.toString)), - "usagePlatform" -> listToCommas(searchParams.usagePlatform) + "usagePlatform" -> listToCommas(searchParams.usagePlatform), + "useAISearch" -> searchParams.useAISearch.map(_.toString), + "aiQuery" -> searchParams.aiQuery, + "vecWeight" -> searchParams.vecWeight.map(_.toString) ).foldLeft(Map[String, String]()) { case (acc, (key, Some(value))) => acc + (key -> value) case (acc, (_, None)) => acc