Merge pull request #66 from aodn/smart-suggestions

Smart suggestions
aodn · Apr 24, 2024 · 6273eaa · 6273eaa
2 parents 0842ee0 + cdd9dc1
commit 6273eaa
Show file tree

Hide file tree

Showing 9 changed files with 66 additions and 53 deletions.
diff --git a/indexer/src/main/java/au/org/aodn/esindexer/service/AodnDiscoveryParameterVocabService.java b/indexer/src/main/java/au/org/aodn/esindexer/service/AodnDiscoveryParameterVocabService.java
@@ -1,8 +1,9 @@
 package au.org.aodn.esindexer.service;
 
-import au.org.aodn.esindexer.utils.CacheArdcVocabsUtils;
+import au.org.aodn.esindexer.utils.VocabsUtils;
 import au.org.aodn.stac.model.ConceptModel;
 import au.org.aodn.stac.model.ThemesModel;
+import co.elastic.clients.elasticsearch.ElasticsearchClient;
 import com.fasterxml.jackson.databind.JsonNode;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.stereotype.Service;
@@ -15,7 +16,10 @@
 @Service
 public class AodnDiscoveryParameterVocabService {
     @Autowired
-    CacheArdcVocabsUtils cacheArdcVocabsUtils;
+    VocabsUtils cacheArdcVocabsUtils;
+
+    @Autowired
+    ElasticsearchClient portalElasticsearchClient;
 
     protected boolean themesMatchConcept(List<ThemesModel> themes, ConceptModel thatConcept) {
         for (ThemesModel theme : themes) {

diff --git a/indexer/src/main/java/au/org/aodn/esindexer/service/IndexerServiceImpl.java b/indexer/src/main/java/au/org/aodn/esindexer/service/IndexerServiceImpl.java
@@ -5,6 +5,7 @@
 import au.org.aodn.esindexer.exception.*;
 import au.org.aodn.esindexer.utils.JaxbUtils;
 import au.org.aodn.metadata.iso19115_3_2018.*;
+import au.org.aodn.stac.model.RecordSuggest;
 import co.elastic.clients.elasticsearch.ElasticsearchClient;
 import co.elastic.clients.elasticsearch._types.ElasticsearchException;
 import co.elastic.clients.elasticsearch.core.*;
@@ -127,14 +128,21 @@ protected StacCollectionModel getMappedMetadataValues(String metadataValues) thr
 
         stacCollectionModel.getSummaries().setScore(score);
 
-        // TODO: in future, blah blah here
-        stacCollectionModel.setTitleSuggest(stacCollectionModel.getTitle());
-
         List<String> aodnDiscoveryCategories = aodnDiscoveryParameterVocabService.getAodnDiscoveryCategories(stacCollectionModel.getThemes());
         if (!aodnDiscoveryCategories.isEmpty()) {
             stacCollectionModel.getSummaries().setDiscoveryCategories(aodnDiscoveryCategories);
         }
 
+
+        // categories suggest using a different index
+
+        // extendable for other aspects of the records data. eg. title, description, etc. something that are unique to the record and currently using "text" type
+        RecordSuggest recordSuggest = RecordSuggest.builder()
+                .title(stacCollectionModel.getTitle())
+                .description(null) // purely for demonstrating the extendability, manage extra suggest field in RecordSuggest class, and the index JSON schema
+                .build();
+        stacCollectionModel.setRecordSuggest(recordSuggest);
+
         return stacCollectionModel;
     }
 

diff --git a/...esindexer/utils/CacheArdcVocabsUtils.java → ...org/aodn/esindexer/utils/VocabsUtils.java b/...esindexer/utils/CacheArdcVocabsUtils.java → ...org/aodn/esindexer/utils/VocabsUtils.java
@@ -32,7 +32,7 @@
 @Component
 // create and inject a stub proxy to self due to the circular reference http://bit.ly/4aFvYtt
 @Scope(proxyMode = ScopedProxyMode.TARGET_CLASS)
-public class CacheArdcVocabsUtils {
+public class VocabsUtils {
     @Value(AppConstants.AODN_DISCOVERY_PARAMETER_VOCAB_API)
     protected String vocabApi;
 
@@ -41,7 +41,7 @@ public class CacheArdcVocabsUtils {
 
     // self-injection to avoid self-invocation problems when calling the cachable method within the same bean
     @Autowired
-    CacheArdcVocabsUtils self;
+    VocabsUtils self;
 
     @Autowired
     ElasticSearchIndexService elasticSearchIndexService;

diff --git a/indexer/src/main/resources/config_files/aodn_discovery_parameter_vocabularies_index.json b/indexer/src/main/resources/config_files/aodn_discovery_parameter_vocabularies_index.json
@@ -1,16 +1,28 @@
 {
+  "settings":{
+    "analysis":{
+      "analyzer":{
+        "custom_analyser":{
+          "type":"custom",
+          "tokenizer":"standard",
+          "filter":[
+            "lowercase",
+            "english_stop"
+          ]
+        }
+      },
+      "filter":{
+        "english_stop":{
+          "type":"stop",
+          "stopwords":"_english_"
+        }
+      }
+    }
+  },
   "mappings": {
     "dynamic": true,
     "properties": {
-      "label": {
-        "type": "text",
-        "fields": {
-          "keyword": {
-            "type": "keyword",
-            "ignore_above": 256
-          }
-        }
-      },
+      "label": { "type":  "search_as_you_type", "analyzer": "custom_analyser" },
       "definition": {
         "type": "text"
       },
@@ -20,15 +32,7 @@
       "broader": {
         "type": "nested",
         "properties": {
-          "label": {
-            "type": "text",
-            "fields": {
-              "keyword": {
-                "type": "keyword",
-                "ignore_above": 256
-              }
-            }
-          },
+          "label": { "type":  "text" },
           "about": {
             "type": "keyword"
           }
@@ -37,30 +41,14 @@
       "narrower": {
         "type": "nested",
         "properties": {
-          "label": {
-            "type": "text",
-            "fields": {
-              "keyword": {
-                "type": "keyword",
-                "ignore_above": 256
-              }
-            }
-          },
+          "label": { "type": "text" },
           "about": {
             "type": "keyword"
           },
           "narrower": {
             "type": "nested",
             "properties": {
-              "label": {
-                "type": "text",
-                "fields": {
-                  "keyword": {
-                    "type": "keyword",
-                    "ignore_above": 256
-                  }
-                }
-              },
+              "label": { "type":  "text" },
               "about": {
                 "type": "keyword"
               }

diff --git a/indexer/src/main/resources/config_files/portal_records_index_schema.json b/indexer/src/main/resources/config_files/portal_records_index_schema.json
@@ -29,9 +29,12 @@
       "stac_version": { "type": "text" },
       "type": { "type": "text" },
       "title": { "type": "text" },
-      "title_suggest": {
-        "type": "search_as_you_type",
-        "analyzer": "custom_analyser"
+      "record_suggest": {
+        "type": "nested",
+        "properties": {
+          "title": { "type": "search_as_you_type", "analyzer": "custom_analyser" },
+          "description": { "type": "search_as_you_type", "analyzer": "custom_analyser" }
+        }
       },
       "discovery_categories" : { "type":  "keyword" },
       "keywords": {

diff --git a/...AodnDiscoveryParameterVocabUtilsTest.java → .../aodn/esindexer/utils/VocabUtilsTest.java b/...AodnDiscoveryParameterVocabUtilsTest.java → .../aodn/esindexer/utils/VocabUtilsTest.java
@@ -1,13 +1,10 @@
 package au.org.aodn.esindexer.utils;
 
-import au.org.aodn.ardcvocabs.configuration.ArdcAutoConfiguration;
 import au.org.aodn.esindexer.service.AodnDiscoveryParameterVocabService;
 import au.org.aodn.stac.model.ConceptModel;
 import au.org.aodn.stac.model.ThemesModel;
 import org.junit.jupiter.api.*;
 import org.springframework.beans.factory.annotation.Autowired;
-import org.springframework.boot.autoconfigure.EnableAutoConfiguration;
-import org.springframework.boot.autoconfigure.ImportAutoConfiguration;
 import org.springframework.boot.test.context.SpringBootTest;
 import org.springframework.test.context.ActiveProfiles;
 
@@ -23,7 +20,7 @@
 @ActiveProfiles("test")
 @TestMethodOrder(MethodOrderer.OrderAnnotation.class)
 @TestInstance(TestInstance.Lifecycle.PER_CLASS)
-public class AodnDiscoveryParameterVocabUtilsTest {
+public class VocabUtilsTest {
     @Autowired
     AodnDiscoveryParameterVocabService aodnDiscoveryParameterVocabService;
 

diff --git a/indexer/src/test/resources/canned/sample4_stac.json b/indexer/src/test/resources/canned/sample4_stac.json
@@ -246,7 +246,9 @@
     "title" : "AODN Discovery Parameter Vocabulary"
   } ],
   "id" : "7709f541-fc0c-4318-b5b9-9053aa474e0e",
-  "title_suggest" : "Ocean acidification historical reconstruction",
+  "record_suggest" : {
+    "title" : "Ocean acidification historical reconstruction"
+  },
   "type" : "Collection",
   "stac_version" : "1.0.0",
   "stac_extensions" : [ "https://stac-extensions.github.io/scientific/v1.0.0/schema.json", "https://stac-extensions.github.io/contacts/v0.1.1/schema.json", "https://stac-extensions.github.io/projection/v1.1.0/schema.json", "https://stac-extensions.github.io/language/v1.0.0/schema.json", "https://stac-extensions.github.io/themes/v1.0.0/schema.json" ]

diff --git a/stacmodel/src/main/java/au/org/aodn/stac/model/RecordSuggest.java b/stacmodel/src/main/java/au/org/aodn/stac/model/RecordSuggest.java
@@ -0,0 +1,11 @@
+package au.org.aodn.stac.model;
+
+import lombok.Builder;
+import lombok.Data;
+
+@Data
+@Builder
+public class RecordSuggest {
+    private String title;
+    private String description;
+}
diff --git a/stacmodel/src/main/java/au/org/aodn/stac/model/StacCollectionModel.java b/stacmodel/src/main/java/au/org/aodn/stac/model/StacCollectionModel.java
@@ -14,8 +14,8 @@ public class StacCollectionModel {
     @JsonProperty("id")
     protected String uuid;
     protected String title;
-    @JsonProperty("title_suggest")
-    protected String titleSuggest;
+    @JsonProperty("record_suggest")
+    protected RecordSuggest recordSuggest;
     protected String description;
     protected ExtentModel extent;
     protected SummariesModel summaries;