Gcmd proxied mapping parameter vocabs

indexer/src/main/java/au/org/aodn/esindexer/service/IndexerServiceImpl.java

@@ -2,6 +2,7 @@
 
 import au.org.aodn.esindexer.configuration.AppConstants;
 import au.org.aodn.esindexer.exception.*;
+import au.org.aodn.esindexer.utils.GcmdKeywordUtils;
 import au.org.aodn.esindexer.utils.JaxbUtils;
 import au.org.aodn.metadata.iso19115_3_2018.MDMetadataType;
 import au.org.aodn.stac.model.StacCollectionModel;
@@ -41,6 +42,8 @@
 import java.io.InputStream;
 import java.util.*;
 import java.util.concurrent.CompletableFuture;
+import java.util.stream.Collectors;
+import java.util.stream.Stream;
 
 @Slf4j
 @Service
@@ -57,7 +60,7 @@ public class IndexerServiceImpl implements IndexerService {
     protected JaxbUtils<MDMetadataType> jaxbUtils;
     protected RankingService rankingService;
     protected VocabService vocabService;
-
+    protected GcmdKeywordUtils gcmdKeywordUtils;
     protected static final long DEFAULT_BACKOFF_TIME = 3000L;
 
     @Lazy
@@ -75,7 +78,8 @@ public IndexerServiceImpl(
             ElasticsearchClient portalElasticsearchClient,
             ElasticSearchIndexService elasticSearchIndexService,
             StacCollectionMapperService stacCollectionMapperService,
-            VocabService vocabService
+            VocabService vocabService,
+            GcmdKeywordUtils gcmdKeywordUtils
     ) {
         this.indexName = indexName;
         this.tokensAnalyserName = tokensAnalyserName;
@@ -87,6 +91,7 @@ public IndexerServiceImpl(
         this.elasticSearchIndexService = elasticSearchIndexService;
         this.stacCollectionMapperService = stacCollectionMapperService;
         this.vocabService = vocabService;
+        this.gcmdKeywordUtils = gcmdKeywordUtils;
     }
 
     public Hit<ObjectNode> getDocumentByUUID(String uuid) throws IOException {
@@ -171,9 +176,11 @@ protected StacCollectionModel getMappedMetadataValues(String metadataValues) thr
         stacCollectionModel.getSummaries().setScore(score);
 
         // parameter vocabs
+        List<String> mappedParameterVocabsFromGcmdKeywords = gcmdKeywordUtils.getMappedParameterVocabsFromGcmdKeywords(stacCollectionModel.getThemes());
         List<String> processedParameterVocabs = vocabService.extractVocabLabelsFromThemes(stacCollectionModel.getThemes(), AppConstants.AODN_DISCOVERY_PARAMETER_VOCABS);
+
         if (!processedParameterVocabs.isEmpty()) {
-            stacCollectionModel.getSummaries().setParameterVocabs(processedParameterVocabs);
+            stacCollectionModel.getSummaries().setParameterVocabs(Stream.concat(mappedParameterVocabsFromGcmdKeywords.stream(), processedParameterVocabs.stream()).distinct().collect(Collectors.toList()));
         }
 
         /*

indexer/src/main/java/au/org/aodn/esindexer/utils/GcmdKeywordUtils.java

@@ -0,0 +1,102 @@
+package au.org.aodn.esindexer.utils;
+
+import au.org.aodn.stac.model.ConceptModel;
+import au.org.aodn.stac.model.ThemesModel;
+import lombok.extern.slf4j.Slf4j;
+import org.springframework.stereotype.Component;
+import org.springframework.util.ResourceUtils;
+
+import javax.annotation.PostConstruct;
+import java.io.*;
+import java.nio.file.Files;
+import java.util.*;
+
+
+@Slf4j
+@Component
+public class GcmdKeywordUtils {
+
+    protected Map<String, String> gcmdMapping = new HashMap<>();
+
+    @PostConstruct
+    public void init() {
+        loadCsvToMap("classpath:config_files/gcmd-mapping.csv");
+    }
+
+    private String getLastWord(String keyword) {
+        String result;
+        if (keyword.contains("|")) {
+            result = keyword.substring(keyword.lastIndexOf("|") + 1).strip();
+        } else if (keyword.contains(">")) {
+            result = keyword.substring(keyword.lastIndexOf(">") + 1).strip();
+        } else {
+            result = keyword.strip();
+        }
+        return result;
+    }
+
+
+    private static String readResourceFile(String path) throws IOException {
+        File f = ResourceUtils.getFile(path);
+        return new String(Files.readAllBytes(f.toPath()));
+    }
+
+    // Load the CSV file into a HashMap
+    private void loadCsvToMap(String path) {
+        try {
+            // Read the file as a single String
+            String fileContent = readResourceFile(path);
+
+            // Split the content into lines
+            String[] lines = fileContent.split("\\r?\\n");
+
+            // Process each line
+            for (String line : lines) {
+                // Split the line into key and value based on comma
+                String[] parts = line.split(",");
+                if (parts.length >= 2) {
+                    String key = parts[0].trim();
+                    String value = parts[1].trim();
+                    gcmdMapping.put(key, value);
+                }
+            }
+        } catch (IOException e) {
+            log.error(e.getMessage());
+        }
+    }
+
+    protected List<String> extractGcmdKeywordLastWords(List<ThemesModel> themes) {
+        Set<String> keywords = new HashSet<>();
+        for (ThemesModel themesModel : themes) {
+            if ((themesModel.getTitle().toLowerCase().contains("gcmd") || themesModel.getTitle().toLowerCase().contains("global change master directory")) && !themesModel.getTitle().toLowerCase().contains("palaeo temporal coverage")) {
+                for (ConceptModel conceptModel : themesModel.getConcepts()) {
+                    if (conceptModel.getId() != null && !conceptModel.getId().isEmpty()) {
+                        keywords.add(getLastWord(conceptModel.getId().replace("\"", "")).toUpperCase());
+                    }
+                }
+            }
+        }
+        return new ArrayList<>(keywords);
+    }
+
+    protected String getParameterVocabByGcmdKeywordLastWord(String gcmdKeywordLastWord) {
+        return gcmdMapping.getOrDefault(gcmdKeywordLastWord, "");
+    }
+
+    public List<String> getMappedParameterVocabsFromGcmdKeywords(List<ThemesModel> themes) {
+        Set<String> results = new HashSet<>();
+
+        List<String> gcmdKeywordLastWords = extractGcmdKeywordLastWords(themes);
+
+        if (!gcmdKeywordLastWords.isEmpty()) {
+            for (String gcmdKeywordLastWord : gcmdKeywordLastWords) {
+                String mappedParameterVocab = getParameterVocabByGcmdKeywordLastWord(gcmdKeywordLastWord);
+                if (!mappedParameterVocab.isEmpty() && !mappedParameterVocab.equalsIgnoreCase("uncategorised")) {
+                    results.add(mappedParameterVocab.toLowerCase());
+                }
+            }
+        }
+
+        return new ArrayList<>(results);
+    }
+}