Revize 6b3bc31e

Re #7918: Zrychlení nahrávání dokumentů
- vytvoření vlastních schémat (schema.xml) pro Solr
- neindexování polí se souřadnicemi

be/fulltextsearch/src/main/java/cz/zcu/kiv/aswi/fulltextsearch/SolrService.java

@@ -6 +6 @@
 import org.apache.solr.client.solrj.SolrQuery;
 import org.apache.solr.client.solrj.SolrServerException;
 import org.apache.solr.client.solrj.impl.HttpSolrClient;
+import org.apache.solr.client.solrj.request.CoreAdminRequest;
 import org.apache.solr.client.solrj.request.schema.SchemaRequest;
 import org.apache.solr.client.solrj.response.QueryResponse;
 import org.apache.solr.client.solrj.response.schema.SchemaResponse;
@@ -47 +48 @@
     private static final String FIELD_DOC_CONTENT = "doc_content";
     private static final String FIELD_IMG_CONTENT = "img_content";
     private static final float IMG_OUTPUT_QUALITY = 0.5f;
-    private static final Logger LOG = LoggerFactory.getLogger(IndexController.class);
+    private static final Logger LOG = LoggerFactory.getLogger(SolrService.class);
 
     private static final int SEARCH_QUERY_ROWS = 100;
 
@@ -61 +62 @@
 
     private void connect() {
         LOG.info("Connecting to Solr");
-        String urlString = URL + ":8983/solr/" + CORE_NAME_DOCUMENTS;
-        solrClientDocuments = new HttpSolrClient.Builder(urlString).build();
+        String urlString = URL + ":8983/solr/";
 
-        urlString = URL + ":8983/solr/" + CORE_NAME_FILES;
-        solrClientFiles = new HttpSolrClient.Builder(urlString).build();
+        try {
+            solrClientDocuments = new HttpSolrClient.Builder(urlString).build();
+            CoreAdminRequest.createCore(CORE_NAME_DOCUMENTS, CORE_NAME_DOCUMENTS, solrClientDocuments, "conf/solrconfig.xml", "documents-schema.xml");
+            solrClientDocuments.setBaseURL(solrClientDocuments.getBaseURL() + "/" +  CORE_NAME_DOCUMENTS);
+
+            solrClientFiles = new HttpSolrClient.Builder(urlString).build();
+            CoreAdminRequest.createCore(CORE_NAME_FILES, CORE_NAME_FILES, solrClientFiles, "conf/solrconfig.xml", "files-schema.xml");
+            solrClientFiles.setBaseURL(solrClientFiles.getBaseURL() + "/"  + CORE_NAME_FILES);
+        } catch (SolrServerException e) {
+            e.printStackTrace();
+        } catch (IOException e) {
+            e.printStackTrace();
+        }
     }
 
     public int uploadFiles(List<MultipartFile> files) throws IOException, SolrServerException, JAXBException {
@@ -210 +221 @@
     private void addBasicFieldsToDocumentsCore() {
         LOG.trace("SolrService method addBasicFieldsToDocumentsCore called");
         LOG.debug("Adding basic fields to Documents core");
-        addFieldToDocumentCore(FIELD_DOC_FILENAME);
-        addFieldToDocumentCore(FIELD_IMG_FILENAME);
-        addFieldToDocumentCore(FIELD_TEXT_REGION);
-        addFieldToDocumentCore(FIELD_TEXT_REGION_COORDS);
+        addFieldToDocumentCore(FIELD_DOC_FILENAME, getAddTextFieldRequest(FIELD_DOC_FILENAME));
+        addFieldToDocumentCore(FIELD_IMG_FILENAME, getAddTextFieldRequest(FIELD_IMG_FILENAME));
+        addFieldToDocumentCore(FIELD_TEXT_REGION, getAddTextFieldRequest(FIELD_TEXT_REGION));
+        addFieldToDocumentCore(FIELD_TEXT_REGION_COORDS, getNotIndexedAddTextFieldRequest(FIELD_TEXT_REGION_COORDS));
     }
 
     private void getFieldNamesFromDocumentCore() throws IOException, SolrServerException {
@@ -229 +240 @@
     }
 
 
-    private void addFieldToDocumentCore(String name) {
+    private void addFieldToDocumentCore(String name, SchemaRequest.Update request) {
         LOG.trace("SolrService method addFieldToDocumentCore called");
 
         if (fieldNamesDocumentCore.contains(name)) return;
-        LOG.debug("Adding field to Documents core");
+        LOG.debug("Adding field \"" + name + "\" to Documents core");
         try {
-            SchemaRequest.Update request = getAddTextFieldRequest(name);
             request.process(solrClientDocuments);
             fieldNamesDocumentCore.add(name);
         } catch (Exception e) {
@@ -244 +254 @@
         }
     }
 
-    private SchemaRequest.Update getAddTextFieldRequest(String name) {
-        LOG.trace("SolrService method getAddTextFieldRequest called");
+    private Map<String, Object> getFieldAttributes(String name) {
         Map<String, Object> fieldAttributes = new LinkedHashMap<>();
         fieldAttributes.put("name", name);
         fieldAttributes.put("type", "text_general");
         fieldAttributes.put("stored", true);
         fieldAttributes.put("multiValued", false);
+        return fieldAttributes;
+    }
+
+    private SchemaRequest.Update getAddTextFieldRequest(String name) {
+        LOG.trace("SolrService method getAddTextFieldRequest called");
+        Map<String, Object> fieldAttributes = getFieldAttributes(name);
+        return new SchemaRequest.AddField(fieldAttributes);
+    }
+
+    private SchemaRequest.Update getNotIndexedAddTextFieldRequest(String name) {
+        LOG.trace("SolrService method getAddTextFieldRequest called");
+        Map<String, Object> fieldAttributes = getFieldAttributes(name);
+        fieldAttributes.put("indexed", false);
         return new SchemaRequest.AddField(fieldAttributes);
     }
 
@@ -439 +461 @@
                 TextLine textLine = textLines[j];
                 String prefix = PREFIX_TEXT_LINE + j;
 
-                addFieldToDocumentCore(prefix);
+                addFieldToDocumentCore(prefix, getAddTextFieldRequest(prefix));
                 solrInputDocument.addField(prefix, textLine.getTextEquiv().getUnicode());
 
-                addFieldToDocumentCore(prefix + SUFFIX_COORDS);
+                addFieldToDocumentCore(prefix + SUFFIX_COORDS, getNotIndexedAddTextFieldRequest(prefix + SUFFIX_COORDS));
                 solrInputDocument.addField(prefix + SUFFIX_COORDS, textLine.getLineCoords().getPointsString());
 
                 TextWord[] textWords = textLine.getTextWords();
@@ -451 +473 @@
                     TextWord textWord = textWords[k];
                     String wordPrefix = PREFIX_TEXT_WORD + k + "_" + prefix;
 
-                    addFieldToDocumentCore(wordPrefix);
+                    addFieldToDocumentCore(wordPrefix, getAddTextFieldRequest(wordPrefix));
                     solrInputDocument.addField(wordPrefix, textWord.getTextEquiv().getUnicode());
 
-                    addFieldToDocumentCore(wordPrefix + SUFFIX_COORDS);
+                    addFieldToDocumentCore(wordPrefix + SUFFIX_COORDS, getNotIndexedAddTextFieldRequest(wordPrefix + SUFFIX_COORDS));
                     solrInputDocument.addField(wordPrefix + SUFFIX_COORDS, textWord.getWordCoords().getPointsString());
                 }
             }