DHFPROD-9050: Additional matching performance updates

Author: Ryan Dew

marklogic-data-hub/src/main/java/com/marklogic/hub/deploy/commands/DeployHubAmpsCommand.java

@@ -52,7 +52,7 @@ protected ResourceManager getResourceManager(CommandContext context) {
         return new HubAmpManager(context.getManageClient());
     }
 
-    class HubAmpManager extends AmpManager {
+    static class HubAmpManager extends AmpManager {
 
         public HubAmpManager(ManageClient client) {
             super(client);

marklogic-data-hub/src/main/java/com/marklogic/hub/deploy/commands/LoadUserModulesCommand.java

@@ -197,7 +197,7 @@ public void execute(CommandContext context) {
         DatabaseClient finalClient = hubConfig.newFinalClient();
 
         Path userModulesPath = hubConfig.getHubPluginsDir();
-        String baseDir = new String();
+        String baseDir = "";
         String EncodedBaseDir = userModulesPath.normalize().toAbsolutePath().toString();
         Path startPath = userModulesPath.resolve("entities");
 

marklogic-data-hub/src/main/java/com/marklogic/hub/flow/impl/FlowRunnerImpl.java

@@ -595,7 +595,7 @@ public Flow getRunningFlow() {
         return this.runningFlow;
     }
 
-    class FlowContext {
+    static class FlowContext {
         boolean jobOutputIsEnabled = true;
         JobService jobService;
 

marklogic-data-hub/src/main/java/com/marklogic/hub/impl/DataHubImpl.java

@@ -459,17 +459,17 @@ public void clearUserArtifacts(){
     public List<Command> buildListOfCommands() {
         Map<String, List<Command>> commandMap = buildCommandMap();
         List<Command> commands = new ArrayList<>();
-        for (String name : commandMap.keySet()) {
-            commands.addAll(commandMap.get(name));
+        for (Map.Entry<String, List<Command>> entry: commandMap.entrySet()) {
+            commands.addAll(entry.getValue());
         }
         return commands;
     }
 
     public List<Command> getSecurityCommandList() {
         Map<String, List<Command>> commandMap = getSecurityCommands();
         List<Command> commands = new ArrayList<>();
-        for (String name : commandMap.keySet()) {
-            commands.addAll(commandMap.get(name));
+        for (Map.Entry<String, List<Command>> entry: commandMap.entrySet()) {
+            commands.addAll(entry.getValue());
         }
         return commands;
     }

marklogic-data-hub/src/main/java/com/marklogic/hub/impl/HubProjectImpl.java

@@ -29,7 +29,15 @@
 import org.springframework.core.io.Resource;
 import org.springframework.core.io.support.PathMatchingResourcePatternResolver;
 import org.springframework.util.FileCopyUtils;
-import java.io.*;
+
+import java.io.File;
+import java.io.FileInputStream;
+import java.io.FileOutputStream;
+import java.io.FileWriter;
+import java.io.IOException;
+import java.io.InputStream;
+import java.io.OutputStream;
+import java.nio.charset.Charset;
 import java.nio.file.Files;
 import java.nio.file.Path;
 import java.nio.file.Paths;
@@ -403,12 +411,11 @@ private void writeResourceFileWithReplace(Map<String, String> customTokens, Stri
                 logger.debug("Getting file with replace: " + srcFile);
                 inputStream = HubProject.class.getClassLoader().getResourceAsStream(srcFile);
 
-                String fileContents = IOUtils.toString(inputStream);
-                for (String key : customTokens.keySet()) {
-
-                    String value = customTokens.get(key);
+                String fileContents = IOUtils.toString(inputStream, Charset.defaultCharset());
+                for (Map.Entry<String, String> entry : customTokens.entrySet()) {
+                    String value = entry.getValue();
                     if (value != null) {
-                        fileContents = fileContents.replace(key, value);
+                        fileContents = fileContents.replace(entry.getKey(), value);
                     }
                 }
                 try(FileWriter writer = new FileWriter(dstFile.toFile())) {

marklogic-data-hub/src/main/java/com/marklogic/hub/impl/ScaffoldingImpl.java

@@ -448,7 +448,7 @@ private String getFileContent(String srcFile, String placeholder) {
         return output.toString();
     }
 
-    public class ContentPlugin extends ResourceManager {
+    public static class ContentPlugin extends ResourceManager {
         private static final String NAME = "mlScaffoldContent";
 
         private RequestParameters params = new RequestParameters();

marklogic-data-hub/src/main/java/com/marklogic/hub/util/json/JSONObject.java

@@ -360,8 +360,8 @@ public String convertMapToJsonString(Map<String, Object> map) throws JsonProcess
      * @param map a map of (String, Object)
      */
     public void putMap(Map<String, Object> map) {
-        for (String key : map.keySet()) {
-            put(key, map.get(key));
+        for (Map.Entry<String, Object> entry : map.entrySet()) {
+            put(entry.getKey(), entry.getValue());
         }
     }
 

marklogic-data-hub/src/main/resources/ml-modules/root/com.marklogic.smart-mastering/algorithms/thesaurus.xqy

@@ -26,7 +26,7 @@ declare function algorithms:thesaurus(
 {
   let $expand-options := fn:head(($expand/options, $expand))
   let $thesaurus := $expand-options/(*:thesaurus|thesaurusURI)
-  let $original-values := $expand-values ! fn:string(.)[. ne '']
+  let $original-values := $expand-values ! fn:string(.) ! (., fn:lower-case(.))[. ne '']
   where fn:exists($thesaurus) and fn:exists($original-values)
   return
     let $expand-options := fn:head(($expand/options, $expand))

marklogic-data-hub/src/main/resources/ml-modules/root/com.marklogic.smart-mastering/impl/process.xqy

@@ -369,7 +369,7 @@ declare function proc-impl:build-match-summary(
   let $archived-collection := coll:archived-collections($match-options-node)
   let $normalized-input :=
     if ($input instance of xs:string*) then
-      for $doc in cts:search(fn:doc(), cts:and-not-query(cts:document-query($input), cts:collection-query($archived-collection)), "unfiltered")
+      for $doc in cts:search(fn:doc(), cts:and-not-query(cts:document-query($input), cts:collection-query($archived-collection)), ("unfiltered", "score-zero"), 0)
       return
         util-impl:build-write-object-for-doc($doc)
     else if ($input instance of map:map*) then
@@ -648,6 +648,8 @@ declare function proc-impl:build-match-summary(
       => map:with(
         "matchSummary",
         map:map()
+        => map:with("matchStepFlow", fn:string($match-options-node/flow))
+        => map:with("matchStepName", fn:string($match-options-node/stepName))
         => map:with("URIsToProcess",
           json:to-array(fn:distinct-values(($no-matches-uris,map:keys($action-details)))))
         => map:with("actionDetails", $action-details)
@@ -682,7 +684,7 @@ declare function proc-impl:process-match-and-merge-with-options(
   let $archived-collection := coll:archived-collections($matching-options)
   let $normalized-input :=
     if ($input instance of xs:string*) then
-      for $doc in cts:search(fn:doc(), cts:and-not-query(cts:document-query($input), cts:collection-query($archived-collection)), "unfiltered")
+      for $doc in cts:search(fn:doc(), cts:and-not-query(cts:document-query($input), cts:collection-query($archived-collection)), ("unfiltered", "score-zero"), 0)
       return
         util-impl:build-write-object-for-doc($doc)
     else if ($input instance of map:map*) then

marklogic-data-hub/src/main/resources/ml-modules/root/com.marklogic.smart-mastering/impl/sm-es-impl.xqy

@@ -181,7 +181,7 @@ declare function es-impl:get-entity-property-info($entity-type-iri as xs:string)
       $entity-properties-by-model-iri => map:get($entity-model-iri) => map:get($entity-type-title)
     else
       let $entity-model-descriptor := fn:head(cts:search(fn:collection("http://marklogic.com/entity-services/models"),
-        cts:triple-range-query(sem:iri($entity-type-iri), sem:curie-expand("rdf:type"), sem:iri("http://marklogic.com/entity-services#EntityType"), "=")))
+        cts:triple-range-query(sem:iri($entity-type-iri), sem:curie-expand("rdf:type"), sem:iri("http://marklogic.com/entity-services#EntityType"), "="), ("score-zero"), 0))
       let $entity-descriptor-maps := map:map()
       let $entity-property-info := es-impl:get-definition-properties($entity-model-descriptor, $entity-descriptor-maps, $entity-type-title, get-entity-type-namespaces($entity-type-iri))
       return (