Merge branch 'feature/update-dropwizard'

Author: lfoppiano

build.gradle

@@ -24,6 +24,7 @@ plugins {
     id 'org.ajoberstar.grgit' version '5.3.0' apply false
     id 'distribution'
     id 'application'
+    id "org.jetbrains.kotlin.jvm" version "1.8.21"
 }
 
 apply plugin: 'jacoco'
@@ -89,29 +90,31 @@ dependencies {
 
     //Apache commons
     implementation group: 'commons-pool', name: 'commons-pool', version: '1.6'
-    implementation group: 'commons-io', name: 'commons-io', version: '2.9.0'
+    implementation group: 'commons-io', name: 'commons-io', version: '2.14.0'
     implementation group: 'org.apache.httpcomponents', name: 'httpclient', version: '4.5.14'
     implementation group: 'org.apache.httpcomponents', name: 'httpmime', version: '4.5.3'
     implementation group: 'org.apache.commons', name: 'commons-lang3', version: '3.6'
     implementation group: 'org.apache.commons', name: 'commons-collections4', version: '4.4'
     implementation group: 'org.apache.commons', name: 'commons-csv', version: '1.5'
 
     //Dropwizard
-    implementation "io.dropwizard:dropwizard-core:1.3.23"
-    implementation "io.dropwizard:dropwizard-assets:1.3.23"
-    implementation "com.hubspot.dropwizard:dropwizard-guicier:1.3.5.0"
-    implementation "io.dropwizard:dropwizard-testing:1.3.23"
-    implementation "io.dropwizard:dropwizard-forms:1.3.23"
-    implementation "io.dropwizard:dropwizard-client:1.3.23"
-    implementation "io.dropwizard:dropwizard-auth:1.3.23"
-    implementation "io.dropwizard.metrics:metrics-core:4.0.0"
-    implementation "io.dropwizard.metrics:metrics-servlets:4.0.0"
-
-    implementation group: 'com.google.guava', name: 'guava', version: '28.2-jre'
+    implementation 'ru.vyarus:dropwizard-guicey:7.0.0'
+
+    implementation 'io.dropwizard:dropwizard-bom:4.0.2'
+    implementation 'io.dropwizard:dropwizard-core:4.0.2'
+    implementation 'io.dropwizard:dropwizard-assets:4.0.2'
+    implementation 'io.dropwizard:dropwizard-testing:4.0.2'
+    implementation 'io.dropwizard:dropwizard-forms:4.0.2'
+    implementation 'io.dropwizard:dropwizard-client:4.0.2'
+    implementation 'io.dropwizard:dropwizard-auth:4.0.2'
+    implementation 'io.dropwizard.metrics:metrics-core:4.2.22'
+    implementation 'io.dropwizard.metrics:metrics-servlets:4.2.22'
+
+    implementation group: 'com.google.guava', name: 'guava', version: '32.0.1-jre'
 
     //Parsing xml/json
-    implementation group: 'com.fasterxml.jackson.core', name: 'jackson-core', version: '2.10.1'
-    implementation group: 'com.fasterxml.jackson.core', name: 'jackson-databind', version: '2.10.1'
+    implementation group: 'com.fasterxml.jackson.core', name: 'jackson-core', version: '2.13.1'
+    implementation group: 'com.fasterxml.jackson.core', name: 'jackson-databind', version: '2.13.4.2'
     implementation group: 'com.fasterxml.jackson.core', name: 'jackson-annotations', version: '2.10.1'
     implementation group: 'xom', name: 'xom', version: '1.3.2'
     implementation group: 'javax.xml.bind', name: 'jaxb-api', version: '2.3.0'
@@ -141,10 +144,27 @@ dependencies {
     }
 
     //Tests
-    testImplementation group: 'junit', name: 'junit', version: '4.12'
-    testImplementation group: 'org.hamcrest', name: 'hamcrest-all', version: '1.3'
+    testImplementation(platform('org.junit:junit-bom:5.10.2'))
+    testRuntimeOnly("org.junit.platform:junit-platform-launcher") {
+        because("Only needed to run tests in a version of IntelliJ IDEA that bundles older versions")
+    }
+    testRuntimeOnly("org.junit.jupiter:junit-jupiter-engine")
+    testImplementation('org.junit.jupiter:junit-jupiter')
+    testRuntimeOnly("org.junit.vintage:junit-vintage-engine") {
+        because 'allows JUnit 3 and JUnit 4 tests to run'
+    }
+
+    testRuntimeOnly("org.junit.platform:junit-platform-launcher") {
+        because 'allows tests to run from IDEs that bundle older version of launcher'
+    }
+    testImplementation 'org.easymock:easymock:5.2.0'
+    testImplementation 'org.hamcrest:hamcrest-all:1.3'
+    testImplementation 'org.hamcrest:hamcrest-library:2.2'
     testImplementation 'org.powermock:powermock-module-junit4:2.0.9'
     testImplementation 'org.powermock:powermock-api-easymock:2.0.9'
+    testImplementation 'org.jetbrains.kotlin:kotlin-test'
+    testImplementation "io.mockk:mockk:1.13.9"
+
 }
 
 configurations.all {
@@ -155,6 +175,7 @@ configurations.all {
     exclude group: 'org.slf4j', module: "slf4j-log4j12"
 }
 
+
 def getJavaLibraryPath = {
     def jepLocalLibraries = ""
     if (Os.isFamily(Os.FAMILY_MAC)) {
@@ -275,6 +296,10 @@ jar {
     enabled true
 }
 
+artifacts {
+    archives shadowJar
+}
+
 distZip.enabled = true
 distTar.enabled = false
 shadowDistZip.enabled = false

resources/config/config-docker.yml

@@ -19,7 +19,7 @@ entityFishingPort: 443
 #entityFishingPort: 8090
 
 # if true we use binary classifiers for the contexts, otherwise use a single multi-label classifier
-# binary classifiers perform better, but havier to use
+# binary classifiers perform better, but heavier to use
 useBinaryContextClassifiers: false
 
 # sequence labeling model (identify data-related sections)
@@ -35,7 +35,7 @@ models:
       window: 20
       nbMaxIterations: 2000
 
-  # classifier model, dataset binary (datset or not dataset in the current sentence)
+  # classifier model, dataset binary (dataset or not dataset in the current sentence)
   - name: "dataseer-binary"
     engine: "delft"
     delft:
@@ -120,7 +120,6 @@ models:
       architecture: "bert"
       transformer: "michiyasunaga/LinkBERT-basecased"
 
-
 # Limit the maximum number of requests (0, no limit)
 maxParallelRequests: 0
 
@@ -131,17 +130,18 @@ corsAllowedHeaders: "X-Requested-With,Content-Type,Accept,Origin"
 
 server:
   type: custom
-  idleTimeout: 120 seconds
   applicationConnectors:
     - type: http
       port: 8060
+      idleTimeout: 120 seconds
+      acceptQueueSize: 2048
   adminConnectors:
     - type: http
       port: 8061
   registerDefaultExceptionMappers: false
   maxThreads: 2048
   maxQueuedRequests: 2048
-  acceptQueueSize: 2048
+
   requestLog:
     appenders: []
 

resources/config/config.yml

@@ -131,17 +131,18 @@ corsAllowedHeaders: "X-Requested-With,Content-Type,Accept,Origin"
 
 server:
   type: custom
-  idleTimeout: 120 seconds
   applicationConnectors:
     - type: http
       port: 8060
+      idleTimeout: 120 seconds
+      acceptQueueSize: 2048
   adminConnectors:
     - type: http
       port: 8061
   registerDefaultExceptionMappers: false
   maxThreads: 2048
   maxQueuedRequests: 2048
-  acceptQueueSize: 2048
+
   requestLog:
     appenders: []
 

src/main/java/org/grobid/core/analyzers/DatastetAnalyzer.java

@@ -1,7 +1,7 @@
 package org.grobid.core.analyzers;
 
-import org.grobid.core.layout.LayoutToken;
 import org.grobid.core.lang.Language;
+import org.grobid.core.layout.LayoutToken;
 
 import java.util.ArrayList;
 import java.util.List;
@@ -21,8 +21,8 @@ public static DatastetAnalyzer getInstance() {
         if (instance == null) {
             //double check idiom
             // synchronized (instanceController) {
-                if (instance == null)
-                    getNewInstance();
+            if (instance == null)
+                getNewInstance();
             // }
         }
         return instance;
@@ -111,7 +111,7 @@ public List<LayoutToken> retokenizeLayoutTokens(List<LayoutToken> tokens) {
         return result;
     }
 
-     public List<LayoutToken> tokenize(LayoutToken chunk) {
+    public List<LayoutToken> tokenize(LayoutToken chunk) {
         List<LayoutToken> result = new ArrayList<>();
         String text = chunk.getText();
         StringTokenizer st = new StringTokenizer(text, DELIMITERS, true);
@@ -127,7 +127,7 @@ public List<LayoutToken> tokenize(LayoutToken chunk) {
         }
 
         return result;
-    } 
+    }
 
     public List<LayoutToken> retokenizeSubdigitsFromLayoutToken(List<LayoutToken> tokens) {
         // already done by this analyzer

src/main/java/org/grobid/core/data/BiblioComponent.java

@@ -1,27 +1,19 @@
 package org.grobid.core.data;
 
-import org.grobid.core.engines.label.TaggingLabel;
-import org.grobid.core.utilities.TextUtilities;
-import org.grobid.core.utilities.OffsetPosition;
-import org.grobid.core.lexicon.DatastetLexicon;
-import org.grobid.core.layout.BoundingBox;
-import org.grobid.core.layout.LayoutToken;
-
 import com.fasterxml.jackson.core.JsonProcessingException;
 import com.fasterxml.jackson.databind.ObjectMapper;
-
-import java.util.List;
-
+import org.grobid.core.layout.BoundingBox;
+import org.grobid.core.utilities.TextUtilities;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 /**
- *  Representation of the bibliographical reference element for a software mention.
- *  The component represent the reference callout (position) and its matched full  
- *  bibliographical reference.
- * 
- *  The bibliographical reference can also be disambiguated against wikidata via
- *  its (inherited) KnowledgeEntity object attributes. 
+ * Representation of the bibliographical reference element for a software mention.
+ * The component represent the reference callout (position) and its matched full
+ * bibliographical reference.
+ * <p>
+ * The bibliographical reference can also be disambiguated against wikidata via
+ * its (inherited) KnowledgeEntity object attributes.
  */
 public class BiblioComponent extends DatasetComponent {
     private static final Logger logger = LoggerFactory.getLogger(BiblioComponent.class);
@@ -62,10 +54,10 @@ public int getRefKey() {
 
     public String toJson() {
         ObjectMapper mapper = new ObjectMapper();
-        
+
         StringBuffer buffer = new StringBuffer();
         buffer.append("{ ");
-                
+
         try {
             buffer.append("\"label\" : " + mapper.writeValueAsString(rawForm));
         } catch (JsonProcessingException e) {
@@ -82,7 +74,7 @@ public String toJson() {
             }
         }*/
         buffer.append(", \"refKey\": " + refKey);
-        
+
         // knowledge information
         if (wikidataId != null) {
             buffer.append(", \"wikidataId\": \"" + wikidataId + "\"");
@@ -99,10 +91,10 @@ public String toJson() {
 
         if (offsets != null) {
             buffer.append(", \"offsetStart\" : " + offsets.start);
-            buffer.append(", \"offsetEnd\" : " + offsets.end);  
+            buffer.append(", \"offsetEnd\" : " + offsets.end);
         }
 
-        if ( (boundingBoxes != null) && (boundingBoxes.size() > 0) ) {
+        if ((boundingBoxes != null) && (boundingBoxes.size() > 0)) {
             buffer.append(", \"boundingBoxes\" : [");
             boolean first = true;
             for (BoundingBox box : boundingBoxes) {
@@ -114,7 +106,7 @@ public String toJson() {
             }
             buffer.append("] ");
         }
-        
+
         buffer.append(" }");
         return buffer.toString();
     }