Merge pull request #457 from wcmc-its/development

Prepare for release of 2.1.0

Dockerfile

@@ -1,4 +1,4 @@
-FROM openjdk:11-jre-slim
+FROM adoptopenjdk/openjdk11:alpine-jre
 
 RUN mkdir -p /app
 WORKDIR /app

README.md

@@ -54,9 +54,11 @@ Please see the [ReCiter wiki](https://github.com/wcmc-its/ReCiter/wiki) for more
 ## Technical 
 
 ### Prerequisites
-- Java 1.8
+- Java 11
 - Latest version of Maven. To install Maven navigate to the directory where ReCiter will be installed, execute `brew install maven` and then `mvn clean install`
 
+If you want to use Java 8 then update `<java.version>1.8</java.version>` in [pom.xml](https://github.com/wcmc-its/ReCiter/blob/a3d5d4665e8692853ca69f2db0caba0eb56f557d/pom.xml#L49)
+
 It is not necessary to install ReCiter in order to use the API.
 
 ### Technological stack
@@ -116,7 +118,7 @@ export SCOPUS_SERVICE=http://localhost:5001
 export PUBMED_SERVICE=http://localhost:5002
 ```
 4. Run `mvn spring-boot:run`. You can add additional options if you want like max and min java memory with `export MAVEN_OPTS=-Xmx1024m`
-5. Go to `http://localhost:<port-number>/swagger-ui.html` to test and run any API.
+5. Go to `http://localhost:<port-number>/swagger-ui/index.html` or `http://localhost:<port-number>/swagger-ui/` (shorthand swagger url)  to test and run any API.
 
 
 ### Amazon AWS

kubernetes/k8-deployment.yaml

@@ -79,14 +79,14 @@ spec:
               name: ENVCONFIG
               key: AWS_REGION
         - name: JAVA_OPTIONS
-          value: "-Xmx2000m"
+          value: "-Xmx2400m"
         ports:
         - containerPort: 5000
           name: reciter
         resources:
           requests:
             memory: 2.6G
-            cpu: 0.8
+            cpu: 1
           limits:
             memory: 2.8G
             cpu: 1

pom.xml

@@ -299,7 +299,7 @@
 	    <dependency>
 	    	<groupId>edu.cornell.weill.reciter</groupId>
 	    	<artifactId>reciter-identity-model</artifactId>
-	    	<version>2.0.7</version>
+	    	<version>2.0.8</version>
 	    </dependency>
 	    <dependency>
 	    	<groupId>edu.cornell.weill.reciter</groupId>
@@ -309,7 +309,7 @@
 	    <dependency>
 	    	<groupId>edu.cornell.weill.reciter</groupId>
 	    	<artifactId>reciter-dynamodb-model</artifactId>
-	    	<version>2.0.6</version>
+	    	<version>2.0.8</version>
 	    </dependency>
 	    <dependency>
 	    	<groupId>com.github.bohnman</groupId>

src/main/java/reciter/algorithm/evidence/targetauthor/articlesize/strategy/ArticleSizeStrategy.java

@@ -18,16 +18,18 @@
  *******************************************************************************/
 package reciter.algorithm.evidence.targetauthor.articlesize.strategy;
 
-import static org.hamcrest.CoreMatchers.sameInstance;
-
+import java.util.HashSet;
 import java.util.List;
+import java.util.Set;
 
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
 import reciter.ApplicationContextHolder;
 import reciter.algorithm.cluster.article.scorer.ReCiterArticleScorer;
 import reciter.algorithm.evidence.targetauthor.AbstractTargetAuthorStrategy;
+import reciter.database.dynamodb.model.ESearchPmid;
+import reciter.database.dynamodb.model.ESearchPmid.RetrievalRefreshFlag;
 import reciter.database.dynamodb.model.ESearchResult;
 import reciter.database.dynamodb.model.QueryType;
 import reciter.engine.Feature;
@@ -38,6 +40,10 @@
 import reciter.model.identity.Identity;
 import reciter.service.ESearchResultService;
 
+/**
+ * @author Sarbajit Dutta
+ * @see <a href= "https://github.com/wcmc-its/ReCiter/issues/228">Article Count Strategy</a>
+ */
 public class ArticleSizeStrategy extends AbstractTargetAuthorStrategy {
 
 	private static final Logger slf4jLogger = LoggerFactory.getLogger(ArticleSizeStrategy.class);
@@ -97,19 +103,37 @@ public double executeStrategy(ReCiterArticle reCiterArticle, Identity identity)
 
 	@Override
 	public double executeStrategy(List<ReCiterArticle> reCiterArticles, Identity identity) {
-
+		//Logic was updated to include lookupType in eSearchPmid so as to only count publications when
+		//ALL_PUBLICATIONS flag is being used and excluding GoldStandard Strategy retrieval type for actual candidate publication count
+		//https://github.com/wcmc-its/ReCiter/issues/455 
 		ESearchResultService eSearchResultService = ApplicationContextHolder.getContext().getBean(ESearchResultService.class);
 		ESearchResult eSearchResult = eSearchResultService.findByUid(identity.getUid());
 		reCiterArticles.forEach(reCiterArticle -> {
 		ArticleCountEvidence articleCountEvidence = new ArticleCountEvidence();
 		if(this.numberOfArticles > 0) {
+			int retrievalArticleCountByLookUpType = 0;
 			if(eSearchResult != null
 					&&
 					eSearchResult.getQueryType() != null 
 					&&
 					(eSearchResult.getQueryType() == QueryType.LENIENT_LOOKUP || eSearchResult.getQueryType() == QueryType.STRICT_COMPOUND_NAME_LOOKUP)) {
-				articleCountEvidence.setCountArticlesRetrieved(this.numberOfArticles);
-				articleCountEvidence.setArticleCountScore(-(this.numberOfArticles - ReCiterArticleScorer.strategyParameters.getArticleCountThresholdScore())/ReCiterArticleScorer.strategyParameters.getArticleCountWeight());
+				if(eSearchResult.getESearchPmids() != null
+						&&
+						!eSearchResult.getESearchPmids().isEmpty()) {
+							Set<Long> uniqueRetrievalArticle = new HashSet<>();
+							eSearchResult.getESearchPmids().stream()
+							.filter(eSearchPmid -> !eSearchPmid.getRetrievalStrategyName().equalsIgnoreCase("GoldStandardRetrievalStrategy") && eSearchPmid.getLookupType() == RetrievalRefreshFlag.ALL_PUBLICATIONS)
+							.map(ESearchPmid::getPmids)
+							.forEach(uniqueRetrievalArticle::addAll);
+							retrievalArticleCountByLookUpType = uniqueRetrievalArticle.size();
+							if(retrievalArticleCountByLookUpType > 0) {
+								articleCountEvidence.setCountArticlesRetrieved(retrievalArticleCountByLookUpType);
+								articleCountEvidence.setArticleCountScore(-(retrievalArticleCountByLookUpType - ReCiterArticleScorer.strategyParameters.getArticleCountThresholdScore())/ReCiterArticleScorer.strategyParameters.getArticleCountWeight());
+							} else {
+								articleCountEvidence.setCountArticlesRetrieved(this.numberOfArticles);
+								articleCountEvidence.setArticleCountScore(-(this.numberOfArticles - ReCiterArticleScorer.strategyParameters.getArticleCountThresholdScore())/ReCiterArticleScorer.strategyParameters.getArticleCountWeight());
+							}
+						}
 			} else if(eSearchResult != null
 					&&
 					eSearchResult.getQueryType() != null 

src/main/java/reciter/pubmed/retriever/PubMedQuery.java

@@ -40,9 +40,9 @@ public String toString() {
             //parts.add(author + " [au]");
         	parts.add(author);
         }
-        //Switched from [DP] to [EDAT] for better capture of pubs : Date of publication - Date added to Entrez
+        //Added both [DP] and [EDAT] for better capture of pubs : Date of publication - Date added to Entrez
         if (start != null && end != null) {
-            parts.add(dt.format(start) + ":" + dt.format(end) + "[EDAT]");
+            parts.add("((" + dt.format(start) + ":" + dt.format(end) + "[EDAT]" + ") OR (" + dt.format(start) + ":" + dt.format(end) + "[DP]))");
         }
         if (strategyQuery != null && !strategyQuery.isEmpty()) {
             parts.add(strategyQuery);

src/main/java/reciter/xml/retriever/engine/AbstractReCiterRetrievalEngine.java

@@ -26,6 +26,7 @@
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.stereotype.Component;
 
+import reciter.api.parameters.RetrievalRefreshFlag;
 import reciter.database.dynamodb.model.ESearchPmid;
 import reciter.database.dynamodb.model.ESearchResult;
 import reciter.database.dynamodb.model.QueryType;
@@ -96,7 +97,7 @@ public abstract class AbstractReCiterRetrievalEngine implements ReCiterRetrieval
 	 * @param pubMedArticles
 	 * @param uid
 	 */
-	protected void savePubMedArticles(Collection<PubMedArticle> pubMedArticles, String uid, String retrievalStrategyName, List<PubMedQueryResult> pubMedQueryResults, QueryType queryType) {
+	protected void savePubMedArticles(Collection<PubMedArticle> pubMedArticles, String uid, String retrievalStrategyName, List<PubMedQueryResult> pubMedQueryResults, QueryType queryType, RetrievalRefreshFlag refreshFlag) {
 		// Save the articles.
 		List<PubMedArticle> pubMedArticleList = new ArrayList<>(pubMedArticles);
 		pubMedService.save(pubMedArticleList);
@@ -107,24 +108,32 @@ protected void savePubMedArticles(Collection<PubMedArticle> pubMedArticles, Stri
 			pmids.add(pubMedArticle.getMedlinecitation().getMedlinecitationpmid().getPmid());
 		}
 		ESearchPmid eSearchPmid = null;
-		if(pmids.size() > 0) {
-			eSearchPmid = new ESearchPmid(pmids, retrievalStrategyName, new Date());
+		if(!pmids.isEmpty()){
+			reciter.database.dynamodb.model.ESearchPmid.RetrievalRefreshFlag eSearchPmidRefreshFlag;
+			if(refreshFlag == RetrievalRefreshFlag.ALL_PUBLICATIONS) {
+				eSearchPmidRefreshFlag = reciter.database.dynamodb.model.ESearchPmid.RetrievalRefreshFlag.ALL_PUBLICATIONS;
+			} else if(refreshFlag == RetrievalRefreshFlag.ONLY_NEWLY_ADDED_PUBLICATIONS) {
+				eSearchPmidRefreshFlag = reciter.database.dynamodb.model.ESearchPmid.RetrievalRefreshFlag.ONLY_NEWLY_ADDED_PUBLICATIONS;
+			} else {
+				eSearchPmidRefreshFlag = reciter.database.dynamodb.model.ESearchPmid.RetrievalRefreshFlag.FALSE;
+			}
+			eSearchPmid = new ESearchPmid(pmids, retrievalStrategyName, new Date(), eSearchPmidRefreshFlag);
 		}
 		ESearchResult eSearchResultDb = eSearchResultService.findByUid(uid);
 		if (eSearchResultDb == null) {
 			List<ESearchPmid> eSearchPmids = new ArrayList<>();
 			if(eSearchPmid != null) {
 				eSearchPmids.add(eSearchPmid);
 			}
-			if(eSearchPmids.size() > 0) {
+			if(!eSearchPmids.isEmpty()) {
 				eSearchResultService.save(new ESearchResult(uid, new Date(), eSearchPmids, queryType));
 			}
 		} else {
 			List<ESearchPmid> eSearchPmids = eSearchResultDb.getESearchPmids();
 			if(eSearchPmid != null) {
 				eSearchPmids.add(eSearchPmid);
 			}
-			if(eSearchPmids.size() > 0) {
+			if(!eSearchPmids.isEmpty()) {
 				eSearchResultService.save(new ESearchResult(uid, new Date(), eSearchPmids, queryType));
 			} else {
 				eSearchResultDb.setRetrievalDate(new Date());