Consolidation.java

package org.grobid.core.utilities;

import com.rockymadden.stringmetric.similarity.RatcliffObershelpMetric;
import org.apache.commons.collections4.CollectionUtils;
import org.apache.commons.lang3.StringUtils;
import org.apache.commons.lang3.exception.ExceptionUtils;
import org.grobid.core.data.BibDataSet;
import org.grobid.core.data.BiblioItem;
import org.grobid.core.data.Funder;
import org.grobid.core.utilities.counters.CntManager;
import org.grobid.core.utilities.crossref.CrossrefClient;
import org.grobid.core.utilities.crossref.CrossrefRequestListener;
import org.grobid.core.utilities.crossref.WorkDeserializer;
import org.grobid.core.utilities.crossref.FunderDeserializer;
import org.grobid.core.utilities.glutton.GluttonClient;
import org.grobid.core.utilities.TextUtilities;
import org.slf4j.Logger;
import org.slf4j.LoggerFactory;
import scala.Option;

import java.util.*;

import static org.apache.commons.lang3.StringUtils.isNotEmpty;
import static org.grobid.core.data.BiblioItem.cleanDOI;

/**
 * Singleton class for managing the extraction of bibliographical information from pdf documents.
 * When consolidation operations are realized, be sure to call the close() method
 * to ensure that all Executors are terminated.
 *
 */
public class Consolidation {
    private static final Logger LOGGER = LoggerFactory.getLogger(Consolidation.class);

    private static volatile Consolidation instance;

    private CrossrefClient client = null;
    private WorkDeserializer workDeserializer = null;
    private FunderDeserializer funderDeserializer = null;
    private CntManager cntManager = null;

    public enum GrobidConsolidationService {
        CROSSREF("crossref"),
        GLUTTON("glutton");

        private final String ext;

        GrobidConsolidationService(String ext) {
            this.ext = ext;
        }

        public String getExt() {
            return ext;
        }

        public static GrobidConsolidationService get(String name) {
            if (name == null) {
                throw new IllegalArgumentException("Name of consolidation service must not be null");
            }

            String n = name.toLowerCase();
            for (GrobidConsolidationService e : values()) {
                if (e.name().toLowerCase().equals(n)) {
                    return e;
                }
            }
            throw new IllegalArgumentException("No consolidation service with name '" + name +
                    "', possible values are: " + Arrays.toString(values()));
        }
    }

    public static Consolidation getInstance() {
        if (instance == null) {
            getNewInstance();
        }
        return instance;
    }

    /**
     * Creates a new instance.
     */
    private static synchronized void getNewInstance() {
        LOGGER.debug("Get new instance of Consolidation");
        instance = new Consolidation();
    }

    /**
     * Hidden constructor
     */
    private Consolidation() {
        if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.GLUTTON)
            client = GluttonClient.getInstance();
        else
            client = CrossrefClient.getInstance();
        workDeserializer = new WorkDeserializer();
        funderDeserializer = new FunderDeserializer();
    }

    public void setCntManager(CntManager cntManager) {
        this.cntManager = cntManager;
    }

    public CntManager getCntManager() {
        return this.cntManager;
    }

    /**
     * After consolidation operations, this need to be called to ensure that all
     * involved Executors are shut down immediatly, otherwise non terminated thread
     * could prevent the JVM from exiting
     */
    public void close() {
        //client.close();
    }

    /**
     * Try to consolidate one bibliographical object with crossref metadata lookup web services based on
     * core metadata. In practice, this method is used for consolidating header metadata.  
     */
    public BiblioItem consolidate(BiblioItem bib, String rawCitation, int consolidateMode) throws Exception {
        final List<BiblioItem> results = new ArrayList<>();

        String theDOI = bib.getDOI();
        if (StringUtils.isNotBlank(theDOI)) {
            theDOI = cleanDOI(theDOI);
        }
        final String doi = theDOI;
        String halId = bib.getHalId();
        String aut = bib.getFirstAuthorSurname();
        String title = bib.getTitle();
        String journalTitle = bib.getJournal();
        String volume = bib.getVolume();
        if (StringUtils.isBlank(volume))
            volume = bib.getVolumeBlock();

        String firstPage = null;
        String pageRange = bib.getPageRange();
        int beginPage = bib.getBeginPage();
        if (beginPage != -1) {
            firstPage = "" + beginPage;
        } else if (pageRange != null) {
            StringTokenizer st = new StringTokenizer(pageRange, "--");
            if (st.countTokens() == 2) {
                firstPage = st.nextToken();
            } else if (st.countTokens() == 1)
                firstPage = pageRange;
        }

        String year = null;
        if ( bib.getNormalizedPublicationDate() != null ) {
            year = "" + bib.getNormalizedPublicationDate().getYear();
        }
        if (year == null)
            year = bib.getYear();

        if (cntManager != null)
            cntManager.i(ConsolidationCounters.CONSOLIDATION);

        long threadId = Thread.currentThread().getId();
        Map<String, String> arguments = null;

        if (StringUtils.isNotBlank(doi)) {
            // call based on the identified DOI
            arguments = new HashMap<String,String>();
            arguments.put("doi", doi);
        } else if (consolidateMode != 3) {
            if (StringUtils.isNotBlank(rawCitation)) {
                // call with full raw string            
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("query.bibliographic", rawCitation);
                }
            }
            if (StringUtils.isNotBlank(halId)) {
                // call based on the identified HAL ID
                if (arguments == null)
                    arguments = new HashMap<String,String>();
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF)
                    arguments.put("halid", halId);
            }
            if (StringUtils.isNotBlank(aut)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("query.author", aut);
                }
            }
            if (StringUtils.isNotBlank(title)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("query.title", title);
                }
            }
            if (StringUtils.isNotBlank(journalTitle)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("query.container-title", journalTitle);
                }
            }
            if (StringUtils.isNotBlank(volume)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("volume", volume);
                }
            }
            if (StringUtils.isNotBlank(firstPage)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("firstPage", firstPage);
                }
            }
            if (StringUtils.isNotBlank(year)) {
                // publication year metadata, CrossRef has no year query field, they are supported by the query.bibliographic 
                // field and filter
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("year", year);
                }
            }
        }
        if (arguments == null || arguments.size() == 0) {
            return null;
        }

        if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF) {
            if (StringUtils.isBlank(doi) && StringUtils.isBlank(rawCitation) &&
                 (StringUtils.isBlank(aut) || StringUtils.isBlank(title)) ) {
                // there's not enough information for a crossref request, which might always return a result
                return null;
            }
        }

        if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF) {
            arguments.put("rows", "1"); // we just request the top-one result
        } else if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.GLUTTON) {
            // GROBID has already parsed the reference, so no need to redo this in glutton
            arguments.put("parseReference", "false");
        }

        final boolean doiQuery;
        try {
            //CrossrefRequestListener<BiblioItem> requestListener = new CrossrefRequestListener<BiblioItem>();
            if (cntManager != null) {
                cntManager.i(ConsolidationCounters.CONSOLIDATION);
            }

            if ( StringUtils.isNotBlank(doi) && (cntManager != null) ) {
                cntManager.i(ConsolidationCounters.CONSOLIDATION_PER_DOI);
                doiQuery = true;
            } else {
                doiQuery = false;
            }

            client.pushRequest("works", arguments, workDeserializer, threadId, new CrossrefRequestListener<BiblioItem>(0) {

                @Override
                public void onSuccess(List<BiblioItem> res) {
                    if ((res != null) && (res.size() > 0) ) {
                        // we need here to post-check that the found item corresponds
                        // correctly to the one requested in order to avoid false positive
                        for(BiblioItem oneRes : res) {
                            /* 
                              Glutton integrates its own post-validation, so we can skip post-validation in GROBID when it is used as 
                              consolidation service.  

                              In case of crossref REST API, for single bib. ref. consolidation (this case comes only for header extraction), 
                              having an extracted DOI matching is considered safe enough, and we don't require further post-validation.

                              For all the other case of matching with CrossRef, we require a post-validation. 
                            */
                            if ((GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.GLUTTON) 
                                ||
                                ( (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF) &&
                                  doiQuery )
                                ||
                                ( (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF) &&
                                   postValidation(bib, oneRes)) 
                               ) {
                                results.add(oneRes);
                                if (cntManager != null) {
                                    cntManager.i(ConsolidationCounters.CONSOLIDATION_SUCCESS);
                                    if (doiQuery)
                                        cntManager.i(ConsolidationCounters.CONSOLIDATION_PER_DOI_SUCCESS);
                                }
                                break;
                            }
                        }
                    }
                }

                @Override
                public void onError(int status, String message, Exception exception) {
                    LOGGER.info("Consolidation service returns error ("+status+") : "+message, exception);
                }
            });
        } catch(Exception e) {
            LOGGER.info("Consolidation error - ", e);
        }

        client.finish(threadId);
        if (results.size() == 0)
            return null;
        else
            return results.get(0);
    }


    /**
     * Try tp consolidate a list of bibliographical objects in one operation with consolidation services.
     * In practice this method is used for consolidating the metadata of all the extracted bibliographical 
     * references. 
     */
    public Map<Integer,BiblioItem> consolidate(List<BibDataSet> biblios) {
        if (CollectionUtils.isEmpty(biblios))
            return null;
        final Map<Integer,BiblioItem> results = new HashMap<Integer,BiblioItem>();
        // init the results
        int n = 0;
        for(n=0; n<biblios.size(); n++) {
            results.put(Integer.valueOf(n), null);
        }
        n = 0;
        long threadId = Thread.currentThread().getId();
        for(BibDataSet bibDataSet : biblios) {
            final BiblioItem theBiblio = bibDataSet.getResBib();

            if (cntManager != null)
                cntManager.i(ConsolidationCounters.TOTAL_BIB_REF);

            // first we get the exploitable metadata
            String doi = theBiblio.getDOI();
            if (StringUtils.isNotBlank(doi)) {
                doi = BiblioItem.cleanDOI(doi);
            }
            // first we get the exploitable metadata
            String halId = theBiblio.getHalId();
            String aut = theBiblio.getFirstAuthorSurname();
            String title = theBiblio.getTitle();
            String journalTitle = theBiblio.getJournal();

            // and the row string
            String rawCitation = bibDataSet.getRawBib();

            Map<String, String> arguments = null;

            String volume = theBiblio.getVolume();
            if (StringUtils.isBlank(volume))
                volume = theBiblio.getVolumeBlock();

            String firstPage = null;
            String pageRange = theBiblio.getPageRange();
            int beginPage = theBiblio.getBeginPage();
            if (beginPage != -1) {
                firstPage = "" + beginPage;
            } else if (pageRange != null) {
                StringTokenizer st = new StringTokenizer(pageRange, "--");
                if (st.countTokens() == 2) {
                    firstPage = st.nextToken();
                } else if (st.countTokens() == 1)
                    firstPage = pageRange;
            }

            String year = null;
            if ( theBiblio.getNormalizedPublicationDate() != null ) {
                year = "" + theBiblio.getNormalizedPublicationDate().getYear();
            }
            if (year == null)
                year = theBiblio.getYear();

            /*if (aut != null) {
                aut = TextUtilities.removeAccents(aut);
            }
            if (title != null) {
                title = TextUtilities.removeAccents(title);
            }
            if (journalTitle != null) {
                journalTitle = TextUtilities.removeAccents(journalTitle);
            }*/

            if (StringUtils.isNotBlank(doi)) {
                // call based on the identified DOI
                arguments = new HashMap<String,String>();
                arguments.put("doi", doi);
            }
            if (StringUtils.isNotBlank(halId)) {
                // call based on the identified HAL ID
                if (arguments == null)
                    arguments = new HashMap<String,String>();
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF)
                    arguments.put("halid", halId);
            }
            if (StringUtils.isNotBlank(rawCitation)) {
                // call with full raw string
                if (arguments == null)
                    arguments = new HashMap<String,String>();
                if ( (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) ||
                     StringUtils.isBlank(doi) )
                    arguments.put("query.bibliographic", rawCitation);
            }
            if (StringUtils.isNotBlank(title)) {
                // call based on partial metadata
                if (arguments == null)
                    arguments = new HashMap<String,String>();
                if ( (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) ||
                     (StringUtils.isBlank(rawCitation) && StringUtils.isBlank(doi)) )
                    arguments.put("query.title", title);
            }
            if (StringUtils.isNotBlank(aut)) {
                // call based on partial metadata
                if (arguments == null)
                    arguments = new HashMap<String,String>();
                if ( (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) ||
                     (StringUtils.isBlank(rawCitation) && StringUtils.isBlank(doi)) )
                    arguments.put("query.author", aut);
            }
            if (StringUtils.isNotBlank(journalTitle)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("query.container-title", journalTitle);
                }
            }
            if (StringUtils.isNotBlank(volume)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("volume", volume);
                }
            }
            if (StringUtils.isNotBlank(firstPage)) {
                // call based on partial metadata
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("firstPage", firstPage);
                }
            }
            if (StringUtils.isNotBlank(year)) {
                // publication year metadata, CrossRef has no year query field, they are supported by the query.bibliographic 
                // field and filter
                if (GrobidProperties.getInstance().getConsolidationService() != GrobidConsolidationService.CROSSREF) {
                    if (arguments == null)
                        arguments = new HashMap<String,String>();
                    arguments.put("year", year);
                }
            }

            if (arguments == null || arguments.size() == 0) {
                n++;
                continue;
            }

            if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF) {
                if (StringUtils.isBlank(doi) && StringUtils.isBlank(rawCitation) &&
                     (StringUtils.isBlank(aut) || StringUtils.isBlank(title)) ) {
                    // there's not enough information for a crossref request, which might always return a result
                    n++;
                    continue;
                }
            }

            if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.CROSSREF)
                arguments.put("rows", "1"); // we just request the top-one result
            else if (GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.GLUTTON) {
                // GROBID has already parsed the reference, so no need to redo this in glutton
                arguments.put("parseReference", "false");
            }

            final boolean doiQuery;
            try {
                //CrossrefRequestListener<BiblioItem> requestListener = new CrossrefRequestListener<BiblioItem>();
                if (cntManager != null) {
                    cntManager.i(ConsolidationCounters.CONSOLIDATION);
                }

                if ( StringUtils.isNotBlank(doi) && (cntManager != null) ) {
                    cntManager.i(ConsolidationCounters.CONSOLIDATION_PER_DOI);
                    doiQuery = true;
                } else {
                    doiQuery = false;
                }

                client.<BiblioItem>pushRequest("works", arguments, workDeserializer, threadId, new CrossrefRequestListener<BiblioItem>(n) {

                    @Override
                    public void onSuccess(List<BiblioItem> res) {
                        if ((res != null) && (res.size() > 0) ) {
                            // for CrossRef API we need here to post-validate if the found item corresponds
                            // to the one requested in order to avoid false positive
                            // Glutton has its own validation mechanisms
                            for(BiblioItem oneRes : res) {
                                if ((GrobidProperties.getInstance().getConsolidationService() == GrobidConsolidationService.GLUTTON) ||
                                    postValidation(theBiblio, oneRes)) {
                                    oneRes.setLabeledTokens(theBiblio.getLabeledTokens());
                                    results.put(Integer.valueOf(getRank()), oneRes);
                                    if (cntManager != null) {
                                        cntManager.i(ConsolidationCounters.CONSOLIDATION_SUCCESS);
                                        if (doiQuery)
                                            cntManager.i(ConsolidationCounters.CONSOLIDATION_PER_DOI_SUCCESS);
                                    }
                                    break;
                                }
                            }
                        }
                    }

                    @Override
                    public void onError(int status, String message, Exception exception) {
                        LOGGER.info("Consolidation service returns error ("+status+") : "+message);
                    }
                });
            } catch(Exception e) {
                LOGGER.info("Consolidation error - ", e);
            }
            n++;
        }
        client.finish(threadId);

        return results;
    }

    /**
     * The public CrossRef API is a search API, and thus returns
     * many false positives. It is necessary to validate return results
     * against the (incomplete) source bibliographic item to block
     * inconsistent results.
     */
    private boolean postValidation(BiblioItem source, BiblioItem result) {
        boolean valid = true;

        // check main metadata available in source with fuzzy matching
        /*if (!StringUtils.isBlank(source.getTitle()) && !StringUtils.isBlank(source.getTitle())) {
//System.out.println(source.getTitle() + " / " + result.getTitle() + " = " + ratcliffObershelpDistance(source.getTitle(), result.getTitle(), false));      
            if (ratcliffObershelpDistance(source.getTitle(), result.getTitle(), false) < 0.8)
                return false;
        }*/

        if (!StringUtils.isBlank(source.getFirstAuthorSurname()) &&
            !StringUtils.isBlank(result.getFirstAuthorSurname())) {
//System.out.println(source.getFirstAuthorSurname() + " / " + result.getFirstAuthorSurname() + " = " + 
//    ratcliffObershelpDistance(source.getFirstAuthorSurname(), result.getFirstAuthorSurname(), false)); 
            if (ratcliffObershelpDistance(source.getFirstAuthorSurname(),result.getFirstAuthorSurname(), false) < 0.8)
                return false;
        }

        /*if (!StringUtils.isBlank(source.getPublicationDate()) && 
            !StringUtils.isBlank(result.getPublicationDate())) {
            if (!source.getPublicationDate().equals(result.getPublicationDate()))
                valid = false;
        }*/

        return valid;
    }

    private double ratcliffObershelpDistance(String string1, String string2, boolean caseDependent) {
        if ( StringUtils.isBlank(string1) || StringUtils.isBlank(string2) )
            return 0.0;
        Double similarity = 0.0;
        if (!caseDependent) {
            string1 = string1.toLowerCase();
            string2 = string2.toLowerCase();
        }
        if (string1.equals(string2)) {
            similarity = 1.0;
        }
        
        if ( isNotEmpty(string1) && isNotEmpty(string2) ) {
            Option<Object> similarityObject =
                RatcliffObershelpMetric.compare(string1, string2);
            if (similarityObject.isDefined()) {
                similarity = (Double) similarityObject.get();
            }
        }

        return similarity;
    }

    public Funder consolidateFunder(Funder funder) {
        final List<Funder> results = new ArrayList<>();

        Map<String, String> arguments = new HashMap<String,String>();

        // CrossRef does not manage stopwords in funder search and has no usable term frequency, so we need
        // to remove basic stopwords in the query to have something manageable from CrossRef
        String funderNameString = funder.getFullName();
        if (funderNameString == null || funderNameString.length() == 0)
            return null;

        funderNameString = TextUtilities.removeFieldStopwords(funderNameString);

        arguments.put("query", funderNameString);
        arguments.put("rows", "10"); // we request the top-10 results, because there are a lot of noise
        // and we need many candidates in the pairwise comparison step

        long threadId = Thread.currentThread().getId();

        try {
            client.pushRequest("funders", arguments, funderDeserializer, threadId, new CrossrefRequestListener<Funder>(0) {
                @Override
                public void onSuccess(List<Funder> res) {
                    if ((res != null) && (res.size() > 0) ) {
                        // we need here to post-check the candidates in a pairwise comparison 
                        // in order to avoid false positive
                        for(Funder oneRes : res) {
                            /* 
                              Glutton integrates its own post-validation, so we can skip post-validation in GROBID when it is used as 
                              consolidation service. However, with CrossRef, post-validation is mandatory to control false positives.  
                            */
                            if (oneRes.getFullName() != null) {
                                String localFullName = oneRes.getFullName();
                                localFullName = TextUtilities.removeFieldStopwords(localFullName);
                                //if (localFullName.toLowerCase().equals(arguments.get("query").toLowerCase())) {
                                if (ratcliffObershelpDistance(localFullName, arguments.get("query"), false)>0.9) {
                                    results.add(oneRes);
                                }
                            }
                        }
                    }
                }

                @Override
                public void onError(int status, String message, Exception exception) {
                    LOGGER.info("Funder consolidation service returns error ("+status+") : "+message, exception);
                }
            });
        } catch(Exception e) {
            LOGGER.info("Funder consolidation error - ", e);
        }
        client.finish(threadId);
        if (results.size() == 0)
            return null;
        else
            return results.get(0);
    }

    public Map<Integer,Funder> consolidateFunders(List<Funder> funders) {
        if (CollectionUtils.isEmpty(funders))
            return null;
        final Map<Integer,Funder> results = new HashMap<Integer,Funder>();
        // init the results
        int n = 0;
        for(n=0; n<funders.size(); n++) {
            results.put(n, null);
        }
        n = 0;
        long threadId = Thread.currentThread().getId();
        for(Funder funder : funders) {
            //final List<Funder> results = new ArrayList<>();

            Map<String, String> arguments = new HashMap<String,String>();

            // CrossRef does not manage stopwords in funder search and has no usable term frequency, so we need
            // to remove basic stopwords in the query to have something manageable from CrossRef
            String funderNameString = funder.getFullName();
            if (funderNameString == null || funderNameString.length() == 0)
                return null;

            funderNameString = TextUtilities.removeFieldStopwords(funderNameString);

            arguments.put("query", funderNameString);
            arguments.put("rows", "10"); // we request the top-10 results, because there are a lot of noise
            // and we need many candidates in the pairwise comparison step

            try {
                client.pushRequest("funders", arguments, funderDeserializer, threadId, new CrossrefRequestListener<Funder>(n) {
                    @Override
                    public void onSuccess(List<Funder> res) {
                        List<Funder> localResults = new ArrayList<>();
                        if ((res != null) && (res.size() > 0) ) {
                            // we need here to post-check the candidates in a pairwise comparison 
                            // in order to avoid false positive
                            for(Funder oneRes : res) {
                                /* 
                                  Glutton integrates its own post-validation, so we can skip post-validation in GROBID when it is used as 
                                  consolidation service. However, with CrossRef, post-validation is mandatory to control false positives.  
                                */
                                if (oneRes.getFullName() != null) {
                                    String localFullName = oneRes.getFullName();
                                    localFullName = TextUtilities.removeFieldStopwords(localFullName);
                                    if (localFullName.toLowerCase().equals(arguments.get("query").toLowerCase())) {
                                        localResults.add(oneRes);
                                        break;
                                    } else if (ratcliffObershelpDistance(localFullName, arguments.get("query"), false)>0.9) {
                                        localResults.add(oneRes);
                                    }
                                }
                            }

                            if (localResults.size() >0)
                                results.put(Integer.valueOf(getRank()), localResults.get(0));
                        } 
                    }

                    @Override
                    public void onError(int status, String message, Exception exception) {
                        LOGGER.info("Funder consolidation service returns error ("+status+") : "+message, exception);
                    }
                });
            } catch(Exception e) {
                LOGGER.info("Funder consolidation error - ", e);
            }
            n++;
        }

        client.finish(threadId);
        return results;
    }

}