diff --git a/grobid-core/src/main/java/org/grobid/core/data/Figure.java b/grobid-core/src/main/java/org/grobid/core/data/Figure.java
index ef4117e93c..0d646ed93d 100644
--- a/grobid-core/src/main/java/org/grobid/core/data/Figure.java
+++ b/grobid-core/src/main/java/org/grobid/core/data/Figure.java
@@ -323,8 +323,12 @@ public String getTeiId() {
         return "fig_" + this.id;
     }
 
+    public boolean isCompleteForTEI() {
+        return (StringUtils.isNotBlank(header) || StringUtils.isNotBlank(caption) || CollectionUtils.isNotEmpty(graphicObjects));
+    }
+
     public String toTEI(GrobidAnalysisConfig config, Document doc, TEIFormatter formatter, List<MarkerType> markerTypes) {
-        if (StringUtils.isEmpty(header) && StringUtils.isEmpty(caption) && CollectionUtils.isEmpty(graphicObjects)) {
+        if (!isCompleteForTEI()) {
             return null;
         }
         Element figureElement = XmlBuilderUtils.teiElement("figure");
@@ -568,4 +572,5 @@ public void setLabel(StringBuilder label) {
     public void setUri(URI uri) {
         this.uri = uri;
     }
+
 }
diff --git a/grobid-core/src/main/java/org/grobid/core/data/Table.java b/grobid-core/src/main/java/org/grobid/core/data/Table.java
index 14d468418c..1016760284 100644
--- a/grobid-core/src/main/java/org/grobid/core/data/Table.java
+++ b/grobid-core/src/main/java/org/grobid/core/data/Table.java
@@ -30,7 +30,6 @@
 import nu.xom.Attribute;
 import nu.xom.Element;
 import nu.xom.Node;
-import nu.xom.Text;
 
 import static org.grobid.core.document.xml.XmlBuilderUtils.teiElement;
 import static org.grobid.core.document.xml.XmlBuilderUtils.addXmlId;
@@ -43,6 +42,7 @@
 public class Table extends Figure {
 	private List<LayoutToken> contentTokens = new ArrayList<>();
 	private List<LayoutToken> fullDescriptionTokens = new ArrayList<>();
+
 	private boolean goodTable = true;
 
     private StringBuilder note = null;
@@ -62,9 +62,13 @@ public Table() {
         note = new StringBuilder();
     }
 
+    public boolean isCompleteForTEI() {
+        return (StringUtils.isNotEmpty(header) && StringUtils.isNotEmpty(caption));
+    }
+
 	@Override
     public String toTEI(GrobidAnalysisConfig config, Document doc, TEIFormatter formatter, List<MarkerType> markerTypes) {
-		if (StringUtils.isEmpty(header) && StringUtils.isEmpty(caption)) {
+		if (!isCompleteForTEI()) {
 			return null;
 		}
 
@@ -104,7 +108,7 @@ public String toTEI(GrobidAnalysisConfig config, Document doc, TEIFormatter form
                 addXmlId(desc, "_" + divID);
             }
 
-            if ( (labeledCaption != null) && (labeledCaption.length() > 0) ) {
+            if (StringUtils.isNotBlank(labeledCaption)) {
                 TaggingTokenClusteror clusteror = new TaggingTokenClusteror(GrobidModels.FULLTEXT, labeledCaption, captionLayoutTokens);
                 List<TaggingTokenCluster> clusters = clusteror.cluster();                
                 for (TaggingTokenCluster cluster : clusters) {
@@ -169,7 +173,7 @@ public String toTEI(GrobidAnalysisConfig config, Document doc, TEIFormatter form
 		}
 
         Element noteNode = null;
-        if (note != null && note.toString().trim().length()>0) {
+        if (StringUtils.isNotBlank(note)) {
 
             noteNode = XmlBuilderUtils.teiElement("note");
             if (config.isGenerateTeiIds()) {
@@ -177,7 +181,7 @@ public String toTEI(GrobidAnalysisConfig config, Document doc, TEIFormatter form
                 addXmlId(noteNode, "_" + divID);
             }
 
-            if ( (labeledNote != null) && (labeledNote.length() > 0) ) {
+            if (StringUtils.isNotBlank(labeledNote)) {
                 TaggingTokenClusteror clusteror = new TaggingTokenClusteror(GrobidModels.FULLTEXT, labeledNote, noteLayoutTokens);
                 List<TaggingTokenCluster> clusters = clusteror.cluster();                
                 for (TaggingTokenCluster cluster : clusters) {
@@ -346,9 +350,14 @@ public String getLabeledNote() {
         return this.labeledNote;
     }
 
-	private boolean validateTable() {
+    /** Check if the table:
+     * - has label, header and content
+     * - header starts with "tab"
+     * - label can be parsed
+     */
+	public boolean validateTable() {
 		CntManager cnt = Engine.getCntManager();
-		if (StringUtils.isEmpty(label) || StringUtils.isEmpty(header) || StringUtils.isEmpty(content)) {
+		if (StringUtils.isAnyBlank(label, header, content)) {
 			cnt.i(TableRejectionCounters.EMPTY_LABEL_OR_HEADER_OR_CONTENT);
 			return false;
 		}
@@ -359,7 +368,8 @@ private boolean validateTable() {
 			cnt.i(TableRejectionCounters.CANNOT_PARSE_LABEL_TO_INT);
 			return false;
 		}
-		if (!getHeader().toLowerCase().startsWith("table")) {
+        // tab covers: table, tabelle, tableu, tabella, etc.
+		if (!StringUtils.startsWithIgnoreCase(getHeader(), "tab")) {
 			cnt.i(TableRejectionCounters.HEADER_NOT_STARTS_WITH_TABLE_WORD);
 			return false;
 		}
@@ -423,4 +433,5 @@ public boolean isGoodTable() {
     public String getTeiId() {
         return "tab_" + this.id;
     }
+
 }
\ No newline at end of file
diff --git a/grobid-core/src/main/java/org/grobid/core/document/Document.java b/grobid-core/src/main/java/org/grobid/core/document/Document.java
index d7dc90b08b..3925470a45 100755
--- a/grobid-core/src/main/java/org/grobid/core/document/Document.java
+++ b/grobid-core/src/main/java/org/grobid/core/document/Document.java
@@ -874,6 +874,7 @@ public static List<GraphicObject> getConnectedGraphics(Block block, Document doc
     public void postProcessTables() {
         for (Table table : tables) {
             if (!table.firstCheck()) {
+                table.setGoodTable(false);
                 continue;
             }
 
@@ -919,7 +920,7 @@ public void postProcessTables() {
             table.getContentTokens().clear();
             table.getContentTokens().addAll(contentResult);
 
-            table.secondCheck();
+            table.setGoodTable(table.secondCheck());
         }
     }
 
diff --git a/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java b/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java
index b102b20795..c94aca004c 100755
--- a/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java
+++ b/grobid-core/src/main/java/org/grobid/core/engines/FullTextParser.java
@@ -11,7 +11,6 @@
 
 import java.nio.charset.StandardCharsets;
 
-import org.apache.lucene.util.CollectionUtil;
 import org.grobid.core.GrobidModels;
 import org.grobid.core.data.*;
 import org.grobid.core.document.Document;
@@ -33,21 +32,16 @@
 import org.grobid.core.features.FeaturesVectorFulltext;
 import org.grobid.core.lang.Language;
 import org.grobid.core.lexicon.Lexicon;
-import org.grobid.core.lexicon.Lexicon.OrganizationRecord;
 import org.grobid.core.layout.*;
 import org.grobid.core.tokenization.TaggingTokenCluster;
 import org.grobid.core.tokenization.TaggingTokenClusteror;
-import org.grobid.core.utilities.LanguageUtilities;
-import org.grobid.core.utilities.TextUtilities;
-import org.grobid.core.utilities.KeyGen;
-import org.grobid.core.utilities.LayoutTokensUtil;
-import org.grobid.core.utilities.GrobidProperties;
-import org.grobid.core.utilities.Consolidation;
+import org.grobid.core.utilities.*;
 import org.grobid.core.utilities.matching.ReferenceMarkerMatcher;
 import org.grobid.core.utilities.matching.EntityMatcherException;
 import org.grobid.core.engines.citations.CalloutAnalyzer;
 import org.grobid.core.engines.citations.CalloutAnalyzer.MarkerType;
 
+import org.jetbrains.annotations.NotNull;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
 
@@ -66,10 +60,13 @@
 import java.util.StringTokenizer;
 import java.util.TreeSet;
 import java.util.regex.Matcher;
+import java.util.stream.Collectors;
+import java.util.stream.IntStream;
 
 import nu.xom.Element;
 
 import static org.apache.commons.lang3.StringUtils.*;
+import static org.grobid.core.utilities.LabelUtils.postProcessFullTextLabeledText;
 
 public class FullTextParser extends AbstractParser {
     private static final Logger LOGGER = LoggerFactory.getLogger(FullTextParser.class);
@@ -252,23 +249,25 @@ else if (config.getConsolidateCitations() == 2)
 
 			// full text processing
 			featSeg = getBodyTextFeatured(doc, documentBodyParts);
-			String resultBody = null;
-			LayoutTokenization layoutTokenization = null;
+			String bodyResults = null;
+			LayoutTokenization bodyLayoutTokens = null;
 			List<Figure> figures = null;
 			List<Table> tables = null;
 			List<Equation> equations = null;
 			if (featSeg != null && isNotBlank(featSeg.getLeft())) {
-				// if featSeg is null, it usually means that no body segment is found in the
+				// if featSeg is null, it usually means that the fulltext body is not found in the
 				// document segmentation
-				String bodytext = featSeg.getLeft();
-				layoutTokenization = featSeg.getRight();
+				String bodyText = featSeg.getLeft();
+				bodyLayoutTokens = featSeg.getRight();
 				//tokenizationsBody = featSeg.getB().getTokenization();
                 //layoutTokensBody = featSeg.getB().getLayoutTokens();
 
-                resultBody = label(bodytext);
+                bodyResults = label(bodyText);
+                //Correct subsequent I-<figure> or I-<table>
+                bodyResults = LabelUtils.postProcessFulltextFixInvalidTableOrFigure(bodyResults);
 
 				// we apply now the figure and table models based on the fulltext labeled output
-				figures = processFigures(resultBody, layoutTokenization.getTokenization(), doc);
+				figures = processFigures(bodyResults, bodyLayoutTokens.getTokenization(), doc);
                 // further parse the caption
                 for(Figure figure : figures) {
                     if (CollectionUtils.isNotEmpty(figure.getCaptionLayoutTokens()) ) {
@@ -277,8 +276,46 @@ else if (config.getConsolidateCitations() == 2)
                         figure.setCaptionLayoutTokens(captionProcess.getRight());
                     }
                 }
-                
-				tables = processTables(resultBody, layoutTokenization.getTokenization(), doc);
+
+                long numberFiguresFulltextModel = Arrays.stream(bodyResults.split("\n"))
+                    .filter(r -> r.endsWith("I-" + TaggingLabels.FIGURE_LABEL))
+                .count();
+
+                List<Figure> badFigures = figures.stream()
+                    .filter(f -> !f.isCompleteForTEI())
+                    .collect(Collectors.toList());
+
+                LOGGER.info("Number of figures badly formatted or incomplete we identified: " + badFigures.size());
+                bodyResults = revertResultsForBadItems(badFigures, bodyResults, TaggingLabels.FIGURE_LABEL,
+                     !(figures.size() > numberFiguresFulltextModel));
+
+                figures = figures.stream()
+                    .filter(f -> !badFigures.contains(f))
+                    .collect(Collectors.toList());
+
+				tables = processTables(bodyResults, bodyLayoutTokens.getTokenization(), doc);
+
+                long numberTablesFulltextModel = Arrays.stream(bodyResults.split("\n"))
+                    .filter(r -> r.endsWith("I-" + TaggingLabels.TABLE_LABEL))
+                .count();
+
+                //We deal with tables considered bad by reverting them as <paragraph>, to reduce the risk them to be
+                // dropped later on.
+
+                //TODO: double check the way the tables are validated
+
+                List<Table> badTables = tables.stream()
+                    .filter(t -> !(t.isCompleteForTEI() && t.validateTable()))
+                    .collect(Collectors.toList());
+
+                LOGGER.info("Number of tables badly formatted or incomplete we identified: " + badTables.size());
+                bodyResults = revertResultsForBadItems(badTables, bodyResults, TaggingLabels.TABLE_LABEL,
+                    !(tables.size() > numberTablesFulltextModel));
+
+                tables = tables.stream()
+                    .filter(t-> !badTables.contains(t))
+                    .collect(Collectors.toList());
+
                 // further parse the caption
                 for(Table table : tables) {
                     if ( CollectionUtils.isNotEmpty(table.getCaptionLayoutTokens()) ) {
@@ -293,7 +330,7 @@ else if (config.getConsolidateCitations() == 2)
                     }
                 }
 
-				equations = processEquations(resultBody, layoutTokenization.getTokenization(), doc);
+				equations = processEquations(bodyResults, bodyLayoutTokens.getTokenization(), doc);
 			} else {
 				LOGGER.debug("Fulltext model: The featured body is empty");
 			}
@@ -301,30 +338,35 @@ else if (config.getConsolidateCitations() == 2)
 			// possible annexes (view as a piece of full text similar to the body)
 			documentBodyParts = doc.getDocumentPart(SegmentationLabels.ANNEX);
             featSeg = getBodyTextFeatured(doc, documentBodyParts);
-			String resultAnnex = null;
-			List<LayoutToken> tokenizationsBody2 = null;
+			String annexResults = null;
+			List<LayoutToken> annexTokens = null;
 			if (featSeg != null && isNotEmpty(trim(featSeg.getLeft()))) {
-				// if featSeg is null, it usually means that no body segment is found in the
+				// if featSeg is null, it usually means that no annex segment is found in the
 				// document segmentation
-				String bodytext = featSeg.getLeft();
-				tokenizationsBody2 = featSeg.getRight().getTokenization();
-				resultAnnex = label(bodytext);
-				//System.out.println(rese);
+				String annexFeatures = featSeg.getLeft();
+				annexTokens = featSeg.getRight().getTokenization();
+				annexResults = label(annexFeatures);
+//				System.out.println(annexResults);
 			}
 
             // post-process reference and footnote callout to keep them consistent (e.g. for example avoid that a footnote
             // callout in superscript is by error labeled as a numerical reference callout)
             List<MarkerType> markerTypes = null;
 
-            if (resultBody != null) 
-                markerTypes = postProcessCallout(resultBody, layoutTokenization);
+            if (bodyResults != null)
+                markerTypes = postProcessCallout(bodyResults, bodyLayoutTokens);
 
             // final combination
             toTEI(doc, // document
-				resultBody, resultAnnex, // labeled data for body and annex
-				layoutTokenization, tokenizationsBody2, // tokenization for body and annex
+				bodyResults,
+                annexResults, // labeled data for body and annex
+                bodyLayoutTokens,
+                annexTokens, // tokenization for body and annex
 				resHeader, // header
-				figures, tables, equations, markerTypes,
+				figures,
+                tables,
+                equations,
+                markerTypes,
 				config);
             return doc;
         } catch (GrobidException e) {
@@ -334,6 +376,129 @@ else if (config.getConsolidateCitations() == 2)
         }
     }
 
+    static String revertResultsForBadItems(List<? extends Figure> badFiguresOrTables, String resultBody, String itemLabel) {
+        return revertResultsForBadItems(badFiguresOrTables, resultBody, itemLabel, true);
+    }
+
+    static String revertResultsForBadItems(List<? extends Figure> badFiguresOrTables, String resultBody, String itemLabel, boolean strict) {
+        //LF: we update the resultBody sequence by reverting these tables as <paragraph> elements
+        if (CollectionUtils.isNotEmpty(badFiguresOrTables)) {
+            List<List<String>> labelledResultsAsList = Arrays.stream(resultBody.split("\n"))
+                .map(l -> Arrays.stream(l.split("\t")).collect(Collectors.toList()))
+                .collect(Collectors.toList());
+
+            for (Figure badItem : badFiguresOrTables) {
+                // Find the index of the first layoutToken of the table in the tokenization
+                List<LayoutToken> layoutTokenItem = badItem.getLayoutTokens();
+                List<Integer> candidateIndexes = findCandidateIndex(layoutTokenItem, labelledResultsAsList,
+                    itemLabel, strict);
+                if (candidateIndexes.isEmpty()) {
+                    LOGGER.info("Cannot find the candidate index for fixing the tables.");
+                    continue;
+                }
+
+                // At this point i have more than one candidate, which can be matched if the same first
+                // token is repeated in the sequence. The next step is to find the matching figure/table
+                // using a large sequence 
+
+                List<String> sequenceTokenItemWithoutSpaces = layoutTokenItem.stream()
+                    .map(LayoutToken::getText)
+                    .map(StringUtils::strip)
+                    .filter(StringUtils::isNotBlank)
+                    .collect(Collectors.toList());
+
+                //TODO: reduce candidate indexes after matching one sequence
+                int resultIndexCandidate = consolidateResultCandidateThroughSequence(candidateIndexes, labelledResultsAsList, sequenceTokenItemWithoutSpaces);
+
+                if (resultIndexCandidate > -1) {
+                    boolean first = true;
+                    for (int i = resultIndexCandidate;i < Math.min(resultIndexCandidate + sequenceTokenItemWithoutSpaces.size(), labelledResultsAsList.size()); i++) {
+                        List<String> line = labelledResultsAsList.get(i);
+                        String label = Iterables.getLast(line);
+                        if (first) {
+                            first = false;
+                        } else {
+                            if (label.startsWith("I-")) {
+                                break;
+                            }
+                        }
+                        line.set(line.size() - 1, label.replace(TaggingLabels.TABLE_LABEL, TaggingLabels.PARAGRAPH_LABEL));
+                    }
+                } else {
+                    LOGGER.warn("Cannot find the result index candidate.");
+                }
+            }
+
+            String updatedResultBody = labelledResultsAsList.stream()
+                .map(l -> String.join("\t", l))
+                .collect(Collectors.joining("\n"));
+
+            resultBody = updatedResultBody;
+        }
+        return resultBody;
+    }
+
+    static int consolidateResultCandidateThroughSequence(List<Integer> candidateIndexes, List<List<String>> splitResult, List<String> tokensNoSpaceItem) {
+        int resultIndexCandidate = -1;
+        if (candidateIndexes.size() == 1){
+            resultIndexCandidate = candidateIndexes.get(0);
+        } else {
+            for (int candidateIndex: candidateIndexes) {
+                List<String> candidateTable = splitResult.subList(candidateIndex, Math.min(candidateIndex + tokensNoSpaceItem.size(), splitResult.size()))
+                    .stream()
+                    .map(i -> i.get(0))
+                    .collect(Collectors.toList());
+
+                String candidateTableText = String.join("", candidateTable);
+                String tokensText = String.join("", tokensNoSpaceItem);
+
+                if (candidateTableText.equals(tokensText)) {
+                    resultIndexCandidate = candidateIndex;
+                    break;
+                }
+            }
+        }
+        return resultIndexCandidate;
+    }
+
+    /**
+     * Find a set of candidates representing the indexes from the labelledResults which could correspond
+     * to the first token of the figure/table
+     *
+     * strict = True then it will check the items related to I-<table> or I-<figure> first
+     * and then the <table> or <figure> only if there are not candidates
+     * strict = False is usually necessary if there are more tables than I- token, this because a figure/table could be
+     * identified within the sequence initially provided by the fulltext model
+     *
+     */
+    @NotNull
+    static List<Integer> findCandidateIndex(List<LayoutToken> layoutTokenItem, List<List<String>> labelledResultsAsList, String itemLabel) {
+        return findCandidateIndex(layoutTokenItem, labelledResultsAsList, itemLabel, true);
+    }
+
+    @NotNull
+    static List<Integer> findCandidateIndex(List<LayoutToken> layoutTokenItem, List<List<String>> labelledResultsAsList, String itemLabel, boolean strict) {
+        LayoutToken firstLayoutTokenItem = layoutTokenItem.get(0);
+
+        List<Integer> candidateIndexes = IntStream.range(0, labelledResultsAsList.size())
+            .filter(i -> labelledResultsAsList.get(i).get(0).equals(firstLayoutTokenItem.getText())
+                && Iterables.getLast(labelledResultsAsList.get(i)).equals("I-" + itemLabel))
+            .boxed()
+            .collect(Collectors.toList());
+
+        if (candidateIndexes.isEmpty() || !strict) {
+            candidateIndexes = IntStream.range(0, labelledResultsAsList.size())
+            .filter(i -> labelledResultsAsList.get(i).get(0).equals(firstLayoutTokenItem.getText())
+                && (
+                    Iterables.getLast(labelledResultsAsList.get(i)).equals(itemLabel)
+                    || Iterables.getLast(labelledResultsAsList.get(i)).equals("I-" + itemLabel))
+            )
+            .boxed()
+            .collect(Collectors.toList());
+        }
+        return candidateIndexes;
+    }
+
 
     /**
      * Machine-learning recognition of full text structures limted to header and funding information.
@@ -518,44 +683,6 @@ public Pair<String, List<LayoutToken>> processShort(List<LayoutToken> tokens, Do
         return Pair.of(res, layoutTokenization);
     }
 
-    /**
-     * Post-process text labeled by the fulltext model on chunks that are known to be text (no table, or figure)
-     * It converts table and figure labels to paragraph labels.
-     */
-    protected static String postProcessFullTextLabeledText(String fulltextLabeledText) {
-        if (fulltextLabeledText == null)
-            return null;
-        StringBuilder result = new StringBuilder();
-
-        String[] lines = fulltextLabeledText.split("\n");
-        String previousLabel = null;
-        for(int i=0; i<lines.length; i++) {
-            String line = lines[i];
-            if (line == null || line.trim().length() == 0)
-                continue;
-            String[] pieces = line.split("\t");
-            String label = pieces[pieces.length-1];
-            if (label.equals("I-"+TaggingLabels.FIGURE.getLabel()) || label.equals("I-"+TaggingLabels.TABLE.getLabel())) {
-                if (previousLabel == null || !previousLabel.endsWith(TaggingLabels.PARAGRAPH.getLabel())) {
-                    pieces[pieces.length-1] = "I-"+TaggingLabels.PARAGRAPH.getLabel();
-                } else {
-                    pieces[pieces.length-1] = TaggingLabels.PARAGRAPH.getLabel();
-                }
-            } else if (label.equals(TaggingLabels.FIGURE.getLabel()) || label.equals(TaggingLabels.TABLE.getLabel())) {
-                pieces[pieces.length-1] = TaggingLabels.PARAGRAPH.getLabel();
-            }
-            for(int j=0; j<pieces.length; j++) {
-                if (j != 0)
-                    result.append("\t");
-                result.append(pieces[j]);
-            }
-            previousLabel = label;
-            result.append("\n");
-        }
-
-        return result.toString();
-    }
-
 	static public Pair<String, LayoutTokenization> getBodyTextFeatured(Document doc,
                                                                        SortedSet<DocumentPiece> documentBodyParts) {
 		if ((documentBodyParts == null) || (documentBodyParts.size() == 0)) {
@@ -1978,17 +2105,17 @@ private static boolean testClosingTag(StringBuilder buffer,
                 buffer.append("</figure>\n\n");
             } else if (lastTag0.equals("<item>")) {
                 buffer.append("</item>\n\n");
-            } else if (lastTag0.equals("<citation_marker>") || 
-                lastTag0.equals("<figure_marker>") || 
-                lastTag0.equals("<table_marker>") || 
+            } else if (lastTag0.equals("<citation_marker>") ||
+                lastTag0.equals("<figure_marker>") ||
+                lastTag0.equals("<table_marker>") ||
                 lastTag0.equals("<equation_marker>")) {
                 buffer.append("</ref>");
 
                 // Make sure that paragraph is closed when markers are at the end of it
-                if (!currentTag0.equals("<paragraph>") && 
-                    (!currentTag0.equals("<citation_marker>") || 
-                     !currentTag0.equals("<figure_marker>") || 
-                     !currentTag0.equals("<table_marker>") || 
+                if (!currentTag0.equals("<paragraph>") &&
+                    (!currentTag0.equals("<citation_marker>") ||
+                     !currentTag0.equals("<figure_marker>") ||
+                     !currentTag0.equals("<table_marker>") ||
                      !currentTag0.equals("<equation_marker>")
                      )
                     ) {
@@ -2152,10 +2279,12 @@ protected Pair<String,String> processTrainingDataFigures(String rese,
 
     	// If there still an open figure
     	if (openFigure) {
-            while((tokenizationsFigure.size() > 0) &&
+            while(CollectionUtils.isNotEmpty(tokenizationsFigure) &&
                 (tokenizationsFigure.get(0).getText().equals("\n") ||
-                    tokenizationsFigure.get(0).getText().equals(" ")) )
+                    tokenizationsFigure.get(0).getText().equals(" "))
+            ) {
                 tokenizationsFigure.remove(0);
+            }
 
             // process the "accumulated" figure
             Pair<String,String> trainingData = parsers.getFigureParser()
@@ -2195,9 +2324,9 @@ protected List<Table> processTables(String rese,
 
             for (Table result : localResults) {
                 List<LayoutToken> localTokenizationTable = result.getLayoutTokens();
-                //result.setLayoutTokens(tokenizationTable);
+//                result.setRawLayoutTokens(tokenizationTable);
 
-                // block setting: we restrict to the tokenization of this particulart table
+                // block setting: we restrict to the tokenization of this particular table
                 SortedSet<Integer> blockPtrs = new TreeSet<>();
                 for (LayoutToken lt : localTokenizationTable) {
                     if (!LayoutTokensUtil.spaceyToken(lt.t()) && !LayoutTokensUtil.newLineToken(lt.t())) {
@@ -2421,7 +2550,7 @@ protected List<Equation> processEquations(String rese,
 	}
 
     /**
-     * Ensure consistent use of callouts in the entire document body  
+     * Ensure consistent use of callouts in the entire document body
      */
     private List<MarkerType> postProcessCallout(String result, LayoutTokenization layoutTokenization) {
         if (layoutTokenization == null)
@@ -2481,7 +2610,7 @@ private List<MarkerType> postProcessCallout(String result, LayoutTokenization la
                     if (figureMarkerSeen.contains(refText)) {
                         // already seen reference marker sequence, we skip it
                         continue;
-                    }                    
+                    }
                     MarkerType localMarkerType = CalloutAnalyzer.getCalloutType(refTokens);
                     if (figureMarkerTypeCounts.get(localMarkerType) == null)
                         figureMarkerTypeCounts.put(localMarkerType, 1);
@@ -2494,7 +2623,7 @@ private List<MarkerType> postProcessCallout(String result, LayoutTokenization la
                     if (tableMarkerSeen.contains(refText)) {
                         // already seen reference marker sequence, we skip it
                         continue;
-                    }  
+                    }
                     MarkerType localMarkerType = CalloutAnalyzer.getCalloutType(refTokens);
                     if (tableMarkerTypeCounts.get(localMarkerType) == null)
                         tableMarkerTypeCounts.put(localMarkerType, 1);
@@ -2507,16 +2636,16 @@ private List<MarkerType> postProcessCallout(String result, LayoutTokenization la
                     if (equationMarkerSeen.contains(refText)) {
                         // already seen reference marker sequence, we skip it
                         continue;
-                    }  
+                    }
                     MarkerType localMarkerType = CalloutAnalyzer.getCalloutType(refTokens);
                     if (equationMarkerTypeCounts.get(localMarkerType) == null)
                         equationMarkerTypeCounts.put(localMarkerType, 1);
                     else
-                        equationMarkerTypeCounts.put(localMarkerType, equationMarkerTypeCounts.get(localMarkerType)+1);  
+                        equationMarkerTypeCounts.put(localMarkerType, equationMarkerTypeCounts.get(localMarkerType)+1);
 
                     if (!equationMarkerSeen.contains(refText))
-                        equationMarkerSeen.add(refText);               
-                } 
+                        equationMarkerSeen.add(refText);
+                }
             }
         }
 
@@ -2550,8 +2679,8 @@ private static MarkerType getBestType(Map<MarkerType,Integer> markerTypeCount) {
      * and body sections.
      */
     private void toTEI(Document doc,
-                       String reseBody,
-                       String reseAnnex,
+                       String bodyLabellingResult,
+                       String annexLabellingResult,
 					   LayoutTokenization layoutTokenization,
                        List<LayoutToken> tokenizationsAnnex,
                        BiblioItem resHeader,
@@ -2577,13 +2706,13 @@ private void toTEI(Document doc,
                 teiFormatter, resCitations, config);
 
             if (acknowledgmentStmt.length() > 0) {
-                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(acknowledgmentStmt.toString(), config);
 
                 if (localResult != null && localResult.getLeft() != null) {
-                    String local_tei = localResult.getLeft().toXML();
-                    local_tei = local_tei.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
-                    annexStatements.add(local_tei);
+                    String localTei = localResult.getLeft().toXML();
+                    localTei = localTei.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
+                    annexStatements.add(localTei);
                 }
                 else {
                     annexStatements.add(acknowledgmentStmt.toString());
@@ -2592,14 +2721,14 @@ private void toTEI(Document doc,
                 if (localResult != null && localResult.getRight() != null) {
                     if (localResult.getRight().getLeft() != null) {
                         List<Funding> localFundings = localResult.getRight().getLeft();
-                        if (localFundings.size()>0) {
+                        if (CollectionUtils.isNotEmpty(localFundings)) {
                             fundings.addAll(localFundings);
                         }
                     }
 
                     if (localResult.getRight().getRight() != null) {
                         List<Affiliation> localAffiliations = localResult.getRight().getRight();
-                        if (localAffiliations.size()>0) {
+                        if (CollectionUtils.isNotEmpty(localAffiliations)) {
                             affiliations.addAll(localAffiliations);
                         }
                     }
@@ -2620,14 +2749,14 @@ private void toTEI(Document doc,
                         resCitations,
                         config);
                 }
-                if (fundingStmt.length() > 0) {
-                    MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                if (StringUtils.isNotBlank(fundingStmt)) {
+                    MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(fundingStmt.toString(), config);
 
                     if (localResult != null && localResult.getLeft() != null) {
-                        String local_tei = localResult.getLeft().toXML();
-                        local_tei = local_tei.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
-                        annexStatements.add(local_tei);
+                        String localTEI = localResult.getLeft().toXML();
+                        localTEI = localTEI.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
+                        annexStatements.add(localTEI);
                     } else {
                         annexStatements.add(fundingStmt.toString());
                     }
@@ -2635,14 +2764,14 @@ private void toTEI(Document doc,
                     if (localResult != null && localResult.getRight() != null) {
                         if (localResult.getRight().getLeft() != null) {
                             List<Funding> localFundings = localResult.getRight().getLeft();
-                            if (localFundings.size()>0) {
+                            if (CollectionUtils.isNotEmpty(localFundings)) {
                                 fundings.addAll(localFundings);
                             }
                         }
 
                         if (localResult.getRight().getRight() != null) {
                             List<Affiliation> localAffiliations = localResult.getRight().getRight();
-                            if (localAffiliations.size()>0) {
+                            if (CollectionUtils.isNotEmpty(localAffiliations)) {
                                 affiliations.addAll(localAffiliations);
                             }
                         }
@@ -2659,13 +2788,13 @@ private void toTEI(Document doc,
                 resCitations,
                 config);
             if (fundingStmt.length() > 0) {
-                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(fundingStmt.toString(), config);
 
                 if (localResult != null && localResult.getLeft() != null){
-                    String local_tei = localResult.getLeft().toXML();
-                    local_tei = local_tei.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
-                    annexStatements.add(local_tei);
+                    String localTEI = localResult.getLeft().toXML();
+                    localTEI = localTEI.replace(" xmlns=\"http://www.tei-c.org/ns/1.0\"", "");
+                    annexStatements.add(localTEI);
                 } else {
                     annexStatements.add(fundingStmt.toString());
                 }
@@ -2673,14 +2802,14 @@ private void toTEI(Document doc,
                 if (localResult != null && localResult.getRight() != null) {
                     if (localResult.getRight().getLeft() != null) {
                         List<Funding> localFundings = localResult.getRight().getLeft();
-                        if (localFundings.size()>0) {
+                        if (CollectionUtils.isNotEmpty(localFundings)) {
                             fundings.addAll(localFundings);
                         }
                     }
 
                     if (localResult.getRight().getRight() != null) {
                         List<Affiliation> localAffiliations = localResult.getRight().getRight();
-                        if (localAffiliations.size()>0) {
+                        if (CollectionUtils.isNotEmpty(localAffiliations)) {
                             affiliations.addAll(localAffiliations);
                         }
                     }
@@ -2689,7 +2818,7 @@ private void toTEI(Document doc,
 
             tei.append(teiFormatter.toTEIHeader(resHeader, null, resCitations, markerTypes, fundings, config));
 
-            tei = teiFormatter.toTEIBody(tei, reseBody, resHeader, resCitations,
+            tei = teiFormatter.toTEIBody(tei, bodyLabellingResult, resHeader, resCitations,
                     layoutTokenization, figures, tables, equations, markerTypes, doc, config);
 
             tei.append("\t\t<back>\n");
@@ -2709,15 +2838,15 @@ private void toTEI(Document doc,
             }
 
             if (affiliations != null && affiliations.size() >0) {
-                
+
                 // check if we have at least one acknowledged research infrastructure here
                 List<Affiliation> filteredInfrastructures = new ArrayList<>();
                 for(Affiliation affiliation : affiliations) {
-                    if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0 && affiliation.isInfrastructure()) 
+                    if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0 && affiliation.isInfrastructure())
                         filteredInfrastructures.add(affiliation);
                     else if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0) {
                         // check if this organization is a known infrastructure
-                        List<Lexicon.OrganizationRecord> localOrganizationNamings = 
+                        List<Lexicon.OrganizationRecord> localOrganizationNamings =
                             Lexicon.getInstance().getOrganizationNamingInfo(affiliation.getAffiliationString());
                         if (localOrganizationNamings != null && localOrganizationNamings.size()>0) {
                             filteredInfrastructures.add(affiliation);
@@ -2729,7 +2858,7 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
                 if (filteredInfrastructures.size() > 0) {
                     tei.append("\n\t\t\t<listOrg type=\"infrastructure\">\n");
                     for(Affiliation affiliation : filteredInfrastructures) {
-                        List<Lexicon.OrganizationRecord> localOrganizationNamings = 
+                        List<Lexicon.OrganizationRecord> localOrganizationNamings =
                             Lexicon.getInstance().getOrganizationNamingInfo(affiliation.getAffiliationString());
                         tei.append("\t\t\t\t<org type=\"infrastructure\">");
                         tei.append("\t\t\t\t\t<orgName type=\"extracted\">");
@@ -2749,7 +2878,7 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
                         }
                         tei.append("\t\t\t\t</org>\n");
                     }
-                    
+
                     tei.append("\t\t\t</listOrg>\n");
                 }
             }
@@ -2761,10 +2890,10 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
                 Pair<String, List<LayoutToken>> headerAvailabilityProcessed = processShort(headerAvailabilityStatementTokens, doc);
                 if (headerAvailabilityProcessed != null) {
                     availabilityStmt = teiFormatter.processTEIDivSection("availability",
-                        "\t\t\t", 
-                        headerAvailabilityProcessed.getLeft(), 
-                        headerAvailabilityProcessed.getRight(), 
-                        resCitations, 
+                        "\t\t\t",
+                        headerAvailabilityProcessed.getLeft(),
+                        headerAvailabilityProcessed.getRight(),
+                        resCitations,
                         config);
                 }
                 if (availabilityStmt.length() > 0) {
@@ -2774,17 +2903,17 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
 
             // availability statements in non-header part
             availabilityStmt = getSectionAsTEI("availability",
-                "\t\t\t", 
-                doc, 
-                SegmentationLabels.AVAILABILITY, 
-                teiFormatter, 
-                resCitations, 
+                "\t\t\t",
+                doc,
+                SegmentationLabels.AVAILABILITY,
+                teiFormatter,
+                resCitations,
                 config);
             if (availabilityStmt.length() > 0) {
                 tei.append(availabilityStmt.toString());
             }
 
-			tei = teiFormatter.toTEIAnnex(tei, reseAnnex, resHeader, resCitations,
+			tei = teiFormatter.toTEIAnnex(tei, annexLabellingResult, resHeader, resCitations,
 				tokenizationsAnnex, markerTypes, doc, config);
 
 			tei = teiFormatter.toTEIReferences(tei, resCitations, config);
@@ -2829,7 +2958,7 @@ private void toTEIHeaderFunding(Document doc,
                 teiFormatter, null, config);
 
             if (acknowledgmentStmt.length() > 0) {
-                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(acknowledgmentStmt.toString(), config);
 
                 if (localResult != null && localResult.getLeft() != null) {
@@ -2873,7 +3002,7 @@ private void toTEIHeaderFunding(Document doc,
                         config);
                 }
                 if (fundingStmt.length() > 0) {
-                    MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                    MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(fundingStmt.toString(), config);
 
                     if (localResult != null && localResult.getLeft() != null) {
@@ -2911,7 +3040,7 @@ private void toTEIHeaderFunding(Document doc,
                 null,
                 config);
             if (fundingStmt.length() > 0) {
-                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult = 
+                MutablePair<Element, MutableTriple<List<Funding>,List<Person>,List<Affiliation>>> localResult =
                     parsers.getFundingAcknowledgementParser().processingXmlFragment(fundingStmt.toString(), config);
 
                 if (localResult != null && localResult.getLeft() != null){
@@ -2957,15 +3086,15 @@ private void toTEIHeaderFunding(Document doc,
             }
 
             if (affiliations != null && affiliations.size() >0) {
-                
+
                 // check if we have at least one acknowledged research infrastructure here
                 List<Affiliation> filteredInfrastructures = new ArrayList<>();
                 for(Affiliation affiliation : affiliations) {
-                    if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0 && affiliation.isInfrastructure()) 
+                    if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0 && affiliation.isInfrastructure())
                         filteredInfrastructures.add(affiliation);
                     else if (affiliation.getAffiliationString() != null && affiliation.getAffiliationString().length()>0) {
                         // check if this organization is a known infrastructure
-                        List<Lexicon.OrganizationRecord> localOrganizationNamings = 
+                        List<Lexicon.OrganizationRecord> localOrganizationNamings =
                             Lexicon.getInstance().getOrganizationNamingInfo(affiliation.getAffiliationString());
                         if (localOrganizationNamings != null && localOrganizationNamings.size()>0) {
                             filteredInfrastructures.add(affiliation);
@@ -2977,7 +3106,7 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
                 if (filteredInfrastructures.size() > 0) {
                     tei.append("\n\t\t\t<listOrg type=\"infrastructure\">\n");
                     for(Affiliation affiliation : filteredInfrastructures) {
-                        List<Lexicon.OrganizationRecord> localOrganizationNamings = 
+                        List<Lexicon.OrganizationRecord> localOrganizationNamings =
                             Lexicon.getInstance().getOrganizationNamingInfo(affiliation.getAffiliationString());
                         tei.append("\t\t\t\t<org type=\"infrastructure\">");
                         tei.append("\t\t\t\t\t<orgName type=\"extracted\">");
@@ -2997,7 +3126,7 @@ else if (affiliation.getAffiliationString() != null && affiliation.getAffiliatio
                         }
                         tei.append("\t\t\t\t</org>\n");
                     }
-                    
+
                     tei.append("\t\t\t</listOrg>\n");
                 }
             }
diff --git a/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java b/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java
index 185f3714d5..2c72ac1600 100755
--- a/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java
+++ b/grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java
@@ -1,5 +1,6 @@
 package org.grobid.core.engines;
 
+import org.apache.commons.collections4.CollectionUtils;
 import org.apache.commons.lang3.StringUtils;
 import org.apache.commons.lang3.tuple.ImmutablePair;
 import org.apache.commons.lang3.tuple.Pair;
@@ -108,9 +109,15 @@ public String processingHeaderSection(GrobidAnalysisConfig config, Document doc,
                 String res = null;
                 if (StringUtils.isNotBlank(header)) {
                     res = label(header);
+
+                    if (GrobidProperties.getGrobidEngineName("header").equals("delft")) {
+                        res = LabelUtils.postProcessFulltextCorrectSequencesWithoutInitialToken(res);
+                    }
                     resHeader = resultExtraction(res, headerTokenization, resHeader);
                 }
 
+
+
                 // language identification
                 StringBuilder contentSample = new StringBuilder();
                 if (resHeader.getTitle() != null) {
@@ -261,7 +268,7 @@ public String processingHeaderSection(GrobidAnalysisConfig config, Document doc,
                     //resHeader.setKeyword(keywords.replace("\n", " ").replace("  ", " "));
                     resHeader.setKeyword(keywords);
                     List<Keyword> keywordsSegmented = BiblioItem.segmentKeywords(keywords);
-                    if ((keywordsSegmented != null) && (keywordsSegmented.size() > 0))
+                    if (CollectionUtils.isNotEmpty(keywordsSegmented))
                         resHeader.setKeywords(keywordsSegmented);
                 }
 
@@ -311,7 +318,7 @@ public String processingHeaderSection(GrobidAnalysisConfig config, Document doc,
                 }
 
                 // copyrights/license identification
-                if (resHeader.getCopyright() != null && resHeader.getCopyright().length()>0) {
+                if (StringUtils.isNotBlank(resHeader.getCopyright())) {
                     if (GrobidProperties.getGrobidEngineName("copyright").equals("delft")) {
                         CopyrightsLicense copyrightsLicense = LicenseClassifier.getInstance().classify(resHeader.getCopyright());
                         if (copyrightsLicense != null) 
@@ -928,6 +935,7 @@ else if (biblio.getPublicationDate() == null)
                     // this will need to be reviewed with more training data, for the moment
                     // avoid concatenation for abstracts as it brings more noise than correct pieces
                     //biblio.setAbstract(biblio.getAbstract() + " " + clusterContent);
+                    //TODO: avoid dumping text on the floor
                 } else {
                     biblio.setAbstract(clusterContent);
                     List<LayoutToken> tokens = cluster.concatTokens();
diff --git a/grobid-core/src/main/kotlin/org/grobid/core/utilities/LabelUtils.kt b/grobid-core/src/main/kotlin/org/grobid/core/utilities/LabelUtils.kt
new file mode 100644
index 0000000000..eecc6d88d9
--- /dev/null
+++ b/grobid-core/src/main/kotlin/org/grobid/core/utilities/LabelUtils.kt
@@ -0,0 +1,132 @@
+package org.grobid.core.utilities
+
+import org.apache.commons.lang3.StringUtils
+import org.grobid.core.engines.label.TaggingLabels
+
+object LabelUtils {
+    /**
+     * Post-process text labeled by the fulltext model on chunks that are known to be text (no table, or figure)
+     * It converts table and figure labels to paragraph labels.
+     */
+    @JvmStatic
+    fun postProcessFullTextLabeledText(fulltextLabeledText: String): String {
+        val result = StringBuilder()
+
+        val lines = fulltextLabeledText
+            .split("\n".toRegex())
+            .dropLastWhile { it.isEmpty() }
+            .toTypedArray()
+        var previousLabel: String? = null
+
+        for (i in lines.indices) {
+            val line = lines[i]
+            if (StringUtils.isBlank(line)) continue
+
+            val pieces = line.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()
+            val label = pieces[pieces.size - 1]
+            if (label == "I-" + TaggingLabels.FIGURE.label || label == "I-" + TaggingLabels.TABLE.label) {
+                if (previousLabel == null || !previousLabel.endsWith(TaggingLabels.PARAGRAPH.label)) {
+                    pieces[pieces.size - 1] = "I-" + TaggingLabels.PARAGRAPH.label
+                } else {
+                    pieces[pieces.size - 1] = TaggingLabels.PARAGRAPH.label
+                }
+            } else if (label == TaggingLabels.FIGURE.label || label == TaggingLabels.TABLE.label) {
+                pieces[pieces.size - 1] = TaggingLabels.PARAGRAPH.label
+            }
+            result.append(pieces.joinToString("\t"))
+            previousLabel = label
+            result.append("\n")
+        }
+
+        return result.toString()
+    }
+
+    /**
+     * This method correct the fulltext sequence when the model has predicted several unlikely
+     * start sequences of table or figures.
+     * For example: I-<figure> followed by another I-<figure> (or table) </figure></figure>
+     **/
+    @JvmStatic
+    fun postProcessFulltextFixInvalidTableOrFigure(fulltextLabeledText: String): String {
+        val result = StringBuilder()
+
+        val lines = fulltextLabeledText
+            .split("\n".toRegex())
+            .dropLastWhile { it.isEmpty() }
+            .toTypedArray()
+
+        var previousLabel: String? = null
+        for (i in lines.indices) {
+            val line = lines[i]
+            if (StringUtils.isBlank(line)) continue
+
+            val pieces = line
+                .split("\t".toRegex())
+                .dropLastWhile { it.isEmpty() }
+                .toTypedArray()
+
+            val label = pieces[pieces.size - 1]
+            if (label == "I-" + TaggingLabels.FIGURE.label) {
+                if (StringUtils.equals(previousLabel, "I-" + TaggingLabels.FIGURE.label)) {
+                    pieces[pieces.size - 1] = TaggingLabels.FIGURE.label
+                }
+            } else if (label == "I-" + TaggingLabels.TABLE.label) {
+                if (StringUtils.equals(previousLabel, "I-" + TaggingLabels.TABLE.label)) {
+                    pieces[pieces.size - 1] = TaggingLabels.TABLE.label
+                }
+            }
+
+            result.append(pieces.joinToString("\t"))
+            previousLabel = label
+            result.append("\n")
+        }
+
+        return result.toString()
+    }
+
+    /**
+     * PostProcess the sequence of labels by reverting change of labels that are not starting with a initial sequence.
+     * For example, in a sequence of I-<abstract> <abstract> followed by <availability>,
+     * we revert all sequence of availability as <abstract>
+     */
+    @JvmStatic
+    fun postProcessFulltextCorrectSequencesWithoutInitialToken(fulltextLabeledText: String): String {
+        val result = StringBuilder()
+
+        val lines = fulltextLabeledText
+            .split("\n".toRegex())
+            .dropLastWhile { it.isEmpty() }
+            .toTypedArray()
+
+        var previousLabel: String? = null
+        for (i in lines.indices) {
+            val line = lines[i]
+            if (StringUtils.isBlank(line)) continue
+            val pieces = line
+                .split("\t".toRegex())
+                .dropLastWhile { it.isEmpty() }
+                .toTypedArray()
+            var label = pieces[pieces.size - 1]
+
+            if (!label.equals(TaggingLabels.OTHER_LABEL)) {
+                if (!label.startsWith("I-")) {
+                    if (previousLabel != null && previousLabel != label) {
+                        pieces[pieces.size - 1] = previousLabel
+                        label = previousLabel
+                    }
+                }
+            }
+
+            result.append(pieces.joinToString("\t"))
+            if (label == TaggingLabels.OTHER_LABEL) {
+                previousLabel = null
+            } else {
+                previousLabel = label.replace("I-", "")
+            }
+            result.append("\n")
+        }
+
+        return result.toString()
+    }
+
+}
diff --git a/grobid-core/src/test/java/org/grobid/core/engines/FullTextParserTest.java b/grobid-core/src/test/java/org/grobid/core/engines/FullTextParserTest.java
deleted file mode 100644
index a4fb60a4e9..0000000000
--- a/grobid-core/src/test/java/org/grobid/core/engines/FullTextParserTest.java
+++ /dev/null
@@ -1,263 +0,0 @@
-package org.grobid.core.engines;
-
-import org.apache.commons.lang3.StringUtils;
-import org.apache.commons.lang3.tuple.Pair;
-import org.grobid.core.analyzers.GrobidAnalyzer;
-import org.grobid.core.factory.GrobidFactory;
-import org.grobid.core.layout.LayoutToken;
-import org.grobid.core.main.LibraryLoader;
-import org.grobid.core.utilities.GrobidProperties;
-import org.junit.AfterClass;
-import org.junit.Before;
-import org.junit.BeforeClass;
-import org.junit.Test;
-
-import java.util.ArrayList;
-import java.util.Arrays;
-import java.util.List;
-import java.util.stream.Collectors;
-
-import static org.hamcrest.CoreMatchers.is;
-import static org.hamcrest.MatcherAssert.assertThat;
-import static org.hamcrest.collection.IsCollectionWithSize.hasSize;
-
-public class FullTextParserTest {
-
-    private FullTextParser target;
-
-    @Before
-    public void setUp() throws Exception {
-        target = new FullTextParser(new EngineParsers());
-    }
-
-    @BeforeClass
-    public static void init() {
-        LibraryLoader.load();
-        GrobidProperties.getInstance();
-    }
-
-    @AfterClass
-    public static void tearDown() {
-        GrobidFactory.reset();
-    }
-
-    @Test
-    public void testProcessTrainingDataFigures_single_figure() throws Exception {
-        String text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO";
-        List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text);
-        String rese = "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
-            "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n\n";
-
-
-        Pair<String, String> stringStringPair = target.processTrainingDataFigures(rese, tokens, "123");
-
-        String tei = stringStringPair.getLeft();
-        String tokenisation = stringStringPair.getRight();
-        String reconstructedText = Arrays.stream(tokenisation.split("\n")).map(l -> l.split("\t")[0]).collect(Collectors.joining(" "));
-
-        assertThat(reconstructedText, is("FIG . 1 . λ ( T ) vs . T for YBCO"));
-        assertThat(tokenisation.split("\n").length, is(13));
-
-    }
-
-    @Test
-    public void testProcessTrainingDataFigures_multiple_figures() throws Exception {
-        String text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO";
-        List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text);
-        String rese = "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
-            "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
-            "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n\n";
-
-
-        Pair<String, String> stringStringPair = target.processTrainingDataFigures(rese, tokens, "123");
-
-        String tei = stringStringPair.getLeft();
-        String tokenisation = stringStringPair.getRight();
-        List<String> output = new ArrayList<>();
-        for (String block : tokenisation.split("\n\n\n")) {
-            String collect = Arrays.stream(block.split("\n")).map(l -> l.split("\t")[0]).collect(Collectors.joining(" "));
-            if (StringUtils.isNotBlank(collect)) {
-                output.add(collect);
-            }
-        }
-
-        assertThat(output, hasSize(2));
-        assertThat(output.get(0), is("FIG . 1 . λ ( T )"));
-        assertThat(output.get(1), is("vs . T for YBCO"));
-        assertThat(tokenisation.split("\n").length, is(15));
-
-    }
-
-    @Test
-    public void testProcessTrainingDataTables_single_table() throws Exception {
-        String text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO";
-        List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text);
-        String rese = "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
-            "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n\n";
-
-
-        Pair<String, String> stringStringPair = target.processTrainingDataTables(rese, tokens, "123");
-
-        String tei = stringStringPair.getLeft();
-        String tokenisation = stringStringPair.getRight();
-        String reconstructedText = Arrays.stream(tokenisation.split("\n")).map(l -> l.split("\t")[0]).collect(Collectors.joining(" "));
-
-        assertThat(reconstructedText, is("FIG . 1 . λ ( T ) vs . T for YBCO"));
-        assertThat(tokenisation.split("\n").length, is(13));
-
-    }
-
-    @Test
-    public void testProcessTrainingDataTable_multiple_tables() throws Exception {
-        String text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO";
-        List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text);
-        String rese = "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
-            "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
-            "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
-            "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n\n";
-
-
-        Pair<String, String> stringStringPair = target.processTrainingDataTables(rese, tokens, "123");
-
-        String tei = stringStringPair.getLeft();
-        String tokenisation = stringStringPair.getRight();
-        List<String> output = new ArrayList<>();
-        for (String block : tokenisation.split("\n\n\n")) {
-            String collect = Arrays.stream(block.split("\n")).map(l -> l.split("\t")[0]).collect(Collectors.joining(" "));
-            if (StringUtils.isNotBlank(collect)) {
-                output.add(collect);
-            }
-        }
-
-        assertThat(output, hasSize(2));
-        assertThat(output.get(0), is("FIG . 1 . λ ( T )"));
-        assertThat(output.get(1), is("vs . T for YBCO"));
-        assertThat(tokenisation.split("\n").length, is(15));
-
-    }
-
-    @Test
-    public void testPostProcessLabeledAbstract_shouldTransformTableLabelInParagraphLabel() {
-        String resultWithTables = "This\tthis\tT\tTh\tThi\tThis\ts\tis\this\tThis\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\tI-<table>\n" +
-            "study\tstudy\ts\tst\tstu\tstud\ty\tdy\tudy\ttudy\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "was\twas\tw\twa\twas\twas\ts\tas\twas\twas\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "supported\tsupported\ts\tsu\tsup\tsupp\td\ted\tted\trted\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "by\tby\tb\tby\tby\tby\ty\tby\tby\tby\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "South\tsouth\tS\tSo\tSou\tSout\th\tth\tuth\touth\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Asian\tasian\tA\tAs\tAsi\tAsia\tn\tan\tian\tsian\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Clinical\tclinical\tC\tCl\tCli\tClin\tl\tal\tcal\tical\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Toxicology\ttoxicology\tT\tTo\tTox\tToxi\ty\tgy\togy\tlogy\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t2\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Collaboration\tcollaboration\tC\tCo\tCol\tColl\tn\ton\tion\ttion\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t2\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tCOMMA\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "which\twhich\tw\twh\twhi\twhic\th\tch\tich\thich\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "is\tis\ti\tis\tis\tis\ts\tis\tis\tis\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "funded\tfunded\tf\tfu\tfun\tfund\td\ted\tded\tnded\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "by\tby\tb\tby\tby\tby\ty\tby\tby\tby\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Wellcome\twellcome\tW\tWe\tWel\tWell\te\tme\tome\tcome\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Trust\ttrust\tT\tTr\tTru\tTrus\tt\tst\tust\trust\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "National\tnational\tN\tNa\tNat\tNati\tl\tal\tnal\tonal\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Health\thealth\tH\tHe\tHea\tHeal\th\tth\tlth\talth\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Medical\tmedical\tM\tMe\tMed\tMedi\tl\tal\tcal\tical\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Council\tcouncil\tC\tCo\tCou\tCoun\tl\til\tcil\tncil\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "International\tinternational\tI\tIn\tInt\tInte\tl\tal\tnal\tonal\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Collaborative\tcollaborative\tC\tCo\tCol\tColl\te\tve\tive\ttive\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t7\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "Grant\tgrant\tG\tGr\tGra\tGran\tt\tnt\tant\trant\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t7\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "GR071669MA\tgr071669ma\tG\tGR\tGR0\tGR07\tA\tMA\t9MA\t69MA\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tCONTAINSDIGITS\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "funding\tfunding\tf\tfu\tfun\tfund\tg\tng\ting\tding\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "bodies\tbodies\tb\tbo\tbod\tbodi\ts\tes\ties\tdies\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "had\thad\th\tha\thad\thad\td\tad\thad\thad\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "no\tno\tn\tno\tno\tno\to\tno\tno\tno\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "role\trole\tr\tro\trol\trole\te\tle\tole\trole\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "analyzing\tanalyzing\ta\tan\tana\tanal\tg\tng\ting\tzing\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "or\tor\to\tor\tor\tor\tr\tor\tor\tor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "interpreting\tinterpreting\ti\tin\tint\tinte\tg\tng\ting\tting\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "data\tdata\td\tda\tdat\tdata\ta\tta\tata\tdata\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "or\tor\to\tor\tor\tor\tr\tor\tor\tor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "writing\twriting\tw\twr\twri\twrit\tg\tng\ting\tting\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            "article\tarticle\ta\tar\tart\tarti\te\tle\tcle\ticle\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
-            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t11\t10\t0\tNUMBER\t0\t0\t<table>";
-        String postprocessed = FullTextParser.postProcessFullTextLabeledText(resultWithTables);
-
-        assertThat(Arrays.stream(StringUtils.split(postprocessed, "\n"))
-            .filter(l -> l.endsWith("<table>"))
-            .count(), is(0L));
-
-        assertThat(Arrays.stream(StringUtils.split(postprocessed, "\n"))
-            .filter(l -> l.endsWith("<paragraph>"))
-            .count(), is (Arrays.stream(StringUtils.split(resultWithTables, "\n"))
-            .filter(l -> l.endsWith("<table>"))
-            .count()));
-
-    }
-
-
-}
\ No newline at end of file
diff --git a/grobid-core/src/test/java/org/grobid/core/utilities/GrobidTestUtils.java b/grobid-core/src/test/java/org/grobid/core/utilities/GrobidTestUtils.java
index f25b263049..9b7db0c685 100644
--- a/grobid-core/src/test/java/org/grobid/core/utilities/GrobidTestUtils.java
+++ b/grobid-core/src/test/java/org/grobid/core/utilities/GrobidTestUtils.java
@@ -10,6 +10,10 @@
 
 public class GrobidTestUtils {
 
+    public static String getWapitiResult(List<String> features, List<Triple<String, Integer, Integer>> labels) {
+        return getWapitiResult(features, labels, " ");
+    }
+
     /**
      * Utility method to generate a hypotetical result from wapiti.
      * Useful for testing the extraction of the sequence labeling.
@@ -17,7 +21,7 @@ public class GrobidTestUtils {
      * @param labels label maps. A list of Triples, containing label (left), start_index (middle) and end_index exclusive (right)
      * @return a string containing the resulting features + labels returned by wapiti
      */
-    public static String getWapitiResult(List<String> features, List<Triple<String, Integer, Integer>> labels) {
+    public static String getWapitiResult(List<String> features, List<Triple<String, Integer, Integer>> labels, String separator) {
 
         List<String> labeled = new ArrayList<>();
         int idx = 0;
@@ -52,7 +56,7 @@ public static String getWapitiResult(List<String> features, List<Triple<String,
             if (features.get(i) == null || features.get(i).startsWith(" ")) {
                 continue;
             }
-            sb.append(features.get(i)).append(" ").append(labeled.get(i)).append("\n");
+            sb.append(features.get(i)).append(separator).append(labeled.get(i)).append("\n");
         }
 
         return sb.toString();
diff --git a/grobid-core/src/test/kotlin/org/grobid/core/engines/FullTextParserTest.kt b/grobid-core/src/test/kotlin/org/grobid/core/engines/FullTextParserTest.kt
new file mode 100644
index 0000000000..bc7a131fba
--- /dev/null
+++ b/grobid-core/src/test/kotlin/org/grobid/core/engines/FullTextParserTest.kt
@@ -0,0 +1,385 @@
+package org.grobid.core.engines
+
+import jnr.posix.BaseIovec.Layout
+import org.apache.commons.lang3.StringUtils
+import org.apache.commons.lang3.tuple.Triple
+import org.grobid.core.analyzers.GrobidAnalyzer
+import org.grobid.core.engines.label.TaggingLabels.TABLE_LABEL
+import org.grobid.core.factory.GrobidFactory
+import org.grobid.core.layout.LayoutToken
+import org.grobid.core.main.LibraryLoader
+import org.grobid.core.utilities.GrobidConfig
+import org.grobid.core.utilities.GrobidProperties
+import org.grobid.core.utilities.GrobidTestUtils
+import org.hamcrest.CoreMatchers
+import org.hamcrest.CoreMatchers.`is`
+import org.hamcrest.MatcherAssert
+import org.hamcrest.MatcherAssert.assertThat
+import org.hamcrest.Matchers.hasSize
+import org.hamcrest.collection.IsCollectionWithSize
+import org.junit.AfterClass
+import org.junit.Before
+import org.junit.BeforeClass
+import org.junit.Test
+import java.util.*
+import java.util.stream.Collectors
+
+class FullTextParserTest {
+    private lateinit var target: FullTextParser
+
+    @Before
+    @Throws(Exception::class)
+    fun setUp() {
+        val modelParameters = GrobidConfig.ModelParameters()
+        modelParameters.name = "bao"
+        GrobidProperties.addModel(modelParameters)
+        target = FullTextParser(EngineParsers())
+    }
+
+
+    companion object {
+        @JvmStatic
+        @BeforeClass
+        @Throws(java.lang.Exception::class)
+        fun init() {
+            LibraryLoader.load()
+            GrobidProperties.getInstance()
+        }
+
+        @JvmStatic
+        @AfterClass
+        @Throws(java.lang.Exception::class)
+        fun tearDown() {
+            GrobidFactory.reset()
+        }
+    }
+
+    @Test
+    @Throws(Exception::class)
+    fun testProcessTrainingDataFigures_single_figure() {
+        val text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO"
+        val tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text)
+        val rese =
+            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n\n"
+
+
+        val stringStringPair = target!!.processTrainingDataFigures(rese, tokens, "123")
+
+        val tei = stringStringPair.left
+        val tokenisation = stringStringPair.right
+        val reconstructedText =
+            Arrays.stream(tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map { l: String -> l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()[0] }
+                .collect(Collectors.joining(" "))
+
+        MatcherAssert.assertThat(reconstructedText, CoreMatchers.`is`("FIG . 1 . λ ( T ) vs . T for YBCO"))
+        MatcherAssert.assertThat(
+            tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray().size,
+            CoreMatchers.`is`(13)
+        )
+    }
+
+    @Test
+    @Throws(Exception::class)
+    fun testProcessTrainingDataFigures_multiple_figures() {
+        val text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO"
+        val tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text)
+        val rese =
+            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<figure>\n\n"
+
+
+        val stringStringPair = target!!.processTrainingDataFigures(rese, tokens, "123")
+
+        val tei = stringStringPair.left
+        val tokenisation = stringStringPair.right
+        val output: MutableList<String> = ArrayList()
+        for (block in tokenisation.split("\n\n\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()) {
+            val collect = Arrays.stream(block.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map { l: String -> l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()[0] }
+                .collect(Collectors.joining(" "))
+            if (StringUtils.isNotBlank(collect)) {
+                output.add(collect)
+            }
+        }
+
+        MatcherAssert.assertThat<List<String>>(output, IsCollectionWithSize.hasSize(2))
+        MatcherAssert.assertThat(output[0], CoreMatchers.`is`("FIG . 1 . λ ( T )"))
+        MatcherAssert.assertThat(output[1], CoreMatchers.`is`("vs . T for YBCO"))
+        MatcherAssert.assertThat(
+            tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray().size,
+            CoreMatchers.`is`(15)
+        )
+    }
+
+    @Test
+    @Throws(Exception::class)
+    fun testProcessTrainingDataTables_single_table() {
+        val text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO"
+        val tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text)
+        val rese =
+            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n\n"
+
+
+        val stringStringPair = target!!.processTrainingDataTables(rese, tokens, "123")
+
+        val tei = stringStringPair.left
+        val tokenisation = stringStringPair.right
+        val reconstructedText =
+            Arrays.stream(tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map { l: String -> l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()[0] }
+                .collect(Collectors.joining(" "))
+
+        MatcherAssert.assertThat(reconstructedText, CoreMatchers.`is`("FIG . 1 . λ ( T ) vs . T for YBCO"))
+        MatcherAssert.assertThat(
+            tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray().size,
+            CoreMatchers.`is`(13)
+        )
+    }
+
+    @Test
+    @Throws(Exception::class)
+    fun testProcessTrainingDataTable_multiple_tables() {
+        val text = "The mechanism for superconductivity FIG. 1. λ(T) vs . T for YBCO"
+        val tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(text)
+        val rese =
+            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "mechanism\tmechanism\tm\tme\tmec\tmech\tm\tsm\tism\tnism\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "superconductivity\tsuperconductivity\ts\tsu\tsup\tsupe\ty\tty\tity\tvity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t4\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t1\t0\t<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "λ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tλ\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "vs\tvs\tv\tvs\tvs\tvs\ts\tvs\tvs\tvs\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "T\tt\tT\tT\tT\tT\tT\tT\tT\tT\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n" +
+                "YBCO\tybco\tY\tYB\tYBC\tYBCO\tO\tCO\tBCO\tYBCO\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t10\t3\t0\tNUMBER\t0\t0\t<table>\n\n"
+
+
+        val stringStringPair = target!!.processTrainingDataTables(rese, tokens, "123")
+
+        val tei = stringStringPair.left
+        val tokenisation = stringStringPair.right
+        val output: MutableList<String> = ArrayList()
+        for (block in tokenisation.split("\n\n\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()) {
+            val collect = Arrays.stream(block.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map { l: String -> l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()[0] }
+                .collect(Collectors.joining(" "))
+            if (StringUtils.isNotBlank(collect)) {
+                output.add(collect)
+            }
+        }
+
+        MatcherAssert.assertThat<List<String>>(output, IsCollectionWithSize.hasSize(2))
+        MatcherAssert.assertThat(output[0], CoreMatchers.`is`("FIG . 1 . λ ( T )"))
+        MatcherAssert.assertThat(output[1], CoreMatchers.`is`("vs . T for YBCO"))
+        MatcherAssert.assertThat(
+            tokenisation.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray().size,
+            CoreMatchers.`is`(15)
+        )
+    }
+
+    @Test
+    fun testFindCandidates_shouldFindMultipleResults() {
+        // I need to prepare a sequence where there might be multiple matches,
+        // and then verify that the sequence is correctly used for discrimination
+        var sequence = "This article solves the problem where some of our interaction are fauly. " +
+            "a 8 9 j 92j 3 3j 9 j 9j Table 1: The reconstruction of the national anthem " +
+            "We are interested in the relation between certain information and " +
+            "a b b d 1 2 3 4 s 3 3 d9 Table 2: The relation between information and noise " +
+            "the related affectionality. " +
+            "a b b d 1 2 3 4 5 6 7 Table 3: The relation between homicides and donuts eating " +
+            "The relation between homicides and donuts eating is a very important one. "
+
+        var tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(sequence)
+
+        // These triples made in following way: label, starting index (included), ending index (excluded)
+        val labels = listOf(
+            Triple.of("I-<paragraph>", 0, 1),
+            Triple.of("<paragraph>", 1, 24),
+            Triple.of("I-<table>", 25, 26),
+            Triple.of("<table>", 26, 61),
+            Triple.of("I-<paragraph>", 62, 63),
+            Triple.of("<paragraph>", 63, 81),
+            Triple.of("I-<table>", 82, 83),
+            Triple.of("<table>", 82, 118),
+            Triple.of("I-<paragraph>", 119, 120),
+            Triple.of("<paragraph>", 120, 129),
+            Triple.of("I-<table>", 130, 131),
+            Triple.of("<table>", 131, 171),
+            Triple.of("I-<paragraph>", 171, 172),
+            Triple.of("<paragraph>", 172, 195),
+        )
+
+        val features = tokens.stream().map { it.text }.collect(Collectors.toList())
+        val wapitiResults = GrobidTestUtils.getWapitiResult(features, labels, "\t")
+
+        val wapitiResultsAsList =
+            Arrays.stream(wapitiResults.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map<List<String>> { l: String ->
+                    Arrays.stream(
+                        l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()
+                    )
+                        .collect(Collectors.toList())
+                }
+                .collect(Collectors.toList())
+
+        val table1Tokens = tokens.subList(25, 61)
+        val foundCandidateIndex = FullTextParser.findCandidateIndex(table1Tokens, wapitiResultsAsList, TABLE_LABEL)
+
+        assertThat(foundCandidateIndex, hasSize(3))
+        assertThat(foundCandidateIndex.get(0), `is`(13))
+        assertThat(foundCandidateIndex.get(1), `is`(42))
+        assertThat(foundCandidateIndex.get(2), `is`(67))
+    }
+
+    @Test
+    fun testConsolidateResultCandidateThroughSequence() {
+        //        var mockDocumentSource = createMock<DocumentSource>(DocumentSource::class.java)
+        //        var document = Document.createFromText("")
+        val sequence = "This article solves the problem where some of our interaction are fauly. " +
+            "a 8 9 j 92j 3 3j 9 j 9j Table 1: The reconstruction of the national anthem " +
+            "We are interested in the relation between certain information and " +
+            "a b b d 1 2 3 4 s 3 3 d9 Table 2: The relation between information and noise " +
+            "the related affectionality. " +
+            "a b b d 1 2 3 4 5 6 7 Table 3: The relation between homicides and donuts eating " +
+            "The relation between homicides and donuts eating is a very important one. "
+
+        val tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(sequence)
+
+        // These triples made in following way: label, starting index (included), ending index (excluded)
+        val labels = listOf(
+            Triple.of("I-<paragraph>", 0, 1),
+            Triple.of("<paragraph>", 1, 24),
+            Triple.of("I-<table>", 25, 26),
+            Triple.of("<table>", 26, 61),
+            Triple.of("I-<paragraph>", 62, 63),
+            Triple.of("<paragraph>", 63, 81),
+            Triple.of("I-<table>", 82, 83),
+            Triple.of("<table>", 82, 118),
+            Triple.of("I-<paragraph>", 119, 120),
+            Triple.of("<paragraph>", 120, 129),
+            Triple.of("I-<table>", 130, 131),
+            Triple.of("<table>", 131, 171),
+            Triple.of("I-<paragraph>", 171, 172),
+            Triple.of("<paragraph>", 172, 195),
+        )
+
+        val features = tokens.stream().map { it.text }.collect(Collectors.toList())
+
+        val wapitiResults = GrobidTestUtils.getWapitiResult(features, labels, "\t")
+        val wapitiResultsAsList =
+            Arrays.stream(wapitiResults.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map<List<String>> { l: String ->
+                    Arrays.stream(
+                        l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()
+                    )
+                        .collect(Collectors.toList())
+                }
+                .collect(Collectors.toList())
+
+        val table1Tokens = tokens.subList(25, 61)
+
+        val sequenceTokenWithoutSpacesTable1: List<String> = table1Tokens.stream()
+            .map { obj: LayoutToken -> obj.text }
+            .map { str: String? -> StringUtils.strip(str) }
+            .filter { cs: String? -> StringUtils.isNotBlank(cs) }
+            .collect(Collectors.toList())
+
+        val candidatesIndexes = Arrays.asList(
+            13, 42, 67
+        )
+        val consolidatedTable1ResultCandidateThroughSequence = FullTextParser.consolidateResultCandidateThroughSequence(
+            candidatesIndexes,
+            wapitiResultsAsList,
+            sequenceTokenWithoutSpacesTable1
+        )
+
+        assertThat(consolidatedTable1ResultCandidateThroughSequence, `is`(13))
+
+        val table2Tokens = tokens.subList(82, 118)
+
+        var sequenceTokenWithoutSpacesTable2: MutableList<String>? = table2Tokens.stream()
+            .map { obj: LayoutToken -> obj.text }
+            .map { str: String? -> StringUtils.strip(str) }
+            .filter { cs: String? -> StringUtils.isNotBlank(cs) }
+            .collect(Collectors.toList())
+
+        val consolidatedTable2ResultCandidateThroughSequence = FullTextParser.consolidateResultCandidateThroughSequence(
+            candidatesIndexes,
+            wapitiResultsAsList,
+            sequenceTokenWithoutSpacesTable2
+        )
+
+        assertThat(consolidatedTable2ResultCandidateThroughSequence, `is`(42))
+
+        val table3Tokens = tokens.subList(130, 171)
+
+        var sequenceTokenWithoutSpacesTable3: MutableList<String>? = table3Tokens.stream()
+            .map { obj: LayoutToken -> obj.text }
+            .map { str: String? -> StringUtils.strip(str) }
+            .filter { cs: String? -> StringUtils.isNotBlank(cs) }
+            .collect(Collectors.toList())
+
+        val consolidatedTable3ResultCandidateThroughSequence = FullTextParser.consolidateResultCandidateThroughSequence(
+            candidatesIndexes,
+            wapitiResultsAsList,
+            sequenceTokenWithoutSpacesTable3
+        )
+
+        assertThat(consolidatedTable3ResultCandidateThroughSequence, `is`(67))
+    }
+}
\ No newline at end of file
diff --git a/grobid-core/src/test/kotlin/org/grobid/core/utilities/LabelUtilsTest.kt b/grobid-core/src/test/kotlin/org/grobid/core/utilities/LabelUtilsTest.kt
new file mode 100644
index 0000000000..0cb3f6a731
--- /dev/null
+++ b/grobid-core/src/test/kotlin/org/grobid/core/utilities/LabelUtilsTest.kt
@@ -0,0 +1,470 @@
+package org.grobid.core.utilities
+
+import org.apache.commons.lang3.StringUtils
+import org.grobid.core.utilities.GrobidConfig.ModelParameters
+import org.hamcrest.CoreMatchers.`is`
+import org.hamcrest.CoreMatchers.not
+import org.hamcrest.MatcherAssert.assertThat
+import org.junit.jupiter.api.BeforeAll
+import java.util.*
+import java.util.stream.Collectors
+import kotlin.test.Test
+
+
+class LabelUtilsTest {
+
+
+    @Test
+    fun testPostProcessLabeledAbstract_shouldTransformTableLabelInParagraphLabel() {
+        val resultWithTables =
+            "This\tthis\tT\tTh\tThi\tThis\ts\tis\this\tThis\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                "study\tstudy\ts\tst\tstu\tstud\ty\tdy\tudy\ttudy\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "was\twas\tw\twa\twas\twas\ts\tas\twas\twas\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "supported\tsupported\ts\tsu\tsup\tsupp\td\ted\tted\trted\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "by\tby\tb\tby\tby\tby\ty\tby\tby\tby\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t0\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "South\tsouth\tS\tSo\tSou\tSout\th\tth\tuth\touth\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Asian\tasian\tA\tAs\tAsi\tAsia\tn\tan\tian\tsian\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Clinical\tclinical\tC\tCl\tCli\tClin\tl\tal\tcal\tical\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Toxicology\ttoxicology\tT\tTo\tTox\tToxi\ty\tgy\togy\tlogy\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t1\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t2\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Collaboration\tcollaboration\tC\tCo\tCol\tColl\tn\ton\tion\ttion\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t2\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tCOMMA\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "which\twhich\tw\twh\twhi\twhic\th\tch\tich\thich\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "is\tis\ti\tis\tis\tis\ts\tis\tis\tis\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "funded\tfunded\tf\tfu\tfun\tfund\td\ted\tded\tnded\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "by\tby\tb\tby\tby\tby\ty\tby\tby\tby\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t3\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Wellcome\twellcome\tW\tWe\tWel\tWell\te\tme\tome\tcome\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Trust\ttrust\tT\tTr\tTru\tTrus\tt\tst\tust\trust\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "National\tnational\tN\tNa\tNat\tNati\tl\tal\tnal\tonal\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t4\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Health\thealth\tH\tHe\tHea\tHeal\th\tth\tlth\talth\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Medical\tmedical\tM\tMe\tMed\tMedi\tl\tal\tcal\tical\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t5\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Council\tcouncil\tC\tCo\tCou\tCoun\tl\til\tcil\tncil\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "International\tinternational\tI\tIn\tInt\tInte\tl\tal\tnal\tonal\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Collaborative\tcollaborative\tC\tCo\tCol\tColl\te\tve\tive\ttive\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t6\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Research\tresearch\tR\tRe\tRes\tRese\th\tch\trch\tarch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t7\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "Grant\tgrant\tG\tGr\tGra\tGran\tt\tnt\tant\trant\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t7\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "GR071669MA\tgr071669ma\tG\tGR\tGR0\tGR07\tA\tMA\t9MA\t69MA\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tCONTAINSDIGITS\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "funding\tfunding\tf\tfu\tfun\tfund\tg\tng\ting\tding\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "bodies\tbodies\tb\tbo\tbod\tbodi\ts\tes\ties\tdies\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "had\thad\th\tha\thad\thad\td\tad\thad\thad\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "no\tno\tn\tno\tno\tno\to\tno\tno\tno\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "role\trole\tr\tro\trol\trole\te\tle\tole\trole\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "analyzing\tanalyzing\ta\tan\tana\tanal\tg\tng\ting\tzing\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "or\tor\to\tor\tor\tor\tr\tor\tor\tor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "interpreting\tinterpreting\ti\tin\tint\tinte\tg\tng\ting\tting\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "data\tdata\td\tda\tdat\tdata\ta\tta\tata\tdata\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t10\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "or\tor\to\tor\tor\tor\tr\tor\tor\tor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "writing\twriting\tw\twr\twri\twrit\tg\tng\ting\tting\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                "article\tarticle\ta\tar\tart\tarti\te\tle\tcle\ticle\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t11\t10\t0\tNUMBER\t0\t0\t<table>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t11\t10\t0\tNUMBER\t0\t0\t<table>"
+
+        val postprocessed = LabelUtils.postProcessFullTextLabeledText(resultWithTables)
+
+        assertThat(
+            Arrays.stream(StringUtils.split(postprocessed, "\n"))
+                .filter { l -> l.endsWith("<table>") }
+                .count(), `is`(0L)
+        )
+
+        assertThat(
+            Arrays.stream(StringUtils.split(postprocessed, "\n"))
+                .filter { l -> l.endsWith("<paragraph>") }
+                .count(), `is`(
+                Arrays.stream(StringUtils.split(resultWithTables, "\n"))
+                    .filter { l -> l.endsWith("<table>") }
+                    .count())
+        )
+    }
+
+//    fun testAdjustInvalidSequenceOfStartLabels() {
+//        val inputStream = javaClass.getResourceAsStream("bodyResults-sample.1.txt")
+//        val bodyResult = inputStream?.bufferedReader().use { it.readText() }
+//
+//        val postProcessed = LabelUtils.postProcessFullTextLabeledText(bodyResult)
+//    }
+
+    @Test
+    fun testPostProcessFulltextFixInvalidTableOrFigure_noChangeNeeded_shouldReturnSameTableOrFigureSequence() {
+        val bodyResult =
+            "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t11\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKEND\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t11\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "014306\t014306\t0\t01\t014\t0143\t6\t06\t306\t4306\tBLOCKSTART\tLINESTART\tLINEINDENT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t8\t11\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "-\t-\t-\t-\t-\t-\t-\t-\t-\t-\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tHYPHEN\t8\t11\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "4\t4\t4\t4\t4\t4\t4\t4\t4\t4\tBLOCKEND\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t11\t0\tNUMBER\t1\t0\t<paragraph>\n" +
+                "FIG\tfig\tF\tFI\tFIG\tFIG\tG\tIG\tFIG\tFIG\tBLOCKSTART\tLINESTART\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t0\tI-<figure_marker>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t3\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "average\taverage\ta\tav\tave\taver\te\tge\tage\trage\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "distances\tdistances\td\tdi\tdis\tdist\ts\tes\tces\tnces\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "between\tbetween\tb\tbe\tbet\tbetw\tn\ten\teen\tween\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "nucleons\tnucleons\tn\tnu\tnuc\tnucl\ts\tns\tons\teons\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t1\t<paragraph>\n" +
+                "Be\tbe\tB\tBe\tBe\tBe\te\tBe\tBe\tBe\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t1\t<paragraph>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tLINEINDENT\tNEWFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t0\t<paragraph>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEEND\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tCOMMA\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t1\t<paragraph>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "3\t3\t3\t3\t3\t3\t3\t3\t3\t3\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t0\t<paragraph>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t1\t<paragraph>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tCOMMA\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t1\tI-<citation_marker>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\tI-<paragraph>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t0\t<paragraph>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t1\t<paragraph>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tCOMMA\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t1\t<paragraph>\n" +
+                "C\tc\tC\tC\tC\tC\tC\tC\tC\tC\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t1\t<paragraph>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t1\t0\t<paragraph>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "states\tstates\ts\tst\tsta\tstat\ts\tes\ttes\tates\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "solid\tsolid\ts\tso\tsol\tsoli\td\tid\tlid\tolid\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "lines\tlines\tl\tli\tlin\tline\ts\tes\tnes\tines\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "denote\tdenote\td\tde\tden\tdeno\te\tte\tote\tnote\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "average\taverage\ta\tav\tave\taver\te\tge\tage\trage\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "distances\tdistances\td\tdi\tdis\tdist\ts\tes\tces\tnces\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "r\tr\tr\tr\tr\tr\tr\tr\tr\tr\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tSAMEFONTSIZE\t0\t1\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "N\tn\tN\tN\tN\tN\tN\tN\tN\tN\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t1\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t1\tALLCAP\tNODIGIT\t1\tCOMMA\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "N\tn\tN\tN\tN\tN\tN\tN\tN\tN\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t1\tALLCAP\tNODIGIT\t1\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "between\tbetween\tb\tbe\tbet\tbetw\tn\ten\teen\tween\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tHIGHERFONT\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "two\ttwo\tt\ttw\ttwo\ttwo\to\two\ttwo\ttwo\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "valence\tvalence\tv\tva\tval\tvale\te\tce\tnce\tence\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "nucleons\tnucleons\tn\tnu\tnuc\tnucl\ts\tns\tons\teons\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t8\t3\t0\tNUMBER\t0\t0\t<paragraph>\n"
+
+        val postProcessed = LabelUtils.postProcessFulltextFixInvalidTableOrFigure(bodyResult)
+
+        assertThat(postProcessed, `is`(bodyResult))
+    }
+
+    @Test
+    fun testPostProcessFulltextFixInvalidTableOrFigure_singleChangeNeeded_shouldCorrectTheTableOrFigureSequence() {
+        val bodyResult =
+            "of\tof\to\tof\tof\tof\tf\tof\tof\tof\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t1\tI-<citation_marker>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t0\t<figure>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t1\t<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t0\t<figure>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "state\tstate\ts\tst\tsta\tstat\te\tte\tate\ttate\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "of\tof\to\tof\tof\tof\tf\tof\tof\tof\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t1\t<figure>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "panels\tpanels\tp\tpa\tpan\tpane\ts\tls\tels\tnels\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "a\ta\ta\ta\ta\ta\ta\ta\ta\ta\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "are\tare\ta\tar\tare\tare\te\tre\tare\tare\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n"
+
+
+        val postProcessed = LabelUtils.postProcessFulltextFixInvalidTableOrFigure(bodyResult)
+
+        assertThat(postProcessed, not(bodyResult))
+
+        val splitResult =
+            Arrays.stream(postProcessed.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map<List<String>> { l: String ->
+                    Arrays.stream(
+                        l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()
+                    )
+                        .collect(Collectors.toList())
+                }
+                .collect(Collectors.toList())
+
+        val countStartingFigure = splitResult.stream()
+            .map { l: List<String> -> l.last() }
+            .filter { l: String -> l.equals("I-<figure>") }
+            .count()
+
+        assertThat(countStartingFigure, `is`(1))
+    }
+
+    @Test
+    fun testPostProcessFulltextFixInvalidTableOrFigure_MultipleChangeNeeded_shouldCorrectTheTableOrFigureSequence() {
+        val bodyResult =
+            "of\tof\to\tof\tof\tof\tf\tof\tof\tof\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<paragraph>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t1\tI-<citation_marker>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t0\t<figure>\n" +
+                "+\t+\t+\t+\t+\t+\t+\t+\t+\t+\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tNEWFONT\tLOWERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t1\t<figure>\n" +
+                "1\t1\t1\t1\t1\t1\t1\t1\t1\t1\tBLOCKIN\tLINESTART\tLINEINDENT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t0\t<figure>\n" +
+                "0\t0\t0\t0\t0\t0\t0\t0\t0\t0\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tNOCAPS\tALLDIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "state\tstate\ts\tst\tsta\tstat\te\tte\tate\ttate\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "of\tof\to\tof\tof\tof\tf\tof\tof\tof\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tLOWERFONT\t0\t0\tNOCAPS\tALLDIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t1\t1\t<figure>\n" +
+                "B\tb\tB\tB\tB\tB\tB\tB\tB\tB\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tHIGHERFONT\t0\t0\tALLCAP\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tDOT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "panels\tpanels\tp\tpa\tpan\tpane\ts\tls\tels\tnels\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tOPENBRACKET\t9\t5\t0\tNUMBER\t0\t0\tI-<figure>\n" +
+                "a\ta\ta\ta\ta\ta\ta\ta\ta\ta\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\tENDBRACKET\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "are\tare\ta\tar\tare\tare\te\tre\tare\tare\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<figure>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\tI-<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n" +
+                "calculated\tcalculated\tc\tca\tcal\tcalc\td\ted\tted\tated\tBLOCKIN\tLINEIN\tLINEINDENT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\tNOPUNCT\t9\t5\t0\tNUMBER\t0\t0\t<table>\n"
+
+
+        val postProcessed = LabelUtils.postProcessFulltextFixInvalidTableOrFigure(bodyResult)
+
+        assertThat(postProcessed, not(bodyResult))
+
+        val splitResult =
+            Arrays.stream<String>(postProcessed.split("\n".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray())
+                .map<List<String>> { l: String ->
+                    Arrays.stream<String>(
+                        l.split("\t".toRegex()).dropLastWhile { it.isEmpty() }.toTypedArray()
+                    )
+                        .collect(Collectors.toList<String>())
+                }
+                .collect(Collectors.toList<List<String>>())
+
+        val countStartingFigure = splitResult.stream()
+            .map { l: List<String> -> l.last() }
+            .filter { l: String -> l.equals("I-<figure>") }
+            .count()
+
+        assertThat(countStartingFigure, `is`(2))
+
+        val countStartingTables = splitResult.stream()
+            .map { l: List<String> -> l.last() }
+            .filter { l: String -> l.equals("I-<table>") }
+            .count()
+
+        assertThat(countStartingTables, `is`(1))
+
+    }
+
+    @Test
+    fun testPostProcessFulltextCorrectSequencesWithoutInitialToken_shouldChangeAbstractLabelInAvailabilityLabel() {
+        val resultHeader = "Data\tdata\tD\tDa\tDat\tData\ta\tta\tata\tData\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tSAMEFONTSIZE\t1\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<other>\n" +
+            "Availability\tavailability\tA\tAv\tAva\tAvai\ty\tty\tity\tlity\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t1\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<other>\n" +
+            "Statement\tstatement\tS\tSt\tSta\tStat\tt\tnt\tent\tment\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t1\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<other>\n" +
+            ":\t:\t:\t:\t:\t:\t:\t:\t:\t:\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t1\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tPUNCT\t0\t0\t1\t0\t<other>\n" +
+            "The\tthe\tT\tTh\tThe\tThe\te\the\tThe\tThe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\tI-<availability>\n" +
+            "raw\traw\tr\tra\traw\traw\tw\taw\traw\traw\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "sequencing\tsequencing\ts\tse\tseq\tsequ\tg\tng\ting\tcing\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "reads\treads\tr\tre\trea\tread\ts\tds\tads\teads\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "metagenomic\tmetagenomic\tm\tme\tmet\tmeta\tc\tic\tmic\tomic\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "samples\tsamples\ts\tsa\tsam\tsamp\ts\tes\tles\tples\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "used\tused\tu\tus\tuse\tused\td\ted\tsed\tused\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "this\tthis\tt\tth\tthi\tthis\ts\tis\this\tthis\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "study\tstudy\ts\tst\tstu\tstud\ty\tdy\tudy\ttudy\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "were\twere\tw\twe\twer\twere\te\tre\tere\twere\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "downloaded\tdownloaded\td\tdo\tdow\tdown\td\ted\tded\taded\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "from\tfrom\tf\tfr\tfro\tfrom\tm\tom\trom\tfrom\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "public\tpublic\tp\tpu\tpub\tpubl\tc\tic\tlic\tblic\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "repositories\trepositories\tr\tre\trep\trepo\ts\tes\ties\tries\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "listed\tlisted\tl\tli\tlis\tlist\td\ted\tted\tsted\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "following\tfollowing\tf\tfo\tfol\tfoll\tg\tng\ting\twing\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "publications\tpublications\tp\tpu\tpub\tpubl\ts\tns\tons\tions\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            ":\t:\t:\t:\t:\t:\t:\t:\t:\t:\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "1038\t1038\t1\t10\t103\t1038\t8\t38\t038\t1038\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "nature11209\tnature11209\tn\tna\tnat\tnatu\t9\t09\t209\t1209\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tCONTAINSDIGITS\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "1038\t1038\t1\t10\t103\t1038\t8\t38\t038\t1038\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "nature11450\tnature11450\tn\tna\tnat\tnatu\t0\t50\t450\t1450\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tCONTAINSDIGITS\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "1016\t1016\t1\t10\t101\t1016\t6\t16\t016\t1016\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "j\tj\tj\tj\tj\tj\tj\tj\tj\tj\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t1\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "cels\tcels\tc\tce\tcel\tcels\ts\tls\tels\tcels\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "2016\t2016\t2\t20\t201\t2016\t6\t16\t016\t2016\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "004\t004\t0\t00\t004\t004\t4\t04\t004\t004\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "1101\t1101\t1\t11\t110\t1101\t1\t01\t101\t1101\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "gr\tgr\tg\tgr\tgr\tgr\tr\tgr\tgr\tgr\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "233940\t233940\t2\t23\t233\t2339\t0\t40\t940\t3940\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "117\t117\t1\t11\t117\t117\t7\t17\t117\t117\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "Data\tdata\tD\tDa\tDat\tData\ta\tta\tata\tData\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "underlying\tunderlying\tu\tun\tund\tunde\tg\tng\ting\tying\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "all\tall\ta\tal\tall\tall\tl\tll\tall\tall\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "figures\tfigures\tf\tfi\tfig\tfigu\ts\tes\tres\tures\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "such\tsuch\ts\tsu\tsuc\tsuch\th\tch\tuch\tsuch\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "as\tas\ta\tas\tas\tas\ts\tas\tas\tas\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "numerical\tnumerical\tn\tnu\tnum\tnume\tl\tal\tcal\tical\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "values\tvalues\tv\tva\tval\tvalu\ts\tes\tues\tlues\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "of\tof\to\tof\tof\tof\tf\tof\tof\tof\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "bar\tbar\tb\tba\tbar\tbar\tr\tar\tbar\tbar\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "plots\tplots\tp\tpl\tplo\tplot\ts\tts\tots\tlots\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "can\tcan\tc\tca\tcan\tcan\tn\tan\tcan\tcan\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "be\tbe\tb\tbe\tbe\tbe\te\tbe\tbe\tbe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "found\tfound\tf\tfo\tfou\tfoun\td\tnd\tund\tound\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "5281\t5281\t5\t52\t528\t5281\t1\t81\t281\t5281\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "zenodo\tzenodo\tz\tze\tzen\tzeno\to\tdo\todo\tnodo\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "10304481\t10304481\t1\t10\t103\t1030\t1\t81\t481\t4481\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "All\tall\tA\tAl\tAll\tAll\tl\tll\tAll\tAll\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "other\tother\to\tot\toth\tothe\tr\ter\ther\tther\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "metadata\tmetadata\tm\tme\tmet\tmeta\ta\tta\tata\tdata\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "as\tas\ta\tas\tas\tas\ts\tas\tas\tas\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "well\twell\tw\twe\twel\twell\tl\tll\tell\twell\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "as\tas\ta\tas\tas\tas\ts\tas\tas\tas\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "source\tsource\ts\tso\tsou\tsour\te\tce\trce\turce\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "code\tcode\tc\tco\tcod\tcode\te\tde\tode\tcode\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "for\tfor\tf\tfo\tfor\tfor\tr\tor\tfor\tfor\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "the\tthe\tt\tth\tthe\tthe\te\the\tthe\tthe\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "sequencing\tsequencing\ts\tse\tseq\tsequ\tg\tng\ting\tcing\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "pipeline\tpipeline\tp\tpi\tpip\tpipe\te\tne\tine\tline\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "downstream\tdownstream\td\tdo\tdow\tdown\tm\tam\team\tream\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "analyses\tanalyses\ta\tan\tana\tanal\ts\tes\tses\tyses\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ",\t,\t,\t,\t,\t,\t,\t,\t,\t,\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tCOMMA\t0\t0\t1\t0\t<abstract>\n" +
+            "and\tand\ta\tan\tand\tand\td\tnd\tand\tand\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "figure\tfigure\tf\tfi\tfig\tfigu\te\tre\ture\tgure\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "generation\tgeneration\tg\tge\tgen\tgene\tn\ton\tion\ttion\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "are\tare\ta\tar\tare\tare\te\tre\tare\tare\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "available\tavailable\ta\tav\tava\tavai\te\tle\tble\table\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "at\tat\ta\tat\tat\tat\tt\tat\tat\tat\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "Zenodo\tzenodo\tZ\tZe\tZen\tZeno\to\tdo\todo\tnodo\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tOPENBRACKET\t0\t0\t1\t0\t<abstract>\n" +
+            "10\t10\t1\t10\t10\t10\t0\t10\t10\t10\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<abstract>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<abstract>\n" +
+            "5281\t5281\t5\t52\t528\t5281\t1\t81\t281\t5281\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "zenodo\tzenodo\tz\tze\tzen\tzeno\to\tdo\todo\tnodo\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<availability>\n" +
+            "10368227\t10368227\t1\t10\t103\t1036\t7\t27\t227\t8227\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tALLDIGIT\t0\t0\t0\t1\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tENDBRACKET\t0\t0\t1\t0\t<availability>\n" +
+            "or\tor\to\tor\tor\tor\tr\tor\tor\tor\tBLOCKEND\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "GitHub\tgithub\tG\tGi\tGit\tGitH\tb\tub\tHub\ttHub\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "(\t(\t(\t(\t(\t(\t(\t(\t(\t(\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tOPENBRACKET\t0\t0\t1\t0\t<availability>\n" +
+            "https\thttps\th\tht\thtt\thttp\ts\tps\ttps\tttps\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            ":\t:\t:\t:\t:\t:\t:\t:\t:\t:\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "github\tgithub\tg\tgi\tgit\tgith\tb\tub\thub\tthub\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tDOT\t0\t0\t1\t0\t<availability>\n" +
+            "com\tcom\tc\tco\tcom\tcom\tm\tom\tcom\tcom\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "zhiru\tzhiru\tz\tzh\tzhi\tzhir\tu\tru\tiru\thiru\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<availability>\n" +
+            "-\t-\t-\t-\t-\t-\t-\t-\t-\t-\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tHYPHEN\t0\t0\t1\t0\t<copyright>\n" +
+            "liu\tliu\tl\tli\tliu\tliu\tu\tiu\tliu\tliu\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<copyright>\n" +
+            "/\t/\t/\t/\t/\t/\t/\t/\t/\t/\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<copyright>\n" +
+            "microbiome_\tmicrobiome_\tm\tmi\tmic\tmicr\t_\te_\tme_\tome_\tBLOCKIN\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t0\t0\t0\t0\t0\t1\tNOPUNCT\t0\t0\t1\t0\t<copyright>\n" +
+            "evolution\tevolution\te\tev\tevo\tevol\tn\ton\tion\ttion\tBLOCKIN\tLINESTART\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<copyright>\n" +
+            ")\t)\t)\t)\t)\t)\t)\t)\t)\t)\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tENDBRACKET\t0\t0\t1\t0\t<copyright>\n" +
+            ".\t.\t.\t.\t.\t.\t.\t.\t.\t.\tBLOCKEND\tLINEEND\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tDOT\t0\t0\t1\t0\t<copyright>\n" +
+            "Funding\tfunding\tF\tFu\tFun\tFund\tg\tng\ting\tding\tBLOCKSTART\tLINESTART\tALIGNEDLEFT\tNEWFONT\tSAMEFONTSIZE\t1\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<other>\n" +
+            ":\t:\t:\t:\t:\t:\t:\t:\t:\t:\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t1\t0\tALLCAP\tNODIGIT\t1\t0\t0\t0\t0\t0\t0\t0\tPUNCT\t0\t0\t1\t0\t<other>\n" +
+            "This\tthis\tT\tTh\tThi\tThis\ts\tis\this\tThis\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tNEWFONT\tSAMEFONTSIZE\t0\t0\tINITCAP\tNODIGIT\t0\t0\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\tI-<funding>\n" +
+            "work\twork\tw\two\twor\twork\tk\trk\tork\twork\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<funding>\n" +
+            "was\twas\tw\twa\twas\twas\ts\tas\twas\twas\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<funding>\n" +
+            "supported\tsupported\ts\tsu\tsup\tsupp\td\ted\tted\trted\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t0\t1\t0\t0\t0\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<funding>\n" +
+            "in\tin\ti\tin\tin\tin\tn\tin\tin\tin\tBLOCKIN\tLINEIN\tALIGNEDLEFT\tSAMEFONT\tSAMEFONTSIZE\t0\t0\tNOCAPS\tNODIGIT\t0\t1\t1\t0\t0\t1\t0\t0\tNOPUNCT\t0\t0\t1\t0\t<funding>"
+        val postprocessed = LabelUtils.postProcessFulltextCorrectSequencesWithoutInitialToken(resultHeader)
+
+        assertThat(
+            Arrays.stream(StringUtils.split(postprocessed, "\n"))
+                .filter { l -> l.endsWith("<abstract>") }
+                .count(), `is`(0L)
+        )
+        assertThat(
+            Arrays.stream(StringUtils.split(postprocessed, "\n"))
+                .filter { l -> l.endsWith("<copyright>") }
+                .count(), `is`(0L)
+        )
+
+        assertThat(
+            Arrays.stream(StringUtils.split(postprocessed, "\n"))
+                .filter { l -> l.endsWith("<availability>") }
+                .count(), `is`(139)
+        )
+    }
+
+
+    companion object {
+        @JvmStatic
+        @BeforeAll
+        @Throws(Exception::class)
+        fun before() {
+            val modelParameters = ModelParameters()
+            modelParameters.name = "bao"
+            GrobidProperties.addModel(modelParameters)
+        }
+    }
+
+}
\ No newline at end of file