Merge pull request #562 from kermitt2/512_PDF_and_authors_with_ORCID

Check whether annotation is orcid and add orcid to author.

grobid-core/src/main/java/org/grobid/core/data/BiblioItem.java

@@ -3838,6 +3838,10 @@ public String toTEIAuthorBlock(int nbTag, GrobidAnalysisConfig config) {
                         TextUtilities.appendN(tei, '\t', nbTag + 1);
                         tei.append("<email>" + TextUtilities.HTMLEncode(author.getEmail()) + "</email>\n");
                     }
+                    if (author.getORCID() != null) {
+                        TextUtilities.appendN(tei, '\t', nbTag + 1);
+                        tei.append("<idno type=\"ORCID\">" + TextUtilities.HTMLEncode(author.getORCID()) + "</idno>\n");
+                    }
 
                     if (author.getAffiliations() != null) {
                         for (Affiliation aff : author.getAffiliations()) {

grobid-core/src/main/java/org/grobid/core/data/Person.java

@@ -27,6 +27,7 @@ public class Person {
     private String title = null;
     private String suffix = null;
     private String rawName = null; // raw full name if relevant/available, e.g. name exactly as displayed
+    private String orcid = null;
     private boolean corresp = false;
     private List<LayoutToken> layoutTokens = new ArrayList<>();
 
@@ -103,6 +104,14 @@ public void setCorresp(boolean b) {
         corresp = b;
     }
 
+    public String getORCID() {
+        return orcid;
+    }
+
+    public void setORCID(String id) {
+        orcid = id;
+    }
+
     public List<String> getAffiliationBlocks() {
         return affiliationBlocks;
     }
@@ -182,6 +191,9 @@ public String toString() {
         if (email != null) {
             res += " (email:" + email + ")";
         }
+        if (orcid != null) {
+            res += " (orcid:" + orcid + ")";
+        }
         if (affiliations != null) {
             for(Affiliation aff : affiliations) {
                 res += " (affiliation: " + aff.toString() + ") ";

grobid-core/src/main/java/org/grobid/core/engines/AuthorParser.java

@@ -9,7 +9,9 @@
 import org.grobid.core.engines.tagging.TaggerFactory;
 import org.grobid.core.exceptions.GrobidException;
 import org.grobid.core.features.FeaturesVectorName;
+import org.grobid.core.layout.BoundingBox;
 import org.grobid.core.layout.LayoutToken;
+import org.grobid.core.layout.PDFAnnotation;
 import org.grobid.core.lexicon.Lexicon;
 import org.grobid.core.tokenization.TaggingTokenCluster;
 import org.grobid.core.tokenization.TaggingTokenClusteror;
@@ -28,6 +30,7 @@
 import java.util.ArrayList;
 import java.util.List;
 import java.util.StringTokenizer;
+import java.util.regex.Matcher;
 
 /**
  * @author Patrice Lopez
@@ -54,14 +57,14 @@ public List<Person> processingCitation(String input) throws Exception {
 
         // for language to English for the analyser to avoid any bad surprises
         List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(input, new Language("en", 1.0));
-        return processing(tokens, false);
+        return processing(tokens, null, false);
     }
 
     public List<Person> processingCitationLayoutTokens(List<LayoutToken> tokens) throws Exception {
         if (CollectionUtils.isEmpty(tokens)) {
             return null;
         }
-        return processing(tokens, false);
+        return processing(tokens, null, false);
     }
 
     /**
@@ -76,11 +79,11 @@ public List<Person> processingHeader(String input) throws Exception {
 
         // for language to English for the analyser to avoid any bad surprises
         List<LayoutToken> tokens = GrobidAnalyzer.getInstance().tokenizeWithLayoutToken(input, new Language("en", 1.0));
-        return processing(tokens, true);
+        return processing(tokens, null, true);
     }
 
-    public List<Person> processingHeaderWithLayoutTokens(List<LayoutToken> inputs) {
-        return processing(inputs, true);
+    public List<Person> processingHeaderWithLayoutTokens(List<LayoutToken> inputs, List<PDFAnnotation> pdfAnnotations) {
+        return processing(inputs, pdfAnnotations, true);
     }
 
     /**
@@ -90,7 +93,7 @@ public List<Person> processingHeaderWithLayoutTokens(List<LayoutToken> inputs) {
      * @param head - if true use the model for header's name, otherwise the model for names in citation
      * @return List of identified Person entites as POJO.
      */
-    public List<Person> processing(List<LayoutToken> tokens, boolean head) {
+    public List<Person> processing(List<LayoutToken> tokens, List<PDFAnnotation> pdfAnnotations, boolean head) {
         if (CollectionUtils.isEmpty(tokens)) {
             return null;
         }
@@ -116,13 +119,35 @@ public List<Person> processing(List<LayoutToken> tokens, boolean head) {
                     continue;
                 }
 
+                if(pdfAnnotations !=null) {
+                    for (LayoutToken authorsToken : cluster.concatTokens()) {
+                        for (PDFAnnotation pdfAnnotation : pdfAnnotations) {
+                            BoundingBox intersectBox = pdfAnnotation.getIntersectionBox(authorsToken);
+                            if (intersectBox != null) {
+                                BoundingBox authorsBox = BoundingBox.fromLayoutToken(authorsToken);
+                                if (intersectBox.equals(authorsBox)) {
+                                } else {
+                                    double pixPerChar = authorsToken.getWidth() / authorsToken.getText().length();
+                                    int charsCovered = (int) ((intersectBox.getWidth() / pixPerChar) + 0.5);
+                                    // !! here we consider the annot is at the tail or end of the names
+                                    String newToken = authorsToken.getText().substring(0, authorsToken.getText().length() - charsCovered);
+                                    Matcher orcidMatcher = TextUtilities.ORCIDPattern.matcher(pdfAnnotation.getDestination());
+                                    if (orcidMatcher.find()) {
+                                        aut.setORCID(orcidMatcher.group(1) + "-"
+                                            + orcidMatcher.group(2) + "-" + orcidMatcher.group(3)+ "-" + orcidMatcher.group(4));
+                                        authorsToken.setText(newToken);
+                                    }
+                                }
+                            }
+                        }
+                    }
+                }
                 TaggingLabel clusterLabel = cluster.getTaggingLabel();
                 Engine.getCntManager().i(clusterLabel);
                 //String clusterContent = LayoutTokensUtil.normalizeText(LayoutTokensUtil.toText(cluster.concatTokens()));
                 String clusterContent = StringUtils.normalizeSpace(LayoutTokensUtil.toText(cluster.concatTokens()));
                 if (clusterContent.trim().length() == 0)
                     continue;
-
                 if (clusterLabel.equals(TaggingLabels.NAMES_HEADER_MARKER)) {
                     // a marker introduces a new author, and the marker could be attached to the previous (usual) 
                     // or following author (rare)

grobid-core/src/main/java/org/grobid/core/engines/HeaderParser.java

@@ -25,7 +25,9 @@
 import org.grobid.core.lang.Language;
 import org.grobid.core.lexicon.Lexicon;
 import org.grobid.core.layout.Block;
+import org.grobid.core.layout.BoundingBox;
 import org.grobid.core.layout.LayoutToken;
+import org.grobid.core.layout.PDFAnnotation;
 import org.grobid.core.tokenization.LabeledTokensContainer;
 import org.grobid.core.tokenization.TaggingTokenCluster;
 import org.grobid.core.tokenization.TaggingTokenClusteror;

grobid-core/src/main/java/org/grobid/core/layout/PDFAnnotation.java

@@ -139,4 +139,27 @@ public boolean cover(LayoutToken token) {
 		}
 		return res;
 	}
+
+    /**
+     * Return the intersection box between token and annotation
+     */
+    public BoundingBox getIntersectionBox(LayoutToken token) {
+        if (token == null)
+            return null;
+        BoundingBox intersectBox = null;
+        int pageToken = token.getPage();
+        if (pageToken == pageNumber) {
+            BoundingBox tokenBox = BoundingBox.fromLayoutToken(token);
+            for(BoundingBox box : boundingBoxes) {
+                if (box.intersect(tokenBox)) {
+                    if (box.contains(tokenBox)) {
+                        intersectBox = tokenBox;
+                        break;
+                    }
+                    intersectBox = box.boundingBoxIntersection(tokenBox);
+                }
+            }
+        }
+        return intersectBox;
+    }
 }

grobid-core/src/main/java/org/grobid/core/utilities/TextUtilities.java

@@ -48,6 +48,10 @@ public class TextUtilities {
     public static final String SLASH = "/";
 
     // note: be careful of catastrophic backtracking here as a consequence of PDF noise! 
+
+    private static final String ORCIDRegex =
+        "^\\s*(?:(?:https?://)?orcid.org/)?([0-9]{4})\\-?([0-9]{4})\\-?([0-9]{4})\\-?([0-9]{3}[\\dX])\\s*$";
+    static public final Pattern ORCIDPattern = Pattern.compile(ORCIDRegex);
 
     // the magical DOI regular expression...
     static public final Pattern DOIPattern = Pattern

grobid-core/src/test/java/org/grobid/core/utilities/TextUtilitiesTest.java

@@ -9,12 +9,12 @@
 
 import java.util.ArrayList;
 import java.util.List;
+import java.util.regex.Matcher;
 
 import static org.hamcrest.CoreMatchers.is;
 import static org.hamcrest.CoreMatchers.startsWith;
 import static org.hamcrest.MatcherAssert.assertThat;
-import static org.junit.Assert.assertEquals;
-import static org.junit.Assert.assertNotNull;
+import static org.junit.Assert.*;
 
 @Ignore
 public class TextUtilitiesTest extends EngineTest {
@@ -390,4 +390,21 @@ public void testIsAllUpperCaseOrDigitOrDot() throws Exception {
         assertThat(TextUtilities.isAllUpperCaseOrDigitOrDot("P.C.T."), is(true));
         assertThat(TextUtilities.isAllUpperCaseOrDigitOrDot("P.C,T."), is(false));
     }
+
+    @Test
+    public void testOrcidPattern() {
+        String[] falseOrcids = {"1234", "1234-5698-137X", "0000-0001-9877-137Y","http://orcid.fr/0000-0001-9877-137X"};
+        String[] trueOrcids = {"0000-0001-9877-137X", "http://orcid.org/0000-0001-9877-137X", "orcid.org/0000-0001-9877-137X"};
+        for(String falseOrcid :  falseOrcids) {
+            Matcher orcidMatcher = TextUtilities.ORCIDPattern.matcher(falseOrcid);
+            assertFalse (orcidMatcher.find());
+        }
+        for(String trueOrcid :  trueOrcids) {
+            Matcher orcidMatcher = TextUtilities.ORCIDPattern.matcher(trueOrcid);
+            if (orcidMatcher.find()) {
+                assertThat(orcidMatcher.group(1) + "-"
+                    + orcidMatcher.group(2) + "-" + orcidMatcher.group(3) + "-" + orcidMatcher.group(4) , is("0000-0001-9877-137X"));
+            }
+        }
+    }
 }