remove parentheses often seen in callouts, for better callout to refe…

…rence match
kermitt2 · Aug 2, 2020 · 11cbbcf · 11cbbcf
1 parent ff10968
commit 11cbbcf
Showing 1 changed file with 2 additions and 1 deletion.
diff --git a/grobid-core/src/main/java/org/grobid/core/utilities/matching/ReferenceMarkerMatcher.java b/grobid-core/src/main/java/org/grobid/core/utilities/matching/ReferenceMarkerMatcher.java
@@ -386,7 +386,8 @@ private static int matchCount(List<LayoutToken> toks, Pattern p) {
     private List<BibDataSet> postFilterMatches(String c, List<BibDataSet> matches) {
         if (c.toLowerCase().contains("et al") || c.toLowerCase().contains(" and ")) {
             String[] sp = c.trim().split(" ");
-            final String author = sp[0].toLowerCase();
+            //callouts often include parentheses as seen in https://grobid.readthedocs.io/en/latest/training/fulltext/
+            final String author = sp[0].replaceAll("[\\(\\[]", "").toLowerCase();
 
             ArrayList<BibDataSet> bibDataSets = Lists.newArrayList(Iterables.filter(matches, new Predicate<BibDataSet>() {
                 @Override