fix problematic dash / minus sign which is not parsed by the NumberFo…

…rmat class
lfoppiano · Dec 13, 2022 · 1b71c43 · 1b71c43
1 parent 484cd0e
commit 1b71c43
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 1 deletion.
diff --git a/src/main/java/org/grobid/core/engines/QuantityParser.java b/src/main/java/org/grobid/core/engines/QuantityParser.java
@@ -34,6 +34,7 @@
 import java.util.ArrayList;
 import java.util.Comparator;
 import java.util.List;
+import java.util.regex.Pattern;
 import java.util.stream.Collectors;
 
 import static org.apache.commons.collections4.CollectionUtils.isEmpty;
@@ -55,6 +56,8 @@ public class QuantityParser extends AbstractParser {
     //    private EnglishTokenizer tokeniser;
     private boolean disableSubstanceParser = false;
 
+    public static Pattern MINUS_SIGN_REGEX= Pattern.compile("\u2212");
+
     public static QuantityParser getInstance(boolean disableSubstance) {
         if (instance == null) {
             instance = getNewInstance(disableSubstance);
@@ -115,6 +118,7 @@ public List<Measurement> process(List<LayoutToken> layoutTokens) {
         //Normalisation
         List<LayoutToken> layoutTokenNormalised = tokens.stream().map(layoutToken -> {
                     layoutToken.setText(UnicodeUtil.normaliseText(layoutToken.getText()));
+                    layoutToken.setText(MINUS_SIGN_REGEX.matcher(layoutToken.getText()).replaceAll("-"));
 
                     return layoutToken;
                 }

diff --git a/src/main/java/org/grobid/core/engines/ValueParser.java b/src/main/java/org/grobid/core/engines/ValueParser.java
@@ -86,6 +86,7 @@ protected BigDecimal parseValueBlock(ValueBlock block, Locale locale) {
                 try {
                     BigDecimal secondPart = null;
                     if (block.getPow() != null && block.getBase() != null) {
+
                         final Number pow = format.parse(block.getPowAsString());
                         String baseAsString = removeSpacesTabsAndBl(block.getBaseAsString());
 
@@ -112,7 +113,7 @@ protected BigDecimal parseValueBlock(ValueBlock block, Locale locale) {
                     }
 
                 } catch (ParseException | ArithmeticException | NumberFormatException e) {
-                    LOGGER.error("Cannot parse " + block.toString() + " with Locale " + locale, e);
+                    LOGGER.error("Cannot parse " + block + " with Locale " + locale, e);
                 }
 
                 break;