Repository: tika
Updated Branches:
  refs/heads/master c93ff3e1c -> c9bc910c5


Grobid Quantities parser extracts types


Project: http://git-wip-us.apache.org/repos/asf/tika/repo
Commit: http://git-wip-us.apache.org/repos/asf/tika/commit/e2e10f9a
Tree: http://git-wip-us.apache.org/repos/asf/tika/tree/e2e10f9a
Diff: http://git-wip-us.apache.org/repos/asf/tika/diff/e2e10f9a

Branch: refs/heads/master
Commit: e2e10f9ac76bd0566a33dae71a666bb2f7746f24
Parents: 9f09a55
Author: cmenekse <[email protected]>
Authored: Mon May 2 10:28:24 2016 -0700
Committer: cmenekse <[email protected]>
Committed: Mon May 2 10:28:24 2016 -0700

----------------------------------------------------------------------
 .../apache/tika/parser/ner/grobid/GrobidNERecogniser.java   | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)
----------------------------------------------------------------------


http://git-wip-us.apache.org/repos/asf/tika/blob/e2e10f9a/tika-parsers/src/main/java/org/apache/tika/parser/ner/grobid/GrobidNERecogniser.java
----------------------------------------------------------------------
diff --git 
a/tika-parsers/src/main/java/org/apache/tika/parser/ner/grobid/GrobidNERecogniser.java
 
b/tika-parsers/src/main/java/org/apache/tika/parser/ner/grobid/GrobidNERecogniser.java
index d2e0bc4..945bbb5 100644
--- 
a/tika-parsers/src/main/java/org/apache/tika/parser/ner/grobid/GrobidNERecogniser.java
+++ 
b/tika-parsers/src/main/java/org/apache/tika/parser/ner/grobid/GrobidNERecogniser.java
@@ -28,6 +28,7 @@ public class GrobidNERecogniser implements NERecogniser{
         add("MEASUREMENT_UNITS");
         add("MEASUREMENTS");
         add("NORMALIZED_MEASUREMENTS");
+        add("MEASUREMENT_TYPES");
     }};
 
 
@@ -140,6 +141,7 @@ public class GrobidNERecogniser implements NERecogniser{
         Set<String> unitSet = new HashSet<String>();
         Set<String> measurementSet = new HashSet<String>();
         Set<String> normalizedMeasurementSet = new HashSet<String>();
+        Set<String> measurementTypeSet = new HashSet<String>();
         
         try {
             String url = restHostUrlStr + readRestEndpoint();
@@ -169,9 +171,12 @@ public class GrobidNERecogniser implements NERecogniser{
                                                        
normalizedMeasurementString.append(normalizedMeasurementNumber);
                                                        
normalizedMeasurementString.append(" ");
                                                }
+                                               if 
(quantity.containsKey("type")) {
+                                                       String measurementType 
= (String) convertToJSONObject(quantity.toString()).get("type");
+                                                       
measurementTypeSet.add(measurementType);
+                                               }
 
                                                JSONObject jsonObj = 
(JSONObject) convertToJSONObject(quantity.toString());
-
                                                if 
(jsonObj.containsKey("rawUnit")) {
                                                        JSONObject rawUnit = 
(JSONObject) jsonObj.get("rawUnit");
                                                        String unitName = 
(String) convertToJSONObject(rawUnit.toString()).get("name");
@@ -195,10 +200,12 @@ public class GrobidNERecogniser implements NERecogniser{
                                        }
                        
                 }
+                
                 entities.put("MEASUREMENT_NUMBERS",measurementNumberSet);
                 entities.put("MEASUREMENT_UNITS",unitSet); 
                 entities.put("MEASUREMENTS",measurementSet);
                 
entities.put("NORMALIZED_MEASUREMENTS",normalizedMeasurementSet);
+                entities.put("MEASUREMENT_TYPES",measurementTypeSet);
                 
             }
         }

Reply via email to