Revision: 18308
http://sourceforge.net/p/gate/code/18308
Author: markagreenwood
Date: 2014-09-11 09:27:18 +0000 (Thu, 11 Sep 2014)
Log Message:
-----------
added code to handle 4 digit years; the parser doesn't touch these as it can't
know if they are a date or not without context, but as we know the text is
under a date annotation we can added the appropriate features in the PR by hand
Modified Paths:
--------------
gate/trunk/plugins/Tagger_DateNormalizer/src/gate/creole/dates/DateAnnotationNormalizer.java
Modified:
gate/trunk/plugins/Tagger_DateNormalizer/src/gate/creole/dates/DateAnnotationNormalizer.java
===================================================================
---
gate/trunk/plugins/Tagger_DateNormalizer/src/gate/creole/dates/DateAnnotationNormalizer.java
2014-09-11 01:20:00 UTC (rev 18307)
+++
gate/trunk/plugins/Tagger_DateNormalizer/src/gate/creole/dates/DateAnnotationNormalizer.java
2014-09-11 09:27:18 UTC (rev 18308)
@@ -23,8 +23,8 @@
import gate.creole.metadata.RunTime;
import java.text.DateFormat;
+import java.util.Calendar;
import java.util.Date;
-import java.util.regex.Pattern;
import mark.util.DateParser;
import mark.util.ParsePositionEx;
@@ -56,6 +56,7 @@
return wholeMatchOnly;
}
+ @SuppressWarnings("deprecation")
@Override
protected void annotate(Date documentDate, DateParser dp, DateFormat df)
throws ExecutionException {
AnnotationSet dates =
document.getAnnotations(getInputASName()).get(getAnnotationName());
@@ -84,8 +85,28 @@
if(d == null) {
//if the text didn't parse skip on to the next character and try
again
- //start++;
- continue;
+ if (text.matches("[0-9]{4}")) {
+ d = new Date(Integer.parseInt(text)-1900,0,1);
+
+ pp.setIndex(4);
+ pp.getFeatures().put("inferred", DateParser.MONTH |
DateParser.DAY);
+
+ Calendar cal = Calendar.getInstance();
+ cal.setTime(documentDate != null ? documentDate : new Date());
+
+ Calendar parsed = Calendar.getInstance();
+ cal.setTime(d);
+
+ String relative = "present";
+ if (parsed.before(cal)) relative = "past";
+ if (parsed.after(cal)) relative = "future";
+
+ pp.getFeatures().put("relative", relative);
+
+ }
+ else {
+ continue;
+ }
}
if (wholeMatchOnly && pp.getIndex() != text.length()) {
This was sent by the SourceForge.net collaborative development platform, the
world's largest Open Source development site.
------------------------------------------------------------------------------
Want excitement?
Manually upgrade your production database.
When you want reliability, choose Perforce
Perforce version control. Predictably reliable.
http://pubads.g.doubleclick.net/gampad/clk?id=157508191&iu=/4140/ostg.clktrk
_______________________________________________
GATE-cvs mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/gate-cvs
