So, what does this mean? Do we have to wait with the 2.2.1 release until this issue (wish) is fixed?
If true, please reopen the issue.

-- Michael

Thilo Goetz wrote:
Yes, this should be fixed.  I'm pretty busy right
now, I'm moving house this week.  If I do it, it
probably won't happen before the middle of next
week.

I also noticed, just by eyeballing the code, that
the code in ArrayFSImpl is not correct: it ignores
the destOffset.  This should be fixed at the same
time.

--Thilo

Marshall Schor wrote:
I just noticed that this change introduced several methods to JCas
version of things like FSArray, which look like stubs that do nothing. For instance, there is a method in FSArray "toStringArray" which calls
an empty stub "copyToArray"

  public void copyToArray(int srcOffset, String[] dest, int destOffset,
int length)
      throws ArrayIndexOutOfBoundsException {
    // TODO Auto-generated method stub
}

  public String[] toStringArray() {
    final int size = size();
    String[] strArray = new String[size];
    copyToArray(0, strArray, 0, size);
    return strArray;
  }

The corresponding methods in ArrayFSImpl are not stubs.

Shouldn't these stubs have real implementations?

-Marshall


Thilo Goetz (JIRA) wrote:
     [ 
https://issues.apache.org/jira/browse/UIMA-301?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Thilo Goetz closed UIMA-301.
----------------------------

       Resolution: Fixed
    Fix Version/s: 2.3

All CAS array types now inherit from org.apache.uima.cas.CommonArrayFS.

Marshall, I had to make some minor changes to JCAS types.  Please check.


CAS APIs should make it easier to deal with arrays of unknown element type
--------------------------------------------------------------------------

                Key: UIMA-301
                URL: https://issues.apache.org/jira/browse/UIMA-301
            Project: UIMA
         Issue Type: Wish
         Components: Core Java Framework
           Reporter: Adam Lally
           Assignee: Thilo Goetz
           Priority: Minor
            Fix For: 2.3


There are several places in tools where we need to display the contents of an 
FS, which could be an array.  Currently we have to iterate over all possible 
primivie-typed arrays in order to access and display their elements.
What would have been nice is a common superinterface of all the
primitive array types, which defines a toStringArray() method. The
toStringArray() methods are already there on the impls, but there's no
superinterface that I can use to get at them.
See UIMA-40 and UIMA-77.

Reply via email to