uima-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Steven Bethard <steven.beth...@gmail.com>
Subject Re: Implementing a Map as a custom feature for annotation
Date Sun, 06 Dec 2009 08:10:39 GMT
On Sat, Dec 5, 2009 at 6:00 AM, Marshall Schor <msa@schor.com> wrote:
> The UIMA CAS does not support storing Java Maps in the CAS.  Of course,
> any individual annotator written in Java could use Java Maps, internally
> within that annotator, as a Java object.  To put it into the CAS, you
> would need to do some kind of "export" of it, into data structures the
> CAS supports.
> One reason for this is to support a wider inter-operability: so that the
> CAS you might produce could be sent to another annotator that was
> written in C++, for instance (which UIMA does support).

Are you saying that support for Maps can't be added to UIMA because of
interoperability concerns? Or just that since Maps aren't there now,
trying to use them will cause problems for interoperability?  It
certainly seems like UIMA could grow Map support - pretty much every
language has a map type of some sort: C++ STL map, Java HashMap, PHP
associative array, Python/Ruby dictionary, etc. Apache Thrift, for
example, supports exchange of all of these, cross-language:



> Anton Shuster wrote:
>> Hello all,
>> I tried my best to search for an answer to this topic but came up
>> short. I'm creating an Annotation type and I want to give it a feature
>> that is a Map (a HashMap or any other implementation).
>> I looked through the documentation, but there is no information on
>> creating your own type such as this. All the documentation seems to
>> talk about is creating features which inherit the built-in types or
>> which inherit other features defined this way.
>> Please let me know what's the right way to go about this.
>> One usage scenario would be for annotating HTML tags. For example, an
>> HTML tag annotation could contain a label String and an attributes
>> Map. There are other scenarios as well, but this one is the most
>> obvious.
>> Thanks for any help,
>> --Anton

Where did you get that preposterous hypothesis?
Did Steve tell you that?
        --- The Hiphopopotamus

View raw message