[ 
https://issues.apache.org/jira/browse/AVRO-237?page=com.atlassian.jira.plugin.system.issuetabpanels:all-tabpanel
 ]

Doug Cutting updated AVRO-237:
------------------------------

    Attachment: AVRO-237.patch

> Map isn't actually a Collection in Java. I think you don't need to 
> special-case it out. 

Good catch.  Fixed.

> but the javadoc says "only fields of the direct class..

Javadoc fixed.

> getFields() is complicated enough that it deserves one line of javadoc

I added javadoc for this and comments explaining why a cache is used in one 
case and not the other.

> it would be nice to at least error out in that edge case.

Good point.  This now throws an exception.

> I don't think you've added a test case where a non-list Collection (or 
> instance thereof) is used.

Good point.  I've now added one.

> I noticed you added Short.class to SpecificData; is there a test case that 
> that should be part of?

I added a test case for this.

> improvements to reflection
> --------------------------
>
>                 Key: AVRO-237
>                 URL: https://issues.apache.org/jira/browse/AVRO-237
>             Project: Avro
>          Issue Type: Improvement
>          Components: java
>            Reporter: Doug Cutting
>            Assignee: Doug Cutting
>             Fix For: 1.3.0
>
>         Attachments: AVRO-237.patch, AVRO-237.patch
>
>
> A few improvements to reflection to better match actual Java usage:
>  - instead of just List implementations, any Collection that's not a Map 
> should be mapped to an Avro array
>  - inherited fields should be included in schemas induced from a class
>  - inherited methods should be included in protocols induced from an interface

-- 
This message is automatically generated by JIRA.
-
You can reply to this email to add a comment to the issue online.

Reply via email to