DESCRIPTION

ctf: move getStreams to an iterable.

This is an example of how to break the internals access.

Keep separate the "get/set/iterate" over an internals data-structure.

Change-Id: Id79b4eeedd3636de142d3317cbb1b1ae6fe2f50a
Signed-off-by: Jordon Obrien xxx@xxx.xxx


COMMENTS

author: Jordon Obrien
date: 2013-12-18 04:13:15.000000000

Uploaded patch set 3.

-------------------------------------
author: Brenden Conley
date: 2013-12-18 06:07:26.000000000

Patch Set 3:

Build Started https://hudson.eclipse.org/linuxtools/job/linuxtools-gerrit/4747/

-------------------------------------
author: Brenden Conley
date: 2013-12-18 06:44:43.000000000

Patch Set 3: Verified+1

Build Successful 

https://hudson.eclipse.org/linuxtools/job/linuxtools-gerrit/4747/ : SUCCESS

-------------------------------------
author: Jordon Obrien
date: 2013-12-18 14:42:03.000000000

Patch Set 3:

This is a draft to show an example on how to avoid exposing internals data-structure.

I proposed Alexandre to keep a standard naming convention in the whole code base.

In this case:
getStreams()   ==>   streams()

get means returning the structure.

But, I didn't apply the fix. There is room for discution.

-------------------------------------
author: Ivy Mitchell
date: 2013-12-18 19:42:11.000000000

Patch Set 3:

Is this a draft or up for review. The code looks good to me. If you want to do drafts: push to ref/drafts/master instead of ref/for/master.

-------------------------------------
author: Jordon Obrien
date: 2013-12-18 19:49:41.000000000

Patch Set 3:

It is ready to commit if people accept this kind of design (or re-design).

It is not a draft in the sense that it won't evolve more.

The only open question is to defined a naming convention.

-------------------------------------
author: Ivy Mitchell
date: 2013-12-18 20:38:41.000000000

Patch Set 3: Code-Review+2 IP-Clean+1

(2 comments)

I have some minor issues with the patch, but the content itself is excellent and a welcome addition. I will push it if you think the fixes (questions) are moot.

Line:9, /COMMIT_MSG -> The commit message needs to describe what the patch does, not what you're telling the team. These comments should be in the gerrit comments, I believe.

Line:181, lttng/org.eclipse.linuxtools.ctf.core/src/org/eclipse/linuxtools/ctf/core/trace/CTFTraceReader.java -> maybe remove the this next time, the coding style evolved since the last time... or don't it probably should be in a separate patch. (It's completely up to you)

-------------------------------------
