Bug 27419 - [PATCH] Add support for rlog and tag ranges to CvsChangeLog
Summary: [PATCH] Add support for rlog and tag ranges to CvsChangeLog
Alias: None
Product: Ant
Classification: Unclassified
Component: Core tasks (show other bugs)
Version: 1.6.1
Hardware: All All
: P3 enhancement with 3 votes (vote)
Target Milestone: 1.8.0
Assignee: Ant Notifications List
Keywords: PatchAvailable
: 27577 (view as bug list)
Depends on:
Reported: 2004-03-03 21:52 UTC by Rob van Oostrum
Modified: 2008-10-17 08:50 UTC (History)
2 users (show)

Code patch (7.58 KB, patch)
2004-03-03 21:52 UTC, Rob van Oostrum
Details | Diff
Documentation patch (2.97 KB, patch)
2004-03-03 21:52 UTC, Rob van Oostrum
Details | Diff

Note You need to log in before you can comment on or make changes to this bug.
Description Rob van Oostrum 2004-03-03 21:52:04 UTC
This patch adds support for rlog and tag ranges to CvsChangeLog
Comment 1 Rob van Oostrum 2004-03-03 21:52:34 UTC
Created attachment 10653 [details]
Code patch
Comment 2 Rob van Oostrum 2004-03-03 21:52:54 UTC
Created attachment 10654 [details]
Documentation patch
Comment 3 Stefan Bodewig 2005-03-16 16:34:05 UTC
why rlog as a parameter instead of command?

tags are supported since 1.6.1, if you use the tag attribute.
Comment 4 Stefan Bodewig 2005-03-16 16:36:26 UTC
*** Bug 27577 has been marked as a duplicate of this bug. ***
Comment 5 Rob van Oostrum 2005-03-16 16:47:28 UTC
Using tags on a working copy will only log changes between those tags on files
that exist in your working copy. For this to accurately display the differences
between those two tags, your working copy would have to be updated to the more
recent tag. Running the command on the repository removed the need for this, in
fact it removes the need for a working copy altogether.

As to the remote option / rlog task question, it seems to me that this is the
more reuse-friendly/transparent way to go (it would also make it easier to
configure your build to use log or rlog (by passing a property value rather than
use ant-contrib if/then/else constructs). But if you'd rather refactor the
common behavior/interface into an abstract superclass, I guess one could make an
argument to that extent.
Comment 6 Stefan Bodewig 2005-03-16 16:58:12 UTC
Rob, I think you've misunderstood my question.  Inside the code, you add rlog as
command argument instead of a command and I don't immediately understand why.

Quite possible that you don't know it either, given the reports is more than a
year old.
Comment 7 Rob van Oostrum 2005-03-16 17:09:07 UTC
Hmmm ... gotcha. I'm foggy on this but it does ring a bell.

I *THINK* it was because of some apparent strange behavior in CVS where it
wasn't parsing the rlog commandline parameters properly and it seemed to be
taking the whole line starting with 'rlog' as one paremeter value. That's the
best I've got right now. Might not be necessary depending on the version of CVS.
I can say that a handful of people here have been using this patch for the past
year and it's been working fine for us. I added a code comment so it must have
been for a good reason. Sorry I can't say with certainty what the reason was/is.
Comment 8 Stefan Bodewig 2008-10-17 08:17:43 UTC
rlog went in with svn revision 705639, I'll look into tags later.

It is no surprise that I had to adjust a few things with a patch four and a half years old.  Many thanks!
Comment 9 Stefan Bodewig 2008-10-17 08:50:16 UTC
tags are in with svn revision 705646.

The startTag tests fail for me, but I guess this is more a sign of my CVS knowledge becoming more than rusty.

Thanks again.