Hi Peter,
Yes it violates the API.
Before fixing this, I would like to write a testcase which could
showcase this violation.
Currently merge_tests.py's 41st testcase has only one such child with
differing mergeinfo.
If I am not successful with my testcase will post a patch that would
call the set_path in a depth first search order.
Thanks
With regards
Kamesh Jayachandran
Peter Lundblad wrote:
> Hi,
>
> In my attempts to get started on the merge tracking branch, I noticed
> a problem in the code that avoids merging subparts of a target tree
> which has its own overlapping merge info.
>
> Specifically, the loop of do_merge in libsvn_client/diff.c that calls
> set_path of the reporter with the target revision. Per the reporter
> API documentation, the calls to a reporter must be done in depth-first
> order and as far as I remember, the reporter of libsvn_repos epends on
> this API restriction.
>
> So, this loop needs to be rewritten to take this into account.
>
> Best regards,
> //Peter
>
> ---------------------------------------------------------------------
> To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
> For additional commands, e-mail: dev-help@subversion.tigris.org
>
>
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jan 25 08:46:06 2007