Perl bindings work differently than documented?

Let me preface this with the fact that I'm not a perl guru. I know enough
to be dangerous, and that's about it.

I'm trying to help a user on the users@ list track down a memory leak. So
I had him trying passing a subpool to the $ctx->merge operation, and it
doesn't work. So, I looked at the Client.pm module and it seems to
indicate that passing a subpool is possible. However, when I tried it in
my own script, I get the following error:
 Type error in argument 10 of svn_client_merge. Expected \
   _p_svn_client_ctx_t

The command I used was:
 $ctx->merge ($url1, $_-1, $url2, $_, $wc, 1, 1, 0, 0, $subpool);

If I leave off the subpool everything is happy (as expected). However, in
the user's case, he's consuming a large quantity of memory while
repeatedly running the merge operation.

I also tried this (which was suggested to me by perl when I got the number
of parameters wrong once):
  SVN::Client::merge($url1, $_-1, $url2, $_, $wc, 1, 1, 0, 0, $ctx,
                     $subpool);

and this (as suggested by the SVN::Client man page):
  SVN::Client::merge($ctx, $url1, $_-1, $url2, $_, $wc, 1, 1, 0, 0,
                     $subpool);

and get the exact same error message.

Does anyone have any idea what I'm doing wrong here? Are we able to use
subpools on client operations?

BTW, I'm using Subversion 1.2.1.

-John

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Thu Jul 14 12:23:35 2005