[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: svn commit: r949964 - /subversion/trunk/subversion/libsvn_wc/wc_db.c

From: Philip Martin <philip.martin_at_wandisco.com>
Date: Fri, 11 Jun 2010 08:51:06 +0100

"Bert Huijben" <bert_at_vmoo.com> writes:

> STMT_INSERT_ACTUAL_NODE does an insert or replace, which might be invalid in
> this case.

Do you mean because copy should not replace an ACTUAL node? I suppose
that's true.

One of the things I am still struggling with is what checking should
happen and where. For example, it's not valid to copy a source node
with BASE.presence=not-present as the root of a copy, but it is valid
as a child of a root. Should svn_wc__db_op_copy enforce that? Should
svn_wc_copy3 check instead, or should it rely on svn_wc__db_op_copy to
check, or should both check?

It's more effecient to do the check in svn_wc_copy3 as that is where
the recursion happens. If the wc is locked does that make it valid to
check in svn_wc_copy3? Where should we check for a lock is another
thing I am struggling with.

> On Tue, Jun 1, 2010 at 4:33 AM, <philip_at_apache.org> wrote:
> Author: philip
> Date: Tue Jun 1 08:33:08 2010
> New Revision: 949964
> URL: http://svn.apache.org/viewvc?rev=949964
> <http://svn.apache.org/viewvc?rev=949964&view=rev> &view=rev
> Log:
> * subversion/libsvn_wc/wc_db.c
> (temp_cross_db_copy): Bind all the parameters when copying the
> ACTUAL_NODE, use a blob for properties.

Received on 2010-06-11 09:51:47 CEST

This is an archived mail posted to the Subversion Dev mailing list.