You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@subversion.apache.org by Brent Burton <br...@io.com> on 2002/10/15 01:32:37 UTC
Re: Making import syntax more consistant (Was Re: making
repositories)
At 04:12 PM 10/14/2002 -0700, Peter Davis wrote:
>-----BEGIN PGP SIGNED MESSAGE-----
>Hash: SHA1
>
>On Monday 14 October 2002 14:33, Brent Burton wrote:
> > But the results I was expecting (for whatever reason) were
> > along the lines of
> > tar cf - some/long/path | (cd ....; tar xf -)
> > which recreates the some/long/path hierarchy under the .... location.
>
>I'm not sure if this is what you were expecting or not, but just to be clear,
>I don't think import should recreate the tree of the second argument in the
>repository. Expanding on your other example:
>
>$ svn import file:///path/to/repos graphics/bsp
>
>would import ./graphics/bsp/* into the repository root. If you had
>graphics/bsp/readme.txt, it would end up as
>file:///path/to/repos/readme.txt.
>That's not the same as:
>
>$ svn import file:///path/to/repos/graphics/bsp graphics/bsp
Right, this latter semantic, and what you propose, seems
to be the best usage. 'svn import' indeed has the same
semantics as 'cp', and your proposal, continuing the analogy,
would essentially place a 'mkdir -p' prior to a 'cp'. This
is a very clean enhancement and clears up the syntax.
The reference to tar was to show what my initial expectation
was given the documentation; the docs don't fully explain
the current semantics (or I simply may have not understood).
A few more examples in the "Subversion Handbook" would help.
>If you expect the results of the latter when specifying the former, then
>consider if the second argument were not relative to the current working
>directory, or if it began with "..". Import couldn't very well create files
>under ".." of the repository root.
Well it's certainly something to check IFF one was to allow that
kind of specification. See httpd as an example of this file handling.
But your point is clear -- at what point does the file spec
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Re: Making import syntax more consistant (Was Re: making
repositories)
Posted by Brent Burton <br...@io.com>.
Hmm, hit the 'send' key early by accident.
At 08:32 PM 10/14/2002 -0500, Brent Burton wrote:
>>If you expect the results of the latter when specifying the former, then
>>consider if the second argument were not relative to the current working
>>directory, or if it began with "..". Import couldn't very well create files
>>under ".." of the repository root.
>
>Well it's certainly something to check IFF one was to allow that
>kind of specification. See httpd as an example of this file handling.
>But your point is clear -- at what point does the file spec
But your point is clear -- how much processing the file spec
does is definable and fairly arbitrary. Perhaps a huge number
of ../ references simply backs up against the repository root,
where ".." there refers to itself, like the filesystem root.
Anyway, I'm digressing.
I support your proposal for specifying new subdirs by simply
including them in the repository path. It's very clean and
unambiguous without impacting the file-import code very much.
-Brent
---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org