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

Re: single file switching (was: svn commit: rev 1163 - trunk/subversion/include trunk/subversion/libsvn_ra_local trunk/subversion/mod_dav_svn trunk/subversion/libsvn_repos)

From: Greg Stein <gstein_at_lyra.org>
Date: 2002-02-05 20:32:15 CET

On Tue, Feb 05, 2002 at 12:00:02PM -0600, Ben Collins-Sussman wrote:
> Sorry, I'm still not understanding you. I think you have a different
> model in your head than I do regarding the workings of libsvn_wc.

Apparently so... :-)

> Granted -- there are zillions of places in libsvn_wc where we loop
> over entries files. But AFAICT, there are only *two* cases where that
> loop notices or cares about URLs at all: commits and updates. And
> we've described how to make those two crawlers detect and deal with
> disjointedness; and the solution doesn't distinguish between disjoint
> dirs or files.


> > The files returned by listdir() may not actually live in the parent you're
> > listing.
> What is listdir()? Do you mean some future RA->listdir() function?
> If so, that irrelevant to a working copy. It's retrieving an ls from
> the server, no?

No. Just a generic reference to getting the files in a directory.

> > Every time a person wants to simply refer to a file, there is a lot
> > more logic involved to see whether it truly lives in the parent, or
> > if it came from somewhere else.
> If libsvn_wc wants to refer to a file or dir, it retrieves the entry
> structure for that object. The entry structure already has the
> correct URL, whether it's disjoint or not. There's no need for extra


> I guess what I'm asking here is: are there any loops over entries
> files -- *other* than commits or updates -- that need to be aware of
> disjoint URLs? I can't think of any.

Not that I know of. My (erroneous) assumption was that we have a lot of
iterations over the entries [that need the URL]. The answer is: yes, we *do*
those iterations quite a bit, but they generally do *not* look at the URL,
so the disjointedness doesn't matter.

> On the flip side, you seem to be saying that there are places where
> libsvn_wc code somehow depends on an assumption that a file's
> entry-URL is directly descended from its parent's entry-URL. But I
> don't see those places. Where are they?

Don't put words in my mouth :-) Never said, but implied/thought. And that's
a leading/baiting question... you know there aren't any :-)

Net result: I confused "numerous directory traversals" with "numerous
directory traversals which need/use the URL". My apologies...


Greg Stein, http://www.lyra.org/
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 21 14:37:04 2006

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

This site is subject to the Apache Privacy Policy and the Apache Public Forum Archive Policy.