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

Fwd: Behaviour of global ignores

From: Simon Large <simon.tortoisesvn_at_googlemail.com>
Date: Tue, 9 Dec 2008 09:39:21 +0000

Please excuse the escalation to dev@, but I would like to put
something more definitive in the TortoiseSVN manual.

Simon

-- 
:       ___
:  oo  // \\      "De Chelonian Mobile"
: (_,\/ \_/ \     TortoiseSVN
:   \ \_/_\_/>    The coolest Interface to (Sub)Version Control
:   /_/   \_\     http://tortoisesvn.net
---------- Forwarded message ----------
From: Simon Large <simon.tortoisesvn_at_googlemail.com>
Date: 2008/12/1
Subject: Re: Behaviour of global ignores
To: Ryan Schmidt <subversion-2008c_at_ryandesign.com>
Cc: users_at_subversion.tigris.org
2008/11/29 Ryan Schmidt <subversion-2008c_at_ryandesign.com>:
> On Nov 28, 2008, at 18:54, Simon Large wrote:
>
>> Just had some strange behaviour reported on the TSVN list, actually 2
>> separate issues. This is with 1.5.
>>
>> 1. On Windows, global ignore patterns seem to use a backslash as a
>> path separator, but everywhere else we have to use a forward slash.
>> Paths not ignored are always reported with forward slashes in status.
>> Why the inconsistency for ignore patterns? I don't remember this being
>> the case previously - is it a recent change? There is no mention of
>> this in the svn book.
>>
>> 2. Said user wants the entire content of a versioned directory to be
>> ignored (the directory itself is versioned, the content within should
>> be ignored). But he has lots of these directories so wants to avoid
>> setting svn:ignore=* on each one. The seemingly obvious solution is to
>> use a global ignore pattern like this: "*\dir\*". But no, that doesn't
>> work. Nor does "*dir*". In fact the only pattern that does work seems
>> to be "*", which is somewhat drastic. It appears that global ignores
>> don't work if they include a path which is already versioned. Am I
>> missing something, is this by design or is it a bug?
>>
>> I guess the answer is that global ignores are not really intended to
>> include path information. The fact is that they work partially with
>> paths, but not consistently. Again, there is no mention of anything
>> other than simple filename use in the book.
>
> It may be an accident that including path components in the global-ignores
> works at all. I've only ever heard of global-ignores (and svn:ignore) used
> to match filenames (not including any path components).
Is this the official answer? Do I need to update the TortoiseSVN
manual to say that path components should not be used in global ignore
patterns?
svn:ignore patterns only make sense applied to immediate child files
and folders. If you tried to use it to ignore grandchildren then it
would fail if the WC was checked out from the child folder instead.
Simon
------------------------------------------------------
http://subversion.tigris.org/ds/viewMessage.do?dsForumId=462&dsMessageId=981569
Received on 2008-12-09 10:39:36 CET

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