Checks for case insensitive and name-mangling filesystems in purge [Was: Cleanup of the purge extension]

Emanuele Aina faina.mail at tiscali.it
Wed Mar 28 14:58:01 CDT 2007


Alexis S. L. Carvalho accettò:

>>> if missing and not force:
>>>    raise util.Abort(_("some nice message that mentions missing files "
>>>                       "and --force"))
>> I'd rather check this before stat'ing the missing files.
> 
> To avoid the extra stats?  /me shrugs.  By the time this runs we'll have
> stat'ed the whole repo, so it shouldn't make much difference.  I
> suggested doing it after the stats just to keep the --force message from
> being displayed before we're sure it'll make a difference.
> 
> But if you prefer it coming before the stats, go for it...

Having tried both I prefer your approach. :)

> Looks OK.  Can you send the patch?

Here it is.

Can you please use the <em at nerd.ocracy.org> address, instead of the one
in the From: field?

I should change the subscription to the list, but I'm just too lazy... :)

I've also updated the automatic tests to check the behavior with missing
files and with the --force option. Obviously I cannot test the behavior
with a case insensitive fs.

To check if the command failed I've used this:

     hg purge -v 2> /dev/null
     if [ $? -ne 0 ]; then
         echo "refused to run"
     fi

Is this ok?

If you think that purge is ready for hgext the old patch I've sent
should still apply. :)

-- 
Buongiorno.
Complimenti per l'ottima scelta.



-------------- next part --------------
A non-text attachment was scrubbed...
Name: purge-abort-on-name-mangling-fs.patch
Type: text/x-patch
Size: 5614 bytes
Desc: not available
Url : http://www.selenic.com/pipermail/mercurial-devel/attachments/20070328/5c2eb9e6/purge-abort-on-name-mangling-fs.bin


More information about the Mercurial-devel mailing list