[PATCH 2 of 2] lfs: migrate most file filtering from threshold to custom filter

[Matt Harbison]

On Sat, 13 Jan 2018 21:11:45 -0500, Yuya Nishihara <yuya at tcha.org> wrote:

> On Sat, 13 Jan 2018 02:23:16 -0500, Matt Harbison wrote:
>> # HG changeset patch
>> # User Matt Harbison <matt_harbison at yahoo.com>
>> # Date 1514706889 18000
>> #      Sun Dec 31 02:54:49 2017 -0500
>> # Node ID 66976f55793cced57929dedc8993204be340c717
>> # Parent  868cc63bfe9d7d7f5b40bc8cd70175cf1a608a95
>> lfs: migrate most file filtering from threshold to custom filter
>
> Queued, thanks.
>
>> -    threshold = repo.ui.configbytes('lfs', 'threshold')
>> +    trackspec = repo.ui.config('lfs', 'track')
>>
>> -    repo.svfs.options['lfsthreshold'] = threshold
>> +    # deprecated config: lfs.threshold
>> +    threshold = repo.ui.configbytes('lfs', 'threshold')
>> +    if threshold:
>> +        trackspec = "(%s) | size('>%s')" % (trackspec, threshold)
>
> I've added fileset.parse(trackspec) to detect bad spec like '..) ..  
> (..', and
> s/%s/%d/ to make sure threshold is integer so no need of string escape.

Thanks.

I assume that your amend obsoleted something local only to your system?   
The reason I ask is because I ran `hg extdiff --patch --hidden -r  
'precursors(c780e06)' -r c780e06`, and it aborted with an empty revision.   
It's the precursors() predicate coming up empty.  I ended up using  
'last(allprecursors(c780e06))', and that found what I submitted.  But that  
seems inconsistent with the obsolete line of that logged revision pointing  
directly to c780e06.

At minimum, it would be nice to have a shorter way to spell that.  But it  
seems weird that the log predicate would care about revisions not in the  
repo.  Boris- it looks like '{precursors}' will skip the obsolete revision  
not on my system, and agrees with 'last(..)'.