[PATCH 3 of 3] diffopts: notice a negated boolean flag in diffopts

[Augie Fackler]

On Mon, Aug 22, 2016 at 10:01 AM, Yuya Nishihara <yuya at tcha.org> wrote:
> On Thu, 18 Aug 2016 12:13:34 -0400, Augie Fackler wrote:
>> # HG changeset patch
>> # User Augie Fackler <augie at google.com>
>> # Date 1471534521 14400
>> #      Thu Aug 18 11:35:21 2016 -0400
>> # Node ID 16491c79cf9b9b61550342690c680ed192506b08
>> # Parent  9ec3aad0bf74ad1c25e87bcd0f536896d1c355ca
>> diffopts: notice a negated boolean flag in diffopts
>>
>> This means that if you have git-diffs enabled by default (pretty
>> common) and you hit the rare (but real) case where a git-diff breaks
>> patch(1) or some other tool, you can easily disable it by just
>> specifying --git=0 on the command line.
>>
>> diff --git a/mercurial/patch.py b/mercurial/patch.py
>> --- a/mercurial/patch.py
>> +++ b/mercurial/patch.py
>> @@ -2142,7 +2142,7 @@ def difffeatureopts(ui, opts=None, untru
>>      def get(key, name=None, getter=ui.configbool, forceplain=None):
>>          if opts:
>>              v = opts.get(key)
>> -            if v:
>> +            if v or v is False:
>
> I'm afraid that False/True/None tristate would cause subtle bugs. It seems
> we choose None or False randomly as the default value of boolean command
> options.

Yeah, the False/None nonsense in the flags layer bugs me. I should do
a survey of the code and see if anyone actually depends on the default
being None rather than False or vice versa.

I think this is solvable, so if people like the feature in general,
I'll tackle the getopt-reinvention we discussed in IRC.