git.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Ben Peart <peartben@gmail.com>
To: Elijah Newren <newren@gmail.com>, Ben Peart <Ben.Peart@microsoft.com>
Cc: "git@vger.kernel.org" <git@vger.kernel.org>,
	"peff@peff.net" <peff@peff.net>,
	"gitster@pobox.com" <gitster@pobox.com>,
	"pclouds@gmail.com" <pclouds@gmail.com>,
	"vmiklos@frugalware.org" <vmiklos@frugalware.org>,
	Kevin Willford <kewillf@microsoft.com>,
	"Johannes.Schindelin@gmx.de" <Johannes.Schindelin@gmx.de>,
	"eckhard.s.maass@googlemail.com" <eckhard.s.maass@googlemail.com>
Subject: Re: [PATCH v3 2/3] merge: Add merge.renames config setting
Date: Thu, 26 Apr 2018 20:54:20 -0400	[thread overview]
Message-ID: <7de8f144-8a37-e471-48e8-0b6f17a7bf29@gmail.com> (raw)
In-Reply-To: <CABPp-BE29rwZCDqFHH-nzrDub6MMdtoiorj0jv3K6B6cmfcaLA@mail.gmail.com>



On 4/26/2018 6:52 PM, Elijah Newren wrote:
> On Thu, Apr 26, 2018 at 1:52 PM, Ben Peart <Ben.Peart@microsoft.com> wrote:
> 
>> +merge.renames::
>> +       Whether and how Git detects renames.  If set to "false",
>> +       rename detection is disabled. If set to "true", basic rename
>> +       detection is enabled.  If set to "copies" or "copy", Git will
>> +       detect copies, as well.  Defaults to the value of diff.renames.
>> +
> 
> We shouldn't allow users to force copy detection on for merges  The
> diff side of the code will detect them correctly but the code in
> merge-recursive will mishandle the copy pairs.  I think fixing it is
> somewhere between big can of worms and
> it's-a-configuration-that-doesn't-even-make-sense, but it's been a
> while since I thought about it.

Color me puzzled. :)  The consensus was that the default value for 
merge.renames come from diff.renames.  diff.renames supports copy 
detection which means that merge.renames will inherit that value.  My 
assumption was that is what was intended so when I reimplemented it, I 
fully implemented it that way.

Are you now requesting to only use diff.renames as the default if the 
value is true or false but not if it is copy?  What should happen if 
diff.renames is actually set to copy?  Should merge silently change that 
to true, display a warning, error out, or something else?  Do you have 
some other behavior for how to handle copy being inherited from 
diff.renames you'd like to see?

Can you write the documentation that clearly explains the exact behavior 
you want?  That would kill two birds with one stone... :)

> 
>> diff --git a/merge-recursive.h b/merge-recursive.h
>> index 80d69d1401..0c5f7eff98 100644
>> --- a/merge-recursive.h
>> +++ b/merge-recursive.h
>> @@ -17,7 +17,8 @@ struct merge_options {
>>          unsigned renormalize : 1;
>>          long xdl_opts;
>>          int verbosity;
>> -       int detect_rename;
>> +       int diff_detect_rename;
>> +       int merge_detect_rename;
>>          int diff_rename_limit;
>>          int merge_rename_limit;
>>          int rename_score;
>> @@ -28,6 +29,11 @@ struct merge_options {
>>          struct hashmap current_file_dir_set;
>>          struct string_list df_conflict_file_set;
>>   };
>> +inline int merge_detect_rename(struct merge_options *o)
>> +{
>> +       return o->merge_detect_rename >= 0 ? o->merge_detect_rename :
>> +               o->diff_detect_rename >= 0 ? o->diff_detect_rename : 1;
>> +}
> 
> Why did you split o->detect_rename into two fields?  You then
> recombine them in merge_detect_rename(), and after initial setup only
> ever access them through that function.  Having two fields worries me
> that people will accidentally introduce bugs by using one of them
> instead of the merge_detect_rename() function.  Is there a reason you
> decided against having the initial setup just set a single value and
> then use it directly?
> 

The setup of this value is split into 3 places that may or may not all 
get called.  The initial values, the values that come from the config 
settings and then any values passed on the command line.

Because the merge value can now inherit from the diff value, you only 
know the final value after you have received all possible inputs.  That 
makes it necessary to be a calculated value.

If you look at diff_rename_limit/merge_rename_limit, detect_rename 
follow the same pattern for the same reasons.  It turns out 
detect_rename was a little more complex because it is used in 3 
different locations (vs just one) which is why I wrapped the inheritance 
logic into the helper function merge_detect_rename().

  reply	other threads:[~2018-04-27  0:54 UTC|newest]

Thread overview: 65+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2018-04-20 13:36 [PATCH v1 0/2] add additional config settings for merge Ben Peart
2018-04-20 13:36 ` [PATCH v1 1/2] merge: Add merge.renames config setting Ben Peart
2018-04-20 17:02   ` Elijah Newren
2018-04-20 17:26     ` Elijah Newren
2018-04-23 12:57       ` Ben Peart
2018-04-20 17:59     ` Ben Peart
2018-04-20 18:34       ` Elijah Newren
2018-04-21  4:23         ` Junio C Hamano
2018-04-23 16:00           ` Ben Peart
2018-04-23 23:23             ` Junio C Hamano
2018-04-24 11:58               ` Johannes Schindelin
2018-04-24 17:47                 ` Elijah Newren
2018-04-25  8:20                   ` Johannes Schindelin
2018-04-22 12:07         ` Eckhard Maaß
2018-04-23 13:15           ` Ben Peart
2018-04-23 21:32             ` Eckhard Maaß
2018-04-24 16:53               ` Ben Peart
2018-04-23 13:22         ` Ben Peart
2018-04-20 13:36 ` [PATCH v1 2/2] merge: Add merge.aggressive " Ben Peart
2018-04-20 17:22   ` Elijah Newren
2018-04-24 16:45     ` Ben Peart
2018-04-24 17:36       ` Elijah Newren
2018-04-24 23:57       ` Junio C Hamano
2018-04-25 14:47         ` Ben Peart
2018-04-20 17:34 ` [PATCH v1 0/2] add additional config settings for merge Elijah Newren
2018-04-20 18:19   ` Ben Peart
2018-04-24 17:11 ` [PATCH v2 " Ben Peart
2018-04-24 17:11   ` [PATCH v2 1/2] merge: Add merge.renames config setting Ben Peart
2018-04-24 18:11     ` Elijah Newren
2018-04-24 18:59     ` Elijah Newren
2018-04-24 20:31       ` Ben Peart
2018-04-25 16:01         ` Elijah Newren
2018-04-24 17:11   ` [PATCH v2 2/2] merge: Add merge.aggressive " Ben Peart
2018-04-25  0:13   ` [PATCH v2 0/2] add additional config settings for merge Junio C Hamano
2018-04-25 15:22     ` Ben Peart
2018-04-26  1:48       ` Junio C Hamano
2018-04-26 20:52 ` [PATCH v3 0/3] add merge.renames config setting Ben Peart
2018-04-26 20:52   ` [PATCH v3 1/3] merge: update documentation for {merge,diff}.renameLimit Ben Peart
2018-04-26 23:11     ` Elijah Newren
2018-04-26 23:23       ` Jonathan Tan
2018-04-26 20:52   ` [PATCH v3 2/3] merge: Add merge.renames config setting Ben Peart
2018-04-26 22:52     ` Elijah Newren
2018-04-27  0:54       ` Ben Peart [this message]
2018-04-27  2:23         ` Junio C Hamano
2018-04-27  3:28           ` Elijah Newren
2018-04-27  7:23             ` Johannes Schindelin
2018-04-27 14:32               ` Elijah Newren
2018-04-27 18:37           ` Eckhard Maaß
2018-04-27 20:23             ` Elijah Newren
2018-04-30  8:03               ` Eckhard Maaß
2018-04-30 16:54                 ` Elijah Newren
2018-04-27  4:17         ` Elijah Newren
2018-04-27 18:19         ` Elijah Newren
2018-04-30 13:11           ` Ben Peart
2018-04-30 16:12             ` Re: Elijah Newren
2018-05-02 14:33               ` Re: Ben Peart
2018-04-26 20:52   ` [PATCH v3 3/3] merge: pass aggressive when rename detection is turned off Ben Peart
2018-04-26 23:00     ` Elijah Newren
2018-04-26 22:08   ` [PATCH v3 0/3] add merge.renames config setting Elijah Newren
2018-05-02 16:01 ` [PATCH v4 0/3] add additional config settings for merge Ben Peart
2018-05-02 16:01   ` [PATCH v4 1/3] merge: update documentation for {merge,diff}.renameLimit Ben Peart
2018-05-02 16:01   ` [PATCH v4 2/3] merge: Add merge.renames config setting Ben Peart
2018-05-04  3:07     ` Junio C Hamano
2018-05-02 16:01   ` [PATCH v4 3/3] merge: pass aggressive when rename detection is turned off Ben Peart
2018-05-02 17:20   ` [PATCH v4 0/3] add additional config settings for merge Elijah Newren

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=7de8f144-8a37-e471-48e8-0b6f17a7bf29@gmail.com \
    --to=peartben@gmail.com \
    --cc=Ben.Peart@microsoft.com \
    --cc=Johannes.Schindelin@gmx.de \
    --cc=eckhard.s.maass@googlemail.com \
    --cc=git@vger.kernel.org \
    --cc=gitster@pobox.com \
    --cc=kewillf@microsoft.com \
    --cc=newren@gmail.com \
    --cc=pclouds@gmail.com \
    --cc=peff@peff.net \
    --cc=vmiklos@frugalware.org \
    --subject='Re: [PATCH v3 2/3] merge: Add merge.renames config setting' \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).