git/list[1] front-page[2] threads[3] people[4] search[5] about
 

Re: [PATCH 2/2] built-in rebase: reinstate `checkout -q` behavior where appropriate

From
Junio C Hamano <gitster@pobox.com>
Date
Nov 12, 2018, 04:04 UTC
Message-ID
<xmqq1s7r3pqy.fsf@gitster-ct.c.googlers.com>
In-Reply-To
<nycvar.QRO.7.76.6.1811091813140.39@tvgsbejvaqbjf.bet>
Johannes Schindelin <Johannes.Schindelin@gmx.de> writes:
Show 16 quoted lines
>> >  static int reset_head(struct object_id *oid, const char *action,
>> > -		      const char *switch_to_branch, int detach_head,
>> > +		      const char *switch_to_branch,
>> > +		      int detach_head, int reset_hard,
>> 
>> It might be worth switching to a single flag variable here. It would
>> make calls like this:
>> 
>> > -	if (reset_head(&options.onto->object.oid, "checkout", NULL, 1,
>> > +	if (reset_head(&options.onto->object.oid, "checkout", NULL, 1, 0,
>> >  	    NULL, msg.buf))
>> 
>> a little more self-documenting (if a little more verbose).
>
> I thought about that, but for just two flags? Well, let me try it and see
> whether I like the result ;-)

My rule of thumb is that repeating three times is definitely when we should consolidate separate ones into a single flag word, and twice is a borderline.

For two-time repetition, it is often worth fixing when somebody notices it during the review, as that is a sign that repetition, even a minor one, disturbed a reader enough to point out. On the other hand, for a file-scope static that is likely to stay as a non-API function but a local helper, it may not be worth it.

So I am OK either way, slightly in favor of fixing it while we remember.

Show 12 quoted lines
>> This one could actually turn into:
>> 
>>   ret = error(...);
>>   goto leave_reset_head;
>> 
>> now. We don't have to worry about an uninitialized desc.buffer anymore
>> (as I mentioned in the previous email), because "nr" would be 0.
>> 
>> It doesn't save any lines, though (but maybe having a single
>> cleanup/exit point would make things easier to read; I dunno).
>
> But you're right, of course. Consistency makes for easier-to-read code.
Yup, that does sound good.
Thanks.
Previous: Johannes SchindelinNext: Johannes Schindelin
Message 10 of 18 in “Fix built-in rebase perf regression”
  1. 0/2 Fix built-in rebase perf regressionJohannes Schindelin via GitGitGadget, Nov 9, 2018
  2. 1/2 rebase: consolidate clean-up code before leaving reset_head()Johannes Schindelin via GitGitGadget, Nov 9, 2018
  3. Jeff KingNov 9, 2018
  4. Johannes SchindelinNov 9, 2018
  5. 2/2 built-in rebase: reinstate `checkout -q` behavior where appropriateJohannes Schindelin via GitGitGadget, Nov 9, 2018
  6. Jeff KingNov 9, 2018
  7. Johannes SchindelinNov 9, 2018
  8. Jeff KingNov 9, 2018
  9. Johannes SchindelinNov 12, 2018
  10. Junio C HamanoNov 12, 2018
  11. Johannes SchindelinNov 12, 2018
  12. Junio C HamanoNov 13, 2018
  13. Junio C HamanoNov 13, 2018
  14. 0/3 Fix built-in rebase perf regressionJohannes Schindelin via GitGitGadget, Nov 12, 2018
  15. 1/3 rebase: consolidate clean-up code before leaving reset_head()Johannes Schindelin via GitGitGadget, Nov 12, 2018
  16. 2/3 rebase: prepare reset_head() for more flagsJohannes Schindelin via GitGitGadget, Nov 12, 2018
  17. 3/3 built-in rebase: reinstate `checkout -q` behavior where appropriateJohannes Schindelin via GitGitGadget, Nov 12, 2018
  18. Johannes SchindelinNov 12, 2018

Read the whole thread, see it on lore, or plain text.

$ cat FOOTERMessages come from the public archive at lore.kernel.org/git, fetched every hour. The front page is chosen and written each morning by an AI editor and can be wrong; the threads themselves are the record. About and API. For agents: an MCP server at https://gitlist.dev/mcp, and any thread, story or person page as Markdown by adding .md to its URL (or sending Accept: text/markdown). Details in /llms.txt.