Re: [PATCH 1/2] completion: improve ls-files filter performance

From: Johannes Schindelin <johannes.schindelin@gmx.de>
Date: Mar 19, 2018, 17:12 UTC
Message-ID: <nycvar.QRO.7.76.6.1803191802140.55@ZVAVAG-6OXH6DA.rhebcr.pbec.zvpebfbsg.pbz>
In-Reply-To: <1521274624-1370-1-git-send-email-drizzd@gmx.net>

Hi drizzd,

first of all: thank you so much for working on this. I am sure it will be noticeable to many Windows users, and also make my life easier.

On Sat, 17 Mar 2018, Clemens Buchacher wrote:

Show 21 quoted lines

> From the output of ls-files, we remove all but the leftmost path
> component and then we eliminate duplicates. We do this in a while loop,
> which is a performance bottleneck when the number of iterations is large
> (e.g. for 60000 files in linux.git).
> 
> $ COMP_WORDS=(git status -- ar) COMP_CWORD=3; time _git
> 
> real    0m11.876s
> user    0m4.685s
> sys     0m6.808s
> 
> Using an equivalent sed script improves performance significantly:
> 
> $ COMP_WORDS=(git status -- ar) COMP_CWORD=3; time _git
> 
> real    0m1.372s
> user    0m0.263s
> sys     0m0.167s
> 
> The measurements were done with mingw64 bash, which is used by Git for
> Windows.

Technically, it is not the *mingw64* bash, but it is an MSYS2 Bash. This does make a little bit of a difference because of the penalty incurred by the POSIX emulation layer provided by the MSYS2 runtime.

(And it also addresses Gabór's question whether you ran the test suite, I guess... it takes multiple hours to run it even once on a regular computer.)

Ciao, Dscho