[bug #68174] dlmread performance can be improved with modern C++

Please use the bug tracker to post updates to a bug report.  The mailing list is intended as a read-only notification stream.  Info posted to this mailing list address won't appear in the tracker database where it is most useful.

Follow-up Comment #1, bug #68174 (group octave):

The CSET (file bug68174_dlmread_fast_20260324_100.cset) is attached. On my
computer (Intel Ultra 9 285) using a large file:

octave:1> a = randn(5000);
octave:2> save -ascii -double test_file.txt a

Before:

octave:3> tic; a = dlmread ("test_file.txt"); toc
Elapsed time is 5.838 seconds.
octave:4> sum(sum(a))
ans = 571.50

after:

octave:3> tic; a = dlmread ("test_file.txt"); toc
Elapsed time is 2.02794 seconds.
octave:4> sum(sum(a))
ans = 571.50


The difference is even bigger (9+ sec before / 2+ sec after) on M4 Mac. I
think `iostream` is less performing in libc++ than in libstdc++, so from_chars
is a bigger winner there.

Dmitri.
-- 


(file #58395)

    _______________________________________________________

Additional Item Attachment:

Name: bug68174_dlmread_fast_20260324_100.cset Size: 12KiB

<https://file.savannah.gnu.org/file/bug68174_dlmread_fast_20260324_100.cset?file_id=58395>


    AGPL NOTICE

These attachments are served by Savane. You can download the corresponding
source code of Savane at
https://savannah.gnu.org/source/savane-9b684d621d606340ec3f8b1a290873daf3ff5474.tar.gz


    _______________________________________________________

Reply to this item at:

  <https://savannah.gnu.org/bugs/?68174>

_______________________________________________
Message sent via Savannah
https://savannah.gnu.org/

-----BEGIN PGP SIGNATURE-----

iHUEABYIAB0WIQQk97aszIMMAvLLwm6qLAuaBUf3TgUCacNRAAAKCRCqLAuaBUf3
TnQ0AQCsWGgBz+1rufrybl9B11v6gakGuGq5OSM6KxSZCERKkQEA00ElImbpmq5N
iZKF85p87spWYfwihFbExb/ogxTZDgE=
=tnBm
-----END PGP SIGNATURE-----