* org-element checks make flyspell prohibitively slow
@ 2014-03-17 14:32 Matt Lundin
2014-03-17 21:22 ` Nicolas Goaziou
0 siblings, 1 reply; 3+ messages in thread
From: Matt Lundin @ 2014-03-17 14:32 UTC (permalink / raw)
To: Emacs-orgmode list
[-- Attachment #1: Type: text/plain, Size: 290 bytes --]
The rewrite of org-mode-flyspell-verify in commit
4a27c2b4b67201e0b23f431bdaeb6460b31e1394 (Nov 21, 2013) makes navigating
org-mode files with large chunks of text very slow.
For instance, I started up a minimal emacs:
/usr/bin/emacs -Q -l ~/config/minimal.el
...where minimal.el is...
[-- Attachment #2: minimal.el --]
[-- Type: application/emacs-lisp, Size: 104 bytes --]
[-- Attachment #3: Type: text/plain, Size: 7317 bytes --]
and (insert "\n\n=> " (emacs-version))
=> GNU Emacs 24.3.1 (x86_64-unknown-linux-gnu, GTK+ Version 3.10.7)
of 2014-01-28 on var-lib-archbuild-extra-x86_64-juergen
and (insert "\n\n=> " (org-version nil t))
=> Org-mode version 8.2.5h (release_8.2.5h-757-gc444e4 @ /home/matt/org-mode/lisp/)
I open a test.org file containing the following.
--8<---------------cut here---------------start------------->8---
* A headline
* Arch packages
* Another headline
--8<---------------cut here---------------end--------------->8---
After opening a line under "Arch Packages" I call...
C-u M-! pacman -Ss [RET]
(Of course, this only works with archlinux.) This inserts a long list of
packages that look like this:
--8<---------------cut here---------------start------------->8---
core/acl 2.2.52-2 [installed]
Access control list utilities, libraries and headers
core/archlinux-keyring 20140220-1 [installed]
Arch Linux PGP keyring
core/attr 2.4.47-1 [installed]
Extended attribute support library for ACL support
core/autoconf 2.69-1 (base-devel) [installed]
A GNU tool for automatically configuring source code
core/automake 1.14.1-1 (base-devel) [installed]
A GNU tool for automatically creating Makefiles
--8<---------------cut here---------------end--------------->8---
All in all, it's 12680 lines....
I navigate to the bottom of the file. I type...
M-x elp-instrument-package [RET] org [RET]
M-x elp-instrument-package [RET] flyspell [RET]
M-x elp-instrument-function [RET] scroll-down-command [RET]
Then I hit M-v three times. This takes a while.
Here are the top elp offenders:
--8<---------------cut here---------------start------------->8---
flyspell-post-command-hook 6 10.753828775 1.7923047959
flyspell-word 6 10.752069764 1.7920116273
org-mode-flyspell-verify 5 8.973166134 1.7946332267
org-element-context 5 8.364142505 1.6728285010
org-element--get-next-object-candidates 699 8.307898326 0.0118854053
org-element-latex-or-entity-successor 5 3.7592736849 0.7518547369
org-element-link-successor 40 1.1079495280 0.0276987382
org-element-sub/superscript-successor 659 1.0986591029 0.0016671610
org-element-line-break-successor 5 0.9729438699 0.194588774
org-element-at-point 5 0.607910786 0.1215821572
org-element--parse-to 5 0.606992172 0.1213984344
org-element--current-element 10 0.4201667370 0.0420166737
org-element-paragraph-parser 10 0.416739094 0.0416739094
org-element-inline-src-block-successor 5 0.3740871620 0.0748174324
org-element-text-markup-successor 10 0.309006309 0.0309006308
org-element-timestamp-successor 5 0.087275674 0.0174551348
org-element-statistics-cookie-successor 5 0.086838821 0.0173677642
org-element-footnote-reference-successor 5 0.0866179840 0.0173235968
org-element-target-successor 5 0.086057234 0.0172114468
org-element-radio-target-successor 5 0.083322691 0.0166645382
org-element-export-snippet-successor 5 0.083078665 0.016615733
org-element-macro-successor 5 0.0828692849 0.0165738569
scroll-down-command 3 0.059660938 0.0198869793
--8<---------------cut here---------------end--------------->8---
Interestingly, after calling elp-results, just trying to navigate to the
org buffer with other-window takes some time. Here's the top of the new
elp list:
--8<---------------cut here---------------start------------->8---
flyspell-post-command-hook 1 1.780324266 1.780324266
flyspell-word 1 1.780091208 1.780091208
org-mode-flyspell-verify 1 1.779600437 1.779600437
org-element-context 1 1.6563819400 1.6563819400
org-element--get-next-object-candidates 137 1.6448783359 0.0120064112
org-element-latex-or-entity-successor 1 0.753972365 0.753972365
org-element-link-successor 8 0.2225488189 0.0278186023
org-element-sub/superscript-successor 129 0.206221951 0.0015986197
org-element-line-break-successor 1 0.19533769 0.19533769
org-element-at-point 1 0.12299361 0.12299361
org-element--parse-to 1 0.122834024 0.122834024
org-element--current-element 2 0.085711871 0.0428559355
org-element-paragraph-parser 2 0.084995934 0.042497967
org-element-inline-src-block-successor 1 0.074795629 0.074795629
org-element-text-markup-successor 2 0.0633956 0.0316978
org-element-export-snippet-successor 1 0.020310027 0.020310027
org-element-timestamp-successor 1 0.017533021 0.017533021
org-element-footnote-reference-successor 1 0.017371898 0.017371898
org-element-statistics-cookie-successor 1 0.017308972 0.017308972
org-element-radio-target-successor 1 0.016797529 0.016797529
org-element-macro-successor 1 0.016773084 0.016773084
org-element-target-successor 1 0.016545404 0.016545404
org-element-subscript-parser 128 0.0039024930 3.048...e-05
org-element-inline-babel-call-successor 1 0.000752604 0.000752604
org-element-link-parser 7 0.000386358 5.5194e-05
flyspell-get-word 1 0.000357585 0.000357585
--8<---------------cut here---------------end--------------->8---
My apologies for the Arch specific example. The indentation of the
output seems to cause particular problems. I'd be happy to provide a
full test file off-list if required.
But this works (more or less) with other very large chunks of text.
E.g.,
C-u M-! w3m -dump http://www.gnu.org/software/emacs/manual/html_mono/emacs.html
Is it possible to speed up org-element-context here? For something
called as often as org-mode-flyspell-verify, do we need all the overhead
of the org-element parser? Or would a hack optimized for speed (which is
what the older version of org-mode-flyspell-verify represented) be
enough? I recall (though my memory may be faulty) discussions on the
list quite some time back in which we decided to prioritize
speed/efficiency over thoroughness/completeness in the checks run by
org-mode-flyspell-verify.
Thanks,
Matt
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: org-element checks make flyspell prohibitively slow
2014-03-17 14:32 org-element checks make flyspell prohibitively slow Matt Lundin
@ 2014-03-17 21:22 ` Nicolas Goaziou
2014-03-21 18:56 ` Matt Lundin
0 siblings, 1 reply; 3+ messages in thread
From: Nicolas Goaziou @ 2014-03-17 21:22 UTC (permalink / raw)
To: Matt Lundin; +Cc: Emacs-orgmode list
Hello,
Matt Lundin <mdl@imapmail.org> writes:
> The rewrite of org-mode-flyspell-verify in commit
> 4a27c2b4b67201e0b23f431bdaeb6460b31e1394 (Nov 21, 2013) makes navigating
> org-mode files with large chunks of text very slow.
[...]
> => Org-mode version 8.2.5h (release_8.2.5h-757-gc444e4 @
> /home/matt/org-mode/lisp/)
Could you update and try again? Parser's cache was inadvertently
disabled. I re-enabled it.
> I open a test.org file containing the following.
>
> * A headline * Arch packages * Another headline
>
> After opening a line under "Arch Packages" I call...
>
> C-u M-! pacman -Ss [RET]
>
> (Of course, this only works with archlinux.) This inserts a long list of
> packages that look like this:
>
> core/acl 2.2.52-2 [installed]
> Access control list utilities, libraries and headers
> core/archlinux-keyring 20140220-1 [installed]
> Arch Linux PGP keyring
> core/attr 2.4.47-1 [installed]
> Extended attribute support library for ACL support
> core/autoconf 2.69-1 (base-devel) [installed]
> A GNU tool for automatically configuring source code
> core/automake 1.14.1-1 (base-devel) [installed]
> A GNU tool for automatically creating Makefiles
>
> All in all, it's 12680 lines....
Note that it is a contrived example: the whole buffer is a single
paragraph containing around 150 objects. The current algorithm for
`org-element-context' is clearly not on par with such a density of
objects per paragraph.
Also, cache cannot help here, because each time you edit a paragraph,
all objects within are removed from the cache (because, AFAIK, there is
no way to know if the edition altered a previously parsed object or not,
so, as a security measure, all of them are wiped out) and you have to
parse them again.
Therefore, navigation should be fast but editing (with flyspell-mode
enabled) is going to be slow.
> But this works (more or less) with other very large chunks of text.
> E.g.,
>
> C-u M-! w3m -dump http://www.gnu.org/software/emacs/manual/html_mono/emacs.html
This one should be reasonably fast in both cases.
> Is it possible to speed up org-element-context here?
Certainly. `org-element-context' is the less optimized part of the
parser code. There is room for improvements.
> For something called as often as org-mode-flyspell-verify, do we need
> all the overhead of the org-element parser?
Of course.
> Or would a hack optimized for speed (which is what the older version
> of org-mode-flyspell-verify represented) be enough?
IMO, the old version of this function was annoying as soon as you
switched to a non-english language. YMMV.
> I recall (though my memory may be faulty) discussions on the list
> quite some time back in which we decided to prioritize
> speed/efficiency over thoroughness/completeness in the checks run by
> org-mode-flyspell-verify.
Why prioritize when we can have both?
I agree that `org-mode-flyspell-verify' is not fast enough at the time
being, but it is quite usable anyway. Also, as a very demanding
function, it is a good benchmark for the parser.
In order to improve the current state, reports (like those in your
message) help a lot. You can also help improving the algorithms.
Thank you.
Regards,
--
Nicolas Goaziou
^ permalink raw reply [flat|nested] 3+ messages in thread
* Re: org-element checks make flyspell prohibitively slow
2014-03-17 21:22 ` Nicolas Goaziou
@ 2014-03-21 18:56 ` Matt Lundin
0 siblings, 0 replies; 3+ messages in thread
From: Matt Lundin @ 2014-03-21 18:56 UTC (permalink / raw)
To: Nicolas Goaziou; +Cc: Emacs-orgmode list
Hi Nicolas,
Nicolas Goaziou <n.goaziou@gmail.com> writes:
> Matt Lundin <mdl@imapmail.org> writes:
>
>> The rewrite of org-mode-flyspell-verify in commit
>> 4a27c2b4b67201e0b23f431bdaeb6460b31e1394 (Nov 21, 2013) makes navigating
>> org-mode files with large chunks of text very slow.
>
> [...]
>
>> => Org-mode version 8.2.5h (release_8.2.5h-757-gc444e4 @
>> /home/matt/org-mode/lisp/)
>
> Could you update and try again? Parser's cache was inadvertently
> disabled. I re-enabled it.
Yes, I can confirm that it is faster now. Thanks.
>> core/acl 2.2.52-2 [installed]
>> Access control list utilities, libraries and headers
>> core/archlinux-keyring 20140220-1 [installed]
>> Arch Linux PGP keyring
>> core/attr 2.4.47-1 [installed]
>> Extended attribute support library for ACL support
>> core/autoconf 2.69-1 (base-devel) [installed]
>> A GNU tool for automatically configuring source code
>> core/automake 1.14.1-1 (base-devel) [installed]
>> A GNU tool for automatically creating Makefiles
>>
>> All in all, it's 12680 lines....
>
> Note that it is a contrived example: the whole buffer is a single
> paragraph containing around 150 objects. The current algorithm for
> `org-element-context' is clearly not on par with such a density of
> objects per paragraph.
Yes, it is indeed a contrived example. I originally thought I had a use
for it --- i.e., analyzing the packages I had installed --- but soon
realized that such a task is better accomplished in a separate text
file.
> Also, cache cannot help here, because each time you edit a paragraph,
> all objects within are removed from the cache (because, AFAIK, there
> is no way to know if the edition altered a previously parsed object or
> not, so, as a security measure, all of them are wiped out) and you
> have to parse them again.
>
> Therefore, navigation should be fast but editing (with flyspell-mode
> enabled) is going to be slow.
Good to know.
Thanks again!
Matt
^ permalink raw reply [flat|nested] 3+ messages in thread
end of thread, other threads:[~2014-03-21 18:57 UTC | newest]
Thread overview: 3+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-03-17 14:32 org-element checks make flyspell prohibitively slow Matt Lundin
2014-03-17 21:22 ` Nicolas Goaziou
2014-03-21 18:56 ` Matt Lundin
Code repositories for project(s) associated with this public inbox
https://git.savannah.gnu.org/cgit/emacs/org-mode.git
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 read-only IMAP folder(s) and NNTP newsgroup(s).