worg-new-exporter successful publish locally

worg-new-exporter successful publish locally

[Jay Kerns]

Hi there,

The new Worg branch publishes for me without error under Org-mode
version 8.0.1 (release_8.0.1-24-g92f372).  I updated the existing
emacs.el (underlying Worg, IIUC) to the Org 8.0 exporting/publishing
syntax. You can see the latest draft in

worgtest-init.el

which lives in the top-level directory. To test the publish yourself
you'd need to customize the base/publishing directories for your local
machine and then do

emacs -Q -l "/path/to/worg/worgtest-init.el"

followed by

M-x publish-worg

There are changes to some files which are necessary for a successful
publish and which are detailed at the bottom of this email. Have a
look in FIXME.

Given the successful publish locally, I went through and updated all
remaining in-buffer settings that I could find (HTML_HEAD, etc.).  To
the best of my knowledge, everything is DONE, notwithstanding the
comments below.

DISCLAIMERS
1. I didn't work on orgmode.org at all.
2. Successful export/publish doesn't say anything about the content
   and/or outdated info in the pages.

Cheers,

-- 
Jay


- sitemap.org
  - Error: Invalid time-stamp specification
  - The error here has to do with an undocumented function that I
    can't figure out.
  - Status: disabled auto-sitemap in publish-alist

- FIXME/how-to-use-Babel-for-R.org
  - content is out of date, but exports without error
  - Status: updated in-buffer settings but otherwise unchanged

- FIXME/test-for-how-to-use-Babel-for-R.org
  - content is out of date, but exports without error
  - Status: updated buffer settings but otherwise unchanged

- org-contrib/babel/header-args.org
  - Error: local variables in header-args.org and header-args.html
    - eval: org-babel-default-header-args:Python
    - if I hit "n" then Emacs hangs until C-g
    - Status: moved to FIXME

- org-tutorials/org-R/variable-popcon.org
  - Error: none, really, it takes a really long time to export but
    succeeds eventually.
  - Status: unchanged, but somebody might investigate the reason for
    this extended export time, maybe it's just because the file is so
    big.

- org-configs/org-customization-survey.org
  - export of this single file takes 1.5 hrs on my laptop
  - I don't know why
  - Status: moved to FIXME

- The following appear to set file local variables
  - code/org-info-js/changes.org
  - code/org-info-js/org-slides/slides.org
  - org-contrib/org-drill.org
  - org-contrib/org-git-link.org
  - org-contrib/babel/examples/drift.org
  - Status updated in buffer settings, but otherwise unchanged.
    Thought you might like to know.

Re: worg-new-exporter successful publish locally

[Bastien]

Hi Jay,

Jay Kerns <gjkernsysu@gmail.com> writes:

> The new Worg branch publishes for me without error under Org-mode
> version 8.0.1 (release_8.0.1-24-g92f372).

*Wow*!  That was super fast!  Thanks a lot for leading this project,
quite an achievement.  I won't have time to test over the week-end but
I hope a few people will, so that I can make the switch on monday when
I have a good connection (not at home right now.)

Thanks again!

-- 
 Bastien

Re: worg-new-exporter successful publish locally

[Carsten Dominik]

On 27.4.2013, at 08:59, Bastien <bzg@gnu.org> wrote:

> Hi Jay,
> 
> Jay Kerns <gjkernsysu@gmail.com> writes:
> 
>> The new Worg branch publishes for me without error under Org-mode
>> version 8.0.1 (release_8.0.1-24-g92f372).
> 
> *Wow*!  That was super fast!  Thanks a lot for leading this project,
> quite an achievement.  I won't have time to test over the week-end but
> I hope a few people will, so that I can make the switch on monday when
> I have a good connection (not at home right now.)

Hi Jay,

I just want to second Bastien's amazement.  Thanks a lot for going through this!

- Carsten

Re: worg-new-exporter successful publish locally

[Thorsten Jolitz]

Jay Kerns <gjkernsysu@gmail.com> writes:

Hi Jay,

I started this page, so I kind of 'own' it, but the local variable has
been added later, not by me, probably by Eric Schulte, because he is
mentioned as second author.

,-------------------------------------------------------------------
| - org-contrib/babel/header-args.org
|   - Error: local variables in header-args.org and header-args.html
|     - eval: org-babel-default-header-args:Python
|     - if I hit "n" then Emacs hangs until C-g
|     - Status: moved to FIXME
`-------------------------------------------------------------------

-- 
cheers,
Thorsten

Re: worg-new-exporter successful publish locally

[Jay Kerns]

HI Thorsten,

On Sat, Apr 27, 2013 at 4:40 AM, Thorsten Jolitz <tjolitz@gmail.com> wrote:
> Jay Kerns <gjkernsysu@gmail.com> writes:
>
> Hi Jay,
>
> I started this page, so I kind of 'own' it, but the local variable has
> been added later, not by me, probably by Eric Schulte, because he is
> mentioned as second author.
>
> ,-------------------------------------------------------------------
> | - org-contrib/babel/header-args.org
> |   - Error: local variables in header-args.org and header-args.html
> |     - eval: org-babel-default-header-args:Python
> |     - if I hit "n" then Emacs hangs until C-g
> |     - Status: moved to FIXME
> `-------------------------------------------------------------------
>
> --
> cheers,
> Thorsten

Sounds good - thanks for the background info. IIRC I didn't modify
that file because I really didn't know anything else to do other than
delete the bit at the bottom which seemed to be causing the trouble,
so instead, it appeared better to ask people more competent than me to
take a look at it.  Certainly Eric (CC:'ed now) would be one of those,
among many, many others.  :-)

-- 
Jay

Re: worg-new-exporter successful publish locally

[Achim Gratz]

Jay Kerns writes:
> The new Worg branch publishes for me without error under Org-mode
> version 8.0.1 (release_8.0.1-24-g92f372).  I updated the existing
> emacs.el (underlying Worg, IIUC) to the Org 8.0 exporting/publishing
> syntax. You can see the latest draft in
>
> worgtest-init.el
>
> which lives in the top-level directory.

Great, thank you.

I've just pushed a change that removes the hardcoded pathnames in this
file and replaces them with three variables.  Set these variables either
directly on the command line or in a local init file (name it whatever
you like and don't push it to Worg) like this:
    
--8<---------------cut here---------------start------------->8---
;; worgtest-local-init.el
;;
;; all paths must end with a slash
(setq worg-base "/path/to/Worg/"
      worg-htmlroot "/path/to/htmlroot/"
      worg-add-load-path (list "/path/to/auctex-11.86/"
			       "/path/to/htmlize-1.39/"
			       "/path/to/org/"))
--8<---------------cut here---------------end--------------->8---

emacs -batch -Q -l worgtest-local-init.el -l worgtest-init.el -f publish-worg

I've also deactivated vc during export (makes it about half a minute
faster) and moved the require for org before the defcustoms stuff,
otherwise I'd get the error

Symbol's function definition is void: org-get-file-contents

This exports OK with Emacs 24.3, freshly compiled today.


Worg: the "Invalid time specification" error is back.  I can push, but
nothing gets published at the moment.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs

Re: worg-new-exporter successful publish locally

[Jay Kerns]

Dear Achim,

On Sat, Apr 27, 2013 at 11:43 AM, Achim Gratz <Stromeko@nexgo.de> wrote:

[snip]

Thank you for looking into this.  The worg-local-init.el is a great
idea! Thanks, also, for confirming successful publish - that's great
news.

I have a question, a comment, and some information to add to my earlier
message.

1. I set up worg-local-init.el as you recommended and everything goes
   smoothly until here:

,----
| Publishing file /home/jay/git/worg/org-hacks.org using
`org-html-publish-to-html'
| Loading cc-langs...
| Symbol's function definition is void: nil
`----

Now, I have opened org-hacks.org (with the same emacs -Q as above but
without batch) and it exports successfully to HTML from the
interactive dispatcher.  Do you have any idea what the problem might
be?

2. If I don't do batch but instead publish interactively with M-x
   publish-worg, the org-hacks.org error doesn't appear but then near
   the end I get

org-publish-get-base-files: Wrong type argument: stringp, (concat
worg-base "org-contrib/bibtex/")

I haven't had time to look deeper into this to see what's going on,
but I thought you might like to know.

3. I've been working on that invalid time format problem we've been having
   and I think I might have identified where the problem is.  I will
   write about it in another message.

Thanks again.

-- 
Jay

Re: worg-new-exporter successful publish locally

[Achim Gratz]

Jay Kerns writes:
> 1. I set up worg-local-init.el as you recommended and everything goes
>    smoothly until here:
>
> ,----
> | Publishing file /home/jay/git/worg/org-hacks.org using
> `org-html-publish-to-html'
> | Loading cc-langs...
> | Symbol's function definition is void: nil
> `----

Yes, I get that, too.

> Now, I have opened org-hacks.org (with the same emacs -Q as above but
> without batch) and it exports successfully to HTML from the
> interactive dispatcher.  Do you have any idea what the problem might
> be?

Something in cc-langs that doe not work correctly in batch mode, it
fails to define or select the correct fontification function for gawk.
In any case, it doesn't happen when you don't export in batch mode and
Emacs 23 does not have that problem, interestingly enough…

> 2. If I don't do batch but instead publish interactively with M-x
>    publish-worg, the org-hacks.org error doesn't appear but then near
>    the end I get
>
> org-publish-get-base-files: Wrong type argument: stringp, (concat
> worg-base "org-contrib/bibtex/")
>
> I haven't had time to look deeper into this to see what's going on,
> but I thought you might like to know.

That's a bug in my code.  This is used inside a backquoted list in some
places and should have a "," prepended there to be evaluated when
defining the list.  I've fixed it on Worg.

> 3. I've been working on that invalid time format problem we've been having
>    and I think I might have identified where the problem is.  I will
>    write about it in another message.

Great, thanks.


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Factory and User Sound Singles for Waldorf Q+, Q and microQ:
http://Synth.Stromeko.net/Downloads.html#WaldorfSounds

Re: worg-new-exporter successful publish locally

[Jay Kerns]

On Sun, Apr 28, 2013 at 3:12 PM, Achim Gratz <Stromeko@nexgo.de> wrote:

[snip]

> I've fixed it on Worg.
>

Confirmed - thanks!

-- 
Jay

Re: worg-new-exporter successful publish locally

[Achim Gratz]

Achim Gratz writes:
> Something in cc-langs that doe not work correctly in batch mode, it
> fails to define or select the correct fontification function for gawk.
> In any case, it doesn't happen when you don't export in batch mode and
> Emacs 23 does not have that problem, interestingly enough…

I've fixed this, not that I would know exactly how… :-)


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Waldorf MIDI Implementation & additional documentation:
http://Synth.Stromeko.net/Downloads.html#WaldorfDocs

Re: worg-new-exporter successful publish locally

[Jay Kerns]

On Sun, Apr 28, 2013 at 4:21 PM, Achim Gratz <Stromeko@nexgo.de> wrote:
> Achim Gratz writes:

>
> I've fixed this, not that I would know exactly how… :-)
>

Confirmed again, and thanks again!  I'm really starting to warm up to
this batch publishing approach, it's quite handy.


-- 
Jay

Re: worg-new-exporter successful publish locally

[Achim Gratz]

Achim Gratz writes:
> Achim Gratz writes:
>> Something in cc-langs that doe not work correctly in batch mode, it
>> fails to define or select the correct fontification function for gawk.
>> In any case, it doesn't happen when you don't export in batch mode and
>> Emacs 23 does not have that problem, interestingly enough…
>
> I've fixed this, not that I would know exactly how… :-)

This is now reported as Emacs bug #14325.  Apparently the workaround
should be good until a better fix is found.  If someone understands
minor modes and font-lock in particular, then maybe Org could include
some fixes for using font-lock in noninteractive application (aka batch
mode).


Regards,
Achim.
-- 
+<[Q+ Matrix-12 WAVE#46+305 Neuron microQkb Andromeda XTk Blofeld]>+

Factory and User Sound Singles for Waldorf Blofeld:
http://Synth.Stromeko.net/Downloads.html#WaldorfSounds