New feature branch: org-xhtml-and-org-odt

New feature branch: org-xhtml-and-org-odt

[Jambunathan K]

Bastien

I have created a new feature branch named "org-xhtml-and-org-odt".

The main "feature" introduced by this branch is:

1. The HTML exporter in lisp/org-html.el is *totally* rewritten. It
   should be on par with the current HTML exporter in the master branch.

   Note: Anyone who depends on HTML export for his day-to-day work
   *must* make sure that nothing is broken.

2. The ODT exporter is now moved to lisp/org-odt.el

Additional Info:

1. testing
   - There is a testfile in contrib/odt/tests/test.org which users can use
     to unit test the changes.

2. info manual 

   - I have updated the ODT section in the info manual and it is worth a
     read for serious users. (Chirstian, I will revise the manual later
     in the day. For now, it is pretty much same as what you had seen
     last time.)

Let me know if there are any questions or comments. 

Jambunathan K.
-- 

Re: New feature branch: org-xhtml-and-org-odt

[Jambunathan K]

> 1. The HTML exporter in lisp/org-html.el is *totally* rewritten. It
>    should be on par with the current HTML exporter in the master
>    branch.

A small note: If you are exercising the new exporter then you will the
following message. (Note the presence of org-lparse in the message)

,----
| Exporting to HTML using org-lparse...
`----

Re: New feature branch: org-xhtml-and-org-odt

[Bastien]

Dear all,

thanks Jambunathan for setting up this branch!

*Please all*: test this branch as much as you can.  We plan to 
integrate the ODT exporter into Org's core for the next release,
and this will only happen if there is no regression between the 
current org-html.el and org-xhtml.el from Jambunathan's branch.

Thanks for your help,

-- 
 Bastien

Re: New feature branch: org-xhtml-and-org-odt

[joakim]

Jambunathan K <kjambunathan@gmail.com> writes:

> Bastien
>
> I have created a new feature branch named "org-xhtml-and-org-odt".
>
> The main "feature" introduced by this branch is:
>
> 1. The HTML exporter in lisp/org-html.el is *totally* rewritten. It
>    should be on par with the current HTML exporter in the master branch.
>
>    Note: Anyone who depends on HTML export for his day-to-day work
>    *must* make sure that nothing is broken.
>
> 2. The ODT exporter is now moved to lisp/org-odt.el
>
> Additional Info:
>
> 1. testing
>    - There is a testfile in contrib/odt/tests/test.org which users can use
>      to unit test the changes.
>
> 2. info manual 
>
>    - I have updated the ODT section in the info manual and it is worth a
>      read for serious users. (Chirstian, I will revise the manual later
>      in the day. For now, it is pretty much same as what you had seen
>      last time.)
>
> Let me know if there are any questions or comments. 

I hope this is not OT but I have a question regarding this new
infrastructure.

I have a project at github called inkmacs, which integrates inkscape,
org-mode, and emacs:

https://github.com/jave/inkmacs

There I support interactive transfer of org nodes to inkscapes text
nodes.

My current code to extract the org nodes is not beautiful and I though
maybe org-lparse improves the situation?

Below is an excerpt from the code. As you can see it's very primitive
and does not support many org constructs. (Its surprisingly useful
anyway though)


-----
(defun org-get-entry-2 ()
  "Get the entry text, after heading, to next heading, or eof."
  (save-excursion
    (org-back-to-heading t)
    (let ((p1 (point-at-bol 2))
          (p2 (progn (forward-line) (search-forward "*" nil t))))
      (setq p2 (if (null p2) (point-max)
                 (1- p2)))
      (buffer-substring p1  p2))))

(defun inkorg-entry-text ()
  "Extract text from current org node.
Return a format suitable to
create an inkscape text node from.
asterisks and properties are removed."
  ;;TODO there ought to be some code in org-exp for this somewhere(org-ascii for example)
  (let ((text  (concat (org-get-heading) "\n" (org-get-entry-2))))
    (set-text-properties 0 (length text) nil text )

    (replace-regexp-in-string "\\([^\n]\\)\n\\([^\n]\\)" "\\1 \\2" 
                              (concat
                               (substring text 0 (string-match org-property-start-re text))
                               (if (string-match org-property-end-re text)
                                   (substring text (progn (string-match org-property-end-re text) (match-end 0)) (length text)))))))
----------

>
> Jambunathan K.

-- 
Joakim Verona

Re: New feature branch: org-xhtml-and-org-odt

[Bastien]

Hi Joakim,

joakim@verona.se writes:

> I have a project at github called inkmacs, which integrates inkscape,
> org-mode, and emacs:
>
> https://github.com/jave/inkmacs
>
> There I support interactive transfer of org nodes to inkscapes text
> nodes.
>
> My current code to extract the org nodes is not beautiful and I though
> maybe org-lparse improves the situation?

Along with org-lparse.el, the next release will contain a proper Org
parser in contrib/ -- it will do what you want.  I let Nicolas publicly
announce this code when he thinks it's mature enough.

Thanks,

-- 
 Bastien

Re: New feature branch: org-xhtml-and-org-odt

[joakim]

Bastien <bzg@altern.org> writes:

> Hi Joakim,
>
> joakim@verona.se writes:
>
>> I have a project at github called inkmacs, which integrates inkscape,
>> org-mode, and emacs:
>>
>> https://github.com/jave/inkmacs>
>> There I support interactive transfer of org nodes to inkscapes text
>> nodes.
>>
>> My current code to extract the org nodes is not beautiful and I though
>> maybe org-lparse improves the situation?
>
> Along with org-lparse.el, the next release will contain a proper Org
> parser in contrib/ -- it will do what you want.  I let Nicolas publicly
> announce this code when he thinks it's mature enough.

Interesting. Is this available in a branch somewhere? Meanwhile I'll
have a look at org-lparse.

>
> Thanks,

-- 
Joakim Verona

Re: New feature branch: org-xhtml-and-org-odt

[Bastien]

Hi Joakim,

joakim@verona.se writes:

> Interesting. Is this available in a branch somewhere? Meanwhile I'll
> have a look at org-lparse.

Not yet.

org-lparse is a line parser, currently used for the ODT exporter, 
and planned to be used for the HTML exporter (which is already based
on line parsing, but does not use an external parser.)

The parser from Nicolas stores an Org buffer (or region) into a list,
reflecting the structure of all the elements and their properties.

Maybe you saw the slides I presented at the GNU Hackers meeting,
I presented an sneak preview of an early prototype by Nicolas - see

http://lumiere.ens.fr/~guerry/emacs-orgmode-gnu-hackers-meeting-2011.html

See slide 19.

HTH,

-- 
 Bastien

Re: New feature branch: org-xhtml-and-org-odt

[joakim]

Bastien <bzg@altern.org> writes:

> Hi Joakim,
>
> joakim@verona.se writes:
>
>> Interesting. Is this available in a branch somewhere? Meanwhile I'll
>> have a look at org-lparse.
>
> Not yet.
>
> org-lparse is a line parser, currently used for the ODT exporter, 
> and planned to be used for the HTML exporter (which is already based
> on line parsing, but does not use an external parser.)
>
> The parser from Nicolas stores an Org buffer (or region) into a list,
> reflecting the structure of all the elements and their properties.
>
> Maybe you saw the slides I presented at the GNU Hackers meeting,
> I presented an sneak preview of an early prototype by Nicolas - see
>
> http://lumiere.ens.fr/~guerry/emacs-orgmode-gnu-hackers-meeting-2011.html
>
> See slide 19.
>
> HTH,

Yes, I saw the presentation and the idea seems nice. However it would be
nice if the code could be put under public scrutiny regardless of its
current state.

For instance, Semantic(part of CEDET in the Emacs core) implements on
the fly parse tree generators for a number of languages and it also has
proof of concept support for emacs outlines, of which org is a
descendant. So it would be nice to know if the org parser re-uses
Semantic or if it chooses another path. If Semantic was used we could
also use SRecode as a template language to implement other formats. I
could certainly use it in my application. Even if the new org parser
doesn't reuse the parser part of semantic, Wisent, we could probably
still hook it into Semantic. Anyway, we are all busy so it would be nice
to reduce wasted effort.





-- 
Joakim Verona

Re: New feature branch: org-xhtml-and-org-odt

[Matt Price]

[-- Attachment #1: Type: text/plain, Size: 393 bytes --]

On Mon, Nov 14, 2011 at 3:13 AM, Jambunathan K <kjambunathan@gmail.com>wrote:

> Bastien
>
> I have created a new feature branch named "org-xhtml-and-org-odt".
>
> hmm, what are the pul instructions for this? sorry, my git-fu is pretty
basic: I tried
git pull origin/org-xhtml-and-org-odt
and
git pull org-xhtml-and-org-odt

but clearly those aren't the syntax I'm looking for...
thanks,
matt

[-- Attachment #2: Type: text/html, Size: 697 bytes --]

Re: New feature branch: org-xhtml-and-org-odt

[suvayu ali]

Hi Matt,

On Mon, Nov 14, 2011 at 21:06, Matt Price <moptop99@gmail.com> wrote:
> hmm, what are the pul instructions for this? sorry, my git-fu is pretty
> basic: I tried
> git pull origin/org-xhtml-and-org-odt
> and
> git pull org-xhtml-and-org-odt

With master checked out (git checkout master), do the following:

$ git pull
$ git checkout -b <some_name_u_like> origin/org-xhtml-and-org-odt

Now you are on the <some_name_u_like> branch which tracks
origin/org-xhtml-and-org-odt. Now you can compile, experiment whatever.
When you have had enough fun, to get back to master do the following:

$ git checkout master

Any subsequent updates to the branch from Jambunathan will be pulled in
automatically when you do `git pull' in the future. However your local
branch tracking origin/org-xhtml-and-org-odt will not be updated. To do
that do the following:

$ git checkout <some_name_u_like>
$ git pull -f

The `-f' is there to deal with situations when Jambunathan rebases his
branch on top of the latest master (this is common practice for topic
branches).

Hope this helps.

-- 
Suvayu

Open source is the future. It sets us free.