Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

[Omid]

Hello All,

When exporting to HTML (Org mode version 8.2.1 ELPA; GNU Emacs version
24.3.1) I see the a strange behavior which can be reproduced with the
following minimal example:

This exports correctly
  #+BEGIN_SRC sh
ls
  #+END_SRC

- This also exports correctly
#+BEGIN_SRC sh
ls
#+END_SRC

- This does NOT export correctly (code block is not detected)
  #+BEGIN_SRC sh
ls
  #+END_SRC

Is there any way around this behavior without removing the indentation
for the code block delimiters?

Thanks,

-- 
Omid

Sent from my Emacs

Re: Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

[Thorsten Jolitz]

Omid <omidlink@gmail.com> writes:

> - This does NOT export correctly (code block is not detected)
>   #+BEGIN_SRC sh
> ls
>   #+END_SRC
>
> Is there any way around this behavior without removing the indentation
> for the code block delimiters?

I think the 'official'regexp to detect a src-block starts like this

,----------------------------------
| "^\\([ \t]*\\)#\\+begin_src [...]
`----------------------------------

so I would guess that this should work, and its a bug if it doesn't.

Inside a list like this

 - First item

   #+begin_src emacs-lisp :exports both
    (+ 3 4)
   #+end_src

 - Second Item

   Some text

it should work too, lets test:

#+begin_src emacs-lisp :exports value
   (with-current-buffer
       (current-buffer)
     (org-export-as 'ascii))
#+end_src

,-----------------------------------------------------------------------
| #+results:
|                            _________________
|
|                                   427
|
|                             Thorsten Jolitz
|                            _________________
|
|
|
|
|
| 1 --text follows this line--
| ============================
|
|   Omid <omidlink@gmail.com> writes:
|
|   > - This does NOT export correctly (code block is not detected)
|   > #+BEGIN_SRC sh ls #+END_SRC Is there any way around this behavior
|   > without removing the indentation for the code block delimiters?
|
|   I think the 'official'regexp to detect a src-block starts like this
|
|   ,----------------------------------
|   "^\\([ \t]*\\)#\\+begin_src [...]
|   `----------------------------------
|
|   so I would guess that this should work, and its a bug if it doesn't.
|
|   Inside a list like like this
|
|   - First item
|
|     ,----
|     | (+ 3 4)
|     `----
|
|   ,----
|   | 7
|   `----
|
|
|   - Second Item
|
|     Some text
|
|   it should work too, lets test:
|
|   ,----
|   | (with-current-buffer
|   |     (current-buffer)
|   |   (org-export-as 'ascii))
|   `----
|
|
|   -- cheers, Thorsten
`-----------------------------------------------------------------------

looks as if it works for me ...

--
cheers,
Thorsten

Re: Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

[Thorsten Jolitz]

PS: 

Omid <omidlink@gmail.com> writes:

> When exporting to HTML (Org mode version 8.2.1 ELPA; GNU Emacs version
> 24.3.1) I see the a strange behavior which can be reproduced with the
> following minimal example:
>
> This exports correctly
>   #+BEGIN_SRC sh
> ls
>   #+END_SRC
>
> - This also exports correctly
> #+BEGIN_SRC sh
> ls
> #+END_SRC
>
> - This does NOT export correctly (code block is not detected)
>   #+BEGIN_SRC sh
> ls
>   #+END_SRC
>
> Is there any way around this behavior without removing the indentation
> for the code block delimiters?

 This exports correctly
#+BEGIN_SRC sh
 ls
#+END_SRC

 - This also exports correctly
#+BEGIN_SRC sh
 ls
#+END_SRC

 - This does NOT export correctly (code block is not detected)
#+BEGIN_SRC sh
 ls
#+END_SRC


#+begin_src emacs-lisp :exports value
   (with-current-buffer
       (current-buffer)
     (org-export-as 'html))
#+end_src

#+results:
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
[...]
This exports correctly
</p>
<div class="org-src-container">

<pre class="src src-sh">ls
</pre>
</div>

<ul class="org-ul">
<li>This also exports correctly
</li>
</ul>
<div class="org-src-container">

<pre class="src src-sh">ls
</pre>
</div>

<ul class="org-ul">
<li>This does NOT export correctly (code block is not detected)
<p>
#+BEGIN_SRC sh
</p>
</li>
</ul>
<p>
ls
</p>
<p>
#+END_SRC
</p>

with html export, the last case does look quite different from the other ones.

-- 
cheers,
Thorsten

Re: Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

[Omid]

Thank you for your quick reply, Thorsten.

Yes, as I mentioned, and as you noticed, the HTML export does not seem
to work correctly for the last case. The #BEGIN_SRC and #END_SRC
delimiters are exported literally.

--
Omid

Sent from my Emacs

On 10/15/2013 10:53 PM, Thorsten Jolitz wrote:
> 
> PS: 
> 
> Omid <omidlink@gmail.com> writes:
> 
>> When exporting to HTML (Org mode version 8.2.1 ELPA; GNU Emacs version
>> 24.3.1) I see the a strange behavior which can be reproduced with the
>> following minimal example:
>>
>> This exports correctly
>>   #+BEGIN_SRC sh
>> ls
>>   #+END_SRC
>>
>> - This also exports correctly
>> #+BEGIN_SRC sh
>> ls
>> #+END_SRC
>>
>> - This does NOT export correctly (code block is not detected)
>>   #+BEGIN_SRC sh
>> ls
>>   #+END_SRC
>>
>> Is there any way around this behavior without removing the indentation
>> for the code block delimiters?
> 
>  This exports correctly
> #+BEGIN_SRC sh
>  ls
> #+END_SRC
> 
>  - This also exports correctly
> #+BEGIN_SRC sh
>  ls
> #+END_SRC
> 
>  - This does NOT export correctly (code block is not detected)
> #+BEGIN_SRC sh
>  ls
> #+END_SRC
> 
> 
> #+begin_src emacs-lisp :exports value
>    (with-current-buffer
>        (current-buffer)
>      (org-export-as 'html))
> #+end_src
> 
> #+results:
> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
> [...]
> This exports correctly
> </p>
> <div class="org-src-container">
> 
> <pre class="src src-sh">ls
> </pre>
> </div>
> 
> <ul class="org-ul">
> <li>This also exports correctly
> </li>
> </ul>
> <div class="org-src-container">
> 
> <pre class="src src-sh">ls
> </pre>
> </div>
> 
> <ul class="org-ul">
> <li>This does NOT export correctly (code block is not detected)
> <p>
> #+BEGIN_SRC sh
> </p>
> </li>
> </ul>
> <p>
> ls
> </p>
> <p>
> #+END_SRC
> </p>
> 
> with html export, the last case does look quite different from the other ones.
> 

Re: Org mode export to HTML not working correctly with code blocks if block is after an item (-) and its delimiters are indented

[Nicolas Goaziou]

Hello,

Omid <omidlink@gmail.com> writes:

> When exporting to HTML (Org mode version 8.2.1 ELPA; GNU Emacs version
> 24.3.1) I see the a strange behavior which can be reproduced with the
> following minimal example:
>
> This exports correctly
>   #+BEGIN_SRC sh
> ls
>   #+END_SRC
>
> - This also exports correctly
> #+BEGIN_SRC sh
> ls
> #+END_SRC
>
> - This does NOT export correctly (code block is not detected)
>   #+BEGIN_SRC sh
> ls
>   #+END_SRC
>
> Is there any way around this behavior without removing the indentation
> for the code block delimiters?

This should now be fixed. Thank you for reporting it.


Regards,

-- 
Nicolas Goaziou

Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Omid]

Hello,

I am using Org-mode version 8.2.5g (8.2.5g-elpa) in GNU Emacs 24.3.1.
I have two questions about the behavior of the fantastic Org
mode+Babel with respect to code and comments:

Here is a minimal example:

  #+BEGIN_SRC f90 :results verbatim :exports both
program main
! This is a very very very very very very very very very very very very
very very very very long comment line.
print *, "Hello, World!"
end program main
  #+END_SRC

Below is the code snippet after M-q (fill-paragraph) on the comment
line. The comment line has been refilled (intended behavior) but all
lines have been indented. This may also be an intended behavior; but

FIRST QUESTION: Is there a way to disable this indentation upon M-Q in
Org Babel code snippets?

  #+BEGIN_SRC f90 :results verbatim :exports both
    program main
    ! This is a very very very very very very very very very very very very
    ! very very very very long comment line.
    print *, "Hello, World!"
    end program main
  #+END_SRC

Below is the code snippet after a C-/ (undo). Note that the
indentation and the refill which were done by the last command above
(M-q) are not being undone. The text in the comment is being removed
which I believe means that the previous (self-insert-command)'s that
created the text are being undone. This is very undesirable since even
a user who is aware of this behavior may by mistake issue the command
M-q, have the code snippet formatted in an undesirable way (e.g.,
sometimes new lines are not respected and code and comments get mixed
up), without any immediate way to undo this reformatting.

SECOND QUESTION: How can one get the usual (undo) behavior in a Babel
code snippet?

  #+BEGIN_SRC f90 :results verbatim :exports both
    program main
    ! This is a very very very very very very very very very very very very
    ! very v comment line.
    print *, "Hello, World!"
    end program main
  #+END_SRC

Please note that these edits are done in the Org mode buffer directly.
If I switch to the native language mode using C-c ' things work as
expected.

Thanks,
Omid

Re: Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Nick Dokos]

Omid <omidlink@gmail.com> writes:

> Hello,
>
> I am using Org-mode version 8.2.5g (8.2.5g-elpa) in GNU Emacs 24.3.1.
> I have two questions about the behavior of the fantastic Org
> mode+Babel with respect to code and comments:
>
> Here is a minimal example:
>
>   #+BEGIN_SRC f90 :results verbatim :exports both
> program main
> ! This is a very very very very very very very very very very very very
> very very very very long comment line.
> print *, "Hello, World!"
> end program main
>   #+END_SRC
>
> Below is the code snippet after M-q (fill-paragraph) on the comment
> line. The comment line has been refilled (intended behavior) but all
> lines have been indented. This may also be an intended behavior; but
>
> FIRST QUESTION: Is there a way to disable this indentation upon M-Q in
> Org Babel code snippets?
>

Try setting org-edit-src-content-indentation to 0. I'm not sure that
it is going to work, but (based on rather flimsy numerological evidence,
namely that its default value is 2 as is the indent below) it might.

>   #+BEGIN_SRC f90 :results verbatim :exports both
>     program main
>     ! This is a very very very very very very very very very very very very
>     ! very very very very long comment line.
>     print *, "Hello, World!"
>     end program main
>   #+END_SRC
>

BTW, I don't get this behavior but I don't use f90 mode, so I'm not sure
whether there is some setup I'm missing. I get (with
org-edit-src-content-indentation set to 0):

  #+BEGIN_SRC f90 :results verbatim :exports both
  program main ! This is a very very very very very very very very very
  very very very very very very very long comment line.  print *,
  "Hello, World!"  end program main
  #+END_SRC

Nick

Re: Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Omid]

Hello Nick,

Thanks for your quick reply. The variable
org-edit-src-content-indentation does indeed have to do with the
indentation I was referring to. But it is relative. According to the
documentation, it sets the "Indentation for the content of a source
code block. This should be the number of spaces added to the
indentation of the #+begin line in order to compute the indentation of
the block content after editing it with M-x org-edit-src-code. Has no
effect if `org-src-preserve-indentation' is non-nil." I wonder if one
can set it to be zero as an absolute value, meaning that if the a line
of code in the source block starts at column n of the buffer, it is
left at column n after M-q.

Regarding the behavior you observe, as I said I get that too from time
to time. In fact, doing M-q on my own example in a fresh Org mode
buffer, I now get what you get:

  #+BEGIN_SRC f90 :results verbatim :exports both
  program main ! This is a very very very very very very very very very
  very very very very very very very long comment line. print *, "Hello,
  World!" end program main
  #+END_SRC

I don't have any special settings for the f90 mode in my .emacs. In
fact, I have observed the same behavior for other languages too when
edited in an Org source block. The "destructive" nature of this
behavior is very special (C-/ (undo) doesn't work). Many a times, I
have tried refilling a line of comments and ended up with a huge mix
of code and comment that was impossible to undo (at least to the best
of my knowledge) and I had to either revert the change if I had the
file under version control, go to an auto backup file, or painfully and
manually separate the code and comments to get back the original code
block.

Omid


On 02/05/2014 04:34 PM, Nick Dokos wrote:
> Omid <omidlink@gmail.com> writes:
> 
>> Hello,
>>
>> I am using Org-mode version 8.2.5g (8.2.5g-elpa) in GNU Emacs 24.3.1.
>> I have two questions about the behavior of the fantastic Org
>> mode+Babel with respect to code and comments:
>>
>> Here is a minimal example:
>>
>>   #+BEGIN_SRC f90 :results verbatim :exports both
>> program main
>> ! This is a very very very very very very very very very very very very
>> very very very very long comment line.
>> print *, "Hello, World!"
>> end program main
>>   #+END_SRC
>>
>> Below is the code snippet after M-q (fill-paragraph) on the comment
>> line. The comment line has been refilled (intended behavior) but all
>> lines have been indented. This may also be an intended behavior; but
>>
>> FIRST QUESTION: Is there a way to disable this indentation upon M-Q in
>> Org Babel code snippets?
>>
> 
> Try setting org-edit-src-content-indentation to 0. I'm not sure that
> it is going to work, but (based on rather flimsy numerological evidence,
> namely that its default value is 2 as is the indent below) it might.
> 
>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>     program main
>>     ! This is a very very very very very very very very very very very very
>>     ! very very very very long comment line.
>>     print *, "Hello, World!"
>>     end program main
>>   #+END_SRC
>>
> 
> BTW, I don't get this behavior but I don't use f90 mode, so I'm not sure
> whether there is some setup I'm missing. I get (with
> org-edit-src-content-indentation set to 0):
> 
>   #+BEGIN_SRC f90 :results verbatim :exports both
>   program main ! This is a very very very very very very very very very
>   very very very very very very very long comment line.  print *,
>   "Hello, World!"  end program main
>   #+END_SRC
> 
> Nick
> 
> 
> 

Re: Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Samuel Wales]

Undo not working is a bug.  I think we identified the bug on the
mailing list, but the fix was never implemented, although I think an
attempted fix was.

I think it has to do with editing a source code block trying to do
things with undo or undo-boundary.  The fix is probably to rip out all
undo-related code there.


On 2/5/14, Omid <omidlink@gmail.com> wrote:
> Hello Nick,
>
> Thanks for your quick reply. The variable
> org-edit-src-content-indentation does indeed have to do with the
> indentation I was referring to. But it is relative. According to the
> documentation, it sets the "Indentation for the content of a source
> code block. This should be the number of spaces added to the
> indentation of the #+begin line in order to compute the indentation of
> the block content after editing it with M-x org-edit-src-code. Has no
> effect if `org-src-preserve-indentation' is non-nil." I wonder if one
> can set it to be zero as an absolute value, meaning that if the a line
> of code in the source block starts at column n of the buffer, it is
> left at column n after M-q.
>
> Regarding the behavior you observe, as I said I get that too from time
> to time. In fact, doing M-q on my own example in a fresh Org mode
> buffer, I now get what you get:
>
>   #+BEGIN_SRC f90 :results verbatim :exports both
>   program main ! This is a very very very very very very very very very
>   very very very very very very very long comment line. print *, "Hello,
>   World!" end program main
>   #+END_SRC
>
> I don't have any special settings for the f90 mode in my .emacs. In
> fact, I have observed the same behavior for other languages too when
> edited in an Org source block. The "destructive" nature of this
> behavior is very special (C-/ (undo) doesn't work). Many a times, I
> have tried refilling a line of comments and ended up with a huge mix
> of code and comment that was impossible to undo (at least to the best
> of my knowledge) and I had to either revert the change if I had the
> file under version control, go to an auto backup file, or painfully and
> manually separate the code and comments to get back the original code
> block.
>
> Omid
>
>
> On 02/05/2014 04:34 PM, Nick Dokos wrote:
>> Omid <omidlink@gmail.com> writes:
>>
>>> Hello,
>>>
>>> I am using Org-mode version 8.2.5g (8.2.5g-elpa) in GNU Emacs 24.3.1.
>>> I have two questions about the behavior of the fantastic Org
>>> mode+Babel with respect to code and comments:
>>>
>>> Here is a minimal example:
>>>
>>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>> program main
>>> ! This is a very very very very very very very very very very very very
>>> very very very very long comment line.
>>> print *, "Hello, World!"
>>> end program main
>>>   #+END_SRC
>>>
>>> Below is the code snippet after M-q (fill-paragraph) on the comment
>>> line. The comment line has been refilled (intended behavior) but all
>>> lines have been indented. This may also be an intended behavior; but
>>>
>>> FIRST QUESTION: Is there a way to disable this indentation upon M-Q in
>>> Org Babel code snippets?
>>>
>>
>> Try setting org-edit-src-content-indentation to 0. I'm not sure that
>> it is going to work, but (based on rather flimsy numerological evidence,
>> namely that its default value is 2 as is the indent below) it might.
>>
>>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>>     program main
>>>     ! This is a very very very very very very very very very very very
>>> very
>>>     ! very very very very long comment line.
>>>     print *, "Hello, World!"
>>>     end program main
>>>   #+END_SRC
>>>
>>
>> BTW, I don't get this behavior but I don't use f90 mode, so I'm not sure
>> whether there is some setup I'm missing. I get (with
>> org-edit-src-content-indentation set to 0):
>>
>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>   program main ! This is a very very very very very very very very very
>>   very very very very very very very long comment line.  print *,
>>   "Hello, World!"  end program main
>>   #+END_SRC
>>
>> Nick
>>
>>
>>
>
>


-- 
The Kafka Pandemic: http://thekafkapandemic.blogspot.com

The disease DOES progress.  MANY people have died from it.  ANYBODY can get it.

Denmark: free Karina Hansen NOW.

Re: Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Samuel Wales]

"[bug] [babel] babel corrupts undo history"

On 2/13/14, Samuel Wales <samologist@gmail.com> wrote:
> Undo not working is a bug.  I think we identified the bug on the
> mailing list, but the fix was never implemented, although I think an
> attempted fix was.
>
> I think it has to do with editing a source code block trying to do
> things with undo or undo-boundary.  The fix is probably to rip out all
> undo-related code there.
>
>
> On 2/5/14, Omid <omidlink@gmail.com> wrote:
>> Hello Nick,
>>
>> Thanks for your quick reply. The variable
>> org-edit-src-content-indentation does indeed have to do with the
>> indentation I was referring to. But it is relative. According to the
>> documentation, it sets the "Indentation for the content of a source
>> code block. This should be the number of spaces added to the
>> indentation of the #+begin line in order to compute the indentation of
>> the block content after editing it with M-x org-edit-src-code. Has no
>> effect if `org-src-preserve-indentation' is non-nil." I wonder if one
>> can set it to be zero as an absolute value, meaning that if the a line
>> of code in the source block starts at column n of the buffer, it is
>> left at column n after M-q.
>>
>> Regarding the behavior you observe, as I said I get that too from time
>> to time. In fact, doing M-q on my own example in a fresh Org mode
>> buffer, I now get what you get:
>>
>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>   program main ! This is a very very very very very very very very very
>>   very very very very very very very long comment line. print *, "Hello,
>>   World!" end program main
>>   #+END_SRC
>>
>> I don't have any special settings for the f90 mode in my .emacs. In
>> fact, I have observed the same behavior for other languages too when
>> edited in an Org source block. The "destructive" nature of this
>> behavior is very special (C-/ (undo) doesn't work). Many a times, I
>> have tried refilling a line of comments and ended up with a huge mix
>> of code and comment that was impossible to undo (at least to the best
>> of my knowledge) and I had to either revert the change if I had the
>> file under version control, go to an auto backup file, or painfully and
>> manually separate the code and comments to get back the original code
>> block.
>>
>> Omid
>>
>>
>> On 02/05/2014 04:34 PM, Nick Dokos wrote:
>>> Omid <omidlink@gmail.com> writes:
>>>
>>>> Hello,
>>>>
>>>> I am using Org-mode version 8.2.5g (8.2.5g-elpa) in GNU Emacs 24.3.1.
>>>> I have two questions about the behavior of the fantastic Org
>>>> mode+Babel with respect to code and comments:
>>>>
>>>> Here is a minimal example:
>>>>
>>>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>>> program main
>>>> ! This is a very very very very very very very very very very very very
>>>> very very very very long comment line.
>>>> print *, "Hello, World!"
>>>> end program main
>>>>   #+END_SRC
>>>>
>>>> Below is the code snippet after M-q (fill-paragraph) on the comment
>>>> line. The comment line has been refilled (intended behavior) but all
>>>> lines have been indented. This may also be an intended behavior; but
>>>>
>>>> FIRST QUESTION: Is there a way to disable this indentation upon M-Q in
>>>> Org Babel code snippets?
>>>>
>>>
>>> Try setting org-edit-src-content-indentation to 0. I'm not sure that
>>> it is going to work, but (based on rather flimsy numerological evidence,
>>> namely that its default value is 2 as is the indent below) it might.
>>>
>>>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>>>     program main
>>>>     ! This is a very very very very very very very very very very very
>>>> very
>>>>     ! very very very very long comment line.
>>>>     print *, "Hello, World!"
>>>>     end program main
>>>>   #+END_SRC
>>>>
>>>
>>> BTW, I don't get this behavior but I don't use f90 mode, so I'm not sure
>>> whether there is some setup I'm missing. I get (with
>>> org-edit-src-content-indentation set to 0):
>>>
>>>   #+BEGIN_SRC f90 :results verbatim :exports both
>>>   program main ! This is a very very very very very very very very very
>>>   very very very very very very very long comment line.  print *,
>>>   "Hello, World!"  end program main
>>>   #+END_SRC
>>>
>>> Nick
>>>
>>>
>>>
>>
>>
>
>
> --
> The Kafka Pandemic: http://thekafkapandemic.blogspot.com
>
> The disease DOES progress.  MANY people have died from it.  ANYBODY can get
> it.
>
> Denmark: free Karina Hansen NOW.
>


-- 
The Kafka Pandemic: http://thekafkapandemic.blogspot.com

The disease DOES progress.  MANY people have died from it.  ANYBODY can get it.

Denmark: free Karina Hansen NOW.

Re: Behavior of Org mode Babel code snippets with respect to M-q (fill-paragraph) and C-/ (undo)

[Bastien]

Hi Samuel,

Samuel Wales <samologist@gmail.com> writes:

> "[bug] [babel] babel corrupts undo history"

I've reread this thread but I currently nothing wrong with
undoing wrt source code editing.

Let me know if I missed something, thanks,

-- 
 Bastien

Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Omid]

Hello,

- I am trying to build a statically-generated website purely with Org
  mode. I would like to use Bootstrap CSS to make the site mobile
  friendly. Bootstrap uses <div class="container"> for the "main" part
  of the <body> of HTML; Org mode export produces <div id="content">
  for that part.

  Is there a variable to set/way to make Org wrap this part in the
  "container" class?

  Looking at several "apropos" results (with all the search keywords I
  could think of), I only see a setting for pre/postamble-class:

  org-html--pre/postamble-class
    Variable: CSS class used for pre/postamble

- Assuming there is no Org-internal way to do this, is there a
  function that is run after a project is published to that I can,
  e.g., run a sed script on the HTML files generated? Again, the only
  remotely relevant result I see in M-x apropos RET org post is

  org-babel-post-tangle-hook
  Variable: Hook run in code files tangled by `org-babel-tangle'.
  Properties: standard-value custom-version custom-type
              custom-requests variable-documentation

- As a "meta question", how would one find the answer to these and
  similar questions without leaving Emacs, since we know Emacs is
  self-documenting and Org mode has a very comprehensive set of
  functions and documentation. Looking in the usual places (C-u M-x
  apropos, C-h v) did not lead me to an answer.

Thanks,

Omid

Sent from my Emacs

Re: Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Eric Abrahamsen]

Omid <omidlink@gmail.com> writes:

> Hello,
>
> - I am trying to build a statically-generated website purely with Org
>   mode. I would like to use Bootstrap CSS to make the site mobile
>   friendly. Bootstrap uses <div class="container"> for the "main" part
>   of the <body> of HTML; Org mode export produces <div id="content">
>   for that part.
>
>   Is there a variable to set/way to make Org wrap this part in the
>   "container" class?
>
>   Looking at several "apropos" results (with all the search keywords I
>   could think of), I only see a setting for pre/postamble-class:
>
>   org-html--pre/postamble-class
>     Variable: CSS class used for pre/postamble

It would help to know your Org mode version, but I'm pretty sure
org-html-divs has been around for a bit. Check out its docstring, and
modify its second element, which is '(content "div" "content") by
default.

> - Assuming there is no Org-internal way to do this, is there a
>   function that is run after a project is published to that I can,
>   e.g., run a sed script on the HTML files generated? Again, the only
>   remotely relevant result I see in M-x apropos RET org post is
>
>   org-babel-post-tangle-hook
>   Variable: Hook run in code files tangled by `org-babel-tangle'.
>   Properties: standard-value custom-version custom-type
>               custom-requests variable-documentation
>
> - As a "meta question", how would one find the answer to these and
>   similar questions without leaving Emacs, since we know Emacs is
>   self-documenting and Org mode has a very comprehensive set of
>   functions and documentation. Looking in the usual places (C-u M-x
>   apropos, C-h v) did not lead me to an answer.

One good thing to know is that Org export variables related to a certain
backend all start with org-BACKEND-*. So you could do C-h v org-html-TAB
and see... a lot of variables. Still, that would (eventually) get you
there. I did apropos-variable for "org html container" and that got me a
similar list as C-h v, albeit slimmed down somewhat, and with
docstrings.

Hope that helps,
Eric

Re: Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Omid]

Thank you, Eric for your quick response.
- I am using Org mode 8.2.6 (from ELPA) and Emacs 24.3.1.
- org-html-divs does almost what I have been looking for, except for
  one, unfortunately crucial, shortcoming: I don't see any way of
  changing the "id" tag to "class". Bootstrap's container is a "class"
  and will not work as an "id". As far as see, I can only do

  (setq org-html-divs (quote ((preamble "div" "preamble")
  (content "div" "container")
  (postamble "div" "postamble"))))
			
  Looking at ox-html.el I see that the "id" tag is hard coded:

  (list :tag "Content"
  (const :format "" content)
  (string :tag "element") (string :tag "     id"))

  Am I missing something here?

- I still could not find a "post publish" hook/variable. Neither
  apropos-variable with org publish, apropos with org publish gives me
  anything useful. C-h v reveals org-export-before-parsing-hook and
  org-export-before-processing-hook but I don't see anything with
  "after" or "post". The only way I see around my problem is to define
  an external publishing-function in org-publish-project-alist (maybe
  a makefile) and do the necessary post processing there. Any better
  suggestion is greatly appreciated.

Thanks,
		
Omid

Sent from my Emacs

On 05/25/2014 10:10 PM, Eric Abrahamsen wrote:
> Omid <omidlink@gmail.com> writes:
> 
>> Hello,
>>
>> - I am trying to build a statically-generated website purely with Org
>>   mode. I would like to use Bootstrap CSS to make the site mobile
>>   friendly. Bootstrap uses <div class="container"> for the "main" part
>>   of the <body> of HTML; Org mode export produces <div id="content">
>>   for that part.
>>
>>   Is there a variable to set/way to make Org wrap this part in the
>>   "container" class?
>>
>>   Looking at several "apropos" results (with all the search keywords I
>>   could think of), I only see a setting for pre/postamble-class:
>>
>>   org-html--pre/postamble-class
>>     Variable: CSS class used for pre/postamble
> 
> It would help to know your Org mode version, but I'm pretty sure
> org-html-divs has been around for a bit. Check out its docstring, and
> modify its second element, which is '(content "div" "content") by
> default.
> 
>> - Assuming there is no Org-internal way to do this, is there a
>>   function that is run after a project is published to that I can,
>>   e.g., run a sed script on the HTML files generated? Again, the only
>>   remotely relevant result I see in M-x apropos RET org post is
>>
>>   org-babel-post-tangle-hook
>>   Variable: Hook run in code files tangled by `org-babel-tangle'.
>>   Properties: standard-value custom-version custom-type
>>               custom-requests variable-documentation
>>
>> - As a "meta question", how would one find the answer to these and
>>   similar questions without leaving Emacs, since we know Emacs is
>>   self-documenting and Org mode has a very comprehensive set of
>>   functions and documentation. Looking in the usual places (C-u M-x
>>   apropos, C-h v) did not lead me to an answer.
> 
> One good thing to know is that Org export variables related to a certain
> backend all start with org-BACKEND-*. So you could do C-h v org-html-TAB
> and see... a lot of variables. Still, that would (eventually) get you
> there. I did apropos-variable for "org html container" and that got me a
> similar list as C-h v, albeit slimmed down somewhat, and with
> docstrings.
> 
> Hope that helps,
> Eric
> 
> 

Re: Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Omid]

I think I found the answer to my second question: in
org-publish-project-alist, there is the property completion-function
that can be set to a (list of) functions to do some post processing.
My first question still stands, though.

Omid

Sent from my Emacs

On 05/26/2014 12:08 AM, Omid wrote:
> Thank you, Eric for your quick response.
> - I am using Org mode 8.2.6 (from ELPA) and Emacs 24.3.1.
> - org-html-divs does almost what I have been looking for, except for
>   one, unfortunately crucial, shortcoming: I don't see any way of
>   changing the "id" tag to "class". Bootstrap's container is a "class"
>   and will not work as an "id". As far as see, I can only do
> 
>   (setq org-html-divs (quote ((preamble "div" "preamble")
>   (content "div" "container")
>   (postamble "div" "postamble"))))
> 			
>   Looking at ox-html.el I see that the "id" tag is hard coded:
> 
>   (list :tag "Content"
>   (const :format "" content)
>   (string :tag "element") (string :tag "     id"))
> 
>   Am I missing something here?
> 
> - I still could not find a "post publish" hook/variable. Neither
>   apropos-variable with org publish, apropos with org publish gives me
>   anything useful. C-h v reveals org-export-before-parsing-hook and
>   org-export-before-processing-hook but I don't see anything with
>   "after" or "post". The only way I see around my problem is to define
>   an external publishing-function in org-publish-project-alist (maybe
>   a makefile) and do the necessary post processing there. Any better
>   suggestion is greatly appreciated.
> 
> Thanks,
> 		
> Omid
> 
> Sent from my Emacs
> 
> On 05/25/2014 10:10 PM, Eric Abrahamsen wrote:
>> Omid <omidlink@gmail.com> writes:
>>
>>> Hello,
>>>
>>> - I am trying to build a statically-generated website purely with Org
>>>   mode. I would like to use Bootstrap CSS to make the site mobile
>>>   friendly. Bootstrap uses <div class="container"> for the "main" part
>>>   of the <body> of HTML; Org mode export produces <div id="content">
>>>   for that part.
>>>
>>>   Is there a variable to set/way to make Org wrap this part in the
>>>   "container" class?
>>>
>>>   Looking at several "apropos" results (with all the search keywords I
>>>   could think of), I only see a setting for pre/postamble-class:
>>>
>>>   org-html--pre/postamble-class
>>>     Variable: CSS class used for pre/postamble
>>
>> It would help to know your Org mode version, but I'm pretty sure
>> org-html-divs has been around for a bit. Check out its docstring, and
>> modify its second element, which is '(content "div" "content") by
>> default.
>>
>>> - Assuming there is no Org-internal way to do this, is there a
>>>   function that is run after a project is published to that I can,
>>>   e.g., run a sed script on the HTML files generated? Again, the only
>>>   remotely relevant result I see in M-x apropos RET org post is
>>>
>>>   org-babel-post-tangle-hook
>>>   Variable: Hook run in code files tangled by `org-babel-tangle'.
>>>   Properties: standard-value custom-version custom-type
>>>               custom-requests variable-documentation
>>>
>>> - As a "meta question", how would one find the answer to these and
>>>   similar questions without leaving Emacs, since we know Emacs is
>>>   self-documenting and Org mode has a very comprehensive set of
>>>   functions and documentation. Looking in the usual places (C-u M-x
>>>   apropos, C-h v) did not lead me to an answer.
>>
>> One good thing to know is that Org export variables related to a certain
>> backend all start with org-BACKEND-*. So you could do C-h v org-html-TAB
>> and see... a lot of variables. Still, that would (eventually) get you
>> there. I did apropos-variable for "org html container" and that got me a
>> similar list as C-h v, albeit slimmed down somewhat, and with
>> docstrings.
>>
>> Hope that helps,
>> Eric
>>
>>

Re: Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Eric Abrahamsen]

Omid <omidlink@gmail.com> writes:

> I think I found the answer to my second question: in
> org-publish-project-alist, there is the property completion-function
> that can be set to a (list of) functions to do some post processing.
> My first question still stands, though.

Yup, that's a bit of a pain.

When you need to really mess with the results of an Org export, the way
to do it is through the "filter" system, not regular hooks. See the
comments under "The Filter System" in lisp/ox.el for details on that.

Filters give you access to the exported strings produced from various
elements in the org buffer. Unfortunately there isn't a filter that
reaches the contents template, which is where ox-html is setting the div
id, so you'll probably need to use the nuclear option:
org-export-filter-final-output-functions. Functions placed in that
variable get access to the whole export string, and you can just do a
search-and-replace for your id/class string there.

Yours,
Eric

> On 05/26/2014 12:08 AM, Omid wrote:
>> Thank you, Eric for your quick response.
>> - I am using Org mode 8.2.6 (from ELPA) and Emacs 24.3.1.
>> - org-html-divs does almost what I have been looking for, except for
>>   one, unfortunately crucial, shortcoming: I don't see any way of
>>   changing the "id" tag to "class". Bootstrap's container is a "class"
>>   and will not work as an "id". As far as see, I can only do
>> 
>>   (setq org-html-divs (quote ((preamble "div" "preamble")
>>   (content "div" "container")
>>   (postamble "div" "postamble"))))
>> 			
>>   Looking at ox-html.el I see that the "id" tag is hard coded:
>> 
>>   (list :tag "Content"
>>   (const :format "" content)
>>   (string :tag "element") (string :tag "     id"))
>> 
>>   Am I missing something here?
>> 
>> - I still could not find a "post publish" hook/variable. Neither
>>   apropos-variable with org publish, apropos with org publish gives me
>>   anything useful. C-h v reveals org-export-before-parsing-hook and
>>   org-export-before-processing-hook but I don't see anything with
>>   "after" or "post". The only way I see around my problem is to define
>>   an external publishing-function in org-publish-project-alist (maybe
>>   a makefile) and do the necessary post processing there. Any better
>>   suggestion is greatly appreciated.
>> 
>> Thanks,
>> 		
>> Omid
>> 
>> Sent from my Emacs
>> 
>> On 05/25/2014 10:10 PM, Eric Abrahamsen wrote:
>>> Omid <omidlink@gmail.com> writes:
>>>
>>>> Hello,
>>>>
>>>> - I am trying to build a statically-generated website purely with Org
>>>>   mode. I would like to use Bootstrap CSS to make the site mobile
>>>>   friendly. Bootstrap uses <div class="container"> for the "main" part
>>>>   of the <body> of HTML; Org mode export produces <div id="content">
>>>>   for that part.
>>>>
>>>>   Is there a variable to set/way to make Org wrap this part in the
>>>>   "container" class?
>>>>
>>>>   Looking at several "apropos" results (with all the search keywords I
>>>>   could think of), I only see a setting for pre/postamble-class:
>>>>
>>>>   org-html--pre/postamble-class
>>>>     Variable: CSS class used for pre/postamble
>>>
>>> It would help to know your Org mode version, but I'm pretty sure
>>> org-html-divs has been around for a bit. Check out its docstring, and
>>> modify its second element, which is '(content "div" "content") by
>>> default.
>>>
>>>> - Assuming there is no Org-internal way to do this, is there a
>>>>   function that is run after a project is published to that I can,
>>>>   e.g., run a sed script on the HTML files generated? Again, the only
>>>>   remotely relevant result I see in M-x apropos RET org post is
>>>>
>>>>   org-babel-post-tangle-hook
>>>>   Variable: Hook run in code files tangled by `org-babel-tangle'.
>>>>   Properties: standard-value custom-version custom-type
>>>>               custom-requests variable-documentation
>>>>
>>>> - As a "meta question", how would one find the answer to these and
>>>>   similar questions without leaving Emacs, since we know Emacs is
>>>>   self-documenting and Org mode has a very comprehensive set of
>>>>   functions and documentation. Looking in the usual places (C-u M-x
>>>>   apropos, C-h v) did not lead me to an answer.
>>>
>>> One good thing to know is that Org export variables related to a certain
>>> backend all start with org-BACKEND-*. So you could do C-h v org-html-TAB
>>> and see... a lot of variables. Still, that would (eventually) get you
>>> there. I did apropos-variable for "org html container" and that got me a
>>> similar list as C-h v, albeit slimmed down somewhat, and with
>>> docstrings.
>>>
>>> Hope that helps,
>>> Eric
>>>
>>>

Re: Is there a variable/way to change <div id="content"> to something else, e.g. <div class="container"> for use with Bootstrap CSS?

[Omid]

Eric,

Thanks a lot for the pointer to the "Filter System." I was not aware
of it and, as you pointed out, the
org-export-filter-final-output-functions variable will help me achieve
what I was looking for.

Cheers,

Omid

Sent from my Emacs

On 05/26/2014 01:11 AM, Eric Abrahamsen wrote:
> Omid <omidlink@gmail.com> writes:
> 
>> I think I found the answer to my second question: in
>> org-publish-project-alist, there is the property completion-function
>> that can be set to a (list of) functions to do some post processing.
>> My first question still stands, though.
> 
> Yup, that's a bit of a pain.
> 
> When you need to really mess with the results of an Org export, the way
> to do it is through the "filter" system, not regular hooks. See the
> comments under "The Filter System" in lisp/ox.el for details on that.
> 
> Filters give you access to the exported strings produced from various
> elements in the org buffer. Unfortunately there isn't a filter that
> reaches the contents template, which is where ox-html is setting the div
> id, so you'll probably need to use the nuclear option:
> org-export-filter-final-output-functions. Functions placed in that
> variable get access to the whole export string, and you can just do a
> search-and-replace for your id/class string there.
> 
> Yours,
> Eric
> 
>> On 05/26/2014 12:08 AM, Omid wrote:
>>> Thank you, Eric for your quick response.
>>> - I am using Org mode 8.2.6 (from ELPA) and Emacs 24.3.1.
>>> - org-html-divs does almost what I have been looking for, except for
>>>   one, unfortunately crucial, shortcoming: I don't see any way of
>>>   changing the "id" tag to "class". Bootstrap's container is a "class"
>>>   and will not work as an "id". As far as see, I can only do
>>>
>>>   (setq org-html-divs (quote ((preamble "div" "preamble")
>>>   (content "div" "container")
>>>   (postamble "div" "postamble"))))
>>> 			
>>>   Looking at ox-html.el I see that the "id" tag is hard coded:
>>>
>>>   (list :tag "Content"
>>>   (const :format "" content)
>>>   (string :tag "element") (string :tag "     id"))
>>>
>>>   Am I missing something here?
>>>
>>> - I still could not find a "post publish" hook/variable. Neither
>>>   apropos-variable with org publish, apropos with org publish gives me
>>>   anything useful. C-h v reveals org-export-before-parsing-hook and
>>>   org-export-before-processing-hook but I don't see anything with
>>>   "after" or "post". The only way I see around my problem is to define
>>>   an external publishing-function in org-publish-project-alist (maybe
>>>   a makefile) and do the necessary post processing there. Any better
>>>   suggestion is greatly appreciated.
>>>
>>> Thanks,
>>> 		
>>> Omid
>>>
>>> Sent from my Emacs
>>>
>>> On 05/25/2014 10:10 PM, Eric Abrahamsen wrote:
>>>> Omid <omidlink@gmail.com> writes:
>>>>
>>>>> Hello,
>>>>>
>>>>> - I am trying to build a statically-generated website purely with Org
>>>>>   mode. I would like to use Bootstrap CSS to make the site mobile
>>>>>   friendly. Bootstrap uses <div class="container"> for the "main" part
>>>>>   of the <body> of HTML; Org mode export produces <div id="content">
>>>>>   for that part.
>>>>>
>>>>>   Is there a variable to set/way to make Org wrap this part in the
>>>>>   "container" class?
>>>>>
>>>>>   Looking at several "apropos" results (with all the search keywords I
>>>>>   could think of), I only see a setting for pre/postamble-class:
>>>>>
>>>>>   org-html--pre/postamble-class
>>>>>     Variable: CSS class used for pre/postamble
>>>>
>>>> It would help to know your Org mode version, but I'm pretty sure
>>>> org-html-divs has been around for a bit. Check out its docstring, and
>>>> modify its second element, which is '(content "div" "content") by
>>>> default.
>>>>
>>>>> - Assuming there is no Org-internal way to do this, is there a
>>>>>   function that is run after a project is published to that I can,
>>>>>   e.g., run a sed script on the HTML files generated? Again, the only
>>>>>   remotely relevant result I see in M-x apropos RET org post is
>>>>>
>>>>>   org-babel-post-tangle-hook
>>>>>   Variable: Hook run in code files tangled by `org-babel-tangle'.
>>>>>   Properties: standard-value custom-version custom-type
>>>>>               custom-requests variable-documentation
>>>>>
>>>>> - As a "meta question", how would one find the answer to these and
>>>>>   similar questions without leaving Emacs, since we know Emacs is
>>>>>   self-documenting and Org mode has a very comprehensive set of
>>>>>   functions and documentation. Looking in the usual places (C-u M-x
>>>>>   apropos, C-h v) did not lead me to an answer.
>>>>
>>>> One good thing to know is that Org export variables related to a certain
>>>> backend all start with org-BACKEND-*. So you could do C-h v org-html-TAB
>>>> and see... a lot of variables. Still, that would (eventually) get you
>>>> there. I did apropos-variable for "org html container" and that got me a
>>>> similar list as C-h v, albeit slimmed down somewhat, and with
>>>> docstrings.
>>>>
>>>> Hope that helps,
>>>> Eric
>>>>
>>>>
> 
>