BIND org-html-style-include-*

BIND org-html-style-include-*

[T.F. Torrey]

Hello,

In my files, this used to work to suppress the default styles and
javascript:

#+BIND: org-export-html-style-include-default nil
#+BIND: org-export-html-style-include-scripts nil

Now, I have these:

#+BIND: org-html-style-include-default nil
#+BIND: org-html-style-include-scripts nil

But they seem to be silently ignored, though if I setq the values ahead
of time, the output is suppressed.

Can these still be set using BIND?  If so, what am I doing wrong?

Also, the value of org-html-mathjax-template seems to be output now by
default, and I don't remember seeing it before.  Is this an intended
change?  If so, how can it be suppressed?  (If not ... ???)

Emacs: GNU Emacs 24.3.50.1 (i686-pc-linux-gnu, GTK+ Version 3.6.0) of
 2012-12-24 on menkib, modified by Debian

Package: Org-mode version 7.9.3e (7.9.3e-970-g728c0e @
/home/tftorrey/.emacs.d/elisp/org/lisp/)

Best regards,
Terry
-- 
T.F. Torrey

Re: BIND org-html-style-include-*

[Nicolas Goaziou]

Hello,

tftorrey@tftorrey.com (T.F. Torrey) writes:

> In my files, this used to work to suppress the default styles and
> javascript:
>
> #+BIND: org-export-html-style-include-default nil
> #+BIND: org-export-html-style-include-scripts nil
>
> Now, I have these:
>
> #+BIND: org-html-style-include-default nil
> #+BIND: org-html-style-include-scripts nil
>
> But they seem to be silently ignored, though if I setq the values ahead
> of time, the output is suppressed.
>
> Can these still be set using BIND?  If so, what am I doing wrong?

What is the value of `org-export-allow-bind-keywords'?

> Also, the value of org-html-mathjax-template seems to be output now by
> default, and I don't remember seeing it before.  Is this an intended
> change?  If so, how can it be suppressed?  (If not ... ???)

I cannot answer you for now, because I don't know enough of the HTML
back-end.


Regards,

-- 
Nicolas Goaziou

Re: BIND org-html-style-include-*

[T.F. Torrey]

Hello,

Nicolas Goaziou <n.goaziou@gmail.com> writes:

> tftorrey@tftorrey.com (T.F. Torrey) writes:
>
>> Now, I have these:
>>
>> #+BIND: org-html-style-include-default nil
>> #+BIND: org-html-style-include-scripts nil
>>
>> But they seem to be silently ignored, though if I setq the values ahead
>> of time, the output is suppressed.
>>
>> Can these still be set using BIND?  If so, what am I doing wrong?
>
> What is the value of `org-export-allow-bind-keywords'?

Aha! That's the one I was looking for.  Thanks.

>> Also, the value of org-html-mathjax-template seems to be output now by
>> default, and I don't remember seeing it before.  Is this an intended
>> change?  If so, how can it be suppressed?  (If not ... ???)
>
> I cannot answer you for now, because I don't know enough of the HTML
> back-end.

I still don't see any similar options regarding MathJax.  Perhaps
everyone else uses MathJax, or the functionality simply hasn't been
implemented yet.  If only the HTML export backend maintainer were still
on the list...

Best regards,
Terry
-- 
T.F. Torrey

Re: BIND org-html-style-include-*

[Nicolas Goaziou]

tftorrey@tftorrey.com (T.F. Torrey) writes:

> I still don't see any similar options regarding MathJax.  Perhaps
> everyone else uses MathJax, or the functionality simply hasn't been
> implemented yet.  If only the HTML export backend maintainer were still
> on the list...

Would you mind describing what is missing? It shouldn't be hard to
implement it.


Regards,

-- 
Nicolas Goaziou

Re: BIND org-html-style-include-*

[Bastien]

Hi,

Nicolas Goaziou <n.goaziou@gmail.com> writes:

> tftorrey@tftorrey.com (T.F. Torrey) writes:
>
>> I still don't see any similar options regarding MathJax.  Perhaps
>> everyone else uses MathJax, or the functionality simply hasn't been
>> implemented yet.  If only the HTML export backend maintainer were still
>> on the list...
>
> Would you mind describing what is missing? It shouldn't be hard to
> implement it.

We need to make `org-html--build-mathjax-config' depends on whether
there is an latex-fragment element in the parse tree.  Otherwise,
including the MathJAX config is useless.

I've been looking at this this morning.

Is there an org-element function to test whether an element/object 
of a certain type is present in the parse tree?

Thanks,

-- 
 Bastien

Re: BIND org-html-style-include-*

[Nicolas Goaziou]

Bastien <bzg@altern.org> writes:

> We need to make `org-html--build-mathjax-config' depends on whether
> there is an latex-fragment element in the parse tree.  Otherwise,
> including the MathJAX config is useless.
>
> I've been looking at this this morning.
>
> Is there an org-element function to test whether an element/object 
> of a certain type is present in the parse tree?

On course, it is `org-element-map' (you can have a look at its
docstring, there is one such example). So, assuming you have access to
communication channel (a.k.a. info argument):

  (org-element-map (plist-get info :parse-tree)
    '(latex-fragment latex-environment) 'identity info t)

will be non-nil when either an exportable `latex-fragment' object or
a `latex-environment' element is found within the parse tree.


Regards,

-- 
Nicolas Goaziou

Re: BIND org-html-style-include-*

[Bastien]

Hi Nicolas,

Nicolas Goaziou <n.goaziou@gmail.com> writes:

>   (org-element-map (plist-get info :parse-tree)
>     '(latex-fragment latex-environment) 'identity info t)

I used this, but only checking against LaTeX fragments.
HTML pages without LaTeX fragments will not include the
MathJax configuration.

Thanks!

-- 
 Bastien

Re: BIND org-html-style-include-*

[Nicolas Goaziou]

Bastien <bzg@altern.org> writes:

> Nicolas Goaziou <n.goaziou@gmail.com> writes:
>
>>   (org-element-map (plist-get info :parse-tree)
>>     '(latex-fragment latex-environment) 'identity info t)
>
> I used this, but only checking against LaTeX fragments.
> HTML pages without LaTeX fragments will not include the
> MathJax configuration.

I don't understand. Latex environments also require Mathjax: see
`org-html-latex-environment'.

Just to be sure we're not mis-communicating, a latex environment is the
following:

  \begin{equation*}
  2 + 2 = 4
  \end{equation*}

This is not to be confused with a latex block, such as:

  #+begin_latex
  \[2 + 2 = 4\]
  #+end_latex


Regards,

-- 
Nicolas Goaziou

Re: BIND org-html-style-include-*

[Bastien]

Nicolas Goaziou <n.goaziou@gmail.com> writes:

> Just to be sure we're not mis-communicating, a latex environment is the
> following:
>
>   \begin{equation*}
>   2 + 2 = 4
>   \end{equation*}
>
> This is not to be confused with a latex block, such as:
>
>   #+begin_latex
>   \[2 + 2 = 4\]
>   #+end_latex

I made the confusion -- I fixed last commit.  Thanks!

-- 
 Bastien

Re: BIND org-html-style-include-*

[T.F. Torrey]

Hello,

Nicolas Goaziou <n.goaziou@gmail.com> writes:

> tftorrey@tftorrey.com (T.F. Torrey) writes:
>
>> I still don't see any similar options regarding MathJax.  Perhaps
>> everyone else uses MathJax, or the functionality simply hasn't been
>> implemented yet.  If only the HTML export backend maintainer were still
>> on the list...
>
> Would you mind describing what is missing? It shouldn't be hard to
> implement it.

By default, the HTML export engine includes some CSS for basic
presentation, some JavaScript for interacting with code in the HTML, and
some JavaScript for the MathJax functionality (among many other things,
of course).

The CSS can be not included by setting org-html-style-include-default to
nil.

The non-MathJax JavaScript can be not included by setting
org-html-style-include-scripts to nil.

However, the new HTML exporter includes the MathJax JavaScript every
time, and I don't see any variable to set to suppress it.

The old exporter did not export it.  IIRC, it was off by default, and
there was a setting to turn it on (don't quote me, I never used it).

Recent posts by Bastien suggest that maybe it is supposed to be included
only if LaTeX is used in the buffer, but I'm not sure he was talking
about this issue.

With the new exporter on my Org, even the most minimal Org file exported
to HTML includes MathJax.  My Org version is 7.9.3e-1032-g791a8d.  The
most recent pulls fail testing (which people must already know).

(By the way, though it may be too late to change them now, the variables
would be better named "org-html-include-style-default" and
"org-html-include-scripts-default" and possibly
"org-html-include-scripts-mathjax".  The HTML exporter has many things
that might be included in the output, and having the variables all
starting with "org-html-include-" would make them easier for everyone to
find, understand, and modify.)

(Similarly, as someone else wrote, #+HTML_STYLE would be much better
named #+HTML_HEAD, given that style is just one of the many things this
directive might put into the head element of the html.)

All the best,
Terry
-- 
T.F. Torrey

Re: BIND org-html-style-include-*

[Sebastien Vauban]

Hi T.F,

T.F. Torrey wrote:
> (Similarly, as someone else wrote, #+HTML_STYLE would be much better
> named #+HTML_HEAD, given that style is just one of the many things this
> directive might put into the head element of the html.)

I vote +1, at least if we can put anything we would put in the HTML HEAD
element -- I mean, if there is no special restriction on the contents we can
stuff there.

Best regards,
  Seb

-- 
Sebastien Vauban

Re: BIND org-html-style-include-*

[Bastien]

Hi Terry,

tftorrey@tftorrey.com (T.F. Torrey) writes:

> However, the new HTML exporter includes the MathJax JavaScript every
> time, and I don't see any variable to set to suppress it.

This is not the case anymore since commit 4d7f4d87.

> Recent posts by Bastien suggest that maybe it is supposed to be included
> only if LaTeX is used in the buffer, but I'm not sure he was talking
> about this issue.

I hope I was -- the MathJax config will now be included if your HTML
page contains LaTeX snippets to display.  It will not be included
otherwise.

> With the new exporter on my Org, even the most minimal Org file exported
> to HTML includes MathJax.  My Org version is 7.9.3e-1032-g791a8d.  The
> most recent pulls fail testing (which people must already know).

You are behind a few commits, please update.

> (By the way, though it may be too late to change them now, the variables
> would be better named "org-html-include-style-default" and
> "org-html-include-scripts-default" and possibly
> "org-html-include-scripts-mathjax".  The HTML exporter has many things
> that might be included in the output, and having the variables all
> starting with "org-html-include-" would make them easier for everyone to
> find, understand, and modify.)

I somewhat agree.  But C-h v org-html-include TAB expands and display 
all variables here, so maybe not such an issue.  It is not too late to
change the names of the variables, but I'd rather stick to these names
because it lowers the effort when switching from 7.9 to 8.0.

> (Similarly, as someone else wrote, #+HTML_STYLE would be much better
> named #+HTML_HEAD, given that style is just one of the many things this
> directive might put into the head element of the html.)

For this one I agree completely.  I will make this change.

-- 
 Bastien

Re: BIND org-html-style-include-*

[Bastien]

[-- Attachment #1: Type: text/plain, Size: 318 bytes --]

Bastien <bzg@altern.org> writes:

> For this one I agree completely.  I will make this change.

I share the patch here to let people discuss this change a bit.

It will break again the configuration of those who take the time
to test the current HEAD (sic), so I'd rather make sure everyone
is fine with it.

Thanks,


[-- Warning: decoded text below may be mangled, UTF-8 assumed --]
[-- Attachment #2: org-html-head.patch --]
[-- Type: text/x-patch, Size: 2787 bytes --]

diff --git a/lisp/ox-html.el b/lisp/ox-html.el
index e1b50f4..963a376 100644
--- a/lisp/ox-html.el
+++ b/lisp/ox-html.el
@@ -118,7 +118,7 @@
    (:html-postamble nil "html-postamble" org-html-postamble)
    (:html-preamble nil "html-preamble" org-html-preamble)
    (:html-style nil nil org-html-style)
-   (:html-style-extra "HTML_STYLE" nil org-html-style-extra newline)
+   (:html-head "HTML_HEAD" nil org-html-head newline)
    (:html-style-include-default nil nil org-html-style-include-default)
    (:html-style-include-scripts nil nil org-html-style-include-scripts)
    (:html-table-tag nil nil org-html-table-tag)
@@ -242,9 +242,9 @@ for the JavaScript code in this tag.
   /*]]>*/-->
 </style>"
   "The default style specification for exported HTML files.
-Please use the variables `org-html-style' and
-`org-html-style-extra' to add to this style.  If you wish to not
-have the default style included, customize the variable
+Please use the variables `org-html-style' and `org-html-head' to
+add to this style.  If you wish to not have the default style
+included, customize the variable
 `org-html-style-include-default'.")
 
 
@@ -947,22 +947,20 @@ If you'd like to refer to an external style file, use something like
 
 As the value of this option simply gets inserted into the HTML <head> header,
 you can \"misuse\" it to add arbitrary text to the header.
-See also the variable `org-html-style-extra'."
+See also the variable `org-html-head'."
   :group 'org-export-html
   :type 'string)
 ;;;###autoload
 (put 'org-html-style 'safe-local-variable 'stringp)
 
-(defcustom org-html-style-extra ""
-  "Additional style information for HTML export.
-The value of this variable is inserted into the HTML buffer right after
-the value of `org-html-style'.  Use this variable for per-file
-settings of style information, and do not forget to surround the style
-settings with <style>...</style> tags."
+(defcustom org-html-head ""
+  "Additional information for the header of the HTML export.
+The value of this variable is inserted into the HTML buffer right
+after the value of `org-html-style'."
   :group 'org-export-html
   :type 'string)
 ;;;###autoload
-(put 'org-html-style-extra 'safe-local-variable 'stringp)
+(put 'org-html-head 'safe-local-variable 'stringp)
 
 
 ;;;; Todos
@@ -1262,7 +1260,7 @@ INFO is a plist used as a communication channel."
    (concat
     (when (plist-get info :html-style-include-default) org-html-style-default)
     (org-element-normalize-string (plist-get info :html-style))
-    (org-element-normalize-string (plist-get info :html-style-extra))
+    (org-element-normalize-string (plist-get info :html-head))
     (when (plist-get info :html-style-include-scripts) org-html-scripts))))
 
 (defun org-html--build-mathjax-config (info)

[-- Attachment #3: Type: text/plain, Size: 14 bytes --]


-- 
 Bastien

Re: BIND org-html-style-include-*

[T.F. Torrey]

Hello,

Bastien <bzg@altern.org> writes:

> Hi Terry,
>
> tftorrey@tftorrey.com (T.F. Torrey) writes:
>
>> However, the new HTML exporter includes the MathJax JavaScript every
>> time, and I don't see any variable to set to suppress it.
>
> This is not the case anymore since commit 4d7f4d87.

Yes.  Thank you for the fix.

>> Recent posts by Bastien suggest that maybe it is supposed to be included
>> only if LaTeX is used in the buffer, but I'm not sure he was talking
>> about this issue.
>
> I hope I was -- the MathJax config will now be included if your HTML
> page contains LaTeX snippets to display.  It will not be included
> otherwise.
>
>> With the new exporter on my Org, even the most minimal Org file exported
>> to HTML includes MathJax.  My Org version is 7.9.3e-1032-g791a8d.  The
>> most recent pulls fail testing (which people must already know).
>
> You are behind a few commits, please update.

"make up2" reports these failures (and thus does not install):

6 unexpected results:
   FAILED  test-org-babel/inline-src_blk-default-results-replace-line-1
   FAILED  test-org-babel/inline-src_blk-results-file
   FAILED  test-org-babel/inline-src_blk-results-raw
   FAILED  test-org-babel/inline-src_blk-results-scalar
   FAILED  test-org-babel/inline-src_blk-results-silent
   FAILED  test-org-babel/inline-src_blk-results-verbatim

I was reluctant to run "make update" when the tests failed, but I did
anyway for the rest of this.

>> (By the way, though it may be too late to change them now, the variables
>> would be better named "org-html-include-style-default" and
>> "org-html-include-scripts-default" and possibly
>> "org-html-include-scripts-mathjax".  The HTML exporter has many things
>> that might be included in the output, and having the variables all
>> starting with "org-html-include-" would make them easier for everyone to
>> find, understand, and modify.)
>
> I somewhat agree.  But C-h v org-html-include TAB expands and display 
> all variables here, so maybe not such an issue.  It is not too late to
> change the names of the variables, but I'd rather stick to these names
> because it lowers the effort when switching from 7.9 to 8.0.

Yes, typing C-h v org-html-include finds these variables:

org-html-style-include-default
org-html-style-include-scripts

It bothers me that the second one is misnamed.  It has nothing to do
with style.  Even for just these two, starting with org-html-include
(that is, org-html-include-style and org-html-include-script --- they
should probably both be singular or plural) would be an improvement, and
who knows what future development brings?

These variables are already changing (from org-export-html-style-*) for
8.0.  Now is a good time to DTRT.

>> (Similarly, as someone else wrote, #+HTML_STYLE would be much better
>> named #+HTML_HEAD, given that style is just one of the many things this
>> directive might put into the head element of the html.)
>
> For this one I agree completely.  I will make this change.

All the people who just changed their files from #+STYLE to #+HTML_STYLE
will curse us, or you anyway, but it would make me happy (the change,
not them cursing you).

Thank you for your time and attention.

Best regards,
Terry
-- 
T.F. Torrey

Re: BIND org-html-style-include-*

[Bastien]

Hi Terry,

tftorrey@tftorrey.com (T.F. Torrey) writes:

> "make up2" reports these failures (and thus does not install):
>
> 6 unexpected results:
>    FAILED  test-org-babel/inline-src_blk-default-results-replace-line-1
>    FAILED  test-org-babel/inline-src_blk-results-file
>    FAILED  test-org-babel/inline-src_blk-results-raw
>    FAILED  test-org-babel/inline-src_blk-results-scalar
>    FAILED  test-org-babel/inline-src_blk-results-silent
>    FAILED  test-org-babel/inline-src_blk-results-verbatim

This is now fixed, thanks.

-- 
 Bastien