* UTILITIES/org-fixup.el (org-make-org-version): Provide feature
'org-version.
* lisp/org-compat.el (org-check-version): New macro. Check if
org-version.el exists and provide autoloads to that. Otherwise check
if org-fixup.el exists and use it to provide definitions. Finally
if nothing worked, complain about a botched installation and provide
fallback definitions.
* lisp/org.el: Use org-check-version.
This should finally provide the correct behaviour in all supported
use-cases as well as a few corner scenarios that were reported on the
mailing list and allow to always build on Git, ELPA and Emacs Bzr
correctly. The file org-version.el must be added to the Emacs sources
and should be re-generated each time orgmode Git is synced to Emacs
Bzr (it would be best to automate this).
Note that the internal functions (org-release) and (org-git-release)
are not autoloaded when org-version is not present and become only
available when org.el is loaded. This should pose no problem since
they are only used by (org-version), which is autoloaded from org.el.
* contrib/lisp/org-element.el (org-element-put-property,
org-element-set-contents): Handle special cases like empty or string
arguments.
(org-element-parse-secondary-string): Correctly set `:parent' property
in objects within the secondary string if element or object containing
it is provided as an optional argument.
(org-element-parse-elements, org-element-parse-objects): Rewrite
functions using setter functions.
* UTILITIES/org-fixup.el: Clean up and add doc strings.
* UTILITIES/org-fixup.el (org-make-autoloads): Add optional arguments
and implementation for comopiling and force compiling.
* UTILITIES/org-fixup.el (org-make-autoloads-compile,
org-make-autoloads-compile-force): Convenience definitions with
pre-selected arguments, avoids complicated quoted on invocation from
command line.
* UTILITIES/org-fixup.el (org-fixup): Add "-git" to git version when
run from a bare Git worktree. Give message instead of returning
string inside quoted progn for easier tracing.
* lisp/org.el: Fix a subtle error resulting in version functions
sometimes not being defined and byte-compiling failing. Always
compile in fallback definitions into org.elc -- org-fixup either
provides re-definitions at compile-time or checks org-version.el and
then the git work tree when run uncompiled. So the fallback
definitions will only come into effect when org-fixup is not
available.
- Clean up work directory and buffers on error.
- Don't use `org-current-export-file'.
- Handle file paths robustly i.e., don't rely on `default-directory'.
- Remove stale code.
* org.texi (Fast access to TODO states): Explicitely says only
letters are supported as fast TODO selection keys.
* org.el (org-todo-keywords): Ditto.
Thanks to Samuel Wales for pointing at this.
* org.el (org-flag-drawer): Add a docstring.
(org-mode-map): Bind ̀org-clock-cancel' to "C-cC-xC-q" and
`org-clock-in-last' to "C-cC-xC-x". This fixes a bug in the
previous keybinding for `org-clock-in-last', which would
override the one for `org-clock-in'.
* org-clock.el (org-clock-in-last): Prevent errors when there
is no clocking history.
(org-clock-cancel): Fix bug when checking against a clock log
in a folded drawer.
Thanks to Bernt Hansen who reported the problem.
* org.el (org-link-expand-abbrev): Implement "%(my-function)"
as a new specifier. Update the docstring.
* org.texi (Link abbreviations): Illustrate the use of the
"%h" specifier. Document the new "%(my-function)" specifier.
Thanks to Takaaki ISHIKAWA who came all the way down from Tokyo
to Paris and raised a not-so-distant issue: "could we translate
emails from the mailing list and have a URI for each translated
email?" See the update in the manual for an answer.
* org.el (org-mode-map): Add `C-c C-x C-I' as a keybinding for
`org-clock-in-last'.
* org-clock.el (org-clock-continuously): New option.
(org-clock-in): Three universal prefix arguments set
`org-clock-continuously' to `t' temporarily.
(org-clock-in-last): Fix call to `org-clock-select-task' and
support continuous clocking.
(org-clock-out-time): New variable.
(org-clock-out): set `org-clock-out-time' when clocking out.
Small docstring rewriting.
(org-clock-remove-empty-clock-drawer): Fix "invalid search
bound" bug when trying to delete empty logbook drawer.
(org-clock-cancel): If the clock log is gone, send a warning
instead of deleting the region that is supposed to contain it.
* org.texi (Clocking commands): New cindex.
(Clocking commands): Update documentation for `org-clock-in'.
Document `org-clock-in-last'. Mention `org-clock-out' and
`org-clock-in-last' as commands that can be globally bound.
(Resolving idle time): Document continuous clocking.
* org.el (org-move-line-down, org-move-line-up): Remove.
(org-metaup, org-metadown): When the region is active, move it
up/down by one line, with no regard to the context.
* org-odt.el (org-odt-cleanup-xml-buffers): Use the new alias.
* org-compat.el: Alias `org-condition-case-unless-debug' to
`condition-case-unless-debug' or `condition-case-no-debug'.
* org-pcomplete.el (org-thing-at-point): Ignore trailing
whitespaces while looking-back at properties.
Thanks to David Coate who reported a related issue.
* org.texi (Top, Introduction): Fix formatting.
(Activation): Add index entries.
(Conventions): Update section.
(Embedded @LaTeX{}): Fix formatting.
Mention that `C-c a' is assumed to be the keybinding for `org-capture'
in the whole manual. Thanks to Vagn Johansen who triggered this.
* UTILITIES/org-fixup.el (org-fixup): Let the user override the
version strings produced in org-fixup whith strings stored in
org-fake-release and org-fake-git-version. Skip loading them from a
pre-existing org-version.el or trying to determine them from git in
this case. This is predominantly useful if Git is not available,
but the user wants to install from a Git tarball which is missing a
pre-configured org-version.el file.
* lisp/org-odt.el (org-odt-cleanup-xml-buffers): Use
`condition-case-no-debug' instead of
`condition-case-unless-debug'. This ensures backward
compatibility with Emacs versions < 24.1.
* org.el (org-mode): Set `indent-region-function'.
(org-indent-region): New function.
(org-fill-paragraph): When in a src block, use `indent-region'
to indent the whole source code instead of falling back on
`fill-paragraph', as this function messes up the code.
Aloha all,
The attached patch changes the link to the languages supported by babel
from a directory listing to the Worg page that uses a table to link to
the language-specific documentation pages. It was suggested by John
Hendy.
All the best,
Tom
John Hendy <jw.hendy@gmail.com> writes:
> On Tue, Jul 10, 2012 at 10:19 AM, Thomas S. Dye <tsd@tsdye.com> wrote:
>> John Hendy <jw.hendy@gmail.com> writes:
>>
>>> On Mon, Jul 9, 2012 at 7:35 PM, Thomas S. Dye <tsd@tsdye.com> wrote:
>>>> John Hendy <jw.hendy@gmail.com> writes:
>>>>
>>>>> I have typically outputted graphics from R via tikzDevice manually,
>>>>> but gave my first whirl to just including things via the :file
>>>>> argument in my block header. In the process, I was trying to get this
>>>>> to work and found I needed to use:
>>>>>
>>>>> -----
>>>>> :results output graphics
>>>>> -----
>>>>>
>>>>> This is listed on Worg ob-R-doc:
>>>>> - http://orgmode.org/worg/org-contrib/babel/languages/ob-doc-R.html#sec-4-3
>>>>>
>>>>> But I didn't originally know about it, as it's not a listed header
>>>>> argument in the manual, which is where I check since they're all
>>>>> listed in one place:
>>>>> - http://orgmode.org/manual/results.html#results
>>>>>
>>>>> Does the manual only list universal babel header arguments and R is in
>>>>> a subset of ones using =:results output graphics= or should it be
>>>>> there?
>>>>>
>>>>
>>>> Hi John,
>>>>
>>>> The babel header arguments are implemented at two levels.
>>>>
>>>> There are general babel arguments that apply to all the supported
>>>> languages. These are described in the manual.
>>>>
>>>> It is also possible for the individual language implementations to
>>>> define their own header arguments. R is a good example, where the very
>>>> many arguments to the various R output devices are built into ob-R.el.
>>>> These language specific header arguments are not described in the
>>>> manual, but instead appear in the language specific documentation on
>>>> Worg.
>>>>
>>>> Re: your specific question about :results output graphics. I believe
>>>> this is an R specific header argument and that it is described in the
>>>> appropriate place in the documentation on Worg (see lines 68 and 230
>>>> of ob-R.el).
>>>>
>>>> The manual is not a one-stop source for answers to all Babel questions.
>>>> Recourse to the language specific documentation is typically necessary,
>>>> which is why it would be great if someone could find time to work up
>>>> documentation for the 20 or so languages that still lack it :)
>>>
>>> Thanks for the clarification and that answers the question. This was a
>>> specific header argument for R. Perhaps the manual could link to this
>>> page?
>>> - http://orgmode.org/worg/org-contrib/babel/languages.html
>>>
>>> Maybe with a diddy: "Some languages feature their own unique header
>>> arguments. The above document only universal, system wide Babel header
>>> arguments and one should consult the specific language page on Worg
>>> for additional features."
>>>
>>> Just a thought.
>>
>> Hi John,
>>
>> Thanks for this thought, which is indeed reasonable. I thought the
>> manual might lack this structure, but when I went back to the
>> manual I found that
>> http://orgmode.org/manual/Specific-header-arguments.html#Specific-header-arguments
>> has this at the bottom of the page: "Additional header arguments are
>> defined on a language-specific basis, see Languages", where "Languages"
>> is a link to another manual page, one with a table of supported
>> languages and a link to the Worg page you cite.
>>
>
> Wow. You see only what you look for. :embarrassed:
>
>> The language is a bit different than your proposal, but the path seems
>> clear enough to me.
>
> Technically the path points to
> http://orgmode.org/worg/org-contrib/babel/languages/; I'd point it to
> http://orgmode.org/worg/org-contrib/babel/languages.html as it looks
> cleaner vs. just looking at the directory of .html files. Either
> way... the manual totally does it's job and I missed it!
>
> Thanks for bearing with me!
> John
>
>>
>> Of course, there is still the issue of the languages that lack
>> language-specific documentation :)
>>
>> All the best,
>> Tom
>>
>> --
>> T.S. Dye & Colleagues, Archaeologists
>> 735 Bishop St, Suite 315, Honolulu, HI 96813
>> Tel: 808-529-0866, Fax: 808-529-0884
>> http://www.tsdye.com
>From 0438f0ae1e23321dbacd0666119205985959433e Mon Sep 17 00:00:00 2001
From: Thomas Dye <tsd@tsdye.com>
Date: Tue, 10 Jul 2012 07:12:13 -1000
Subject: [PATCH] Changed babel languages link following a suggestion by John
Hendy
* org-clock.el (org-clocktable-defaults): Revert extra layer of quoting.
It appears the goal in 9b23f082 was to allow defaulting a plist member
to a variable; however the plist was being constructed with LIST and
therefore the `, combination was unnecessary.
TINYCHANGE
