* lisp/ob-exp.el (org-babel-exp-code): Fix export of src blocks with
flags.
* testing/lisp/test-ob-exp.el (ob-export/export-src-block-with-flags):
New test.
Thanks to Giuseppe Lipari for reporting it.
* ox.el (org-export--get-subtree-options): When using the
headline as a title for a subtree export, only take the true
heading, no TODO keyword, no priority cookie, no tag.
Thanks to Robert Klein for reporting this.
* lisp/ob-exp.el (org-babel-exp-process-buffer): Make processing more
robust when results are inserted before source block or when source
block is followed by multiple blank lines.
* testing/lisp/test-ob-exp.el (ob-export/export-with-results-before-block):
Add test.
* lisp/ox.el (org-export-insert-default-template): Only insert
keywords and options relatives to the selected back-end. Ignore
those relatives to its parent in the case of a derived back-end.
* contrib/lisp/ox-s5.el (org-s5-title-slide-template): Per Arne
Babenhauserheide, the heading levels on the default title slide
template were wrong per the S5 spec.
* lisp/ox-ascii.el (org-ascii--current-text-width): Convert `length'
to `string-width'.
(org-ascii--build-title, org-ascii--build-toc)
(org-ascii--list-listings, org-ascii--list-tables)
(org-ascii-template--document-title)
(org-ascii-inner-template, org-ascii-format-inlinetask-default)
(org-ascii-format-inlinetask-default, org-ascii-item
(org-ascii--table-cell-width, org-ascii-table-cell)
(org-ascii--current-text-width): Likewise.
I've checked all occurrences of the function `length' in ox-ascii.el.
It turns out that the most of them are calculating the width of given
string. To support fullwidth characters, we better use `string-width'
instead of `length'.
Some characters in UCS are categorized as "East Asian Ambiguous"[1].
The return value of `string-width' with those characters depends on
how Emacs is setup. We leave those ambiguous character handling to
Emacs.
Two usages of `length' in `ox-ascii.el' were left as-is, because those
were used for:
- bullet depth calculation in `org-ascii-headline', and
- cell position calculation in `org-ascii--table-cell-width'.
[1]: http://www.unicode.org/reports/tr11/#Ambiguous
* lisp/ox-latex.el (org-latex-headline, org-latex-item): Fix items
starting with a square bracket.
Thanks to Jon Degenhardt, Vladimir Lomov and Eric S Fraga for
contributing to the discussion.
* org.el (org-demote): Ignore invisible text when aligning
tags.
(org-set-tags): When JUST-ALIGN is 'ignore-column, ignore
invisible text when restoring the cursor to the correct
column.
This fixes a bug about demoting hidden headlines.
If org-move-to-column temporarily ignore visibility
specs, this will prevent org-demote to work correctly
in hidden regions.
Thanks to Susan Cragin for reporting this bug.
* ob-python.el (org-babel-python-var-to-python): Bugfix: Strip
properties before formatting the results.
Thanks to Daniel Gerber for reporting this and providing a fix,
and to others who contributed.
* org-agenda.el (org-agenda-regexp-filter-preset): Fix
typo in docstring.
(org-agenda-reapply-filters): New function.
(org-agenda-drag-line-forward): Rewrite to fix a bug when used
in filtered agendas.
(org-agenda-drag-line-backward): Rewrite using
`org-agenda-drag-line-forward'.
Thanks to Thomas Morgan for reporting this.
* lisp/org.el (org-get-previous-line-level): Do not call
`org-current-level' twice unless necessary. Also, avoid using
`line-number-at-pos' when the information needed is to know if point
is in the first line of the visible part of the buffer.
* org-entities.el (org-entities-help): Prevent the display of
pretty entities, as this help buffer is meant to list literal
strings, not utf-8 representations.
Thanks to Brice Waegenire for reporting this.
* lisp/ox-beamer.el (org-beamer-select-environment): Function doesn't
work if fast tag selection is disabled, so make sure it is always
on, independently on user's configuration.
Reported-by: Anders Johansson <mejlaandersj@gmail.com>
* doc/Makefile (%.pdf): Ensure that LANG and LC_ALL are set to "C",
not "C " (with trailing whitespace).
Make evaluation order produces spurious whitespace in variables when a
variable definition is followed by whitespace plus a comment.