This fixes a bug in set-tags-command excluding a tag that is both set
locally and inherited from the initial minibuffer input by modifying
org-get-tags to prefer keeping the locally set tag over the inherited
tag, as this behavior is more intuitive for org-get-tags anyway.
* lisp/org.el (org-get-tags): Keep local tags over inherited.
* testing/lisp/test-org.el (test-org/set-tags-command): Add test.
* lisp/org.el (org-get-cursor-date): Fix regular expression.
Previous regular expression assumed the time grid string will have two
digits in the hour portion of the time string. However, the time grid
string does not always have two digits. For example:
" 8:00......"
* lisp/org-faces.el (org-block): Set background extension beyond
end-of-line.
* lisp/org-compat.el (org--set-faces-extend): New function to
temporarily (re)set :extend for Emacs≥27.
* lisp/org.el (org-mode): Call it to set the extend attribute of
relevant faces to the correct value.
* lisp/org.el (org-todo): Use nil instead of
org-loop-over-headlines-in-active-region as the MATCH argument.
org-loop-over-headlines-in-active-region isn't related to the MATCH
argument; it's related to SCOPE and decides the value of the let-bound
cl variable that's passed as SCOPE. Note, though, that
org-loop-over-headlines-in-active-region is let-bound to nil around
the org-map-entries call, so this didn't cause any issues.
* lisp/org.el (org-fontify-meta-lines-and-blocks-1): Rewrite the "line
begins with *" check so that it does not assume that match data has
remained intact.
The upstream call to org-src-font-lock-fontify-block overwrites the
match data in some cases. As a result, (match-string 1) signals an
error and #+end_src is not fontified. Wrapping the call to
org-src-font-lock-fontify-block with save-match-data would fix the
reported issue. However, the position at the beginning of the line is
already stored in a variable, so use that information instead to check
whether the line starts with "*".
Reported-by: John Ciolfi <ciolfi@mathworks.com>
<r9r1ulpctz.fsf@ah-ciolfi-lah-ciolfi-l.dhcp.mathworks.com>
* lisp/org.el (org-mark-element): Do not set mark past narrowed part
of the buffer.
* testing/lisp/test-org.el (test-org/mark-element): New test.
Reported-by: Kevin Liu <kevin@nivekuil.com>
<http://lists.gnu.org/r/emacs-orgmode/2020-05/msg00812.html>
* lisp/org.el (org-set-tags-command): Only fix cursor position in very
specific circumstances (i.e., when cursor is on an empty headline).
* testing/lisp/test-org.el (test-org/set-tags-command): Add tests
related to point position.
Modified-by: Kyle Meyer <kyle@kyleam.com>
Adjusted for review comments and added tests.
* lisp/org.el (org-dynamic-block-insert-dblock): Add an interactive-p
argument that is non-nil for interactive calls and signals to call the
dynamic block function interactively.
org-clock-report takes a prefix argument. When it had a regular
binding, this was easy to access. As of 34b71a0ca (Add a dispatcher
command for inserting dynamic blocks, 2018-12-23), its only "binding"
is through the dynamic block dispatcher. Make it possible to supply a
prefix argument in that context too.
* lisp/org-id.el (org-id-overriding-file-name): New variable.
(org-id-get): Prefer org-id-overriding-file-name over the buffer's
file name if set.
(org-id-locations-load): Give a more informative error when file is
nil.
* lisp/org.el (org-clone-subtree-with-time-shift): Let-bind
org-id-overriding-file-name, enabling an ID to be created for a cloned
subtree rather than crashing with a type error.
Note that, before 9865e6bd8 (org-id: Speedup, minor functional change
and fix, 2019-08-01), this wouldn't fail with a type error. However,
the ID would not be added correctly because org-id-add-location simply
wouldn't process the ID at all if the passed file was nil.
Reported-by: Ian Eure <ian@retrospec.tv>
<87a73caayj.fsf@phaktory>
* lisp/org.el (org-time-stamp-inactive): Drop incorrect statements in
docstring about differences between inactive and active timestamps.
S-cursor commands work on inactive time stamp, and clicking on an
inactive time stamp visits the day in the agenda, just as it does for
an active time stamp.
Reported-by: Jorge P. de Morais Neto <jorge@disroot.org>
<87a73gstpz.fsf@disroot.org>
These both show up when compiling files separately (e.g., 'make
single').
ob-core already assumes ob-tangle is loaded based on declared
functions, presumably because ob.el is given responsibility for
collectively loading these pieces. org.el loads org-attach within the
body of a function before calling org-attach-dir.
* lisp/org.el (org-set-tags-command): Also avoid using `looking-back',
in particular without a LIMIT argument. This silences the
byte-compiler.
(cherry picked from commit 83518daf11)
* lisp/org.el (org-set-tags-command): Fix point position after
setting tag at the beginning of a blank heading.
(org-kill-line): Prevent from throwing an error when killing
the headline while point is right after the * chars.
Since 'add-to-list', being a plain function, cannot access lexical
variables, such use must be rewritten for correctness.
(Some instances actually do work thanks to a compiler macro,
but it's not something code should rely on.)
* lisp/org.el (org-reload): Replace add-to-list with push.
Replace add-to-list to lexical variable with push (bug#39373)
32763dac46e61cc34e8fe4d19df4905d09c1a27f
Mattias Engdegård
Sat Feb 1 22:27:23 2020 +0100
* lisp/org.el (org-do-emphasis-faces): Only fontify emphasis
markers when they are visible.
Thanks to d.williams for provided a bug report and this fix.
* lisp/org.el (org-refresh-property)
(org-refresh-category-properties): Tiny docstring or comment
fixes.
(org-set-effort): Fail early when the effort value is wrong.
Thanks to Firmin Martin for reporting this.
* lisp/org.el (org-cycle-internal-local): Handle
`org-cycle-include-plain-lists'.
Thanks to Bruce Tulloch for reporting this.
This fixes a bug introduced in commit c2aebcee.
* lisp/org.el (org-indent-region): Use `save-window-excursion'
to restore point after indenting region in source blocks.
Thanks to Gustavo Barros for reporting this bug.
* lisp/org.el (org-cycle-internal-local): When cycling visibility
on a subtree, don't take a list has a child.
These two subtrees should cycle the same way:
* A subtree
A Paragraph.
* A subtree
- A list.
Before the fix, the second subtree would cycle through three steps
instead of two.
Thanks to Samuel Wales for reporting this.
* lisp/org.el (org-open-at-point): Don't forceably try ot open
attachments.
This commit fixes a bug: when trying to open links in a subtree with
no attachments, `org-open-at-point' would create an ID property with
no apparent reason.
