| Commit message (Collapse) | Author | Age |
... | |
| |/
| |
| |
| |
| |
| | |
As discussed elsewhere, handling the mouse was not done before, and
breaks work flows that include using the mouse for other things
(middle-click paste). This only changes the default option to False.
|
| |
| |
| |
| |
| |
| |
| |
| | |
This creates a deferred that runs on a user configurable timer and is
provided the ui. This deferred will re-fire after the number of seconds
specified by the timer are eclipsed. By default this is 300 seconds.
Fixes #522
|
| |
| |
| |
| |
| | |
This is done to give more emphasis to the fact that there might be header
lines at the top of the string.
|
| | |
|
|/ |
|
|\
| |
| | |
Use absolute_imports from __future__
|
| | |
|
| |
| |
| |
| |
| |
| |
| |
| | |
The old regex that was suggested to capture addresses for completion from the
command `notmuch address` was bogus. The json output format is much more
predictable.
Reported in #990.
|
|/
|
|
|
|
|
|
|
|
|
|
|
| |
Currently if global.quit_on_last_bclose is set to True (not the default),
and the last buffer is an envelope, then when 'd' or ':bclose' is
called, the user will be prompted twice to 'close without sending?'.
This patch fixes that by skipping the prompt in BufferCloseCommand, if
there is 1 buffer, and global.quit_on_last_bclose is True. It does this
by adding a private keyword to the ExitCommand() constructor, telling
that classes apply method to not prompt.
Fixes #979
|
|
|
| |
Fix typo
|
|
|
|
|
|
|
|
|
|
| |
Mouse support is enabled by default. To disable mouse support, set
```
handle_mouse = False
```
in your config file.
|
| |
|
|
|
|
|
|
|
|
| |
this is to make sure that sphinx's autodoc works
for `alot.UI` and other urwid-derived classes
when building the docs on readthedocs.net
fixes #938
|
|
|
|
| |
This is just whitespace changes.
|
| |
|
| |
|
|
|
|
| |
https://www.quantifiedcode.com/app/issue_class/AiJMd9EB
|
| |
|
| |
|
| |
|
|
|
|
|
|
| |
When checking out files from git the timestamps might not give enough
information to make which files to remake. So this new target can be used to
force a remake of all generated files on the next run.
|
|
|
|
|
|
|
|
|
| |
The path to the present development version of alot should be prepended to
sys.path in order to shadow other versions of alot installed on the system.
Also it should be a directory from which the alot module is loadable. That is
the root of the git repository and not one level above this.
This was overlooked in #912.
|
|
|
|
|
| |
`sphinx-build -W` was complaining as this is not a option of alot and can not
be found in our documentation.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
On Arch Linux python points to python3 and hence the makefile fails. With
this change users can tell the makefile which python executable to use:
make PYTHON=/some/version2/python ...
or simply
make PYTHON=python2 ...
For sake of completeness the same is done for rm.
|
| |
|
|
|
|
|
|
|
|
| |
api/settings.rst previously erroneously pluralized alot.addressbook,
named one class that does not exist, named one from a sub module, and
omitted two sub modules (with classes).
This was causing build errors from sphinx
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Default value 'first_message' lists authors in the order they joined the
conversation. Value 'latest_message' order authors by their latest
message, which makes it easier to see which authors who wrote the most
recent messages.
Note that authors with duplicate emails were previously filtered. We now
keep all authors where name + email is unique. This is behavior is
desired for some email notification services. For example, Jira will
set the From header to "Joe User <jira@company.com>" meaning you will
only see one author if you only keep unique email addresses.
|
| |
|
|
|
|
| |
This should save some manual deleting and regenerating of these files.
|
|
|
|
| |
How the SIGUSR1 signal is handled
|
| |
|
| |
|
|
|
|
|
|
|
| |
Auto-detect the '--list' for you automagically when you just ':reply'
on a list, it is based on a config param that give you the control,
by default it is off, you have to setup your config file, like:
auto_replyto_mailinglist = True
|
|
|
|
|
|
|
|
|
|
|
| |
Provide an option to the reply command ':reply --list':
* In reply to someone to a list (not you), it will take if available
(Reply-To, X-BeenThere or X-Mailing-List).
* In the rare reply to you sending to a list, it will take the To which
is the list mail.
This work start/rebase/refactor from this 'old' (3 years) PR on github:
https://github.com/pazz/alot/pull/479#issuecomment-152750321
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|