Currently, the 'search.exclude_tags' option is automatically
set to "deleted;spam;" if it's missing from the config file.
This violates the Principle of Least Surprise, so *only* set
'search.exclude_tags' to "deleted;spam;" if we didn't find a
configuration file at all.
This patch is actually Austin Clements' work:
id:"20120117203211.GQ16740@mit.edu"
Currently, the 'search.exclude_tags' option is automatically set to
"deleted;spam;" if it's missing from the config file.
This violates the Principle of Least Surprise, so update the tests to
*only* expect the exclusion of messages which are tagged "deleted" if the
'search.exclude_tags' option is explicitly set *and* contains that tag.
All other config-related functions and args include the section
title in their name, so for the sake of consistency, mirror that.
Also, the "auto"matic part is a given, so that was dropped.
Before 3434d1940 the return values of libnotmuch functions were
declared as c_void_p and the code checking for errors compared the
returned value to None, which is the ctypes equivalent of a NULL
pointer.
But said commit wrapped all the data types in python classes and the
semantic changed in a subtle way. If a function returns NULL, the
wrapped python value is falsish, but no longer equal to None.
Backported from master to 0.11.
Previously, top-level message headers were printed as Subject, From,
To, Date, while embedded message headers were printed From, To,
Subject, Date. This makes both cases use the former order and updates
the tests accordingly.
Strangely, the raw format also uses this function, so this also fixes
the two raw format tests affected by this change.
emacsclient --eval '(kill-emacs)' makes emacs versions 23.1
and 23.2 ask user input from running emacs. Redefining
yes-or-no-p function when kill-emacs is executed for these
emacs versions in test-lib.el avoids this test problem.
Used emacs (whitespace-cleanup) function to "cleanup blank problems"
in test files where that could be done without breaking tests;
test/emacs was partially, and test/multipart was fully reverted.
JSON does not support hex literals (0x..) so numbers must be formatted
as %d instead of %x.
Currently, the possible values for the gmime error code are 1 (expired
signature), 2 (no public key), 4 (expired key) and 8 (revoked key).
The other possible value is 16 (unsupported algorithm) but obviously
it is much more rare. If this happens, the current code will add
'"errors": 10'. This is valid JSON (it looks like a decimal number)
but it is incorrect (should be 16, not 10).
Since this is just an issue in the JSON encoder, no changes are needed
on the Emacs side (or in other UIs using the JSON output).
Before 3434d1940 the return values of libnotmuch functions were
declared as c_void_p and the code checking for errors compared the
returned value to None, which is the ctypes equivalent of a NULL
pointer.
But said commit wrapped all the data types in python classes and the
semantic changed in a subtle way. If a function returns NULL, the
wrapped python value is falsish, but no longer equal to None.
Uncrustify is a free (as in GPL2+) tool that indents and beautifies
C/C++ code. It is similar to GNU indent in functionality although
probably more configurable (in fairness, indent has better
documentation). Uncrustify does not have the indent mis-feature of
needing to have every typedef'ed type defined in the
configuration (even standard types like size_t).
This configuration starts with the linux-kernel style from the
uncrustify config, disables aggressive re-indenting of structs,
and fine tunes the handling 'else' and braces.
In an ideal situation, running uncrustify on notmuch code would be
NOP; currently this is not true for all files because 1) the
configuration is not perfect 2) the coding style of notmuch is not
completely consistent; in particular the treatment of braces after
e.g. for (_) is not consistent.
Some fine tuning by Tomi Olilla.
Define a keymap for attachment buttons to allow multiple actions.
Define 3 possible actions:
save attachment: exactly as currently,
view attachment: uses mailcap entry,
view attachment with user chosen program
Keymap on a button is: s for save, v for view and o for view with
other program. Default (i.e. enter or mouse button) is save but this
is configurable in notmuch customize.
One implementation detail: the view attachment function forces all
attachments to be "displayed" using mailcap even if emacs could
display them itself. Thus, for example, text/html appears in a browser
and text/plain asks whether to save (on a standard debian setup)
There are lots of API changes in gmime 2.6 crypto handling. By adding
preprocessor directives, it is however possible to add gmime 2.6 compatibility
while preserving compatibility with gmime 2.4 too.
This is mostly based on id:"8762i8hrb9.fsf@bookbinder.fernseed.info".
This was tested against both gmime 2.6.4 and 2.4.31. With gmime 2.4.31, the
crypto tests all work fine (as expected). With gmime 2.6.4, one crypto test is
currently broken (signature verification with signer key unavailable), most
likely because of a bug in gmime which will hopefully be fixed in a future
version.
This makes `show-trailing-whitespace' happy, i.e. it does not mark the
whole search box line as trailing spaces.
Since the dot is invisible, this change makes no visible difference
for `notmuch-hello'.
Edited-by: Pieter Praet <pieter@praet.org> to fix the tests.
This ignores the results of the two writes in sigint handlers even
harder than before.
While my libc lacks the declarations that trigger these warnings, this
can be tested by adding the following to notmuch.h:
__attribute__((warn_unused_result))
ssize_t write(int fd, const void *buf, size_t count);
For showing a message in raw format, rather than silently succeeding
when a read or a write fails (or, probably, looping if a read fails),
try to print an error message and exit with a non-zero status.
This silences one of the buildbot warnings about unused results. While
my libc lacks the declarations that trigger these warnings, this can
be tested by adding the following to notmuch.h:
__attribute__((warn_unused_result))
size_t fwrite(const void *ptr, size_t size, size_t nmemb, FILE *stream);
'message contains options relevant to 'notmuch-send, not the other way around.
Thanks to Austin for suggesting `custom-add-to-group'.
id:"20120118184408.GD16740@mit.edu"
Add various functions to print notmuch messages and tie them together
with a simple frontend.
Add a binding ('#') in `notmuch-show-mode' to print the current
message.
one trailing space removed by db.
Add an explicit note to the README explaining what programs are
necessary and the perhaps-surprising behavior of skipping tests if
they aren't present.
Signed-off-by: Ethan Glasser-Camp <ethan@betacantrips.com>
Enable the truncation of lines in `notmuch-show-mode' to avoid visual
noise caused by the wrapping of the header lines.
Don't enable `visual-line-mode' because it disables line truncation.
The benefits of `visual-line-mode' were that it wrapped long lines
in received messages. With `notmuch-wash-wrap-long-lines' now default
behaviour, this is no longer required.
To allow for expansion whilst keeping everything tidy and organized,
move all defcustom/defface variables to the following subgroups,
defined in notmuch-lib.el:
- Hello
- Search
- Show
- Send
- Crypto
- Hooks
- External Commands
- Appearance
As an added benefit, defcustom keyword args are now consistently
ordered as they appear @ defcustom's docstring (OCD much?).
Proper defgroup docstrings and various other improvements
by courtesy of Austin Clements.
Previously, the dependency file list was generated before the CLI
sources were added to SRCS, so dependency files weren't generated for
CLI sources. This moves that code to after the CLI sources are added.
This reverts commit 4b256ff557.
According to id:"87aa5nlwwg.fsf@praet.org" and followup messages, the
assumptions of the patch seem not to hold in emacs 23.
In NEWS file, indentation for item descriptions is generally 2 spaces
but in a few cases there were 3 or 4 (4 caused different markdown
handling) space indentations. Indentation in those lines are brought
to consistent 2-space indentation.
This adds a "search" section to the config file and an
"auto_tag_exclusions" setting in that section. The search and count
commands pass tag tags from the configuration to the library.
This is useful for tags like "deleted" and "spam" that people
generally want to exclude from query results. These exclusions will
be overridden if a tag is explicitly mentioned in a query.
If a message was received to the user's address that was in a named
group list, notmuch reply does not use that address for picking the
from address.
Groups lists are of the form: foo:bar@example.com,baz@example.com;
Signed-off-by: Jani Nikula <jani@nikula.org>
All strings are unicode strings in python 3 and the basestring and
unicode types are removed hence the need for a specialized version.
Signed-off-by: Justus Winter <4winter@informatik.uni-hamburg.de>
When running the Emacs tests in verbose mode, only the first missing
prereq is reported because the `run_emacs' function is short-circuited
early:
#+begin_example
emacs: Testing emacs interface
missing prerequisites: [0] emacs(1)
skipping test: [0] Basic notmuch-hello view in emacs
SKIP [0] Basic notmuch-hello view in emacs
#+end_example
This can lead to situations reminiscent of "dependency hell", so instead
of returning based on each individual `test_require_external_prereq's exit
status, we now do so only after checking all the prereqs:
#+begin_example
emacs: Testing emacs interface
missing prerequisites: [0] dtach(1) emacs(1) emacsclient(1)
skipping test: [0] Basic notmuch-hello view in emacs
SKIP [0] Basic notmuch-hello view in emacs
#+end_example
Also added missing prereq for dtach(1).
As suggested by j4ni in #notmuch, rename
`notmuch-jump-to-recent-buffer' as `notmuch-cycle-notmuch-buffers' and
have it behave accordingly.
Consider `message-mode' buffers to be of interest.
The tags were coloured using text properties. Unfortunately that text
(the header line) also has an overlay, which overrides the text
properties. There's not point in applying text properties that will
never be seen.
Provide reply to sender counterparts to the search and show reply
functions. Add key binding 'R' to reply to sender, while keeping 'r' as
reply to all, both in search and show views.
Signed-off-by: Jani Nikula <jani@nikula.org>
