Moved static functions _config_get_list () and _config_set_list ()
closer to the beginning of file so that their definition is known
(without adding forward declarations) in upcoming changes.
`mail-header-parse-address' expects un-decoded mailbox parts, which is
not what we have at this point. Replace it with simple string
deconstruction.
Mark the corresponding test as no longer broken.
Minor whitespace cleanup.
The `mm-inlinable-p' function works better if it has access to the
data of the relevant part, so load that content before calling it.
Don't load the content for parts that the user has indicated no desire
to inline.
This fixes the display of attached image/jpeg parts, for example.
There are two ways to do search in Emacs UI: search widget in
notmuch-hello buffer and `notmuch-search' function bound to "s".
Before the change, these search mechanisms used different history
lists. The patch makes notmuch-hello search use the same history list
as `notmuch-search' function.
Before the change, "s" in notmuch-hello buffer would jump to the
search box. The patch changes the binding to `notmuch-search' which
is consistent with all other notmuch buffers.
Adjusted some uncrustify variables to get closer to prevailing style:
* Label indent (for goto) relative to current indentation.
* Registered GMimeObject and mime_node_t being as types.
* Space after ! (not) operator.
* No space after 'stringify' (#) preprosessor token.
* No spacing change around ## (option not versatile enough).
There are at least 3 cases where attention needs to be paid:
* If there is newline between function name and open paren in function
call, the paren (and args) are indented too far right.
* #define HOUR (60 *MINUTE) -- i.e. no space after star (*).
* void (*foo)(args) -- i.e no space between (name) and (args).
Add a new test function to allow simpler testing of emacs
functionality.
`test_emacs_expect_t' takes one argument - a lisp expression to
evaluate. The test passes if the expression returns `t', otherwise it
fails and the output is reported to the tester.
When checking for a running emacs, test_emacs evaluates the empty list
'()'. This returns 'nil' when emacs is running, which is then
prepended to the actual test result. Given that it is not part of the
actual test output the test harness can incorrectly report test
failure (or success).
This callback is the gateway to the new mime_node_t-based formatters.
This maintains backwards compatibility so the formatters can be
transitioned one at a time. Once all formatters are converted, the
formatter structure can be reduced to only message_set_{start,sep,end}
and part, most of show_message can be deleted, and all of
show-message.c can be deleted.
This makes the part numbers readily accessible to formatters.
Hierarchical part numbering would be a more natural and efficient fit
for MIME and may be the way to go in the future, but depth-first
numbering maintains compatibility with what we currently do.
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"
