David Bremner [Tue, 18 Aug 2020 10:58:28 +0000 (07:58 -0300)]
update changelog for 0.31~rc1-1
David Bremner [Tue, 18 Aug 2020 00:02:30 +0000 (21:02 -0300)]
version: bump to 0.31~rc1
Tomi Ollila [Mon, 17 Aug 2020 20:39:22 +0000 (23:39 +0300)]
NEWS: notmuch-mutt: system(shell pipeline) replaced internally
David Bremner [Sun, 16 Aug 2020 17:04:36 +0000 (14:04 -0300)]
test: fix uninitialized variable use in T562-lib-database
Fix a copy paste error of using the boolean ret as a notmuch_status_t,
and uninitialized.
David Bremner [Sun, 16 Aug 2020 15:27:11 +0000 (12:27 -0300)]
build: clean up sphinx.config
Follow the existing practice and remove it under "distclean", same as
sh.config and Makefile.config
David Bremner [Sun, 16 Aug 2020 14:32:28 +0000 (11:32 -0300)]
devel/release-checks.sh: use grep to find copyright year.
This is quite fragile, but it works for now, unlike the python
version.
In general it seems conf.py is not intended to be evaluated outside of
sphinx, as it assumes certain global names (in particular "tags") are
defined.
David Bremner [Sun, 16 Aug 2020 14:12:59 +0000 (11:12 -0300)]
debian: update symbols for 0.31
Two new API entries for better error handling
David Bremner [Sun, 16 Aug 2020 14:09:18 +0000 (11:09 -0300)]
debian: start changelog for 0.31~rc0-1
David Bremner [Sun, 16 Aug 2020 14:06:13 +0000 (11:06 -0300)]
version: bump to 0.31~rc0
Start the release process for 0.31
William Casarin [Tue, 11 Aug 2020 17:36:53 +0000 (10:36 -0700)]
emacs/tree: add notmuch-tree-archive-thread-then-next
Now that notmuch-tree-next-thread acts more like its notmuch-show
counterpart, let's update the binding to move to the next thread after
archiving.
Signed-off-by: William Casarin <jb55@jb55.com>
William Casarin [Tue, 11 Aug 2020 17:36:52 +0000 (10:36 -0700)]
emacs/tree: enable moving to next thread in search results
This introduces a new function called
notmuch-tree-next-thread-from-search which is analogous to
notmuch-show-next-thread. It will switch to the next or previous
thread from the parent search results.
We rename notmuch-tree-{prev,next}-thread to a more descriptive
notmuch-tree-{prev,next}-thread-in-tree to reflect the fact that it
only moves to the next thread in the current tree.
notmuch-tree-next-thread now switches to the next thread in the
current tree first, but if there are none, it looks for the next tree
in the search results.
This makes notmuch-tree feel more like notmuch-show when using the
M-Enter, M-n and M-p bindings.
Signed-off-by: William Casarin <jb55@jb55.com>
William Casarin [Tue, 11 Aug 2020 17:36:51 +0000 (10:36 -0700)]
emacs/tree: introduce notmuch-tree-parent-buffer variable
This variable will be used in a similar fashion to
notmuch-show-parent-buffer. It will be used to navigate between
threads from the parent search buffer.
Signed-off-by: William Casarin <jb55@jb55.com>
David Bremner [Sat, 15 Aug 2020 12:15:05 +0000 (09:15 -0300)]
test: update README to reflect dropping upgrade tests
These test databases have been unneeded since
ee897cab8.
Teemu Likonen [Sat, 15 Aug 2020 06:28:29 +0000 (09:28 +0300)]
Emacs: Indent first header line only when indentation is turned on
Previously in message-show mode message's first header line (From
header) was always indented, even if user had turned thread
indentation off with "<" (notmuch-show-toggle-thread-indentation)
command.
This change modifies notmuch-show-insert-headerline function so that
it doesn't indent the first header line if notmuch-show-indent-content
variable is nil.
This change also modifies tests so that they expect this new output
format:
test/emacs-show.expected-output/notmuch-show-indent-thread-content-off
Tomi Ollila [Mon, 27 Jul 2020 19:38:33 +0000 (22:38 +0300)]
notmuch-mutt: replace shell pipeline with internal pipe processing
The shell pipeline used to symlink files based in search results
to "cache" directory for mutt(1) to use was prone to portability
problems (due to /bin/sh differences).
The replacement executes `notmuch search` without intermediate shell
(so shell_quote was removed in this case), reads the filenames from
piped output and symlinks files internally.
Tomi Ollila [Tue, 11 Aug 2020 07:38:19 +0000 (10:38 +0300)]
configure: Check if emacs >= 25 (instead of >= 24) is available
"The minimum supported major version of GNU Emacs is now 25.1."
25.1 is the first "released" version of Emacs 25.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:09 +0000 (13:50 +0200)]
try-emacs-mua: Trim `require' advice for Emacs 25
- Since Emacs 25 comes with `load-prefer-newer' we can remove the
complicated variant of the advice, which implemented a poorman's
version of that.
- Since Emacs 25 comes with the new advice mechanism, we can use
that now for the simple variant of the advice, which just informs
about the library that is being required.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:08 +0000 (13:50 +0200)]
emacs: Use new advice mechanism do advice mm-shr
Also because we now only support Emacs >= 25,
we can remove the check for Emacs >= 24.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:07 +0000 (13:50 +0200)]
emacs: Do not abuse advice to monkey patch while testing
Use `cl-letf*' instead.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:06 +0000 (13:50 +0200)]
emacs: Drop old advices that were only need for Emacs 23
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:05 +0000 (13:50 +0200)]
emacs: Remove notmuch-read-char-choice
Just use `read-char-choice', which existed since Emacs 24.1.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:04 +0000 (13:50 +0200)]
emacs: Remove notmuch-setq-local
Just use setq-local, which existed since Emacs 24.3.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:03 +0000 (13:50 +0200)]
emacs: Use cl-incf where appropriate
It's shorter. That's it pretty much.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:02 +0000 (13:50 +0200)]
NEWS: At least Emacs 25.1 is required now
Some backward incompatible changes follow in the next few commits
and going forward contributors don't have to worry about Emacs 24
at all anymore.
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:01 +0000 (13:50 +0200)]
NEWS: Add stub for 0.31
Jonas Bernoulli [Sat, 8 Aug 2020 11:50:00 +0000 (13:50 +0200)]
emacs: Add end-of-file line to libraries that lack it
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:59 +0000 (13:49 +0200)]
emacs: Provide 'rstdoc' feature at end of file
Features should nearly always be provided at the very end of their
libraries. This feature isn't one of the rare exceptions.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:58 +0000 (13:49 +0200)]
.gitignore: Sort using sort-lines
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:57 +0000 (13:49 +0200)]
test: Fix indentation
Fix it to consistently match the style we have configured in
".dir-locals.el".
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:56 +0000 (13:49 +0200)]
.dir-locals.el: Set variables for correct "shell" mode
The major mode used for shell scripts is named 'sh-mode'.
'shell-mode' on the other hand implements an interactive
shell in emacs-lisp.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:55 +0000 (13:49 +0200)]
Fix typos
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:54 +0000 (13:49 +0200)]
emacs: Increase consistency of library headers
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:53 +0000 (13:49 +0200)]
emacs: Various cosmetic changes
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:52 +0000 (13:49 +0200)]
emacs: Autoload notmuch-jump using an autoload cookie
Doing that is better than using an `autoload' form because the latter
may result in dependencies getting hidden and indeed it turns out we
have to declare `notmuch-jump' in "notmuch-tag.el".
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:51 +0000 (13:49 +0200)]
emacs: Autoload notmuch-jump-search only once
This function is being autoloaded using an autoload cookie, so it
shouldn't additionally be autoloaded using an `autoload' form.
When building libraries we don't actually load the autoloads file and
dropping the `autoload' form results in an error, which reveals a so
far unspecified dependency: `notmuch-tree' needs `notmuch-jump'.
Before this commit compiling (or even just loading) `notmuch-tree'
resulted in `notmuch-jump' being loaded because the former requires
`notmuch-lib', which autoloaded `notmuch-jump-search'.
The bug was that this dependency was not explicitly specified, which
we fix by adding the respective `require' form.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:50 +0000 (13:49 +0200)]
emacs: Improve doc-strings
- The first sentence should fit on the first line in full. This is
even the case when that causes the line to get a bit long. If it
gets very long, then it should be made shorter.
- Even even the second sentence would fit on the first line, if it
just provides some details, then it shouldn't be done.
- Symbols are quoted like `so'.
- There is no clear rule on how to (not) quote non-atomic
s-expressions, but quoting like '(this) is definitely weird.
- It is a good idea to remember that \" becomes " and to take
that in mind when adjusting the automatic filling by hand.
- Use the imperative form.
- Arguments are written in all uppercase.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:49 +0000 (13:49 +0200)]
emacs: Use makefile-gmake-mode in Makefile*s
Use `makefile-gmake-mode' instead of `makefile-mode' because the
former also highlights ifdef et al. while the latter does not.
"./Makefile.global" and one "Makefile.local" failed to specify any
major mode at all but doing so is necessary because Emacs does not
automatically figure out that these are Makefiles (of any flavor).
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:48 +0000 (13:49 +0200)]
emacs: notmuch-poll: Let the user know we are polling
It is done synchronously and it can take a while,
so we should let the user know what is going on.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:47 +0000 (13:49 +0200)]
emacs: No longer define notmuch-hello-mode-map as a function
It was defined as such for a decade; ever since
a56010ac8b89a2489eee5c78469f05cee85ec858 but there
wasn't a reason to do that then nor is there now.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:46 +0000 (13:49 +0200)]
emacs: Fix some function declarations
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:45 +0000 (13:49 +0200)]
emacs: Extend face to window edge again
Since Emacs 27 each face has to be explicitly configured to "extend
to the edge of the window". Without doing that the face used for
the newline character only has an effect that spans "one character"
(i.e. it looks like there is a single trailing space character).
We don't want that so extend the face in Emacs 27, so that it looks
the same as it did in older Emacs releases. We have to do this
conditionally, otherwise older Emacsen would choke on it.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:44 +0000 (13:49 +0200)]
emacs: Use one or three lines for 'if' forms
Putting the COND and THEN parts on the same line but ELSE on a
separate line makes it harder to determine if there actually is
an ELSE part.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:43 +0000 (13:49 +0200)]
emacs: Use 'when' instead of 'if' when there is no ELSE part
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:42 +0000 (13:49 +0200)]
emacs: Use 'unless' instead of 'when' and 'not'
Also use 'unless' in a few cases where previously 'if' was used with
'not' but without an ELSE part.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:41 +0000 (13:49 +0200)]
emacs: Use 'and' instead of 'when' when the return value matters
Also do so for some 'if' forms that lack an ELSE part.
Even go as far as using 'and' and 'not' instead of 'unless'.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:40 +0000 (13:49 +0200)]
emacs: Only set one variable per setq form
It's a bit weird to avoid having to write the "(setq ... )" more than
once, just because we can. In a language that uses '=' for the same
purpose we also happily use that once per assignment.
While there are no benefit to using just one 'setq' there are some
drawbacks. It is not always clear on first what is a key and what a
value and as a result it is easy to make a mistake. Also it becomes
harder to comment out just one assignment.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:39 +0000 (13:49 +0200)]
emacs: Closing parenthesis go on the same line
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:38 +0000 (13:49 +0200)]
emacs: Fix indentation
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:37 +0000 (13:49 +0200)]
emacs: Remove excess empty lines
Most people who write lots of lisp tend to only sparsely use empty
"separator" lines within forms. In lisp they feel unnecessary and
since most files stick to this convention we get a bit confused
when there are extra empty lines. It feels like the s-expressions
are falling into pieces.
All of this is especially true between a function's doc-string and
body because the doc-string is colored differently, which visually
already separates it quite sufficiently from the code that follows.
Jonas Bernoulli [Sat, 8 Aug 2020 11:49:36 +0000 (13:49 +0200)]
emacs: Shorten long lines
Teemu Likonen [Fri, 7 Aug 2020 04:46:41 +0000 (07:46 +0300)]
Emacs: Ensure left-to-right display for message headers
In notmuch-show buffer insert invisible U+200E LEFT-TO-RIGHT MARK
character at the beginning of message header paragraph if the From
header contains a right-to-left character. This ensures that the
header paragraph is always rendered in left-to-right mode.
See Emacs Lisp reference manual section "(elisp) Bidirectional
Display" for more info.
Đoàn Trần Công Danh [Mon, 3 Aug 2020 15:40:57 +0000 (22:40 +0700)]
T355: specify hash algorithm explicitly
On some systems (notably, the one shipped with LibreSSL),
default fingerprint digest algorithm is SHA256.
On other systems, users can change default digest algorithm by changing
default_md in /etc/ssl/default_md.
Let's ask openssl to provide us specific algorithm to make the test
more deterministic.
Signed-off-by: Đoàn Trần Công Danh <congdanhqx@gmail.com>
Đoàn Trần Công Danh [Mon, 3 Aug 2020 15:40:56 +0000 (22:40 +0700)]
configure: drop check for default xapian backend
Starting from xapian 1.3.5, xapian switched default backend to glass.
From
00cdfe10 (build: drop support for xapian versions less than 1.4,
2020-04-22), we only support xapian 1.4.0+. Effectively, we don't need
to check for default xapian backend anymore.
Further more, from
99a7aac8 (test: drop use of db_ending, 2020-07-29),
our test framework has become independence from default xapian.
Let's drop it.
Signed-off-by: Đoàn Trần Công Danh <congdanhqx@gmail.com>
Đoàn Trần Công Danh [Mon, 3 Aug 2020 15:40:55 +0000 (22:40 +0700)]
configure: check for pytest with python -m pytest
On different distro, pytest is suffixed with different patterns.
On the other hand, `python3-pytest' could be invoked correctly,
via `python3 -m pytest', the latter is used by our tests, now.
Switch to `$python -m pytest` to fix address all incompatible naming.
Signed-off-by: Đoàn Trần Công Danh <congdanhqx@gmail.com>
David Bremner [Sat, 1 Aug 2020 12:26:03 +0000 (09:26 -0300)]
test: regression tests for n_indexopts_{get,set}_decrypt_policy
The main criteria is that they don't crash. Working with a closed
database is a bonus.
David Bremner [Sat, 1 Aug 2020 12:26:02 +0000 (09:26 -0300)]
test: regression test for traversing config list with closed db
Also mention error return in API docs
David Bremner [Sat, 1 Aug 2020 12:26:01 +0000 (09:26 -0300)]
test: regression test for n_d_get_config_list on closed db.
Exception is caught.
David Bremner [Sat, 1 Aug 2020 12:26:00 +0000 (09:26 -0300)]
test: regression test for n_directory_{get,set}_mtime
The mtime is cached, so closing the db is not a problem. Writing the
mtime throws an exception, which is caught.
David Bremner [Sat, 1 Aug 2020 12:25:59 +0000 (09:25 -0300)]
lib: fix return value for n_directory_delete
Falling out of the catch meant the error return was lost
David Bremner [Sat, 1 Aug 2020 12:25:58 +0000 (09:25 -0300)]
test: known broken test for n_directory_delete with closed db.
There is a return value bug in notmuch_directory_delete that is hiding
the exception.
David Bremner [Sat, 1 Aug 2020 12:25:57 +0000 (09:25 -0300)]
lib: catch exceptions in n_directory_get_child_files
Also clarify API in error case.
David Bremner [Sat, 1 Aug 2020 12:25:56 +0000 (09:25 -0300)]
test: add known broken test for n_directory_get_child_files
This is a clone of the one for get_child_directories
David Bremner [Sat, 1 Aug 2020 12:25:55 +0000 (09:25 -0300)]
lib: catch exceptions in n_directory_get_child_directories
Also clarify API in error case.
David Bremner [Sat, 1 Aug 2020 12:25:54 +0000 (09:25 -0300)]
test: add known broken test for n_directory_get_child_directories
Start a new test file (for the notmuch_directory_* API group) to hold
this test.
David Bremner [Sat, 1 Aug 2020 12:25:53 +0000 (09:25 -0300)]
test: regression test for n_m_get_filenames
Closing the database after the iterator is created is not a problem.
David Bremner [Sat, 1 Aug 2020 12:25:52 +0000 (09:25 -0300)]
test: split header for lib-message tests.
This allows finer control over when to close the database.
David Bremner [Sat, 1 Aug 2020 12:25:51 +0000 (09:25 -0300)]
test: add regression test for n_messages_collect_tags
Also test n_messages_destroy.
David Bremner [Sat, 1 Aug 2020 12:25:50 +0000 (09:25 -0300)]
lib: return NULL from n_d_get_default_indexopts on error
This is a rare and probably serious programming error, so better not
to silently return a default value.
David Bremner [Thu, 16 Jul 2020 11:11:26 +0000 (08:11 -0300)]
build: upload html docs as part of release process
Use a URL https://notmuchmail.org/doc/latest to leave room for a
future more ambitious scheme deploying multiple versions.
This also forces the html docs to built as part of the release
process. In the future this should be updated to tolerate generating a
release without sphinx installed. This needs a new target analogous to
build-info and build-man that does nothing if sphinx is not installed.
David Bremner [Thu, 30 Jul 2020 00:01:30 +0000 (21:01 -0300)]
test: drop use of db_ending
This will allow the dropping of the test for the default ending from
configure.
David Bremner [Thu, 30 Jul 2020 00:01:29 +0000 (21:01 -0300)]
test: drop NOTMUCH_DEFAULT_XAPIAN_BACKEND from T360-symbol-test
Inspired by the suggestion of
id:
20200727154108.16269-3-congdanhqx@gmail.com
to drop the configuration test for the default backend. This version
is hopefully robust against backend changes.
David Bremner [Fri, 24 Jul 2020 11:14:06 +0000 (08:14 -0300)]
test: destroy thread from closed database
Check for (non)-crash.
David Bremner [Fri, 24 Jul 2020 11:14:05 +0000 (08:14 -0300)]
test: regression test for n_thread_get_tags
Code is taken from the API docs, with the twist that the underlying
database is closed. Not crashing is the main point.
David Bremner [Fri, 24 Jul 2020 11:14:04 +0000 (08:14 -0300)]
test: add regression tests for oldest and newest dates
These are strictly to watch for crashes, so don't be too fussy about
the actual timestamps.
David Bremner [Fri, 24 Jul 2020 11:14:03 +0000 (08:14 -0300)]
test: add regression tests for n_t_get_{authors, subject}
This is returning explicitely cached data, so no database access is needed.
David Bremner [Fri, 24 Jul 2020 11:14:02 +0000 (08:14 -0300)]
test: add regression test for n_t_get_messages
This is similar to the case of toplevel messages. Currently everything
is cached, so no database access is necessary. This might change in
the future, but it should not crash in either case.
David Bremner [Fri, 24 Jul 2020 11:14:01 +0000 (08:14 -0300)]
test: add regression tests for n_thread_get_toplevel_messages
Include a test for the previously omitted non-trivial code path for
notmuch_thread_get_replies.
David Bremner [Fri, 24 Jul 2020 11:14:00 +0000 (08:14 -0300)]
test: add regression test for n_thread_get_total_{messages,files}
This is returning cached info, so does not need to access the (closed)
database.
David Bremner [Fri, 24 Jul 2020 11:13:59 +0000 (08:13 -0300)]
test: regression test for n_t_get_thread_id
This is just cached data, so as long as we don't prematurely free
anything, it should be fine.
David Bremner [Wed, 22 Jul 2020 10:51:27 +0000 (07:51 -0300)]
test: regression test for notmuch_query_destroy
As with other void API entries, not crashing counts as success.
David Bremner [Wed, 22 Jul 2020 10:51:26 +0000 (07:51 -0300)]
test: regression tests for n_q_count_{messages, threads}
At least these exceptions are caught.
David Bremner [Wed, 22 Jul 2020 10:51:25 +0000 (07:51 -0300)]
test: regression test for n_q_search_messages
Exception handling matches notmuch_query_search_threads, at least for
this case.
David Bremner [Wed, 22 Jul 2020 10:51:24 +0000 (07:51 -0300)]
test: regression test for n_q_search_threads
At least this exception is caught.
David Bremner [Wed, 22 Jul 2020 10:51:23 +0000 (07:51 -0300)]
test: add regression test for n_q_add_tag_exclude
This relies on the change to not tear down the auxilary Xapian
objects, in particular the query parser, when the database is closed.
David Bremner [Wed, 22 Jul 2020 10:51:22 +0000 (07:51 -0300)]
test: regression test for n_q_{set, get}_sort
We just want to make sure accessing the query owned by a closed db
does not cause a crash.
David Bremner [Wed, 22 Jul 2020 10:51:21 +0000 (07:51 -0300)]
test: regression test for set_omit_excluded
David Bremner [Wed, 22 Jul 2020 10:51:20 +0000 (07:51 -0300)]
test: regression test for notmuch_query_get_database
Current functionality is too trivial to really justify a test, but
it's simpler just to test the complete API.
David Bremner [Wed, 22 Jul 2020 10:51:19 +0000 (07:51 -0300)]
test: add regression test for n_q_{create,_get_query_string}
Start a new file of tests, to keep to the (emerging) scheme of one
notmuch_foo group per file
David Bremner [Wed, 22 Jul 2020 10:51:18 +0000 (07:51 -0300)]
test: move notmuch_message_* tests to their own file
This is for consistency with the recently added tests for
notmuch_database_*.
David Bremner [Sun, 26 Jul 2020 23:31:36 +0000 (20:31 -0300)]
lib: replace use of static_cast for writable databases
static_cast is a bit tricky to understand and error prone, so add a
second pointer to (potentially the same) Xapian database object that
we know has the right subclass.
David Bremner [Sun, 26 Jul 2020 23:31:35 +0000 (20:31 -0300)]
lib: encapsulate the use of notmuch_database_t field 'mode'
The plan is to change the underlying representation.
David Bremner [Sun, 26 Jul 2020 23:31:34 +0000 (20:31 -0300)]
lib: drop two gratuitous assignments to database mode
I'm not sure what the point of modifying that right before destroying
the object is. In a future commit I want to remove that element of the
object, so simplify that task.
Jonas Bernoulli [Mon, 27 Jul 2020 15:25:03 +0000 (17:25 +0200)]
test: Explicitly state that we want to sign with sender
Since Emacs 27 'mml-secure-epg-sign' errors out if we don't opt-in to
signing as the sender using 'mml-secure-openpgp-sign-with-sender'.
Jonas Bernoulli [Mon, 27 Jul 2020 15:25:02 +0000 (17:25 +0200)]
test: Remove misguided emacs testing utilities
The goal of this abstraction was to save space. But that failed as
the result actually was that four trivial lines got replace with 15
fairly complicated lines. The opposite of what it was supposed to
do.
Also it made it harder to come up with the fix in the previous commit;
simply grepping for the relevant symbols did not work because they get
constructed at run-time instead of appearing in the source file.
Jonas Bernoulli [Mon, 27 Jul 2020 15:25:01 +0000 (17:25 +0200)]
test: Deal with Emacs 27 switching to lexical scope by default
Starting with Emacs 27 undeclared variables in evaluated interactive
code uses lexical scope. This includes code passed with '--eval' as
we do in the Emacs tests, which also happen to assume dynamic scope.
- This can affect variables defined by libraries that we use. We
let-bind such variables to change the behavior of functions which we
then call with these bindings in effect. If these libraries are not
loaded beforehand, then the bindings are lexical and fail to have
the effect we intended.
At this time only 'smtpmail' has to be loaded explicitly (for the
variables let-bound in emacs_deliver_message and emacs_fcc_message).
'message' doesn't have to be loaded explicitly, because loading
'notmuch' (in 'run_emacs') already takes care of that, indirectly.
- Our own testing-only variables also have to be declared explicitly.
We should have done that anyway, but because of how and where these
variables are used it was very easy to overlook that (i.e. it isn't
something the byte-compiler ever looks at). Not so in Emacs 27
anymore; here this oversight caused four tests to fail.
The numeric values of these variables get incremented by functions
that we add to hooks that are run by many tests, not just the tests
where we actually inspect the value and therefore take care to let-
bind the values to 0 before we begin. The global values therefore
have to be numeric values as well. I have chosen -100 instead of 0
as the default in case someone writes a test that inspects the value
but forgets to let-bind the value. I hope that the unusual negative
value that one is going to see in such a case will help debugging
the issue.
Jonas Bernoulli [Mon, 27 Jul 2020 15:25:00 +0000 (17:25 +0200)]
gitignore: Ignore generated sphinx.config
David Bremner [Sun, 19 Jul 2020 13:11:13 +0000 (10:11 -0300)]
test: add known broken test for n_d_get_default_indexopts
Xapian exceptions are swallowed and turned into default return value.
David Bremner [Sun, 19 Jul 2020 12:51:03 +0000 (09:51 -0300)]
lib: fix error return bug with n_d_set_config.
The catch block either needs to return, or the function needs to
return "status". Choose the latter for consistency with
n_d_get_config.
David Bremner [Sun, 19 Jul 2020 12:44:05 +0000 (09:44 -0300)]
test: add known broken test for n_d_set_config
Error status is currently lost.
David Bremner [Sun, 19 Jul 2020 12:39:16 +0000 (09:39 -0300)]
test: add regression test for n_d_get_config
Xapian exceptions seem to handled OK, at least for this case.