David Bremner [Sun, 14 Aug 2022 14:49:54 +0000 (11:49 -0300)]
changelog for 0.37~rc0-3
David Bremner [Sun, 14 Aug 2022 13:56:15 +0000 (10:56 -0300)]
debian: changelog for 0.37~rc0-2
David Bremner [Sun, 14 Aug 2022 10:28:40 +0000 (07:28 -0300)]
debian: upload release candidate to experimental, not unstable
Oops, tagged "upstream" version is incorrect, but it's a minor issue.
David Bremner [Sun, 14 Aug 2022 09:55:40 +0000 (06:55 -0300)]
debian: changelog stanza for 0.37~rc0-1
David Bremner [Sun, 14 Aug 2022 09:53:41 +0000 (06:53 -0300)]
version: update to 0.37~rc0
David Bremner [Sun, 31 Jul 2022 11:33:23 +0000 (08:33 -0300)]
doc/emacs: use define-key instead of hard-coding markup
Add keys to index. Allow indexing of both keys and commands. Use
standard typesetting for multiple keybindings.
David Bremner [Sun, 31 Jul 2022 11:33:22 +0000 (08:33 -0300)]
doc/emacs: replace use of emacsvar with el:defcustom
Mainly this allows dropping the ad-hoc directive emacsvar
David Bremner [Sun, 31 Jul 2022 11:33:21 +0000 (08:33 -0300)]
doc/emacs: replace emacscmd with el:define-key
This allows us to eliminate the ad-hoc emacscmd directive / role, and
improve indexing.
doc/emacs: convert remaining uses of emacscmd to define-key
This allows us to remove the no-longer used object type.
David Bremner [Sun, 31 Jul 2022 11:33:20 +0000 (08:33 -0300)]
doc/emacs: split index category from label
A more verbose category is needed for the index, but not the
description in the manual.
David Bremner [Sun, 31 Jul 2022 11:33:19 +0000 (08:33 -0300)]
doc/emacs: drop domain label
This makes the generate info files somewhat less verbose, so that they
say e.g. "Key Binding" instead of "Emacs Key Binding"
David Bremner [Sun, 31 Jul 2022 11:33:18 +0000 (08:33 -0300)]
doc/elisp: replace label for Emacs commands and key bindings.
In the notmuch doc index it is important to be clear which parts are
Emacs related.
David Bremner [Sun, 31 Jul 2022 11:33:17 +0000 (08:33 -0300)]
doc: copy elisp.py from flycheck, make available
This corresponds to flycheck commit
1702d2db3d8ba9bcb9b2bf810e791e907fcc3adc, which is apparently the last
time this file was modified (2018).
Embedding seems defensible since the file is evidently not changing
that much, and we need to change some labels to make it fit with the
not-just-emacs nature of notmuch docs.
David Bremner [Sat, 30 Jul 2022 23:45:51 +0000 (20:45 -0300)]
doc: add some documentation about duplicate files.
This is mainly motivated by the new emacs-UI features to select
duplicates, but hopefully it will help demystify things for users of
the CLI as well.
David Bremner [Sun, 31 Jul 2022 12:46:51 +0000 (09:46 -0300)]
emacs: add docstring for notmuch-show-choose-duplicate
It should have one in any case, but in particular it is needed for the
manual.
David Bremner [Mon, 1 Aug 2022 11:39:18 +0000 (08:39 -0300)]
emacs/show: restrict inlined mimetypes on refresh.
This fixes the bug reported by Al [1]. Essentially apply the same fix
as [2] in a different place.
[1]: id:877d41nmr1.fsf@gmail.com
[2]:
90a7c1af368a527700dcde9b0dcbd760afc7bd92
David Bremner [Mon, 1 Aug 2022 11:39:17 +0000 (08:39 -0300)]
test/emacs: add known broken test for inlining on refresh.
This duplicates the bug reported by Al [1]
[1]: id:877d41nmr1.fsf@gmail.com
David Bremner [Thu, 4 Aug 2022 11:48:08 +0000 (08:48 -0300)]
emacs: add duplicate support to notmuch-show-view-raw-message
This fixes one of the bugs reported by Gregor in [1]
[1]: id:87edxw8jp4.fsf@no.workgroup
David Bremner [Thu, 4 Aug 2022 11:48:07 +0000 (08:48 -0300)]
emacs/show: provide a more friendly function to get duplicate num
There is not much code here, but at least we can fix the indexing as 1
based.
David Bremner [Thu, 4 Aug 2022 11:48:06 +0000 (08:48 -0300)]
test/emacs: add known broken test for viewing raw duplicate
This duplicates one of the bugs reported by Gregor in [1].
[1]: id:87edxw8jp4.fsf@no.workgroup
David Bremner [Thu, 4 Aug 2022 11:08:33 +0000 (08:08 -0300)]
emacs: update notmuch-show-get-filename to handle duplicates
This fixes one of the bugs reported by Gregor in [1].
[1]: id:87edxw8jp4.fsf@no.workgroup
David Bremner [Thu, 4 Aug 2022 11:08:32 +0000 (08:08 -0300)]
test/emacs: add known broken test for stashing duplicate filename
This duplicates (no pun intended) one of the bugs reported by Gregor
in id:87edxw8jp4.fsf@no.workgroup
David Bremner [Mon, 1 Aug 2022 13:14:15 +0000 (10:14 -0300)]
NEWS: initial set of NEWS for 0.37.
There are one or two more changes to apply (and potentially document)
before release, but this covers most of my commits.
David Bremner [Sat, 30 Jul 2022 23:45:49 +0000 (20:45 -0300)]
doc/notmuch: merge two sections on external commands.
When adding the description of the propagation of NOTMUCH_CONFIG, I
missed that there was already a section on external commands, with a
different title.
David Bremner [Fri, 29 Jul 2022 12:19:56 +0000 (09:19 -0300)]
tests: remove dead code from T055-path-config.sh
This case statement does nothing.
Tomi Ollila [Sun, 31 Jul 2022 17:32:11 +0000 (20:32 +0300)]
notmuch.c: add missing trailing newlines in two error messages
Removed duplicate error check (and the message) in 3rd case where
the same error message (w/o trailing newline) was present.
In case of test/T040-setup.sh, command substitution deletes
trailing newlines, so related test there cannot be changed
(and therefore could not notice this user experience flaw).
David Bremner [Sat, 16 Jul 2022 21:22:28 +0000 (17:22 -0400)]
doc/emacs: add docstring for notmuch-unthreaded-result-format.
The main change is actually updating cross references.
David Bremner [Sat, 16 Jul 2022 21:22:27 +0000 (17:22 -0400)]
emacs: update defcustom for notmuch-unthreaded-result-format
This is essentially a copy of that for notmuch-tree-result-format,
aside from the default value.
David Bremner [Sat, 16 Jul 2022 21:22:26 +0000 (17:22 -0400)]
doc/emacs: add docstring and example for n-tree-result-format
This example is based on one originally by Jon Hurst.
David Bremner [Sat, 16 Jul 2022 21:22:25 +0000 (17:22 -0400)]
emacs: update defcustom for notmuch-search-result-format.
It seems redundant to have the previous example, since the default
value is always show by describe variable.
Enforce more restrictions on the keys in the alist, since arbitrary
strings don't work as field names.
Document that functions can be used in lieu of field names.
David Bremner [Sat, 16 Jul 2022 21:22:24 +0000 (17:22 -0400)]
doc/emacs: add docstring and example for n-search-result-format
When the ability to use functions was added, this example was confined
to a commit message, which is not user discoverable.
David Bremner [Sat, 16 Jul 2022 21:22:23 +0000 (17:22 -0400)]
emacs: update defcustom for notmuch-search-result-format.
It seems redundant to have the previous example, since the default
value is always show by describe variable.
Enforce more restrictions on the keys in the alist, since arbitrary
strings don't work as field names.
Document that functions can be used in lieu of field names.
David Bremner [Fri, 1 Jul 2022 21:45:48 +0000 (18:45 -0300)]
emacs/reply: reply to correct duplicate
Essentially we just need to arrange to pass the right --duplicate
argument to notmuch reply.
As a side-effect, correct the previously unused value of EXPECTED in
T453-emacs-reply.sh.
David Bremner [Fri, 1 Jul 2022 21:45:47 +0000 (18:45 -0300)]
CLI/reply: support --duplicate argument
We want the reply used to match that shown e.g. in the emacs
interface. As a first step provide that functionality on the command
line.
Schema does not need updating as the duplicate key was already
present (with a constant value of 1).
David Bremner [Fri, 1 Jul 2022 21:45:46 +0000 (18:45 -0300)]
emacs/show: display count of duplicates in headerline
There is no real cost here, except screen real estate. Some people
might prefer hiding the duplicate count, but we leave that for a
future commit.
David Bremner [Fri, 1 Jul 2022 21:45:45 +0000 (18:45 -0300)]
emacs/show: provide notmuch-show-choose-duplicate
This new command allows the user to interactively choose a different
duplicate (file) to display for a given message in
notmuch-show-mode. Since both tree and unthreaded view use
notmuch-show-mode, this provides the same facility there.
David Bremner [Fri, 1 Jul 2022 21:45:44 +0000 (18:45 -0300)]
CLI/show: support --duplicate for structured output
This introduces a new mandatory key for message structures, namely
"duplicate". Per convention in devel/schemata this does _not_ increase
the format version. This means that clients are responsible for
checking that it exists, and not crashing if it does not.
The main functional change is teaching mime_node_open to understand a
'duplicate' argument.
Support for --duplicate in notmuch-reply would make sense, but we
defer it to a later commit.
David Bremner [Fri, 1 Jul 2022 21:45:43 +0000 (18:45 -0300)]
CLI/show: initial support for --duplicate for (raw output only)
Add command line argument --duplicate, analogous with that already
supported for notmuch-search.
Use of a seperate function for _get_filename is mainly a form of
documentation at this point.
md5sum is of course a weak hash, but it is good enough for
this (non-adversarial) test suite use.
David Bremner [Fri, 15 Jul 2022 14:23:21 +0000 (10:23 -0400)]
CLI/git: opportunistically use bindings to check for known messages
If the bindings are installed, use them to avoid one exec of notmuch
search per message.
David Bremner [Sat, 2 Jul 2022 23:45:11 +0000 (20:45 -0300)]
CLI/git: current cache contents of index
Rather than shelling out once per message to get the list of files
corresponding to tags, it is much faster (although potentially a bit
memory intensive) to read them all at once.
David Bremner [Sat, 2 Jul 2022 19:11:58 +0000 (16:11 -0300)]
perf-test: add tests notmuch-git
The main focus of these initial tests is the (currently unacceptably
slow) checkout performance.
David Bremner [Fri, 1 Jul 2022 16:14:29 +0000 (13:14 -0300)]
doc/emacs: add notmuch-show-empty-saved-searches to manual
Answering a user question, I had to dig for this variable, but I think
it is a reasonably common customization wish, particularly for users
with custom count-functions.
David Bremner [Fri, 1 Jul 2022 10:40:39 +0000 (07:40 -0300)]
doc/emacs: define and use emacsvar and emacscmd markup
This makes it easier to navigate the merged index of all of the
notmuch documentation.
David Bremner [Fri, 1 Jul 2022 10:52:35 +0000 (07:52 -0300)]
doc: do not generate python module index
This currently seems unhelpful, and not generating it removes if from
the navigation bar, which prevents mis-clicking.
David Bremner [Fri, 1 Jul 2022 09:52:04 +0000 (06:52 -0300)]
doc: drop module index from ToC
There is only one module listed, so the index is not useful.
David Bremner [Thu, 30 Jun 2022 12:27:10 +0000 (09:27 -0300)]
doc: retitle notmuch-emacs manual
This makes the HTML TOC read more naturally. Add a label to make cross
referencing easier.
David Bremner [Thu, 30 Jun 2022 12:00:43 +0000 (09:00 -0300)]
doc: reorganize sphinx toctree
This mainly affects the html output (although users generating epub or
pdf would also notice a change). The goal is twofold: make the TOC a
bit friendler and easier to navigate by introducing some hierarchy,
and allow links for nmbug and notmuch-setup.
David Bremner [Fri, 1 Jul 2022 11:23:24 +0000 (08:23 -0300)]
doc: improve config item labels in info output
Commit
306b7028d added the nmconfig role / directive. Unfortunately
the default of using the directive name in texinfo output is pretty
ugly, so attempt to make it more human readable by passing `objname`
to add_object_type invocation.
David Bremner [Tue, 28 Jun 2022 01:02:05 +0000 (22:02 -0300)]
doc/emacs: add discussion about notmuch-show performance.
Most of the information is already in the docstrings, but hopefully
by grouping it is more discoverable.
David Bremner [Tue, 28 Jun 2022 01:02:04 +0000 (22:02 -0300)]
emacs/show: introduce notmuch-show-height-limit
Analogous to notmuch-show-depth-limit, it requires a bit more
calculation, but that cost seems dominated by the rendering cost.
David Bremner [Tue, 28 Jun 2022 01:02:03 +0000 (22:02 -0300)]
emacs/show: obey notmuch-show-max-text-part-size for part 0
This parameter was originally introduced to hide large attachements
that happened to be text/plain. From a performance point of view,
there is no reason not to also hide large message bodies.
David Bremner [Tue, 28 Jun 2022 01:02:02 +0000 (22:02 -0300)]
emacs/show: optionally insert bodies of "deep messages" lazily
This leverages the machinery already there to insert buttons for
attachments.
A potential use-case is browsing the top layers of the tree to decide
which of the lower subtrees to read.
David Bremner [Tue, 28 Jun 2022 01:02:01 +0000 (22:02 -0300)]
perf-test/emacs: test showing longish threads
This particular thread takes about 100 times longer to display in
emacs than on the command line.
The parameter notmuch-show-max-text-part-size exists, but is currently
ineffective for this task because it only hides attachments, not
part 0.
In future commits, the parameter notmuch-show-{depth,height}-limit
will trigger a similar kind of hiding for bodies as
n-s-m-text-part-size already does for attachments.
David Bremner [Sun, 3 Jul 2022 15:11:00 +0000 (12:11 -0300)]
debian: add git as a build-dependency, for the test suite
This is needed to run (and test) notmuch-git.
David Bremner [Thu, 23 Jun 2022 12:30:45 +0000 (09:30 -0300)]
CLI/git: add --format-version argument to init subcommand
This is primarily intended to support testing upward compatibility
with legacy repos.
David Bremner [Thu, 23 Jun 2022 12:30:44 +0000 (09:30 -0300)]
CL/git: add format version 1
The original nmbug format (now called version 0) creates 1
subdirectory of 'tags/' per message. This causes problems for more
than (roughly) 100k messages.
Version 1 introduces 2 layers of hashed directories. This scheme was
chose to balance the number of subdirectories with the number of extra
directories (and git objects) created via hashing.
This should be upward compatible in the sense that old repositories
will continue to work with the updated notmuch-git.
David Bremner [Mon, 4 Jul 2022 15:33:48 +0000 (12:33 -0300)]
CLI/git: replace most mentions of nmbug
Particularly in help messages, nmbug is confusing for users who may
have never heard of it.
David Bremner [Sun, 16 Jan 2022 15:25:10 +0000 (11:25 -0400)]
test: known broken test for top level .notmuch in split configs
In split configurations there is no special significance to a top
level directory called .notmuch in the mail root. Users should
therefore be able to have mail stored underneath it.
David Bremner [Sun, 3 Jul 2022 20:26:46 +0000 (17:26 -0300)]
bindings/python-cffi: fix docstring
the method Database.get_message does exist any more (if it ever
did). This makes the docstring unhelpful as an example.
David Bremner [Fri, 1 Jul 2022 21:45:42 +0000 (18:45 -0300)]
test: add new corpus of duplicate messages
This corpus will be used to test a new --duplicate option for notmuch-show
David Bremner [Fri, 1 Jul 2022 21:45:41 +0000 (18:45 -0300)]
test: define and use notmuch_sexp_*_sanitize functions
These are based on the equivalent functions for json. Like those,
these are pretty simple-minded, and don't really understand the
syntax.
David Bremner [Fri, 1 Jul 2022 21:45:40 +0000 (18:45 -0300)]
test: use notmuch_json_show_sanitize more places
This makes the tests more robust against changing output formats, by
allowing us to centralize fixes in the sanitization function. It is
not appropriate for all cases, in particular it is unneeded when using
test_json_nodes, and unhelpful when testing filenames.
David Bremner [Fri, 1 Jul 2022 01:14:20 +0000 (22:14 -0300)]
doc: replace symlink with copies for nmbug, notmuch-setup
Previously only man page aliases were being added as symlinks. The
addition to man_pages in conf.py automatically propagates to the list
of generated info pages.
Installation of the new pages is handled by existing recipes.
David Bremner [Tue, 28 Jun 2022 11:50:55 +0000 (08:50 -0300)]
emacs: mark notmuch-query.el as obsolete
The only functionality actually used by notmuch is the base function
notmuch-query-get-threads; the other functions in this file have
nothing to do with that (single) use. Move that function into
notmuch-lib.el and rename to reflect use. Deprecate the other
functions in notmuch-query.el.
David Bremner [Wed, 15 Jun 2022 12:14:47 +0000 (09:14 -0300)]
lib/sexp: add parameter expansion for regex and wildcard
Fix the bug reported at [1].
The parameter expansion for regex and wildcard modifiers has to be
done a bit differently, because their arguments are not s-expressions
defining complete Xapian queries.
[1]: id:87o7yxqxy6.fsf@code.pm
David Bremner [Wed, 15 Jun 2022 12:14:46 +0000 (09:14 -0300)]
test/sexp: add known broken tests for macro param inside rx/wildcard
These tests replicate the problem reported by Eric Colson [1] (for the
regex case).
[1]: id:87o7yxqxy6.fsf@code.pm
Michael J Gruber [Wed, 29 Jun 2022 11:04:35 +0000 (13:04 +0200)]
make git ignore new build products
nmbug and notmuch-git are new build products. Make git ignore them just
like other build products.
Signed-off-by: Michael J Gruber <git@grubix.eu>
David Bremner [Sat, 4 Jun 2022 21:53:59 +0000 (18:53 -0300)]
doc/sexp: document range queries.
Give examples for date fields, as these are commonly useful for user
queries.
David Bremner [Sat, 4 Jun 2022 21:53:58 +0000 (18:53 -0300)]
lib/sexp: allow * as alias for "" in range searches.
It can be tedious to use "" inside of a string, e.g. in a shell script.
David Bremner [Sat, 4 Jun 2022 21:53:57 +0000 (18:53 -0300)]
lib/sexp: special case "" as an argument in lastmod ranges.
Support this syntax for constincy with (data from to) ranges.
David Bremner [Sat, 4 Jun 2022 21:53:56 +0000 (18:53 -0300)]
test/sexp: one sided range searches.
The date range parsing machinery already knows how to do something
appropriate with an empty string, but the lastmod parsing blindly
tries to parse each atom as a number.
David Bremner [Wed, 25 May 2022 10:51:17 +0000 (07:51 -0300)]
lib: check for writable db in n_m_tags_maildir_flags
The database needs to be writable because the list of stored file
names will change in general.
David Bremner [Wed, 25 May 2022 10:51:16 +0000 (07:51 -0300)]
lib: add NOTMUCH_STATUS_CLOSED_DATABASE, use in _n_d_ensure_writable
In order for a database to actually be writeable, it must be the case that it
is open, not just the correct type of Xapian object. By explicitely
checking, we are able to provide better error reporting, in particular
for the previously broken test in T566-lib-message.
David Bremner [Wed, 25 May 2022 10:51:15 +0000 (07:51 -0300)]
lib: Add missing private status values.
These were missed when the corresponding status codes were added.
David Bremner [Wed, 25 May 2022 10:51:14 +0000 (07:51 -0300)]
lib/tag: handle NULL argument to notmuch_tags_valid
Make the behaviour when passed NULL consistent with
notmuch_filenames_valid. The library already passes the result of
notmuch_message_get_tags without checking for NULL, so it should be
handled.
David Bremner [Wed, 25 May 2022 10:51:13 +0000 (07:51 -0300)]
test: add known broken test for notmuch_tags_valid (NULL)
This should return false, but currently segfaults.
Start a new file for tags library API related tests. This is maybe
overkill, but new C boilerplate which doesn't corrupt the database is
needed anyway.
David Bremner [Wed, 25 May 2022 10:51:12 +0000 (07:51 -0300)]
test: error handling for n_m_tags_to_maildir_flags
The closed database case should fail gracefully, but currently it
segfaults.
David Bremner [Mon, 23 May 2022 23:39:01 +0000 (20:39 -0300)]
test: error handling in _n_message_{add,remove}_filename
Use a closed database to force throwing a Xapian exception. Leave the
check vague to accomodate a potential explicit check for open database.
David Bremner [Mon, 23 May 2022 23:39:00 +0000 (20:39 -0300)]
lib/message: check return status from _n_m_add_{path,folder}_terms
Mainly to propagate information about Xapian exceptions.
David Bremner [Mon, 23 May 2022 23:38:59 +0000 (20:38 -0300)]
lib/message: check return status of _n_m_{add,remove}_term
Xapian exceptions are not something that can be ignored, in general.
David Bremner [Mon, 23 May 2022 23:38:58 +0000 (20:38 -0300)]
lib: define macro NODISCARD
In either C++17 (or later) mode, or when running cppcheck, this can be
used to selectively generate warnings about discarded return values.
David Bremner [Mon, 23 May 2022 23:38:57 +0000 (20:38 -0300)]
lib/message: drop _notmuch_message_get_thread_id_only
This function has been unused since commit
4083fd8.
David Bremner [Mon, 23 May 2022 23:38:56 +0000 (20:38 -0300)]
test: _notmuch_message_remove_term catches exceptions.
Unfortunately we can't differentiate between the two distinct error
conditions for Document::remove_term.
David Bremner [Mon, 23 May 2022 23:38:55 +0000 (20:38 -0300)]
lib/message: catch exceptions in _n_m_add_term
Some code movement is needed to make sure the cache is only
invalidated when the Xapian operation succeeds.
David Bremner [Mon, 23 May 2022 23:38:54 +0000 (20:38 -0300)]
test: define test_private_C
When testing error handling, it is sometimes difficult to cover a
particular error path deterministically. Introduce a test function to
allow calling lower level functions directly.
David Bremner [Sat, 4 Jun 2022 11:11:19 +0000 (08:11 -0300)]
CLI: set NOTMUCH_CONFIG in hooks.
This addresses a bug report / feature request of Uwe Kleine-König. The
assumption is that we always load a config file in the CLI (i.e. we
never pass "" as the config file argument to
notmuch_database_open_with_config).
[1]: id:
8baa58c3-7ab9-ec03-1bbd-
28aa5be838f2@kleine-koenig.org
David Bremner [Mon, 30 May 2022 11:38:19 +0000 (08:38 -0300)]
test: add known broken tests for setting NOTMUCH_CONFIG in hooks
Setting this according to --config was requested by Uwe
Kleine-König. There are some other ways that the configuration file
might be found in current notmuch, so check those as well.
As a bonus, fix a bug in the hook tests that left NOTMUCH_CONFIG set
even though a config file was provided via NOTMUCH_PROFILE.
[1]: id:
8baa58c3-7ab9-ec03-1bbd-
28aa5be838f2@kleine-koenig.org
David Bremner [Thu, 7 Apr 2022 22:43:15 +0000 (17:43 -0500)]
debian: install notmuch-git
Use a separate binary package to avoid dragging in dependencies on
python and git for those that do not want them.
David Bremner [Sat, 14 May 2022 18:25:47 +0000 (15:25 -0300)]
CLI/git: add safety checks for checkout and commit
Commits or checkouts that modify a large fraction of the messages in
the database should be relatively rare (and in some automated process,
probably non-existent). For initial setup, where such operations are
expected, the user can pass --force.
David Bremner [Tue, 10 May 2022 10:32:48 +0000 (07:32 -0300)]
CLI/git: support configuration for repo location / prefix
This is probably more convenient than always passing a command line
argument.
Use notmuch-config for consistency with other notmuch CLI tools.
Now that there is something relevant in the config files, test the
--config option.
David Bremner [Sat, 7 May 2022 13:32:18 +0000 (10:32 -0300)]
CLI/git: change defaults for repo and prefix
The previous defaults were not suitable for personal (i.e. not
bugtracking for notmuch development) use.
Provide two ways for the user to select nmbug compatible defaults;
command line argument and checking the name of the script.
David Bremner [Fri, 8 Apr 2022 00:41:38 +0000 (19:41 -0500)]
doc/notmuch-git: initial documentation
This is mainly derived from the various help outputs from the script,
with some massaging of markup and addition of links.
David Bremner [Tue, 19 Apr 2022 11:32:14 +0000 (08:32 -0300)]
CLI/git: cache git indices
If the private index file matches a previously known revision of the
database, we can update the index incrementally using the recorded
lastmod counter. This is typically much faster than a full update,
although it could be slower in the case of large changes to the
database.
The "git-read-tree HEAD" is also a bottleneck, but unfortunately
sometimes is needed. Cache the index checksum and hash to reduce the
number of times the operation is run. The overall design is a
simplified version of the PrivateIndex class.
David Bremner [Sat, 16 Apr 2022 22:44:10 +0000 (19:44 -0300)]
CLI/git: add @timed decorator, time a few functions
Perf will show which binaries are using the CPU cycles, and standard
python profilers will show which python functions, but neither is
great at finding which call to an external binary is taking time, or
locating I/O hotspots.
David Bremner [Sun, 10 Apr 2022 17:22:57 +0000 (14:22 -0300)]
CLI/git: replace enumeration of tags with sexp query.
Unlike the (current) infix query parser provided by Xapian, the
notmuch specific sexp query parser supports prefixed wildcard queries,
so use those. In addition to being somewhat faster, this avoids
needing to escape all of the user's tags to pass via the shell.
David Bremner [Sun, 10 Apr 2022 17:06:14 +0000 (14:06 -0300)]
test/git: add known broken test for tag with quotes.
There is current insufficient sanitization and/or escaping of tag names
internally in notmuch-git.
David Bremner [Sun, 3 Apr 2022 22:17:16 +0000 (16:17 -0600)]
CLI/git: suppress warnings about initial branch name
The canonical nmbug repository still uses "master" as the main branch
name, so defer any potential switch away from that name.
David Bremner [Sun, 15 May 2022 16:47:54 +0000 (13:47 -0300)]
CLI/git: rename environment variables.
Although the code required to support both new and old environment
variables is small, it complicates the semantics of configuration, and
make the documentation harder to follow.
David Bremner [Sun, 20 Mar 2022 21:33:54 +0000 (18:33 -0300)]
test: initial tests for notmuch-git
Exercise the main functionality of notmuch-git. add_git_repos() will
hopefully be simplifed when an init subcommand is added.
W. Trevor King [Mon, 19 Jan 2015 00:40:10 +0000 (16:40 -0800)]
CLI/git: Add an 'init' command
For folks that want to start versioning a new tag-space, instead of
cloning one that someone else has already started.
The empty-blob hash-object call avoids errors like:
$ nmbug commit
error: invalid object 100644
e69de29bb2d1d6434b8b29ae775ad8c2e48c5391 for
'tags/...'
fatal: git-write-tree: error building trees
'git HASH(0x9ef3eb8) write-tree' exited with nonzero value
David Bremner suggested [1]:
$ git hash-object -w /dev/null
instead of my Python version of:
$ git hash-object -w --stdin <&-
but I expect that closing stdin is more portable than the /dev/null
path (which doesn't exist on Windows, for example).
The --bare init and use of NMBGIT as the work tree (what could go
wrong with an empty commit?) are suggestions from Michal Sojka [2].
[1]: id:87y4vu6uvf.fsf@maritornes.cs.unb.ca
http://thread.gmane.org/gmane.mail.notmuch.general/18626/focus=18720
[2]: id:87a93a5or2.fsf@resox.2x.cz
http://thread.gmane.org/gmane.mail.notmuch.general/19495/focus=19767
David Bremner [Mon, 4 Apr 2022 03:05:36 +0000 (21:05 -0600)]
CLI/git: make existance of config branch optional on clone
This branch is actually only used by an associated
utility (notmuch-report), and notmuch-git works fine without it.