| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
All compose commands were made available in the message viewer context
by mistake. Remove them.
This fixes a crash when the :send command is invoked while viewing
a message:
Fixes: 159fb38daf53 ("commands: refactor registration")
Fixes: https://lists.sr.ht/~rjarry/aerc-devel/%3CCZ01NTGWNUBV.14A9POG9ITJ6F%40posteo.net%3E
Changelog-fixed: Fixed crash when running `:send` with a `:preview`
tab focused.
Reported-by: Sertonix <sertonix@posteo.net>
Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Maarten Aertsen <maarten@nlnetlabs.nl>
|
| |
|
|
|
|
|
| |
This has nothing to do at the root of the source tree.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Acked-by: Bence Ferdinandy <bence@ferdinandy.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a new :menu command that can be used to spawn a shell command in an
interactive dialog and execute aerc commands for each of the lines
printed on its standard output like xargs(1) would do in a shell.
The shell command can be configured in aerc.conf under
[general].default-menu-cmd and overridden via the :menu -c <cmd> option.
There are two flags to provide input text to the shell command which can
be useful in combination with other aerc commands:
-a: All account names, one per line:
'<account>' LF
-d: All current account directory names, one per line:
'<directory>' LF
-ad: All directories of all accounts, one per line:
'<account>' '<directory>' LF
Here are some examples:
:menu -adc fzf :cf -a
:menu -c 'fzf --multi' :attach
:menu -dc 'fzf --multi' :cp
And also for key bindings:
<C-p> = :menu -adc fzf :cf -a<Enter>
Changelog-added: New `:menu` command to invoke other aerc commands
based on a shell command output.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Bence Ferdinandy <bence@ferdinandy.com>
Tested-by: Johannes Thyssen Tishman <johannes@thyssentishman.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Register all commands with the same function and store them in the same
map.
Use bit flags to determine in which contexts each command should be
available.
Remove duplicate commands now that the same command can be exposed in
multiple contexts.
Refactor API to allow executing commands from other commands without
import cycles.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Bence Ferdinandy <bence@ferdinandy.com>
Tested-by: Johannes Thyssen Tishman <johannes@thyssentishman.com>
|
| |
|
|
|
|
|
|
| |
Create the sub-command structure for the :patch command. Make the :patch
command accessible from any context.
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Remove CompletionFromList which is a trivial wrapper around FilterList.
Remove the prefix, suffix and isFuzzy arguments from FilterList.
Replace prefix, suffix by an optional callback to allow post processing
of completion results before presenting them to the user.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
|
| |
Implement command completion with complete struct field tags from the
get-opt library introduced earlier.
Changelog-changed: Improved command completion.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
Tested-by: Moritz Poldrack <moritz@poldrack.dev>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
|
| |
Do not expose the completion of a command via its command set. Instead,
require a single command object to be resolved in order to execute it.
Extract the command names and the template terms completions in main.go.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
Tested-by: Moritz Poldrack <moritz@poldrack.dev>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use the argument parsing framework introduced earlier to unify the
parsing of (almost) all command options. Remove custom parsing code and
to avoid extraneous types, add fields with `opt` tags on command structs
that have options and arguments. Commands that take no argument do not
need anything.
Since the command objects now carry data, create a new temporary
instance of them before passing them to opt.ArgsToStruct when executing
a command.
A few of the commands use specific semantics for parsing (:choose), or
are delegating argument parsing to another function (:sort, :search,
:filter). For these commands, simply add a dummy "-" passthrough
argument. Since all commands still have the argument list (after split)
nothing needs to be changed in this area.
There should be no functional change besides the Usage strings and
reported errors which are now generated automatically.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
Tested-by: Moritz Poldrack <moritz@poldrack.dev>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
| |
Do not expose the execution of a command via its command set. Instead,
require a single command object to be resolved in order to execute it.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
Tested-by: Moritz Poldrack <moritz@poldrack.dev>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
| |
Move the template expansion before the command name is looked up. It
saves a few void renderings and will be required for the next commits.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Some commands need to invoke others and/or run shell commands. For this,
we need the raw command line as entered by the user. Pass it down the
call chain just before it is split to invoke the command Execute method.
Remove unit tests for the template expand() test which does have any
added value now that it is performed on a single string without any
quote juggling.
Update all code to handle a single string instead of a list of
arguments.
Introduce a new dependency on git.sr.ht/~rjarry/go-opt to deal with
shell splitting. This is in preparation for using opt.ArgsToStruct to
parse arguments for all aerc commands.
There should be no functional change after this patch.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
Tested-by: Moritz Poldrack <moritz@poldrack.dev>
Tested-by: Inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
The single Aerc object is passed around in almost all command functions.
This hinders readability.
Store the single Aerc instance as a global variable. Export public
functions from the app package to access methods of that object. Remove
all explicit references to *app.Aerc and replace them with calls to
these functions. For references to private/unexported fields and
functions from within the app package, directly access the global aerc
object.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Acked-by: Moritz Poldrack <moritz@poldrack.dev>
|
| |
|
|
|
|
|
|
|
| |
This is the central point of all aerc. Having it named widgets is
confusing. Rename it to app. It will make a cleaner transition when
making the app.Aerc object available globally in the next commit.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Acked-by: Moritz Poldrack <moritz@poldrack.dev>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Fix a panic in the template execution when the template evalutes to nil.
Check for length of returned arguments after the template code is
expanded.
To reproduce, run :{{exec "cat > /dev/null" .MessageId}}
Fixes: 42cd4157 ("commands: execute commands with templates")
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Use the template interface consistently. Before, we have exported the
state.TemplateData struct and used it in most places instead of the
models.TemplateData interface. This lead to some inconsistencies, i.e.
Role() has been defined on the exported struct but not on the interface.
Unexport the state.TemplateData struct, add a DataSetter interface to
set the data needed for the template data and call the Data() method
which returns a models.TemplateData interface when the template data is
needed.
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Tim Culverhouse <tim@timculverhouse.com>
|
| |
|
|
|
|
|
|
|
|
|
|
| |
Execute template code before running a command.
With this, we can use templates in our keybinds like:
[messages]
E = :'{{if match (.Flags|join "") "O"}}envelope{{else}}view{{end}}'<Enter>
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Expand and complete template code. The exline does understand template
code and can evaluate it. With this patch, the completion systems
supports writing the templates by showing the available template data
fields and template functions. Pressing <Tab> after }} will show the
evaluated template in the completion menu. Complex template code, such
as if-else statements work as well.
Examples:
:filter -f {{<Tab>
will show the possible template data fields and functions
:filter -f {{index (.To|email) 0}}<Tab>
will show the value of the template expression in the completion list
Pressing <Tab> twice after a completed template expression will
substitute the expression with its value.
:{{if match .Folder "INBOX"}}check-mail{{else}}cf INBOX{{end}}
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Improve command completion by supporting option flags and option
arguments completion. Option completion is activated when the command
implements the OptionsProvider interface. Implementing the
OptionCompleter allows the completion of individual option arguments.
The completion interfaces harmonizes the completion behavior in aerc,
makes the completion code clearer and simplifies the completion
functionality.
With this patch, the Complete method on commands will only have to deal
with the actual completion, i.e. paths, folders, etc and not worry about
options. To remove all options and its mandatory arguments from args,
use commands.Operands().
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
| |
Add .Role as a template field for use in distinguishing between
mailboxes with a given IANA role, or custom role defined by aerc
("query" for notmuch queries, for example).
Signed-off-by: Tim Culverhouse <tim@timculverhouse.com>
Acked-by: Robin Jarry<robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Since previous commit, all commands now support expanding text/template
markup. Reuse that for the new-email trigger command.
Update commands.ExecuteCommand to take optional *AccountConfig and
*MessageInfo arguments. If these are nil, fallback to using the
currently selected account and message (if any).
Pass the proper *AccountConfig and *MessageInfo objects when firing the
trigger command so that these are used instead of the currently selected
ones.
If new-email contains % placeholders, try to convert them to template
markup reusing the same conversion added in commit 535300cfdbfc
("config: add columns based index format"). Warn the user that they need
to update their configuration file.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Moritz Poldrack <moritz@poldrack.dev>
|
| |
|
|
|
|
|
|
|
| |
Interpret go template constructs in all aerc command arguments based on
the currently selected account, folder and message (if any).
Signed-off-by: Aivars Vaivods <aivars@vaivods.lv>
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Tim Culverhouse <tim@timculverhouse.com>
|
| |
|
|
|
|
|
|
|
| |
A previous commit introduced a trace loglevel, but aerc doesn't make
much use of it. Logging is not very informative either. Add a log.Tracef
to every command and log the provided arguments as well.
Signed-off-by: Bence Ferdinandy <bence@ferdinandy.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch implements :prompt completion.
The completion mechanism only provides completions when there is at least
one argument specified (prompt text).
The mechanism is based on other commands' completions and works as follows:
1. Attempts to look up a command by the name specified in args[1].
2.a On success it uses command.Complete.
2.b Otherwise, if total arguments count is lesser or equals than 2
(i.e. no command arguments specified), it attempts to complete
the command's name.
Additional effort is made to preserve prompt text, which often contains
spaces and formatting.
Signed-off-by: Sergey Smirnykh <sergey.smirnykh@siborgium.xyz>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
| |
Panic might occur if aerc.SelectedAccount() is nil
Signed-off-by: kt programs <ktprograms@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
|
|
| |
Fix panic in the GetCompletions function when using the keybind for the
terminal command.
Fixes: https://todo.sr.ht/~rjarry/aerc/29
Reported-by: inwit <inwit@sindominio.net>
Signed-off-by: Koni Marti <koni.marti@gmail.com>
Acked-by: Robin Jarry <robin@jarry.cc>
|
| |
|
|
|
|
|
|
| |
Change the option to enable fuzzy completion to be fuzzy-complete, since
it's no longer only used for folders
Signed-off-by: Kt Programs <ktprograms@gmail.com>
Acked-by: Koni Marti <koni.marti@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
aerc.SelectedAccount() is used in lots of places. Most of them without
checking the return value.
In some cases, the currently selected tab is not related to any account
(widget.Terminal for example). This can lead to unexpected crashes when
accessing account specific configuration.
When possible, return an error when no account is currently selected.
If no error can be returned, fallback to non-account specific
configuration.
Signed-off-by: Robin Jarry <robin@jarry.cc>
Reviewed-by: Koni Marti <koni.marti@gmail.com>
|
| |
|
|
|
|
|
|
|
|
|
| |
Provide an option to change the completion style when selecting a folder
from completing with folders starting with the input string to
completing with folders in which the input string is a substring present
at any point in the folder name.
References: https://lists.sr.ht/~sircmpwn/aerc/%3C20201129181020.186984-1-inwit%40sindominio.net%3E
References: https://lists.sr.ht/~sircmpwn/aerc/%3C20210223202536.199355-1-clayton%40craftyguy.net%3E
Signed-off-by: inwit <inwit@sindominio.net>
|
| |
|
|
|
|
|
| |
I'm not sure what are the implications but it seems required.
Link: https://github.com/golang/go/issues/20883
Signed-off-by: Robin Jarry <robin@jarry.cc>
|
| | |
|
| | |
|
| |
|
|
|
|
| |
Before, lower_only was not being correctly set and was only considering
whether the string ended with a lowercase sequence. Refactored this with
some more explicit functions as the logic is a little confusing.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
This adds tab completion to textinput components. They can be configured
with a completion function. This function is called when the user
presses <tab>. The first completion is initially shown to the user
inserted into the text. Repeated presses of <tab> or <backtab> cycle
through the completions list. The completions list is invalidated when
any other non-tab-like key is pressed.
Also changed is some logic for current completion generation so that
all available commands are returned when <tab> is pressed with no
current text and similarly for arguments of commands.
|
| |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
This patch sets up the trigger config section of aerc.conf.
Each trigger has its own function which is called from the place where
it is triggered. Currently only the new-email trigger is implemented.
The triggers make use of format strings. For instance, in the new-email
trigger this allows the user to select the trigger command and also the
information extracted from the command and placed into their command.
To actually execute the trigger commands the keypresses are simulated.
Further triggers can be implemented in the future.
Formatting of the command is moved to a new package.
|
| | |
|
| |
|
|
|
| |
Credit for this fix goes to Reto; I guess if we're not gonna be mutt
we should probabaly do things correctly.
|
| |
|
|
|
| |
Tab completion currently only works on commands. Contextual completion
will be added in the future.
|
| | |
|
| | |
|
| | |
|
| | |
|
| |
|
