App::Sqitch::Command - Sqitch Command support
my $cmd = App::Sqitch::Command->load( deploy => \%params ); $cmd->run;
App::Sqitch::Command is the base class for all Sqitch commands.
ENGINES
Returns the list of supported engines, currently:
firebird
mysql
oracle
pg
sqlite
vertica
exasol
options
my @spec = App::Sqitch::Command->options;
Returns a list of Getopt::Long options specifications. When load loads the class, any options passed to the command will be parsed using these values. The keys in the resulting hash will be the first part of each option, with dashes converted to underscores. This hash will be passed to configure along with a App::Sqitch::Config object for munging into parameters to be passed to the constructor.
load
configure
Here's an example excerpted from the config command:
config
sub options { return qw( get unset list global system config-file=s ); }
This will result in hash keys with the same names as each option except for config-file=s, which will be named config_file.
config-file=s
config_file
my $params = App::Sqitch::Command->configure($config, $options);
Takes two arguments, an App::Sqitch::Config object and the hash of command-line options as specified by options. The returned hash should be the result of munging these two objects into a hash reference of parameters to be passed to the command subclass constructor.
By default, this method converts dashes to underscores in command-line options keys, and then merges the configuration values with the options, with the command-line options taking priority. You may wish to override this method to do something different.
my $cmd = App::Sqitch::Command->load( \%params );
A factory method for instantiating Sqitch commands. It loads the subclass for the specified command, uses the options returned by options to parse command-line options, calls configure to merge configuration with the options, and finally calls new with the resulting hash. Supported parameters are:
new
sqitch
The App::Sqitch object driving the whole thing.
An App::Sqitch::Config representing the current application configuration state.
command
The name of the command to be executed.
args
An array reference of command-line arguments passed to the command.
my $cmd = App::Sqitch::Command->new(%params);
Instantiates and returns a App::Sqitch::Command object. This method is not designed to be overridden by subclasses; they should implement BUILDARGS or BUILD, instead.
BUILDARGS
BUILD
my $sqitch = $cmd->sqitch;
Returns the App::Sqitch object that instantiated the command. Commands may access its properties in order to manage global state.
These methods should be overridden by all subclasses.
execute
$cmd->execute;
Executes the command. This is the method that does the work of the command. Must be overridden in all subclasses. Dies if the method is not overridden for the object on which it is called, or if it is called against a base App::Sqitch::Command object.
my $command = $cmd->command;
The name of the command. Defaults to the last part of the package name, so as a rule you should not need to override it, since it is that string that Sqitch uses to find the command class.
These methods are mainly provided as utilities for the command subclasses to use.
default_target
my $target = $cmd->default_target;
This method returns the default target. It should only be used by commands that don't use a parse_args() to find and load a target.
parse_args()
This method should always return a target option, never undef. If the --engine option or core.engine configuration option has been set, then the target will support that engine. In the latter case, if engine.$engine.target is set, that value will be used. Otherwise, the returned target will have a URI of db: and no associated engine; the engine method will throw an exception. This behavior should be fine for commands that don't need to load the engine.
undef
--engine
core.engine
engine.$engine.target
db:
engine
parse_args
my ($name1, $name2, $targets, $changes) = $cmd->parse_args( names => \@names, target => $target_name, args => \@args );
Examines each argument to determine whether it's a known change spec or identifies a target. Unrecognized arguments will replace false values in the names array reference. Any remaining unknown arguments will trigger an error.
names
Returns a list consisting all the desired names, followed by an array reference of target objects and an array reference of change specs.
This method is useful for commands that take a number of arguments where the order may be mixed.
The supported parameters are:
An array reference of the command arguments.
target
The name of a target, if any. Useful for commands that offer their own --target option. This target will be the default target, and the first returned in the targets array.
--target
An array reference of names. If any is false, its place will be taken by an otherwise unrecognized argument. The number of values in this array reference determines the number of values returned as names in the return values. Such values may still be false or undefined; it's up to the caller to decide what to do about that.
all
In the event that no targets are recognized (or changes that implicitly recognize the default target), if this parameter is true, then all known targets from the configuration will be returned.
no_changes
If true, the parser will not check to see if any argument corresponds to a change. The last value returned will be undef instead of the usual array reference. Any argument that might have been recognized as a change will instead be included in either the targets array -- if it's recognized as a target -- or used to set names to return. Any remaining are considered unknown arguments and will result in an exception.
targets
no_default
If true, no default target will be returned, even if no other targets are found. See below for details.
If a target parameter is passed, it will always be instantiated and returned as the first item in the "target" array, and arguments recognized as changes in the plan associated with that target will be returned as changes.
If no target is passed or appears in the arguments, a default target will be instantiated based on the command-line options and configuration -- unless the no_default parameter is true. Unlike the target returned by default_target, this target must have an associated engine specified by the --engine option or configuration. This is on the assumption that it will be used by commands that require an engine to do their work. Of course, any changes must be recognized from the plan associated with this target.
Changes are only recognized if they're found in the plan of the target that precedes them. If no target precedes them, the target specified by the target parameter or the default target will be searched. Such changes can be specified in any way documented in sqitchchanges.
Targets may be recognized by any one of these types of arguments:
Target Name
Database URI
Engine Name
Plan File
In the case of plan files, parse_args() will return the first target it finds for that plan file, even if multiple targets use the same plan file. The order of precedence for this determination is the default project target, followed by named targets, then engine targets.
run
$cmd->run('echo hello');
Runs a system command and waits for it to finish. Throws an exception on error.
capture
my @files = $cmd->capture(qw(ls -lah));
Runs a system command and captures its output to STDOUT. Returns the output lines in list context and the concatenation of the lines in scalar context. Throws an exception on error.
STDOUT
probe
my $git_version = $cmd->capture(qw(git --version));
Like capture, but returns just the chomped first line of output.
chomp
verbosity
my $verbosity = $cmd->verbosity;
Returns the verbosity level.
trace
Send trace information to STDOUT if the verbosity level is 3 or higher. Trace messages will have trace: prefixed to every line. If it's lower than 3, nothing will be output.
trace:
debug
$cmd->debug('Found snuggle in the crib.');
Send debug information to STDOUT if the verbosity level is 2 or higher. Debug messages will have debug: prefixed to every line. If it's lower than 2, nothing will be output.
debug:
info
$cmd->info('Nothing to deploy (up-to-date)');
Send informational message to STDOUT if the verbosity level is 1 or higher, which, by default, it is. Should be used for normal messages the user would normally want to see. If verbosity is lower than 1, nothing will be output.
comment
$cmd->comment('On database flipr_test');
Send comments to STDOUT if the verbosity level is 1 or higher, which, by default, it is. Comments have # prefixed to every line. If verbosity is lower than 1, nothing will be output.
#
emit
$cmd->emit('core.editor=emacs');
Send a message to STDOUT, without regard to the verbosity. Should be used only if the user explicitly asks for output, such as for sqitch config --get core.editor.
sqitch config --get core.editor
vent
$cmd->vent('That was a misage.');
Send a message to STDERR, without regard to the verbosity. Should be used only for error messages to be printed before exiting with an error, such as when reverting failed changes.
STDERR
page
$sqitch->page('Search results:');
Like emit(), but sends the output to a pager handle rather than STDOUT. Unless there is no TTY (such as when output is being piped elsewhere), in which case it is sent to STDOUT. Meant to be used to send a lot of data to the user at once, such as when display the results of searching the event log:
emit()
$iter = $sqitch->engine->search_events; while ( my $change = $iter->() ) { $cmd->page(join ' - ', @{ $change }{ qw(change_id event change) }); }
warn
$cmd->warn('Could not find nerble; using nobble instead.');
Send a warning messages to STDERR. Warnings will have warning: prefixed to every line. Use if something unexpected happened but you can recover from it.
warning:
usage
$cmd->usage('Missing "value" argument');
Sends the specified message to STDERR, followed by the usage sections of the command's documentation. Those sections may be named "Name", "Synopsis", or "Options". Any or all of these will be shown. The doc used to display them will be the first found of:
sqitch-$command-usage
sqitch-$command
App::Sqitch::Command::$command
App::Sqitch::Command
For an ideal usage messages, sqitch-$command-usage.pod should be created by all command subclasses.
sqitch-$command-usage.pod
The Sqitch command-line client.
David E. Wheeler <david@justatheory.com>
Copyright (c) 2012-2018 iovation Inc.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
To install App::Sqitch, copy and paste the appropriate command in to your terminal.
cpanm
cpanm App::Sqitch
CPAN shell
perl -MCPAN -e shell install App::Sqitch
For more information on module installation, please visit the detailed CPAN module installation guide.