deprecation notice for original
The original author's
nomnom
is deprecated; this fork is not: we continue to update this library.When you are unsure about using this library's fork (or others), you may wish to check out https://github.com/tj/commander.js.
Personally, I prefer using
nomnom
overcommander
or other option parsers, so I will keep this one. :-)
nomnom
nomnom is an option parser for node. It noms your args and gives them back to you in a hash.
var opts = require("@gerhobbelt/nomnom")
.option('debug', {
abbr: 'd',
flag: true,
help: 'Print debugging info'
})
.option('config', {
abbr: 'c',
default: 'config.json',
help: 'JSON file with tests to run'
})
.option('version', {
flag: true,
help: 'print version and exit',
callback: function() {
return "version 1.2.4";
}
})
.parse();
if (opts.debug)
// do stuff
You don't have to specify anything if you don't want to:
var opts = require("@gerhobbelt/nomnom").parse();
var url = opts[0]; // get the first positional arg
var file = opts.file // see if --file was specified
var verbose = opts.v // see if -v was specified
var extras = opts._ // get an array of the unmatched, positional args
Install
npm install @gerhobbelt/nomnom
More Details
Nomnom supports args like -d
, --debug
, --no-debug
, --file=test.txt
, --file test.txt
, -f test.txt
, -xvf
, and positionals. Positionals are arguments that don't fit the -a
or --atomic
format and aren't attached to an option.
Values are JSON parsed, so --debug=true --count=3 --file=log.txt
would give you:
{
"debug": true,
"count": 3,
"file": "log.txt"
}
Commands
Nomnom supports command-based interfaces (e.g. with git: git add -p
and git rebase -i
where add
and rebase
are the commands):
var parser = require("@gerhobbelt/nomnom");
parser.command('browser')
.callback(function(opts, command) {
runBrowser(opts.url);
})
.help("run browser tests");
parser.command('sanity')
.option('outfile', {
abbr: 'o',
help: "file to write results to"
})
.option('config', {
abbr: 'c',
default: 'config.json',
help: "json manifest of tests to run"
})
.callback(function(opts, command) {
runSanity(opts.filename);
})
.help("run the sanity tests")
parser.parse();
Each command generates its own usage message when -h
or --help
is specified with the command.
Usage
Nomnom prints out a usage message if --help
or -h
is an argument. If no commands are called (the script is simply run without arguments), -h
will be called automatically unless you've disabled this feature using the .autoShowUsage(false)
API call.
Usage for these options in test.js
:
var opts = require("@gerhobbelt/nomnom")
.script("runtests")
.options({
path: {
position: 0,
help: "Test file to run",
list: true
},
config: {
abbr: 'c',
metavar: 'FILE',
help: "Config file with tests to run"
},
debug: {
abbr: 'd',
flag: true,
help: "Print debugging info"
}
}).parse();
...would look like this:
usage: runtests <path>... [options]
path Test file to run
options:
-c FILE, --config FILE Config file with tests to run
-d, --debug Print debugging info
Options
You can either add a specification for an option with nomnom.option('name', spec)
or pass the specifications to nomnom.options()
as a hash keyed on option name. Each option specification can have the following fields:
abbr and full
abbr
is the single character string to match to this option, full
is the full-length string (defaults to the name of the option).
This option matches -d
and --debug
on the command line:
nomnom.option('debug', {
abbr: 'd'
})
This option matches -n 3
, --num-lines 12
on the command line:
nomnom.option('numLines', {
abbr: 'n',
full: 'num-lines'
})
flag
If this is set to true, the option acts as a flag and doesn't swallow the next value on the command line. Default is false
, so normally if you had a command line --config test.js
, config
would get a value of test.js
in the options hash. Whereas if you specify:
nomnom.option('config', {
flag: true
})
config
would get a value of true
in the options hash, and test.js
would be a free positional arg.
metavar
metavar
is used in the usage printout e.g. "PATH"
in "-f PATH, --file PATH"
.
string
A shorthand for abbr
, full
, and metavar
. For example, to attach an option to -c
and --config
use a string: "-c FILE, --config=FILE"
help
A string description of the option for the usage printout.
default
The value to give the option if it's not specified in the arguments.
type
If you don't want the option JSON-parsed, specify type "string"
.
callback
A callback that will be executed as soon as the option is encountered. If the callback returns a string it will print the string and exit:
nomnom.option('count', {
callback: function(count) {
if (count != parseInt(count)) {
return "count must be an integer";
}
}
})
position
The position of the option if it's a positional argument. If the option should be matched to the first positional arg use position 0
, etc.
list
Specifies that the option is a list. Appending can be achieved by specifying the arg more than once on the command line:
node test.js --file=test1.js --file=test2.js
If the option has a position
and list
is true
, all positional args including and after position
will be appended to the array.
required
If this is set to true
and the option isn't in the args, a message will be printed and the program will exit.
choices
A list of the possible values for the option (e.g. ['run', 'test', 'open']
). If the parsed value isn't in the list a message will be printed and the program will exit.
transform
A function that takes the value of the option as entered and returns a new value that will be seen as the value of the option.
nomnom.option('date', {
abbr: 'd',
transform: function(timestamp) {
return new Date(timestamp);
}
})
hidden
Option won't be printed in the usage
Parser interface
require("nomnom")
will give you the option parser. You can also make an instance of a parser with require("nomnom")()
. You can chain any of these functions off of a parser:
option
Add an option specification with the given name:
nomnom.option('debug', {
abbr: 'd',
flag: true,
help: "Print debugging info"
})
options
Add options as a hash keyed by option name, good for a cli with tons of options like this example:
nomnom.options({
debug: {
abbr: 'd',
flag: true,
help: "Print debugging info"
},
fruit: {
help: "Fruit to buy"
}
})
usage
The string that will override the default generated usage message.
help
A string that is appended to the usage.
autoShowUsage
The enable
argument determines if nomnom will automatically print the usage when no commandline argument has been specified at all; the default is true
i.e. the feature is enabled. Pass false
to disable this feature.
getUsage
Return the generated usage message. When .command(...)
has been called before, the usage message for that last 'active' command will be produced.
script
Nomnom can't detect the alias used to run your script. You can use script
to provide the correct name for the usage printout instead of e.g. node test.js
.
printer
Overrides the usage printing function.
command
Takes a command name and gives you a command object on which you can chain command options.
When you provide a callback function via the .callback()
API, then that callback will be invoked when the commandline has been parsed and the given command has been found there.
See .callback()
for details.
When you did not provide a name argument or a null/undefined name, this API is equivalent to invoking .nocommand()
.
nocommand
Gives a command object that will be used when no command is called.
When you provide a callback function via the .callback()
API, then that callback will be invoked when the commandline has been parsed and
either no command has been found there
or a command was found for whicch no callback has been specified.
See .callback()
for details.
nocolors
Disables coloring of the usage message.
setColors
Override the help/report formatting object with your own to provide custom coloring schemes. It should provide these functions:
{
usageHeadingColor: f1,
usageStringColor: f2,
positionalHelpColor: f3,
optionsHeaderColor: f4,
helpColor: f5,
requiredArgColor: f6
};
where f1
through f6
are function references for functions which receive a single string and should return a (processed/transformed) string including the desired coloring console escape sequences.
parse
Parses node's process.argv
and returns the parsed options hash. You can also provide argv:
var opts = nomnom.parse(["-xvf", "--atomic=true"])
nom
The same as parse()
.
Command interface
A command is specified with nomnom.command('name')
. All these functions can be chained on a command:
option
Add an option specifically for this command.
options
Add options for this command as a hash of options keyed by name.
callback
A callback that will be called with the parsed options when the command is used. The arguments passed to the callback are a filled options
object and the command name.
When the .command()
section of the setup code does not specify a callback, then the "fallback callback" will be invoked instead. This callback is the one defined in the .nocommand()
a.k.a. empty .command()
(without any name
argument). Then the command
name parameter passed to the provided callback will be null
.
help
A help string describing the function of this command.
usage
Override the default generated usage string for this command.