Option parser with generated usage and commands
npm install @gerhobbelt/nomnom>
> # 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 over commander or other option parsers, so I will keep this one. :-)
>
javascript
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:
`javascript
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
for node.js and npm:
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):
`javascript
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:
`javascript
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 ... [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:
`javascript
nomnom.option('debug', {
abbr: 'd'
})
`
This option matches -n 3, --num-lines 12 on the command line:
`javascript
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:
`javascript
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:
`javascript
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.
`javascript
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:
`javascript
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:
`javascript
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:
`javascript
{
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:
`javascript
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`.