README
get-wild
- NAME
- FEATURES
- INSTALLATION
- SYNOPSIS
- DESCRIPTION
- TYPES
- EXPORTS
- OPTIONS
- DEVELOPMENT
- COMPATIBILITY
- SEE ALSO
- VERSION
- AUTHOR
- COPYRIGHT AND LICENSE
NAME
get-wild - extract nested properties from an object with support for wildcards
FEATURES
- configurable wildcard support, e.g.
"foo.*.bar"
- support for negative array indices, e.g.
"foo[-1].bar"
- pluggable path parser
- no dependencies
- ~750 B minified + gzipped
- curried (data last) versions for functional programming
- fully typed (TypeScript)
- CDN builds (UMD) - jsDelivr, unpkg
INSTALLATION
$ npm install get-wild
SYNOPSIS
import { get } from 'get-wild'
const obj = { foo: { bar: { baz: 'quux' } } }
get(obj, 'foo.bar.baz') // "quux"
get(obj, 'foo.fizz.buzz') // undefined
get(obj, 'foo.fizz.buzz', 42) // 42
get(obj, ['foo', 'bar', 'baz']) // "quux"
get(obj, ['foo', 'fizz', 'buzz'], 42) // 42
Negative array indices
const array = [
[{ value: 1 }, { value: 2 }, { value: 3 }],
[{ value: 4 }, { value: 5 }, { value: 6 }],
[{ value: 7 }, { value: 8 }, { value: 9 }],
]
get(array, '[1][-2].value') // 5
get(array, [-1, -1, 'value']) // 9
Wildcards
const data = {
users: {
'abc123': {
name: 'John Doe',
homepage: 'https://example.com/john-doe',
hobbies: ['eating', 'sleeping'],
},
'def345': {
name: 'Jane Doe',
homepage: 'https://example.com/jane-doe',
},
'ghi567': {
name: 'Nemo',
hobbies: ['singing', 'dancing'],
}
}
}
get(data, 'users.*.name')
// ["John Doe", "Jane Doe", "Nemo"]
get(data, 'users.*.homepage')
// ["https://example.com/john-doe", "https://example.com/jane-doe", undefined]
// also works with arrays
get(array, '1.*.value') // [4, 5, 6]
get(array, '[-1].*.value') // [7, 8, 9]
Flatten results
get(data, 'users.*.hobbies')
// ["eating", "sleeping", undefined, "singing", "dancing"]
Remove missing results
get(data, 'users.*.hobbies', [])
// ["eating", "sleeping", "singing", "dancing"]
Raw results
get(data, 'users.**.hobbies')
// [["eating", "sleeping"], undefined, ["singing", "dancing"]]
DESCRIPTION
This module exports a function which can be used to extract nested properties
from an object, including arrays and any other non-falsey values. This is
similar to the get
function provided by Lodash (and many other libraries),
but it adds the following features:
- wildcard support, e.g.
"foo.*.bar.*.baz"
- support for negative array indices, e.g.
"foo[-1][-2]"
In addition to the default get
implementation, get-wild exports a
builder function (getter
) which can be used to create a custom
get
with different options baked in, as well as curried
(data last) versions of get
and getter
to
facilitate functional composition and reuse.
Why?
I needed a small, dependency-free version of Lodash.get
with wildcard
support, and preferably with an option to filter out/exclude missing values.
Although there are a huge number of get
implementations on NPM, I could only
find one or two with wildcard support and none that are
standalone/dependency-free.
Why not?
If you don't need support for wildcards, negative array-indices, or other options, there are smaller implementations, e.g. dlv.
TYPES
The following types are referenced in the descriptions below.
type Options = {
collect?: (value: {}) => Array<any>;
default?: any;
flatMap?: PropertyKey | false;
map?: PropertyKey | false;
parser?: Options['split'];
split?: string | Parser;
};
type Parser = (path: string) => Array<PropertyKey>;
type Path = PropertyKey | Array<PropertyKey>;
EXPORTS
get-wild
get
- Type:
(obj: any, path: Path, $default?: any) => any
import { get } from 'get-wild'
get(obj, 'foo.*.bar', [])
Takes an object, a path and an optional default value, and returns the value(s) found in the object at the specified path, or the default value (which is undefined by default) if the path doesn't exist or the value is undefined. The path can be supplied as a dotted expression (string), symbol or number, or an array of strings, symbols or numbers.
The syntax of dotted path expressions mostly matches that of regular JavaScript path expressions, with a few additions.
If there are no steps in the path, the object itself is returned (or the default value if the object is undefined).
Wildcard matching is performed by collect
ing an array of values
at the wildcard's location and recursively get
ting the remainder of
the path from each value. Wildcards can be used at any locations in a path to
turn a single lookup into an array of lookup results for values at that
location.
The values returned by wildcard matches can be customized. By default, *
flattens the results (using flatMap
), while **
uses
map
, which returns the results verbatim, though this mapping can be
configured (or disabled) via the map
and flatMap
options.
The get
export is generated by a builder function, getter
, which
can be used to create a custom get
function with different options.
getter
- Type:
(options?: Options) => (obj: any, path: Path, $default?: any) => any
import { getter } from 'get-wild'
const get = getter({ split: '.' })
const obj = { '': { '': 42 } }
get(obj, '.') // 42
getter
is a function which is used to build get
functions. The default
get
export is generated by calling getter
with no arguments, which uses the
following default options:
{
collect: Object.values,
default: undefined,
flatMap: '*',
map: '**',
split: defaultParser,
}
The behavior of get
can be configured by generating a custom version which
overrides these defaults, e.g.:
import { getter } from 'get-wild'
const split = path => path.split('/')
const get = getter({ split })
get(obj, 'foo/bar/*/quux')
split
- Type:
Parser
- Aliases: parse, parser
import { getter, split } from 'get-wild'
import memoize from '@example/memoizer'
const memoized = memoize(split)
const get = getter({ split: memoized })
get(obj, '...')
The default function used by get
to turn a path expression (string)
into an array of steps (strings, symbols or numbers).
The array is not mutated, so, e.g., the function can be memoized to avoid re-parsing long/frequently-used paths. Alternatively, the path can be pre-parsed into an array (this is done automatically if the curried versions are used).
Syntax
The parser supports an extended version of JavaScript's native path syntax, e.g. the path:
`a[-1].b[42]["c.d"].e['f g'].*.h["i \\"j\\" k"][""]`
is parsed into the following steps:
["a", -1, "b", 42, "c.d", "e", "f g", "*", "h", 'i "j" k', ""]
Property names are either unquoted (strings), bracketed integers, or bracketed
single or double-quoted strings. Unquoted names can contain any characters
apart from spaces (\s
), "
, '
, `
, [
, ]
, .
or \
.
Unquoted property names must be preceded by a dot unless the name is at the start of the path, in which case the dot must be omitted. Bracketed values must not be preceded by a dot.
If the path is an empty string, an empty array is returned.
get-wild/fp
get
- Type:
(path: Path, $default?: any) => (obj: any, $default?: any) => any
import { get } from 'get-wild/fp'
import R from 'ramda'
const followers = get('followers.*.name', [])
followers(user) // get(user, path, [])
followers(user, '<anon>') // get(user, path, "<anon>")
const allFollowers = users.flatMap(R.unary(followers))
A curried version of get
which takes a path and optional default
value and returns a function which takes an object and an optional default
value and returns the value(s) located at the path.
Note that, when the generated function is passed to map
, the arity needs to
be fixed to avoid misinterpreting map
's second argument (the index) as a
default value. To avoid this, the pluck
variant can be used, which
does this automatically.
getter
- Type:
(options?: Options) => (path: Path, $default?: any) => (obj: any, $default?: any) => any
import { getter } from 'get-wild/fp'
import R from 'ramda'
const get = getter({ default: [], split: '.' })
const followers = get('followers.*.name')
followers(user) // get(user, path, [])
followers(user, '<anon>') // get(user, path, "<anon>")
const allFollowers = users.flatMap(R.unary(followers))
A variant of getter
which takes an optional Options
object and returns a curried version of get
with the options baked
in.
Note that, when the generated function is passed to map
, the arity needs to
be fixed to avoid misinterpreting map
's second argument (the index) as a
default value. To avoid this, the plucker
variant can be used,
which does this automatically.
pluck
- Type:
(path: Path, $default?: any) => (obj: any) => any
import { pluck } from 'get-wild/fp'
const followers = pluck('followers.*.name', [])
followers(user) // get(user, path, [])
followers(user, '<anon>') // get(user, path, [])
const allFollowers = users.flatMap(followers)
A variant of the curried version of get
without support for the
optional default-value override.
This is useful in situations where additional arguments may be misinterpreted
or forbidden, e.g. when the generated function is passed to map
.
plucker
- Type:
(options?: Options) => (path: Path, $default?: any) => (obj: any) => any
import { plucker } from 'get-wild/fp'
const pluck = plucker({ default: [], split: '.' })
const followers = pluck('followers.*.name')
followers(user) // get(user, path, [])
followers(user, '<anon>') // get(user, path, [])
const allFollowers = users.flatMap(followers)
A variant of getter
which returns a version of pluck
with the options baked in.
This is useful in situations where additional arguments may be misinterpreted
or forbidden, e.g. when the generated function is passed to map
.
OPTIONS
collect
- Type:
(value: {}) => Array<any>
- Default:
Object.values
import { getter } from 'get-wild'
const collect = value => {
return (value instanceof Map || value instanceof Set)
? Array.from(value.values())
: Object.values(value)
}
const map = new Map([
[1, { value: 'foo' }],
[2, { value: 'bar' }],
[3, { value: 'baz' }],
[4, { value: 'quux' }],
])
const obj = { map }
const get = getter({ collect })
get(obj, 'map.*.value') // ["foo", "bar", "baz", "quux"]
The collect
function is used to convert a value under a wildcard token into
an array of values. If not supplied, it defaults to
Object.values
, which works with objects, arrays, and other
non-nullish values. It can be overridden to add support for traversable values
that aren't plain objects, e.g. ES6 Map and Set instances.
Note that the value passed to collect
is not falsey and not an array, as both
are handled without calling collect
.
default
- Type:
any
- Default:
undefined
const get = getter({ default: [] })
get(data, 'users.*.hobbies')
// ["eating", "sleeping", "singing", "dancing"]
get(data, 'users.*.hobbies', [])
// ["eating", "sleeping", "singing", "dancing"]
get(data, 'users.*.hobbies', undefined)
// ["eating", "sleeping", undefined, "singing", "dancing"]
This option allows the default value to be baked into a get
function. If no
third argument is passed to the function, this value is returned for
missing/undefined properties. Otherwise, the supplied value is returned.
flatMap
- Type:
PropertyKey | false
- Default:
"*"
import { getter } from 'get-wild'
const get = getter({ flatMap: '**', map: '*' })
get(obj, 'foo.**.bar.baz')
The token used to map values at the specified location and flatten the results.
If an empty array is supplied as the third argument to get
,
missing/undefined values are removed from the result.
If set to false, wildcard matching with flatMap
is disabled and the token is
treated as a regular property name.
Usage
Wildcard matching with flatMap
behaves in a similar way to basic
directory/filename matching with globs (minus the pattern matching). The
selected properties (at the end of the path) are returned as direct children of
the resulting array (wildcard matches always return an array), either as
matched results or as default values if there's no match.
For example, with the default mapping, a path such as
accounts.active.*.followers.*.name
, which extracts the names of all followers
of active accounts, would return an array of account names interspersed with
default values where an account doesn't have any followers (or if a follower's
name is undefined), e.g.:
get(data, 'accounts.active.*.followers.*.name')
// ["john", "paul", undefined, "george", undefined, "ringo"]
This can be reduced to just the names by setting the default value to an empty array, e.g.:
get(data, 'accounts.active.*.followers.*.name', [])
// ["john", "paul", "george", "ringo"]
Syntax
Note that with the default parser, the token must be a syntactically-valid name, e.g. this doesn't work:
import { getter } from 'get-wild'
const obj = { foo: [{ bar: 1 }, { bar: 2 }] }
const get = getter({ flatMap: '[]' })
get(obj, 'foo.[].bar') // SyntaxError: Invalid step @ 3: "foo.[].bar"
If a custom parser is supplied, any token can be used:
const get = getter({ flatMap: '[]', split: '.' })
get(obj, 'foo.[].bar') // [1, 2]
map
- Type:
PropertyKey | false
- Default:
"**"
import { getter } from 'get-wild'
const get = getter({ map: '*', flatMap: '**' })
get(obj, 'foo.*.bar.baz')
The token used to map values at the specified location without flattening the results.
Matching with map
selects the same values as flatMap
, but they remain
nested inside arrays, with each enclosing map
in the path adding another
layer of wrapping.
If set to false, wildcard matching with map
is disabled and the token is
treated as a regular property name.
split
- Type:
string | Parser
- Default:
split
- Alias: parser
import { getter } from 'get-wild'
const split = path => path.split('.')
const get = getter({ split })
const obj = { '': { '': 42 } }
get(obj, '.') // 42
A function which takes a path expression (string) and parses it into an array of property names (strings, symbols or numbers). If not supplied, a default parser which supports an extended version of JavaScript's native path syntax is used.
As a shortcut, if the value is a string, a function which splits the path on that string is used, i.e. the following are equivalent:
const split = path => path.split('.')
const get = getter({ split })
const get = getter({ split: '.' })
DEVELOPMENT
NPM Scripts
The following NPM scripts are available:
- build - compile the library for testing and save to the target directory
- build:doc - generate the README's TOC (table of contents)
- build:release - compile the library for release and save to the target directory
- clean - remove the target directory and its contents
- rebuild - clean the target directory and recompile the library
- repl - launch a node REPL with the library loaded
- test - recompile the library and run the test suite
- test:run - run the test suite
- typecheck - sanity check the library's type definitions
COMPATIBILITY
- Environments with support for ES6 and
Array#flat
/Array#flatMap
(polyfill)
SEE ALSO
VERSION
1.5.0
AUTHOR
COPYRIGHT AND LICENSE
Copyright © 2020-2021 by chocolateboy.
This is free software; you can redistribute it and/or modify it under the terms of the MIT license.