A lightweight query string parser/stringifier with support for nesting and some configurability.
Built on top of fast-querystring.
npm i -S picoquery
2.x
and above are ESM only (i.e. your project will need type: "module"
in the package.json
).
If you cannot yet move to ESM, you may continue to use 1.x
which will still
receive non-breaking changes, backported from main
.
Parsing a query string:
import {parse} from 'picoquery';
parse('foo.bar=abc&baz=def');
/*
{
foo: {
bar: 'abc'
},
baz: 'def'
}
*/
Stringifying an object:
import {stringify} from 'picoquery';
stringify({
foo: {
bar: 123
}
});
/*
foo.bar=123
*/
Converts the given object into a query string, optionally with configured options.
Parses the given query string into an object, optionally with configured options.
The default options are as follows:
{
nesting: true,
nestingSyntax: 'dot',
arrayRepeat: false,
arrayRepeatSyntax: 'repeat',
delimiter: '&'
}
When true, nested objects are supported.
For example, when parsing:
parse('foo.bar=baz', {nesting: true});
// {foo: {bar: 'baz'}}
When stringifying:
stringify({foo: {bar: 'baz'}}, {nesting: true});
// foo.bar=baz
This also results in arrays being supported:
parse('foo.0=bar', {nesting: true});
// {foo: ['bar']}
stringify({foo: ['bar']}, {nesting: true});
// foo.0=bar
Sets which style of nesting syntax should be used. The choices are:
dot
(e.g.foo.bar=baz
)index
(e.g.foo[bar]=baz
)js
(e.g.foo.bar[0]=baz
, i.e. arrays are indexed and properties are dotted)
If true
, this will treat repeated keys as arrays.
For example:
parse('foo=x&foo=y', {arrayRepeat: true});
// {foo: ['x', 'y']}
stringify({foo: ['x', 'y']}, {arrayRepeat: true});
// foo=x&foo=y
Sets which style of array repetition syntax should be used. The choices are:
bracket
(e.g.foo[]=x&foo[]=y
)repeat
(e.g.foo=x&foo=y
)
Sets the delimiter to be used instead of &
.
For example:
parse('foo=x;bar=y', {delimiter: ';'});
// {foo: 'x', bar: 'y'}
stringify({foo: 'x', bar: 'y'}, {delimiter: ';'});
// foo=x;bar=y
Can be set to a function which will be used to deserialize each value during parsing.
It will be called with the value
and the already deserialized
key
(i.e. (value: string, key: PropertyKey) => *
).
For example:
parse('foo=300', {
valueDeserializer: (value) => {
const asNum = Number(value);
return Number.isNaN(asNum) ? value : asNum;
}
});
// {foo: 300}
Can be set to a function which will be used to deserialize each key during parsing.
It will be called with the key
from the query string
(i.e. (key) => PropertyKey
).
For example:
parse('300=foo', {
keyDeserializer: (key) => {
const asNum = Number(key);
return Number.isNaN(asNum) ? key : asNum;
}
});
// {300: 'foo'}
Can be set to a function which determines if an object-like value should be serialized instead of being treated as a nested object.
All non-object primitives will always be serialized.
For example:
// Assuming `StringifableObject` returns its constructor value when `toString`
// is called.
stringify({
foo: new StringifiableObject('test')
}, {
shouldSerializeObject(val) {
return val instanceof StringifableObject;
},
valueSerializer: (value) => {
return String(value);
}
});
// foo=test
If you want to fall back to the default logic, you can import the default function:
import {defaultShouldSerializeObject, stringify} from 'picoquery';
stringify({
foo: new StringifiableObject('test')
}, {
shouldSerializeObject(val) {
if (val instanceof StringifableObject) {
return true;
}
return defaultShouldSerializeObject(val);
}
});
Can be set to a function which will be used to serialize each value during stringifying.
It will be called with the value
and the key
(i.e. (value: unknown, key: PropertyKey) => string
).
For example:
stringify({foo: 'bar'}, {
valueSerializer: (val) => String(val) + String(val)
});
// foo=barbar
Note that you can import the default serializer if you only want to handle some cases.
For example:
import {defaultValueSerializer, stringify} from 'picoquery';
stringify({foo: 'bar'}, {
valueSerializer: (val, key) => {
if (val instanceof Date) {
return val.toISOString();
}
// Call the original serializer otherwise
return defaultValueSerializer(val, key);
}
});
IMPORTANT: there are a few things to take into account with these benchmarks:
fast-querystring
is not capable of parsing or stringifying nested objects, so the results are incomparible (but here as reference)- all libraries have their own level of configurability. It will be possible to increase perf in each of them by disabling various features, but these are just the 'happy path'
Benchmark: Basic (no nesting)
┌─────────┬─────────────────────────────────┬─────────────┬────────────────────┬──────────┬─────────┐
│ (index) │ Task Name │ ops/sec │ Average Time (ns) │ Margin │ Samples │
├─────────┼─────────────────────────────────┼─────────────┼────────────────────┼──────────┼─────────┤
│ 0 │ 'picoquery' │ '2,393,539' │ 417.79130492907393 │ '±0.42%' │ 1196770 │
│ 1 │ 'qs' │ '382,933' │ 2611.4212945313157 │ '±1.04%' │ 191467 │
│ 2 │ 'fast-querystring (no nesting)' │ '2,633,148' │ 379.77342802356833 │ '±1.58%' │ 1316575 │
└─────────┴─────────────────────────────────┴─────────────┴────────────────────┴──────────┴─────────┘
Benchmark: Dot-syntax nesting
┌─────────┬─────────────────────────────────┬─────────────┬───────────────────┬──────────┬─────────┐
│ (index) │ Task Name │ ops/sec │ Average Time (ns) │ Margin │ Samples │
├─────────┼─────────────────────────────────┼─────────────┼───────────────────┼──────────┼─────────┤
│ 0 │ 'picoquery' │ '1,406,039' │ 711.2176054736389 │ '±0.56%' │ 703020 │
│ 1 │ 'qs' │ '205,611' │ 4863.536155478235 │ '±0.98%' │ 102806 │
│ 2 │ 'fast-querystring (no nesting)' │ '2,588,560' │ 386.3151031345139 │ '±0.66%' │ 1294281 │
└─────────┴─────────────────────────────────┴─────────────┴───────────────────┴──────────┴─────────┘
Benchmark: Basic (no nesting)
┌─────────┬─────────────────────────────────┬─────────────┬────────────────────┬──────────┬─────────┐
│ (index) │ Task Name │ ops/sec │ Average Time (ns) │ Margin │ Samples │
├─────────┼─────────────────────────────────┼─────────────┼────────────────────┼──────────┼─────────┤
│ 0 │ 'picoquery' │ '4,163,092' │ 240.20605684139227 │ '±0.16%' │ 2081547 │
│ 1 │ 'qs' │ '843,630' │ 1185.352608720127 │ '±0.76%' │ 421816 │
│ 2 │ 'fast-querystring (no nesting)' │ '3,795,565' │ 263.46536774751036 │ '±0.24%' │ 1897783 │
└─────────┴─────────────────────────────────┴─────────────┴────────────────────┴──────────┴─────────┘
Benchmark: Dot-syntax nesting
┌─────────┬─────────────────────────────────┬─────────────┬────────────────────┬──────────┬─────────┐
│ (index) │ Task Name │ ops/sec │ Average Time (ns) │ Margin │ Samples │
├─────────┼─────────────────────────────────┼─────────────┼────────────────────┼──────────┼─────────┤
│ 0 │ 'picoquery' │ '1,768,770' │ 565.3644818387182 │ '±1.08%' │ 884387 │
│ 1 │ 'qs' │ '332,254' │ 3009.743667534287 │ '±1.38%' │ 166128 │
│ 2 │ 'fast-querystring (no nesting)' │ '7,227,031' │ 138.36940410118828 │ '±1.15%' │ 3613517 │
└─────────┴─────────────────────────────────┴─────────────┴────────────────────┴──────────┴─────────┘
MIT