npm package discovery and stats viewer.

Discover Tips

  • General search

    [free text search, go nuts!]

  • Package details

    pkg:[package-name]

  • User packages

    @[username]

Sponsor

Optimize Toolset

I’ve always been into building performant and accessible sites, but lately I’ve been taking it extremely seriously. So much so that I’ve been building a tool to help me optimize and monitor the sites that I build to make sure that I’m making an attempt to offer the best experience to those who visit them. If you’re into performant, accessible and SEO friendly sites, you might like it too! You can check it out at Optimize Toolset.

About

Hi, 👋, I’m Ryan Hefner  and I built this site for me, and you! The goal of this site was to provide an easy way for me to check the stats on my npm packages, both for prioritizing issues and updates, and to give me a little kick in the pants to keep up on stuff.

As I was building it, I realized that I was actually using the tool to build the tool, and figured I might as well put this out there and hopefully others will find it to be a fast and useful way to search and browse npm packages as I have.

If you’re interested in other things I’m working on, follow me on Twitter or check out the open source projects I’ve been publishing on GitHub.

I am also working on a Twitter bot for this site to tweet the most popular, newest, random packages from npm. Please follow that account now and it will start sending out packages soon–ish.

Open Software & Tools

This site wouldn’t be possible without the immense generosity and tireless efforts from the people who make contributions to the world and share their work via open source initiatives. Thank you 🙏

© 2024 – Pkg Stats / Ryan Hefner

@sagold/json-query

v6.2.0

Published

json-pointer utilities for querying and transforming data

Downloads

268,292

Readme

install

yarn add @sagold/json-query

Features

  • json-pointer syntax #/list/0/id
  • glob-patterns for properties (*, **)
  • regex-support for properties {any.*}
  • pattern-support for inifinite recursion /tree(/nodes/*)+/value
  • or-patterns /node((/left), (/right))
  • finite search in circular-data **
  • lookahead-rules to test selected property ?property:value and regex values ?property:{\d+}
  • and typechecks /value?:array

Quick introduction

Basically, a query is a json-pointer, which describes a path of properties into the json-data

import { get } from "@sagold/json-query";
const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/object/a/id"); // ["id-a"]

But each property may also be a glob-pattern or a regular expression:

* selects all direct children

const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/object/*/id"); // ["id-a", "id-b"]

** selects all values

const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/object/**");
// [ { a: { id: "id-a" }, b: { id: "id-b" } }, { id: "id-a" }, "id-a", { id: "id-b" }, "id-b" ]

{} calls a regular expression

const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/{obj.*}/{.*}/id"); // ["id-a", "id-b"]

Note. Regular expressions within strings, have to escape any backslashes, e.g. instead of {\d} you need to pass {\\d}

lookahead rules are used to validate the current value based on its properties

?child tests if a childProperty is defined

const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/object/*?id"); // [{ id: "id-a" }, { id: "id-b" }]

?child:value tests if a childProperty matches a value

const input = { object: { a: { id: "id-a" }, b: { id: "id-b" } } };

const values = get(input, "/object/*?id:id-b"); // [{ id: "id-b" }]

lookahead rules can also be negated ?child:!value, tested by regex ?child:{^re+}, combined ?child&&other or joined ?child||other. Undefined may be tested with ?property:undefined, per default undefined is excluded from matches.

typechecks can be used to query certain data-types

?:<type>, where <type> may be any of ["boolean", "string", "number", "object", "array", "value"]

const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

const values = get(input, "/**?:string"); // ["id-b"]

?:value will match all types except objects and arrays

const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

const values = get(input, "/**?:value"); // [33, "id-b"]

patterns can be used to combine queries into a single result (OR) and to build up results from recursive queries (+)

Queries can be grouped by parenthesis, where /a/b/c = /a(/b)(/c) = /a(/b/c).

((/a), (/b)) resolves both queries on the previous result

const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

const values = get(input, "/object((/a), (/b))"); // [{ id: 33 }, { id: "id-b" }]

and the result may be queried further

get(input, "/object((/a), (/b))/id"); // [33, "id-b"]
get(input, "/object((/a), (/b))/id?:number"); // [33]

(/a)+ will repeat the grouped query for all possible results

const input = {
    id: 1,
    a: { // first iteration
        id: 2,
        a: { // second iteration
            id: 3
            a: 4 // last iteration
        }
    }
};

const values = get(input, "/(/a)+"); // [{ id: 2, a: { id: 3, a: 4 } }, { id: 3, a: 4 }, 4]

escaping properties In case you have special characters in property-names or values, you can escape any value using doubled-quotes "<value>":

  • escape property-name: '/root/*/"strange/property"' is split to ["root", "*", "strange/property"]
  • escape query-property '/root/*?"strange/property":42'
  • escape query-value '/root/*?id:"#/pointer/value"'

API

json-query exposes get, set, remove and a split-helper

method | signature | description --------|-------------------------------------------------------------------|------------------------------ get | (input:any, query:string, returnType?:string|function) | query data, returns results set | (input:any, query:string, value:string|function, replace?:string)| set value, returns modified input split | (query: string) | returns a list properties and queries remove | (input:any, query: string, returnRemoved?:boolean) | delete query targets, returns input

get

per default, get returns a list of all values

import { get } from "@sagold/json-query";
const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };
const values = get(input, "/**?:value"); // [33, "id-b"]

Using the optional value returnType you can change the result type to the following options ["all", "value", "pointer", "map"]. The string values can also be accessed as property on get: get.ALL, get.VALUE, get.POINTER, get.MAP:

returnType | description ------------|------------------------------------------------------------------ "value" | returns all matched values of the query [33, "id-b"] "pointer" | returns json-pointer to results ["#/object/a", "#/object/b"] "map" | returns an pairs of jsonPointer: resultValue as an object "all" | returns a list, where each result is an array of [value, keyToValue, parentObject, jsonPointer] function | callback with (value, keyToValue, parentObject, jsonPointer) => {}. If a value is returned, the result will be replaced by the return-value

import { get } from "@sagold/json-query";
const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

get(input, "/**?:value", get.VALUE); // [33, "id-b"]
get(input, "/**?:value", get.POINTER); // ["#/object/a/id", "#/object/b/id"]
get(input, "/**?:value", get.MAP); // { "#/object/a/id": 33, "#/object/b/id": "id-b" }

get(input, "/**?:value", get.ALL);
// [
//    [33, "id", { id: 33 }, "#/object/a/id"],
//    ["id-b", "id", { id: "id-b" }, "#/object/b/id"]
// ]

get(input, "/**?:value", (value, key, parent, pointer) => `custom-${pointer}`);
// ["custom-#/object/a/id", "custom-#/object/b/id"]

remove

remove deletes any match from the input data. Note: the input will be modified. If this is unwanted behaviour, copy your data up front.

import { remove } from "@sagold/json-query";
const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

remove(input, "/object/*/id"); // { object: { a: {}, b: {} } };

Per default, the input object is returned. Setting the optional argument returnRemoved = true, will return a list of the removed items

import { remove } from "@sagold/json-query";
const input = { object: { a: { id: 33 }, b: { id: "id-b" } } };

remove(input, "/object/*/id", true); // [ 33, "id-b" ]

set

set inserts given input-value on result and creates missing properties and arrays. Note: Any expanding queries like * or patterns will not create any intermediate values

set has the following signature

set(input:any, query:string, value:string\|function, force?:string): any

instead of value, you can also pass a function to generate the values to set:

value(pointerOfParent:string, lastPropertyName:string, parentObject:string, pointerAtValue:string): any

Create data from simple properties

import { set } from "@sagold/json-query";

const result = set({}, "/object/id", 42); // { object: { id: 42 }}

Add properties to multiple existing objects

import { set } from "@sagold/json-query";

const result = set({ list: [ { id: 1 }, { id: 2 } ] }, "/list/*/index", 42);
// { list: [ { id: 1, index: 42 }, { id: 2, index: 42 } ] }

Or using a value-function

import { set } from "@sagold/json-query";

const result = set({ list: [ { id: 1 }, { id: 2 } ] }, "/list/*/index",
    ( _, _, parent) => `id-${parent.id}`
);
// { list: [ { id: 1, index: "id-1" }, { id: 2, index: "id-2" } ] }

Currently, set will not override simple values

import { set } from "@sagold/json-query";

const result = set({ value: 2 }, "/value/id", 3);
// { value: 2 }

And queries will not add values to the data

import { set } from "@sagold/json-query";

const result = set({ a: { id: 2 } }, "((/a), (/b))/id", true);
// { a: { id: true } }

When working with arrays, you have to choose between the following actions

  • insert item at index 1: /list/[1]/id
  • replace item at index 1: /list/1/id
  • append item /list/[]/id

Using the force option, you can enforce insertion or replacement, independent of the syntax (same for the whole query)

set(data, "/list/[1]/id", 42, set.REPLACE_ITEMS); // will always replace index
// and
set(data, "/list/1/id", 42, set.INSERT_ITEMS); // will always insert at index

Numbers will always be interpreted as arrays

set({}, "/list/0/id", 42); // { list: [{ id: 42 }]}
set({}, "/list/[]/id", 42); // { list: [{ id: 42 }]}
set({}, "/list/[0]/id", 42); // { list: [{ id: 42 }]}

// but setting an index is respected
set({}, "/list/2/id", 42); // { list: [undefined, undefined, { id: 42 }]}

In order to treat numbers as objects, escape them using double-quotes

set({}, '/list/"2"/id', 42); // { list: { 2: { id: 42 } } }
// or "/list/\"2\"/id"

About patterns

Pattern-queries enable selection of recursive patterns and offer a way to build up a collection of data for further filterung. A pattern uses brackets () to identify repeatable structures and offers multiple selections for the same data-entry.

Using a pattern-query like #/tree((/left),(/right))* will recursively select all left and right-nodes. e.g.

const data = {
  tree: {
    left: {
      id: "1",
      left: { id: "2" },
      right: { id: "3" }
    },
    right: {
      id: "4"
    }
  }
};

const result = get(data, "#/tree((/left),(/right))*/id");
// ["1", "2", "3", "4"]

Note that each pattern-queries is resovled using query.get and thus supports all mentioned features.

One use-case for pattern-queries can be found in json-schema specification. Any definition in #/defs may reference itself or be referenced circular. A linear query cannot describe the corresponding data, but pattern-queries might be sufficient.

details

A pattern is a simple group defined by brackets: #/a(/b)/c, which is identical to #/a/b/c. But a group may also have a quantifier +: #/a(/b)+/c. Using a quantifier, the query within the pattern will be applied as long as it matches any data. Its combined result will then be passed to /c.

e.g. applying the pattern #/a(/b)+/c on the following input data:

const input = {
  a: {
    b: {
      c: "1",
      b: {
        c: "2",
        b: {}
      }
    }
  }
};

will first select property a and then repeatedly select property b: [a/b, a/b/b, a/b/b/b]. This result is filtered by c, which will return ["1", "2"] (the last b-object has no property c).

Patterns can also be used for OR-operations. An OR is identified by a semicolon , and must be within and between patterns, like ((/a/b),(/c)). Not valid patterns are (/a/b, /c) and r/(/a/b),(/c)/f.

Currently, using OR is commutative in a sense that ((/a),(/b)) = ((/b),(/a)), (with a different ordering of the resulting set), distributive so that /a((/b), (/c)) = ((/a/b), (/a/c)). Parenthesis without a quantifier are associative, e.g. #/a/b/c = #/a(/b)/c = #/a(/b/c) = #/a(/b)(/c). Thus, a pattern ((/b)(/c))+ can also be written like (/b/c)+.

further examples

for further examples refer to the unit tests

Breaking Changes

  • with version v5.0.0 package has been rename to @sagold/json-query
  • with version v4.0.0 (2019/10/01)
    • the api has been simplified to methods query.get and query.delete (removed run and pattern)
  • with version v3.0.0
    • the syntax has changed to es6, which might require code transpilation
    • queries for root-pointer (#, #/, /) now callback root object with (rootObject, null, null, "#")
  • with v2.0.0 a negated filter (lookahead), e.g. *?valid:!true will not return objects where valid === undefined. To match objects with missing properties you can still query them explicitly with *?valid:!true||valid:undefined