@loadingio/ldquery

lightweight DOM Helpers in Vanilla JS.

Installation

install via npm:

npm install @loadingio/ldquery

or, download it from github:

https://github.com/loadingio/ldquery/releases/latest/download/ldq.js
https://github.com/loadingio/ldquery/releases/latest/download/ldq.min.js

or use it directly via CDN:

https://cdn.jsdelivr.net/gh/loadingio/ldquery@v1.1.3/dist/ldq.js
https://cdn.jsdelivr.net/gh/loadingio/ldquery@v1.1.3/dist/ldq.min.js

in you html, include it directly with

Usage

ldquery provides following functions:

* find(node, selector, index):
shorthand for querySelector and querySelectorAll. return array when index is omitted.
* index(node): index of node in childNodes of its parent.
* child(node): children of node, in Array.
* parent(node, selector, endNode):
search parent(including node) for which matching the selector, until endNode is reached. return null if not found.
* attr(node, name, value): get attribute "name" if value is omitted. set attribute if value is provided.
* on(node, name, callback): addEventListener on event name.
* remove(node): remove node from parent.
* insertAfter(node, newNode, oldNode): insert newNode after oldNode under node.
* fetch: fetch API helper.
* xhr: XMLHttpRequest API helper.
* json(data): handy function to convert things to object.
* cls(o,p,n): handy function for setting classes.
- o: either object or value.
- object: class name / on,off pair
- value: true to toggle classes in p on, and classes in n off. vice versa.
- p: positive class list ( toggled when o is true )
- n: negative class list ( toggled when o is false )
invoke these functions with ld$, e.g.,

ld$.find(document.body, '.btn', 0)

or ignore the first parameter if it's document:

ld$.find('.btn', 0) / equivalent to ld$.find(document, '.btn', 0) /


It also provides in wrapper style:

ld$(document.body).find('.btn', 0)


While it's not a good idea but you can pollute the native DOM for a more intuitive way to use these functions:

# in livescript syntax
HTMLElement.prototype <<< ld$obj.prototype

This could be enabled by remove the comment mark in corresponding line inside ldq.ls.

Fetch

ldquery wraps fetch api with promise-based error handling and some additional parameter.

ld$.fetch(, , )
.then -> ...
.catch (e) -> # use e.data to get raw response from fetch

Common raw options:

* method: HTTP method. e.g., "POST"
* body: things to send to server. pass JSON by stringify it.
- ld$.fetch "url", {body: JSON.stringify({data: 1})}
* headers: header hash. customize your own header here.

Common ldquery Options:

* type: indicate the response data type. could be text or json.
* json: json object to pass. shorthand for manually setting headers and stringify object.
- this:
ld$.fetch("url", {}, {json: {data: 1}})
- is the same with this:
ld$.fetch("url", {body: "{data: 1}", headers: {'Content-Type': 'application/json; charset=UTF-8'}})
* params: shorthand for setting get params. expect an object with key/value pair. will convert it to querystring.
- this:
ld$.fetch("url", {}, {params: {qs: "some text"}}
- is the same with this:
ld$.fetch("url?qs=some%20text", {}, {})
* timeout: elapsed time ( in milliseconds ) for a timeout error to be fired. default 40s
* no-default-headers: default null. set this to true to bypass default headers in ld$.fetch.headers.


You can also config global headers by updating values in ld$.fetch.headers:

ld$.fetch.headers["X-CSRF-TOKEN"] = ...

XMLHttpRequest

While fetch is newer than XMLHttpRequest, for now XMLHttpRequest does better in handling progress information. Thus you can use ld$.xhr instead of ld$.fetch if such information is needed.

Using xhr is almost identical with using fetch:

ld$.xhr(, , )
.then -> ...
.catch (e) -> # use e.data to get raw response from fetch

Yet it accepts one additional ldquery option:

* progress({percent, val, len}): a callback function accepting progress information as an object
- percent: 0 ~ 1, 1 = finished.
- val: current progress.
- len: expected total progress.
- this function has no expected return value.

$3

ld$.fetch doesn't depend on ldError, but it indeed returns ldError compatible errors, with following fields:

* name: ldError
* id: error code
* code: http code. ( optional )
* data: returned data
* json: parsed json from returned data, if any ( optional )

When request failed, ld$.fetch will try to parse returned data as json, and check if there is a field "name" with value "ldError". If "ldError" name field exists, this object will be used directly to construct the Error object.

Compatibility

ldquery uses following modern web features:

* fetch api - not support in IE, Older Edge ( <= 13). Use polyfill to support to IE >= 11
* classList - not support well in IE. Use Polyfill to fix this.
* Array.from - not supported in IE. Use Polyfill to fix this.

License

MIT License.