WrapChars.js

V2.2.4

Contents

• About
• Setup
• Use
• Known issues
• Version history

About

Tiny script containing just one, static method that takes any element and wraps each inline character in a given tag, with an optional class. Very small (<3KB minified, no dependencies).

Version 1 of this was very blunt in its approach, stripping out any nested tags entirely. This version is much more polite, and will do its best to preserve existing markup. It has several ways to control the wrapping process.

Homepage/Demos

Setup

Vanilla

Download one of the JavaScript files from the dist folder, probably the minified one. Import it into your HTML:

<script src="WrapChars.min.js"></script>

You can also embed from UNPKG:

<script src="https://www.unpkg.com/wrap-chars"></script>

Ensure the DOM has loaded completely before continuing.

npm/browserify/webpack etc

tbc

Use

WrapChars.wrap(elementReference, [paramsObject]);

Say we have the following HTML:

<blockquote cite="Lewis Carroll" id="poem">
    <span class="line">’Twas brillig, and the slithy toves</span><br>
    <span class="line">Did gyre and gimble in the wabe:</span><br>
    <span class="line">All mimsy were the borogoves,</span><br>
    <span class="line">And the mome raths outgrabe.</span>
</blockquote>

Grab a reference to the element you want to affect:

let domElement = document.getElementById ('poem');

To wrap the characters using the default settings, call the WrapChars.wrap() static method using the reference as the argument:

WrapChars.wrap (domElement);

You could also iterate over a NodeList if you prefer:

let lines = document.querySelectorAll ('*.lines');
lines.forEach (line => WrapChars.wrap (line));

Note that this is a destructive process. If you think you may want to revert to the original HTML structure at any point you will need to store a copy of the original markup before applying the method.

Any whitespace is reduced down to a single space. Text nodes containing only whitespace are ignored completely.

Configuration

WrapChars.wrap() accepts two arguments:

Argument	Type	Description
element	HTMLElement	A reference to a DOM element, e.g. as obtained via querySelector() or getElementById().
params	Object	(optional) An object containing key/value pairs to configure the method (see below).

params object

The params object can contain the following properties, all optional:

Property	Type	Description
split	string	Defines how text nodes will be subdivided for wrapping. Currently takes a value of "letter" or "word". Default is "letter". split replaces type since v2.2, though type still exists as an alias.
tagName	string	The type of element that will be wrapped around each character. Default is "span".
className	string	The class name that can be applied to each wrapped element. Default is undefined.
deep	Boolean	Boolean value where, if true, will parse the entire DOM tree of the element. If false will only wrap inline text of the element itself. Default is true.
wrapSpaces	Boolean	Boolean value to specify if spaces should be wrapped. If a value is set for spaceChar those characters are also subject to this setting. Default is false
skipClass	string	If specified, any element with a matching class name will be ignored. Default is undefined.
spaceChar	string	The character to replace spaces with, if specified. The value is standardised internally, so can be a raw character, an escaped hexcode, a Unicode code or an HTML entity. HTML elements are stripped out, to limit exploits. WrapChars.wrap() intentionally does not convert existing HTML entities, so certain results can be achieved by adding them beforehand. Default is undefined.

Example

WrapChars.wrap(myElement, {
    split: "word",
    tagName: "div",
    className: "wrapped_content",
    deep: false,
    wrapSpaces: true,
    skipClass: "skip_this",
    spaceChar: " "
});

Known issues

• word type can't differentiate between words and punctuation. To work around this, pre-wrap characters and make use of skipClass.
• Certain (unlikely) combinations of HTML mark-up can cause empty elements to be added when wrapSpaces is true and split is equal to word. Can be fixed with minor HTML reformatting.
• CodeQL flags a couple of lines for passing DOM text directly into HTML. Since elements are removed first I am not convinced this is an issue.

Version history

• v2.2 - Added wrapSpaces property, renamed type to split.
• v2.1 - Added deep and skipClass properties, improved whitespace handling.
• v2.0 - Near complete rewrite to play nicer with other people's markup
• v1.0 - Original version