How to format/tidy/beautify in JavaScript

user_268074 0
Viewed 68644

How can I format/tidy/beautify HTML in JavaScript? I have tried doing a search/replace for angle brackets (<, >) and indenting accordingly. But of course it does not take into account when the is JS or CSS etc inside the HTML.

The reason I want to do this is I have made a content editor (CMS) which has both WYSIWYG and source code views. The problem the code written by the WYSIWYG editor is normally a single line. So I would like a JavaScript that could format this into a more readable form on demand.

Here what I have so far:

function getIndent(level) {
    var result = '',
        i = level * 4;
    if (level < 0) {
        throw "Level is below 0";
    }
    while (i--) {
        result += ' ';
    }
    return result;
}

function style_html(html) {
    html = html.trim();
    var result = '',
        indentLevel = 0,
        tokens = html.split(/</);
    for (var i = 0, l = tokens.length; i < l; i++) {
        var parts = tokens[i].split(/>/);
        if (parts.length === 2) {
            if (tokens[i][0] === '/') {
                indentLevel--;
            }
            result += getIndent(indentLevel);
            if (tokens[i][0] !== '/') {
                indentLevel++;
            }

            if (i > 0) {
                result += '<';
            }

            result += parts[0].trim() + ">\n";
            if (parts[1].trim() !== '') {
                result += getIndent(indentLevel) + parts[1].trim().replace(/\s+/g, ' ') + "\n";
            }

            if (parts[0].match(/^(img|hr|br)/)) {
                indentLevel--;
            }
        } else {
            result += getIndent(indentLevel) + parts[0] + "\n";
        }
    }
    return result;
}
html jquery javascript
9 Answers

I use this method to format HTML. Simple, but does the job:

function format(html) {
    var tab = '\t';
    var result = '';
    var indent= '';

    html.split(/>\s*</).forEach(function(element) {
        if (element.match( /^\/\w/ )) {
            indent = indent.substring(tab.length);
        }

        result += indent + '<' + element + '>\r\n';

        if (element.match( /^<?\w[^>]*[^\/]$/ ) && !element.startsWith("input")  ) { 
            indent += tab;              
        }
    });

    return result.substring(1, result.length-3);
}
user_24707650

I needed something similar and here is my solution, inspired by method provided by michal.jakubeczy. It is slightly complicated in order to preserve formatting within <pre> tags. Hope this will help someone.

function formatHTML(html) {
    var indent = '\n';
    var tab = '\t';
    var i = 0;
    var pre = [];

    html = html
        .replace(new RegExp('<pre>((.|\\t|\\n|\\r)+)?</pre>'), function (x) {
            pre.push({ indent: '', tag: x });
            return '<--TEMPPRE' + i++ + '/-->'
        })
        .replace(new RegExp('<[^<>]+>[^<]?', 'g'), function (x) {
            var ret;
            var tag = /<\/?([^\s/>]+)/.exec(x)[1];
            var p = new RegExp('<--TEMPPRE(\\d+)/-->').exec(x);

            if (p) 
                pre[p[1]].indent = indent;

            if (['area', 'base', 'br', 'col', 'command', 'embed', 'hr', 'img', 'input', 'keygen', 'link', 'menuitem', 'meta', 'param', 'source', 'track', 'wbr'].indexOf(tag) >= 0) // self closing tag
                ret = indent + x;
            else {
                if (x.indexOf('</') < 0) { //open tag
                    if (x.charAt(x.length - 1) !== '>')
                        ret = indent + x.substr(0, x.length - 1) + indent + tab + x.substr(x.length - 1, x.length);
                    else 
                        ret = indent + x;
                    !p && (indent += tab);
                }
                else {//close tag
                    indent = indent.substr(0, indent.length - 1);
                    if (x.charAt(x.length - 1) !== '>')
                        ret =  indent + x.substr(0, x.length - 1) + indent + x.substr(x.length - 1, x.length);
                    else
                        ret = indent + x;
                }
            }
            return ret;
        });

    for (i = pre.length; i--;) {
        html = html.replace('<--TEMPPRE' + i + '/-->', pre[i].tag.replace('<pre>', '<pre>\n').replace('</pre>', pre[i].indent + '</pre>'));
    }

    return html.charAt(0) === '\n' ? html.substr(1, html.length - 1) : html;
}

function unformatHTML(html) {
    var i = 0;
    var pre = [];

    html = html.replace(new RegExp('<pre>((.|\\t|\\n|\\r)+)?</pre>'), function (x) {
        pre.push(x);
        return '<--TEMPPRE' + i++ + '/-->'
    }).replace(/\n/g, '').replace(/\t/g, '');

    for (i = pre.length; i--;) {
        html = html.replace('<--TEMPPRE' + i + '/-->', pre[i]);
    }

    html = html.replace(new RegExp('<pre>\\n'), '<pre>').replace(new RegExp('\\n\\t*</pre>'), '</pre>');
    return html;
}
user_54141370

I find js-beautify far superior to any solution posted so far.

Add the script to your lib folder:

Bring inside header as usual:

<script src="libs/beautify.js"></script>

Target code wherever it is on your page (e.g. pre or code tag) and use the js_beautify function to format as needed:

$(".my_class").text(js_beautify($(".my_class").text()))

This will format as needed. All kinds of config options available on the repo.

user_16395940
Related