Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

Pretty printing XML with javascript

From the text of the question I get the impression that a string result is expected, as opposed to an HTML-formatted result.

If this is so, the simplest way to achieve this is to process the XML document with the identity transformation and with an <xsl:output indent="yes"/> instruction:

<xsl:stylesheet version="1.0"
 xmlns:xsl="http://www.w3.org/1999/XSL/Transform">
 <xsl:output omit-xml-declaration="yes" indent="yes"/>

    <xsl:template match="node()|@*">
      <xsl:copy>
        <xsl:apply-templates select="node()|@*"/>
      </xsl:copy>
    </xsl:template>
</xsl:stylesheet>

When applying this transformation on the provided XML document:

<root><node/></root>

most XSLT processors (.NET XslCompiledTransform, Saxon 6.5.4 and Saxon 9.0.0.2, AltovaXML) produce the wanted result:

<root>
  <node />
</root>

This can be done using native javascript tools, without 3rd party libs, extending the @Dimitre Novatchev's answer:

var prettifyXml = function(sourceXml)
{
    var xmlDoc = new DOMParser().parseFromString(sourceXml, 'application/xml');
    var xsltDoc = new DOMParser().parseFromString([
        // describes how we want to modify the XML - indent everything
        '<xsl:stylesheet xmlns:xsl="http://www.w3.org/1999/XSL/Transform">',
        '  <xsl:strip-space elements="*"/>',
        '  <xsl:template match="para[content-style][not(text())]">', // change to just text() to strip space in text nodes
        '    <xsl:value-of select="normalize-space(.)"/>',
        '  </xsl:template>',
        '  <xsl:template match="node()|@*">',
        '    <xsl:copy><xsl:apply-templates select="node()|@*"/></xsl:copy>',
        '  </xsl:template>',
        '  <xsl:output indent="yes"/>',
        '</xsl:stylesheet>',
    ].join('\n'), 'application/xml');

    var xsltProcessor = new XSLTProcessor();    
    xsltProcessor.importStylesheet(xsltDoc);
    var resultDoc = xsltProcessor.transformToDocument(xmlDoc);
    var resultXml = new XMLSerializer().serializeToString(resultDoc);
    return resultXml;
};

console.log(prettifyXml('<root><node/></root>'));

Outputs:

<root>
  <node/>
</root>

JSFiddle

Note, as pointed out by @jat255, pretty printing with <xsl:output indent="yes"/> is not supported by firefox. It only seems to work in chrome, opera and probably the rest webkit-based browsers.


Slight modification of efnx clckclcks's javascript function. I changed the formatting from spaces to tab, but most importantly I allowed text to remain on one line:

var formatXml = this.formatXml = function (xml) {
        var reg = /(>)\s*(<)(\/*)/g; // updated Mar 30, 2015
        var wsexp = / *(.*) +\n/g;
        var contexp = /(<.+>)(.+\n)/g;
        xml = xml.replace(reg, '$1\n$2$3').replace(wsexp, '$1\n').replace(contexp, '$1\n$2');
        var pad = 0;
        var formatted = '';
        var lines = xml.split('\n');
        var indent = 0;
        var lastType = 'other';
        // 4 types of tags - single, closing, opening, other (text, doctype, comment) - 4*4 = 16 transitions 
        var transitions = {
            'single->single': 0,
            'single->closing': -1,
            'single->opening': 0,
            'single->other': 0,
            'closing->single': 0,
            'closing->closing': -1,
            'closing->opening': 0,
            'closing->other': 0,
            'opening->single': 1,
            'opening->closing': 0,
            'opening->opening': 1,
            'opening->other': 1,
            'other->single': 0,
            'other->closing': -1,
            'other->opening': 0,
            'other->other': 0
        };

        for (var i = 0; i < lines.length; i++) {
            var ln = lines[i];

            // Luca Viggiani 2017-07-03: handle optional <?xml ... ?> declaration
            if (ln.match(/\s*<\?xml/)) {
                formatted += ln + "\n";
                continue;
            }
            // ---

            var single = Boolean(ln.match(/<.+\/>/)); // is this line a single tag? ex. <br />
            var closing = Boolean(ln.match(/<\/.+>/)); // is this a closing tag? ex. </a>
            var opening = Boolean(ln.match(/<[^!].*>/)); // is this even a tag (that's not <!something>)
            var type = single ? 'single' : closing ? 'closing' : opening ? 'opening' : 'other';
            var fromTo = lastType + '->' + type;
            lastType = type;
            var padding = '';

            indent += transitions[fromTo];
            for (var j = 0; j < indent; j++) {
                padding += '\t';
            }
            if (fromTo == 'opening->closing')
                formatted = formatted.substr(0, formatted.length - 1) + ln + '\n'; // substr removes line break (\n) from prev loop
            else
                formatted += padding + ln + '\n';
        }

        return formatted;
    };

Found this thread when I had a similar requirement but I simplified OP's code as follows:

function formatXml(xml, tab) { // tab = optional indent value, default is tab (\t)
    var formatted = '', indent= '';
    tab = tab || '\t';
    xml.split(/>\s*</).forEach(function(node) {
        if (node.match( /^\/\w/ )) indent = indent.substring(tab.length); // decrease indent by one 'tab'
        formatted += indent + '<' + node + '>\r\n';
        if (node.match( /^<?\w[^>]*[^\/]$/ )) indent += tab;              // increase indent
    });
    return formatted.substring(1, formatted.length-3);
}

works for me!


Personnaly, I use google-code-prettify with this function :

prettyPrintOne('<root><node1><root>', 'xml')

Or if you'd just like another js function to do it, I've modified Darin's (a lot):

var formatXml = this.formatXml = function (xml) {
    var reg = /(>)(<)(\/*)/g;
    var wsexp = / *(.*) +\n/g;
    var contexp = /(<.+>)(.+\n)/g;
    xml = xml.replace(reg, '$1\n$2$3').replace(wsexp, '$1\n').replace(contexp, '$1\n$2');
    var pad = 0;
    var formatted = '';
    var lines = xml.split('\n');
    var indent = 0;
    var lastType = 'other';
    // 4 types of tags - single, closing, opening, other (text, doctype, comment) - 4*4 = 16 transitions 
    var transitions = {
        'single->single'    : 0,
        'single->closing'   : -1,
        'single->opening'   : 0,
        'single->other'     : 0,
        'closing->single'   : 0,
        'closing->closing'  : -1,
        'closing->opening'  : 0,
        'closing->other'    : 0,
        'opening->single'   : 1,
        'opening->closing'  : 0, 
        'opening->opening'  : 1,
        'opening->other'    : 1,
        'other->single'     : 0,
        'other->closing'    : -1,
        'other->opening'    : 0,
        'other->other'      : 0
    };

    for (var i=0; i < lines.length; i++) {
        var ln = lines[i];
        var single = Boolean(ln.match(/<.+\/>/)); // is this line a single tag? ex. <br />
        var closing = Boolean(ln.match(/<\/.+>/)); // is this a closing tag? ex. </a>
        var opening = Boolean(ln.match(/<[^!].*>/)); // is this even a tag (that's not <!something>)
        var type = single ? 'single' : closing ? 'closing' : opening ? 'opening' : 'other';
        var fromTo = lastType + '->' + type;
        lastType = type;
        var padding = '';

        indent += transitions[fromTo];
        for (var j = 0; j < indent; j++) {
            padding += '    ';
        }

        formatted += padding + ln + '\n';
    }

    return formatted;
};