174

How would you convert from XML to JSON and then back to XML?

The following tools work quite well, but aren't completely consistent:

Has anyone encountered this situation before?

Bhargav Rao
  • 50,140
  • 28
  • 121
  • 140
Jason Suárez
  • 2,445
  • 3
  • 18
  • 20

14 Answers14

127

I think this is the best one: Converting between XML and JSON

Be sure to read the accompanying article on the xml.com O'Reilly site, which goes into details of the problems with these conversions, which I think you will find enlightening. The fact that O'Reilly is hosting the article should indicate that Stefan's solution has merit.

mikemaccana
  • 110,530
  • 99
  • 389
  • 494
Josh Stodola
  • 81,538
  • 47
  • 180
  • 227
  • 1
    thanks for the reply! In my case, the JSON is the canonical representation, and XML is just used for XSLT.. the use of which is not my idea! :) – Jason Suárez Nov 20 '09 at 23:29
  • 1
    This is only in the browser. Doesn't apply for node.js or non-browser environments. Any other ideas? – Homer6 May 02 '13 at 17:47
  • 1
    In regards to @JasonDenizac comment to his post, I am not sure to understand how this link helps fixing the issue of having an object instead of an array of one item... – guiomie Jul 25 '13 at 15:27
  • 1
    I found that if you start from json-xml-json, this library works well, but if you want xml-json-xml there is a problem with reversibility since it adds metadata xml elements like and – vishr Dec 26 '13 at 18:37
  • 3
    Please note that this is a copyleft licensed solution. It is only an option when you are writing open source software. – Jasper Jun 26 '14 at 11:45
  • @Jasper: That isn't true re the LGPL. Straight from the gnu's mouth, as it were: https://www.gnu.org/philosophy/why-not-lgpl.html. Using the LGPL is generally only a problem if you change the code and even then you only need to release your changes. – 0x1mason Oct 20 '14 at 23:02
  • @0x1mason LGPL is tricky. Using a minifier (which is common for javascript) might be in breach of it (though you would need a lawyer to be sure). I was wrong to say you can't use it when not writing FOSS, but if you aren't, you will have to be careful not to step into the legal minefield that LGPL creates... – Jasper Oct 27 '14 at 09:37
  • For me this alerts some "unhandled types" in Firefox v30 on Win7 – dude Dec 17 '14 at 16:12
  • It is Really the Best. – Ahsan Apr 16 '16 at 06:31
  • Is there a "Standard" tool. i dread downloading the best tool only to find two years later its been abandoned. – dank8 Sep 20 '19 at 02:26
  • @Dan The tools xml2json and json2xml _have been_ abandoned for 13 years. But who cares? The scripts will continue to work. It is vanilla JS. – René K May 19 '20 at 07:21
55

https://github.com/abdmob/x2js - my own library (updated URL from http://code.google.com/p/x2js/):

This library provides XML to JSON (JavaScript Objects) and vice versa javascript conversion functions. The library is very small and doesn't require any other additional libraries.

API functions

  • new X2JS() - to create your instance to access all library functionality. Also you could specify optional configuration options here
  • X2JS.xml2json - Convert XML specified as DOM Object to JSON
  • X2JS.json2xml - Convert JSON to XML DOM Object
  • X2JS.xml_str2json - Convert XML specified as string to JSON
  • X2JS.json2xml_str - Convert JSON to XML string

Online Demo on http://jsfiddle.net/abdmob/gkxucxrj/1/

var x2js = new X2JS();
function convertXml2JSon() {
    $("#jsonArea").val(JSON.stringify(x2js.xml_str2json($("#xmlArea").val())));
}

function convertJSon2XML() {
    $("#xmlArea").val(x2js.json2xml_str($.parseJSON($("#jsonArea").val())));
}

convertXml2JSon();
convertJSon2XML();
$("#convertToJsonBtn").click(convertXml2JSon);
$("#convertToXmlBtn").click(convertJSon2XML);
Community
  • 1
  • 1
abdolence
  • 2,369
  • 1
  • 21
  • 29
  • 1
    Hi, how did you overcome the issue where if you have one object in an object it is in object litteral, where if there is n > 1 objects, you have an array. This makes it hard to use xml to json objects in templates... – guiomie Jul 25 '13 at 15:23
  • Yes, you should use some trick and it depends on your knowledge about XML structure (because no XSD here). Use ..._asArray syntax to access your node always as array(sequence) – abdolence Jul 25 '13 at 16:41
  • 1
    Sample:// XML string to JSON var xmlText = "Successddsfgdsdgfdgfd"; var jsonObj = X2JS.xml_str2json( xmlText ); alert (jsonObj.MyOperation.test); alert (jsonObj.MyOperation.test_asArray[0]); – abdolence Jul 25 '13 at 16:45
  • My main problem, is when I transform my json to xml back, the json is full of extra properties, and when the string xml version keeps all the useless stuff. It has all kind of commas, and whitespaces... – guiomie Jul 26 '13 at 17:59
  • could you submit your sample to https://code.google.com/p/x2js/issues I'll check it – abdolence Jul 27 '13 at 11:52
  • Is this your own library? If so, please add a note to your answer. It's good form. – Jasper Jun 26 '14 at 12:08
  • Well done, I can easily parse an entire OData $metadata Saved me days of refactor – Raffaeu Jan 31 '17 at 16:03
  • @abdolence: you saved my time. thank you so much. it really worked. – Running Rabbit Mar 21 '18 at 05:28
  • @abdolence Is that repo still maintained? It looks kinda dead :-( – Potherca Aug 18 '20 at 10:21
  • @Potherca I'm really sorry, x2js is really an old lib and I stopped using it personally many years ago. I saw a lot of forks though and think you can find something more updated for nowadays browsers/environments. – abdolence Aug 19 '20 at 13:06
38

These answers helped me a lot to make this function:

function xml2json(xml) {
  try {
    var obj = {};
    if (xml.children.length > 0) {
      for (var i = 0; i < xml.children.length; i++) {
        var item = xml.children.item(i);
        var nodeName = item.nodeName;

        if (typeof (obj[nodeName]) == "undefined") {
          obj[nodeName] = xml2json(item);
        } else {
          if (typeof (obj[nodeName].push) == "undefined") {
            var old = obj[nodeName];

            obj[nodeName] = [];
            obj[nodeName].push(old);
          }
          obj[nodeName].push(xml2json(item));
        }
      }
    } else {
      obj = xml.textContent;
    }
    return obj;
  } catch (e) {
      console.log(e.message);
  }
}

As long as you pass in a jquery dom/xml object: for me it was:

Jquery(this).find('content').eq(0)[0]

where content was the field I was storing my xml in.

KyleMit
  • 30,350
  • 66
  • 462
  • 664
Ryan Conrad
  • 499
  • 4
  • 3
20

I've created a recursive function based on regex, in case you don't want to install library and understand the logic behind what's happening:

const xmlSample = '<tag>tag content</tag><tag2>another content</tag2><tag3><insideTag>inside content</insideTag><emptyTag /></tag3>';
console.log(parseXmlToJson(xmlSample));

function parseXmlToJson(xml) {
    const json = {};
    for (const res of xml.matchAll(/(?:<(\w*)(?:\s[^>]*)*>)((?:(?!<\1).)*)(?:<\/\1>)|<(\w*)(?:\s*)*\/>/gm)) {
        const key = res[1] || res[3];
        const value = res[2] && parseXmlToJson(res[2]);
        json[key] = ((value && Object.keys(value).length) ? value : res[2]) || null;

    }
    return json;
}

Regex explanation for each loop:

  • res[0] - return the xml (as is)
  • res[1] - return the xml tag name
  • res[2] - return the xml content
  • res[3] - return the xml tag name in case the tag closes itself. In example: <tag />

You can check how the regex works here: https://regex101.com/r/ZJpCAL/1

Note: In case json has a key with an undefined value, it is being removed. That's why I've inserted null at the end of line 9.

MasterPiece
  • 445
  • 7
  • 13
  • 1
    Not working as expected deep nested xml string values. – sankar muniyappa Jul 15 '22 at 18:55
  • @sankarmuniyappa please write an example that didn't work so I'll fix it :) – MasterPiece Jul 17 '22 at 13:04
  • how to make it work with this XML? ``` tag contentanother contentinside content tag contentanother contentinside content ``` – timy Sep 19 '22 at 13:00
  • The function returns the correct output for your case: `{"chan":{"item":{"tag":"tag content","tag2":"another content","tag3":{"insideTag":"inside content","emptyTag":null}}}}` – MasterPiece Sep 19 '22 at 18:19
  • @MasterPiece hmm, in https://docs.scriptable.app, the function only returns the last one item: ```{"tag":"tag2 content","tag2":"another content 2","tag3":{"insideTag":"inside content 2","emptyTag":null}}``` – timy Sep 20 '22 at 10:54
  • if I write inside content in multiple lines using Template Literals it does not return nested json instead It separates all keys. – yograj tandel Apr 03 '23 at 07:15
5

I was using xmlToJson just to get a single value of the xml.
I found doing the following is much easier (if the xml only occurs once..)

let xml =
'<person>' +
  ' <id>762384324</id>' +
  ' <firstname>Hank</firstname> ' +
  ' <lastname>Stone</lastname>' +
'</person>';

let getXmlValue = function(str, key) {
  return str.substring(
    str.lastIndexOf('<' + key + '>') + ('<' + key + '>').length,
    str.lastIndexOf('</' + key + '>')
  );
}


alert(getXmlValue(xml, 'firstname')); // gives back Hank
Nebulosar
  • 1,727
  • 3
  • 20
  • 46
5

You can also use txml. It can parse into a DOM made of simple objects and stringify. In the result, the content will be trimmed. So formating of the original with whitespaces will be lost. But this could be used very good to minify HTML.

const xml = require('txml');
const data = `
<tag>tag content</tag>
<tag2>another content</tag2>
<tag3>
  <insideTag>inside content</insideTag>
  <emptyTag />
</tag3>`;

const dom = xml(data); // the dom can be JSON.stringified

xml.stringify(dom); // this will return the dom into an xml-string

Disclaimer: I am the author of txml, the fastest xml parser in javascript.

Tobias Nickel
  • 512
  • 5
  • 7
  • 1
    This is a great drop-in replacement for `fast-xml-parser`, using `simplify`, that doesn't convert strings to numbers arbitrarily (like `"1.0"` to `1` or `1.`) – Daniel Nov 12 '20 at 07:41
4

A while back I wrote this tool https://bitbucket.org/surenrao/xml2json for my TV Watchlist app, hope this helps too.

Synopsys: A library to not only convert xml to json, but is also easy to debug (without circular errors) and recreate json back to xml. Features :- Parse xml to json object. Print json object back to xml. Can be used to save xml in IndexedDB as X2J objects. Print json object.

surya
  • 1,351
  • 1
  • 13
  • 29
  • @kleopatra this link points to the tool which converts xml to json. Its not a reference but the actual link to the resource. Not sure how else i should do it :) – surya Sep 02 '13 at 02:25
3

In 6 simple ES6 lines:

xml2json = xml => {                                                                                                                                                     
  var el = xml.nodeType === 9 ? xml.documentElement : xml                                                                                                               
  var h  = {name: el.nodeName}                                                                                                                                          
  h.content    = Array.from(el.childNodes || []).filter(e => e.nodeType === 3).map(e => e.textContent).join('').trim()                                                  
  h.attributes = Array.from(el.attributes || []).filter(a => a).reduce((h, a) => { h[a.name] = a.value; return h }, {})                                                 
  h.children   = Array.from(el.childNodes || []).filter(e => e.nodeType === 1).map(c => h[c.nodeName] = xml2json(c))                                                    
  return h                                                                                                                                                              
}  

Test with echo "xml2json_example()" | node -r xml2json.es6 with source at https://github.com/brauliobo/biochemical-db/blob/master/lib/xml2json.es6

brauliobo
  • 5,843
  • 4
  • 29
  • 34
1

Disclaimer: I've written fast-xml-parser

Fast XML Parser can help to convert XML to JSON and vice versa. Here is the example;

var options = {
    attributeNamePrefix : "@_",
    attrNodeName: "attr", //default is 'false'
    textNodeName : "#text",
    ignoreAttributes : true,
    ignoreNameSpace : false,
    allowBooleanAttributes : false,
    parseNodeValue : true,
    parseAttributeValue : false,
    trimValues: true,
    decodeHTMLchar: false,
    cdataTagName: "__cdata", //default is 'false'
    cdataPositionChar: "\\c",
};
if(parser.validate(xmlData)=== true){//optional
    var jsonObj = parser.parse(xmlData,options);
}

If you want to parse JSON or JS object into XML then

//default options need not to set
var defaultOptions = {
    attributeNamePrefix : "@_",
    attrNodeName: "@", //default is false
    textNodeName : "#text",
    ignoreAttributes : true,
    encodeHTMLchar: false,
    cdataTagName: "__cdata", //default is false
    cdataPositionChar: "\\c",
    format: false, 
    indentBy: "  ",
    supressEmptyNode: false
};
var parser = new parser.j2xParser(defaultOptions);
var xml = parser.parse(json_or_js_obj);
Amit Kumar Gupta
  • 7,193
  • 12
  • 64
  • 90
1

Here' a good tool from a documented and very famous npm library that does the xml <-> js conversions very well: differently from some (maybe all) of the above proposed solutions, it converts xml comments also.

var obj = {name: "Super", Surname: "Man", age: 23};

var builder = new xml2js.Builder();
var xml = builder.buildObject(obj);
SimoneMSR
  • 368
  • 1
  • 6
  • 16
0

I would personally recommend this tool. It is an XML to JSON converter.

It is very lightweight and is in pure JavaScript. It needs no dependencies. You can simply add the functions to your code and use it as you wish.

It also takes the XML attributes into considerations.

var xml = ‘<person id=”1234” age=”30”><name>John Doe</name></person>’;
var json = xml2json(xml); 

console.log(json); 
// prints ‘{“person”: {“id”: “1234”, “age”: “30”, “name”: “John Doe”}}’

Here's an online demo!

0

There is an open sourced library Xml-to-json with methods jsonToXml(json) and xmlToJson(xml).

Here's an online demo!

Valentyn Kolesnikov
  • 2,029
  • 1
  • 24
  • 31
0

This function directly reads the DOM properties of the XMLDocument (or document node/element) to build the JSON completely and accurately without trying to guess or match. Pass it responseXML, not responseText from XMLHttpRequest.

xml2json(xmlDoc)

If you only have a string of XML and not an XMLDocument, jQuery will convert your text to one.

xml2json($(xmlString)[0])
  • Each node becomes an object. (All elements are nodes, not all nodes are elements (e.g. text within an element).)
  • Every object contains the node name and type.
  • If it has attributes, they appear as properties in an attributes object.
  • If it has children, they appear recursively as node->objects in a children array.
  • If it's a Text, CDATA, or Comment node (bare text between element tags) or a comment, it shouldn't have attributes or children but the text will be in a text property.
{
  // Always present
  "name": "FancyElement",
  "type": "Element",

  // If present
  "attributes: {
    "attr1": "val1",
    "attr2": "val2"
  },
  "children": [...],
  "text": "buncha fancy words"
}

Caveat: I'm not familiar with all the node types. It's probably not grabbing needed/useful info from all of them. It was tested on and behaves as expected for

  • Element
  • Text
  • CDATA
  • Comment
  • Document

function xml2json(xml) {
  try {
    const types = [null,
      "Element",
      "Attribute",
      "Text",
      "CDATA",
      "EntityReference", // Deprecated
      "Entity", // Deprecated
      "ProcessingInstruction",
      "Comment",
      "Document",
      "DocumentType",
      "DocumentFragment",
      "Notation" // Deprecated
    ];

    var o = {};
    o.name = xml.nodeName;
    o.type = types[xml.nodeType];

    if (xml.nodeType == 3 ||
        xml.nodeType == 4 ||
        xml.nodeType == 8 ) {
      o.text = xml.textContent;
    } else {
      if (xml.attributes) {
        o.attributes = {};
        for (const a of xml.attributes) {
          o.attributes[a.name] = a.value;
        }
      }

      if (xml.childNodes.length) {
        o.children = [];
        for (const x of xml.childNodes) {
          o.children.push(xml2json(x))
        }
      }
    }
    return (o);
  } catch (e) {
    alert('Error in xml2json. See console for details.');
    console.log('Error in xml2json processing node:');
    console.log(o);
    console.log('Error:');
    console.log(e);
  }
}

var doc = document.getElementById("doc");
var out = document.getElementById("out");

out.innerText = JSON.stringify(xml2json(doc), null, 2);
/* Let's process the whole Code Snippet #document, why not?
 * Yes, the JSON we just put in the document body and all 
 * this code is encoded in the JSON in the console.
 * In that copy you can see why the XML DOM will all be one line.
 * The JSON in the console has "\n" nodes all throughout.
 */
console.log(xml2json(document));
#doc,
#out {
  border: 1px solid black;
}
<div id="doc"><!-- The XML DOM will all be on one line --><div personality="bubbly" relevance=42>This text is valid for HTML.<span>But it probably shouldn't be siblings to an element in XML.</span></div></div>

<pre id="out"></pre>
Travis Bemrose
  • 379
  • 1
  • 12
-5

The best way to do it using server side as client side doesn't work well in all scenarios. I was trying to build online json to xml and xml to json converter using javascript and I felt almost impossible as it was not working in all scenarios. Ultimately I ended up doing it server side using Newtonsoft in ASP.MVC. Here is the online converter http://techfunda.com/Tools/XmlToJson

Sheo Narayan
  • 1,196
  • 2
  • 14
  • 17