NodeXmlDeserializer

NOTE: The NPM package for this project is called XmlDeserializer.

Description: A simple and somewhat crude XML de-serializer package for Node.

shut up and show me the code!

    var xml = "<root>";
    xml += "<childelement someProp=\"somevalue\">";
    xml += "contents of child element";
    xml += "</childelement>";
    xml += "</root>";

    var ds = require("xmldeserializer");
    var elements = ds.deserialize(xml);
    var jsonString = ds.getJson(elements);
    console.log(jsonString);

Output:

{
  "root":{
    "properties" : {},
    "childelement":{
        properties : { "someProp":"somevalue" },
        "value":"contents of child element"}
    }
}

Example showing an array and other edge cases.

    var xml = "<Root>";
    xml += "<ListOfItems ItemCategory=\"blue\">";
    xml += "<Item>lagoon</Item>";
    xml += "<Item>man group</Item>";
    xml += "<Item>christmas</Item>";
    xml += "<Item>light special</Item>";
    xml += "</ListOfItems>";
    xml += "<EmptyXmlNode foo=\"bar\" />";
    xml += "<XmlNodeWithEmptyValue></XmlNodeWithEmptyValue>";
    xml += "</Root>";

    var ds = require("xmldeserializer");
    var elements = ds.deserialize(xml);
    var jsonString = ds.getJson(elements);
    console.log(jsonString);

Output:

• Notice that the array is wrapped in its own hash named "arrayOfX", where X is the name of the XML node.
• Also, to accomodate properties, each array item is its own hash.
• Last, notice that "EmtpyXmlNode is self-closing. Be definition its value is null.
• However if a node has a separate closing tag, but no value, then the value in the json will be an empty string.

{
  "root":{
    "properties" : {},
    "listOfItems":{
        "properties" : { "itemCategory":"blue" },
        "arrayOfItem" : [
            { "properties": {},
              "value": "lagoon"
            },
            { "properties": {},
              "value": "man group"
            },
            { "properties": {},
              "value": "christmas"
            },
            { "properties": {},
              "value": "light special"
            }
        ]
    },
    "emptyXmlNode" : {
        "properties" : { "foo":"bar" },
        "value": null
    },
    "xmlNodeWithEmptyValue" : {
        "properties": {},
        "value": ""
    }
  }
}

More examples in the app.js test file.

Features:

• No schema or object definitions! Just send it XML and out comes JSON!
• Supports parsing of namespaces. You can (optionally) add the namespace or namespace abbreviation to the contents of the subsequent hash.
• Supports CDATA, including special characters (newlines, etc).
• Auto detects multiple element names and creates an array, or you can specify the node name to always be interpreted as an array.

Limitations:

• No support for streams. You must pass in the entire XML block in a string.
• XML properties must have values enclosed in quotes. e.g. (This means numeric and boolean property values aren't currently supported.)
• The getJson method currently doesn't take namespaces into consideration. If your xml schema has duplicate names in different namespaces this may be an issue for you.
• Only supports the 5 standard entities (quot, amp, apos, lt, gt). No support for custom defined entities... it will probably cause the parser to blow up.

Using xml deserializer with Express:

var express = require("express");
var app = express();

// This will put the deserialized data into req.body for your route handlers to access.
app.use(function(req, res, next) {
    if (req.headers["content-type"] == "application/xml") {
        req.rawBody = "";
        req.setEncoding("utf8");
        req.on('data', function(chunk) { 
            req.rawBody += chunk;
        });
        req.on('end', function() {
            var elements = xmldeserializer.deserialize(req.rawBody);
            req.body = JSON.parse(xmldeserializer.getJson(elements));
            next();
        });
    }
    else {
        next();
    }
});