Reads XML documents and emits JavaScript objects with a simple, easy to use structure.
- Small, fast and simple
- Runs everywhere (browser, node.js, React Native, ServiceWorkers, WebWorkers...)
- Event driven and synchronous API
- Can process input piece-by-piece in a serial fashion
- Stream mode (low memory usage)
- Reads CDATA sections
npm install --save xml-readerObjects emitted by the reader have the following structure:
/**
* typedef {Object} XmlNode
* @property {string} name - element name (empty for text nodes)
* @property {string} type - node type ('element' or 'text')
* @property {string} value - value of a text node
* @property {XmlNode} parent - reference to parent node
* @property {Object} attributes - attributes {name: value, ...}
* @property {XmlNode[]} children - array of children nodes
*/Basic example. Read and parse a XML document.
const XmlReader = require('xml-reader');
const reader = XmlReader.create();
const xml =
`<?xml version="1.0" encoding="UTF-8"?>
<message>
<to>Alice</to>
<from>Bob</from>
<heading color="blue">Hello</heading>
<body color="red">This is a demo!</body>
</message>`;
reader.on('done', data => console.log(data));
reader.parse(xml);
/*
Console output:
{ name: 'message',
type: 'element',
children: [
{ name: 'to',
type: 'element',
children: [{ type: 'text', value: 'Alice' }]},
{ name: 'from',
type: 'element',
children: [{ type: 'text', value: 'Bob' }]},
{ name: 'heading',
type: 'element',
attributes: { color: 'blue' },
children: [{ type: 'text', value: 'Hello' }]},
{ name: 'body',
type: 'element',
attributes: { color: 'red' },
children: [{ type: 'text', value: 'This is a demo!' }]}]}
Note: empty values and references to parent nodes removed for brevity!
*/This mode is only valid for reading complete documents (root node must be closed).
const XmlReader = require('xml-reader');
const xml = '<doc>Hello!</doc>';
const result = XmlReader.parseSync(xml/*, options*/);In stream mode, nodes are removed from root as they are emitted. This way memory usage does not increase.
const XmlReader = require('xml-reader');
const reader = XmlReader.create({stream: true});
const xml =
`<root>
<item v=1/>
<item v=2/>
<item v=3/>
</root>`;
reader.on('tag:item', (data) => console.log(data));
// {name: 'item', type: 'element', value: '', attributes: {v: '1'}, children: []}
// {name: 'item', type: 'element', value: '', attributes: {v: '2'}, children: []}
// {name: 'item', type: 'element', value: '', attributes: {v: '3'}, children: []}
reader.on('done', (data) => console.log(data.children.length));
// 0
reader.parse(xml);You can also listen to all tags:
reader.on('tag', (name, data) => console.log(`received a ${name} tag:`, data));In this example we are calling multiple times to the parser. This is useful if your XML document is a stream that comes from a TCP socket or WebSocket (for example XMPP streams).
Simply feed the parser with the data as it arrives. As you can see, the result is exactly the same as the previous one.
const XmlReader = require('xml-reader');
const reader = XmlReader.create({stream: true});
const xml =
`<root>
<item v=1/>
<item v=2/>
<item v=3/>
</root>`;
reader.on('tag:item', (data) => console.log(data));
// {name: 'item', type: 'element', value: '', attributes: {v: '1'}, children: []}
// {name: 'item', type: 'element', value: '', attributes: {v: '2'}, children: []}
// {name: 'item', type: 'element', value: '', attributes: {v: '3'}, children: []}
reader.on('done', (data) => console.log(data.children.length));
// 0
// Note that we are calling the parse function providing just one char each time
xml.split('').forEach(char => reader.parse(char));Use the reset() method to reset the reader. This is useful if a stream gets interrupted and you want to start a new one or to use the same reader instance to parse multiple documents (just reset the reader between them).
Example:
const doc1 = '<document>...</document>';
const doc2 = '<document>...</document>';
reader.parse(doc1);
// when the document ends, the reader stops emitting events
reader.reset();
// now you can parse a new document
reader.parse(doc2);Default options are:
{
stream: false,
parentNodes: true,
tagPrefix: 'tag:',
doneEvent: 'done',
emitTopLevelOnly: false,
}If true (default), each node of the AST has a parent node which point to its parent. If false the parent node is always null.
Enable or disable stream mode. In stream mode nodes are removed from root after being emitted. Default false.
Ignored in parseSync;
Default value is 'done'. This is the name of the event emitted when the root node is closed and the parse is done.
Ignored in parseSync;
Default value is 'tag:'. The event driven API emits an event each time a tag is read. Use this option to set a name prefix.
Ignored in parseSync;
Default value is false. When true, tag events are only emitted by top level nodes (direct children from root). This is useful for XMPP streams like XMPP where each top level child is a stanza.
For example, given the following XML stream:
<stream>
<message from="alice" to="bob">
<body>hello</body>
<date>2016-10-06</date>
</message>
<message from="alice" to="bob">
<body>bye</body>
<date>2016-10-07</date>
</message>tags emitted with emitTopLevelOnly=false
body
date
message
body
date
message
tags emitted with emitTopLevelOnly=true
message
message
MIT