Perfect-XML

XML & HTML parsing support, modernized for Swift 6. This is real, load-bearing infrastructure: Perfect-FileMaker imports PerfectXML to parse the FileMaker XML Data API.

It implements most of the DOM Core level 2 read-only APIs plus XPath support, and also includes a SAX push-parser, a streaming reader (XMLStream), and Codable integration for decoding XML directly into model types (see Additional APIs below).

Requirements: Swift tools version 6.2, macOS 12+ (platforms: [.macOS(.v12)] in Package.swift). No formal Linux platform entry is declared in Package.swift, but the underlying libxml2 systemLibrary target ships an .apt(["libxml2-dev"]) provider for Linux builds.

Concurrency: this is a synchronous, non-async/await library. Every public class (document/node/parser types) is @unchecked Sendable over an underlying libxml2 pointer, each with an explicit rationale comment, as part of the Swift 6.2 strict-concurrency modernization.

Security: DOM parsing passes XML_PARSE_NONET to defend against XXE / external-entity / DTD-based injection. If you're evaluating this package as a parsing trust boundary, that's the relevant flag.

The pre-Swift-6 version of this package is preserved on the legacy branch.

Building

Perfect-XML has zero external Swift package dependencies — the only dependency is a local libxml2 .systemLibrary target (pkgConfig: "libxml-2.0") inlined directly into this repo's Package.swift, which replaces what used to be a separate Perfect-libxml2 package. System libxml2 (via brew/apt) is still required at build time; see the platform notes below.

.package(url: "https://github.com/PerfectlySoft/Perfect-XML.git", branch: "main")

macOS Build Notes

If you receive a compile error that says the following, you need to install and link libxml2

note: you may be able to install libxml-2.0 using your system-packager:

    brew install libxml2

Compile Swift Module 'PerfectXML' (2 sources)
<module-includes>:1:9: note: in file included from <module-includes>:1:
#import "libxml2.h"

To install and link libxml2 with homebrew, use the following two commands

brew install libxml2
brew link --force libxml2

Linux Build Notes

Ensure that you have installed libxml2-dev and pkg-config.

sudo apt-get install libxml2-dev pkg-config

Usage

Instantiate an XDocument object with your XML string

import PerfectXML
let document = XDocument(fromSource: xmlString)

Instantiate an HTMLDocument object with your HTML string

import PerfectXML
let document = HTMLDocument(fromSource: htmlString)

Now you can get the root node of the document structure by using the documentElement property

print(document.documentElement?.string(pretty: true))

Each node has several important properties

nodeValue
nodeName
parentNode
childNodes

Each node also has a getElementsByTagName: method that recursively searches through it and its children to return an array of all nodes that have that name. This method makes it easy to find a single value in the XML file.

import PerfectXML

let serverResponseXML = "<?xml version=\"1.0\" encoding=\"utf-8\"?>\n" +
"<Autodiscover xmlns=\"http://schemas.microsoft.com/exchange/autodiscover/responseschema/2006\">" + 
"  <Response>" + 
"    <Error Time=\"14:52:25.4524532\" Id=\"3280124998\">" + 
"      <ErrorCode>600</ErrorCode>" + 
"      <Message>Invalid Request</Message>" +
"      <DebugData />" +
"    </Error>" + 
"  </Response>" +
"</Autodiscover>"

let serverResponseDocument = XDocument(fromSource: serverResponseXML)
let serverMessage = serverResponseDocument?.documentElement?.getElementsByTagName("Message").first?.nodeValue
print(serverMessage) // prints Optional("Invalid Request")

Constructing model types from XML feeds is easy with getElementsByTagName

import PerfectXML

let rssXML = "<?xml version=\"1.0\" encoding=\"UTF-8\" ?>" +
"<rss version=\"2.0\">" +
"  <channel>" +
"    <title>W3Schools Home Page</title>" +
"    <link>http://www.w3schools.com</link>" +
"    <description>Free web building tutorials</description>" +
"    <item>" +
"      <title>RSS Tutorial</title>" +
"      <link>http://www.w3schools.com/xml/xml_rss.asp</link>" +
"      <description>New RSS tutorial on W3Schools</description>" +
"    </item>" +
"    <item>" +
"      <title>XML Tutorial</title>" +
"      <link>http://www.w3schools.com/xml</link>" +
"      <description>New XML tutorial on W3Schools</description>" +
"    </item>" +
"  </channel>" +
"</rss>"

let rssDocument = XDocument(fromSource: rssXML)
let feedItems = rssDocument?.documentElement?.getElementsByTagName("item")
print(feedItems?.count) // prints 2
let items = feedItems?.map({ MyCustomStruct(xmlNode: $0) })

Sometimes tag names are a bit too generic in order to meaningfully search for them, such as "title" in the example above. If we wanted to get the Channel title, link, and description, we could visit each of its children in a more deliberate way.

import PerfectXML

let rssXML = "<?xml version=\"1.0\" encoding=\"UTF-8\" ?>" +
"<rss version=\"2.0\">" +
"  <channel>" +
"    <title>W3Schools Home Page</title>" +
"    <link>http://www.w3schools.com</link>" +
"    <description>Free web building tutorials</description>" +
"    <item>" +
"      <title>RSS Tutorial</title>" +
"      <link>http://www.w3schools.com/xml/xml_rss.asp</link>" +
"      <description>New RSS tutorial on W3Schools</description>" +
"    </item>" +
"    <item>" +
"      <title>XML Tutorial</title>" +
"      <link>http://www.w3schools.com/xml</link>" +
"      <description>New XML tutorial on W3Schools</description>" +
"    </item>" +
"  </channel>" +
"</rss>"

let rssDocument = XDocument(fromSource: rssXML)
let channelNode = rssDocument?.documentElement?.getElementsByTagName("channel").first
let channelTitle = channelNode?.childNodes.filter({ $0.nodeName == "title" }).first?.nodeValue
let channelLink = channelNode?.childNodes.filter({ $0.nodeName == "link" }).first?.nodeValue
let channelDescription = channelNode?.childNodes.filter({ $0.nodeName == "description" }).first?.nodeValue

print(channelTitle) // Optional("W3Schools Home Page")
print(channelLink) // Optional("http://www.w3schools.com")
print(channelDescription) // Optional("Free web building tutorials")

Parse XML Source

This snippet will parse an XML source string and then convert it back into a string.

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b><c a=\"attr1\">HI</c><d/></b></a>\n"
let doc = XDocument(fromSource: docSrc)
let str = doc?.string(pretty: false)
XCTAssert(str == docSrc, "\(str)")

Parse HTML Source

This snippet will parse an HTML source string.

let docSrc = "<html>\n<head>\n<title>title</title></head>\n<body>\n<div>hi</div>\n</body>\n</html>\n"
let doc = HTMLDocument(fromSource: docSrc)
let nodeName = doc?.documentElement?.nodeName
XCTAssert(nodeName == "html")

Inspect Node Names

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b/><c/><d/></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
let names = ["b", "c", "d"]
for (n, v) in zip(children.childNodes, names) {
	guard let _ = n as? XElement else {
		return XCTAssert(false)
	}
	XCTAssert(n.nodeName == v, "\(n.nodeName) != \(v)")
}

Inspect Text Node

let value = "ABCD"
let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a>\(value)</a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
do {
	let children = children.childNodes
	XCTAssert(children.count == 1)
	guard let textChild = children.first as? XText else {
		return XCTAssert(false)
	}
	XCTAssert(textChild.nodeValue == value)
}

Check Node Type

let value = "ABCD"
let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a>\(value)</a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
let nodeType = children.nodeType
if case .elementNode = nodeType {
	XCTAssert(true)
} else {
	XCTAssert(false, "\(nodeType)")
}

First & Last Child

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b/><c/><d/></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
guard let firstChild = children.firstChild else {
	return XCTAssert(false)
}
guard let lastChild = children.lastChild else {
	return XCTAssert(false)
}
XCTAssert(firstChild.nodeName == "b")
XCTAssert(lastChild.nodeName == "d")

Next & Previous Sibling

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b/><c/><d/></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
guard let firstChild = children.firstChild else {
	return XCTAssert(false)
}
XCTAssert(firstChild.nodeName == "b")
guard let nextSib = firstChild.nextSibling else {
	return XCTAssert(false)
}
guard let prevSib = nextSib.previousSibling else {
	return XCTAssert(false)
}
XCTAssert(nextSib.nodeName == "c")
XCTAssert(prevSib.nodeName == "b")

Element Attributes

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b atr1=\"the value\" atr2=\"the other value\"></b></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
guard let firstChild = children.firstChild as? XElement else {
	return XCTAssert(false)
}
XCTAssert(firstChild.nodeName == "b")
guard let atr1 = firstChild.getAttribute(name: "atr1") else {
	return XCTAssert(false)
}
XCTAssert(atr1 == "the value")
guard let atr2 = firstChild.getAttributeNode(name: "atr2") else {
	return XCTAssert(false)
}
XCTAssert(atr2.value == "the other value")

With namespaces:

let names = ["atr1", "atr2"]
let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a xmlns:foo=\"foo:bar\"><b foo:atr1=\"the value\" foo:atr2=\"the other value\"></b></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let children = doc?.documentElement else {
	return XCTAssert(false, "No children")
}
XCTAssert(children.nodeName == "a")
guard let firstChild = children.firstChild else {
	return XCTAssert(false)
}
XCTAssert(firstChild.nodeName == "b")
guard let attrs = firstChild.attributes else {
	return XCTAssert(false, "nil attributes")
}
XCTAssert(attrs.length == 2)
for name in names {
	guard let item = attrs.getNamedItemNS(namespaceURI: "foo:bar", localName: name) else {
		return XCTAssert(false)
	}
	XCTAssert(item.nodeName == name)
}

Get Elements By Name

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b/><a><b><b/></b></a></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let elements = doc?.documentElement?.getElementsByTagName("b") else {
	return XCTAssert(false)
}
XCTAssert(elements.count == 3)
for node in elements {
	XCTAssert(node.nodeName == "b")
}

With namespaces:

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a xmlns:foo=\"foo:bar\"><b/><foo:a><b>FOO<b/></b></foo:a></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
do {
	guard let elements = doc?.getElementsByTagNameNS(namespaceURI: "foo:bar", localName: "a") else {
		return XCTAssert(false)
	}
	XCTAssert(elements.count == 1)
	for node in elements {
		XCTAssert(node.nodeName == "a")
		XCTAssert(node.localName == "a")
		XCTAssert(node.prefix == "foo")
		XCTAssert(node.namespaceURI == "foo:bar")
	}
}
do {
	guard let elements = doc?.documentElement?.getElementsByTagNameNS(namespaceURI: "foo:bar", localName: "a") else {
		return XCTAssert(false)
	}
	XCTAssert(elements.count == 1)
	for node in elements {
		XCTAssert(node.nodeName == "a")
		XCTAssert(node.localName == "a")
		XCTAssert(node.prefix == "foo")
		XCTAssert(node.namespaceURI == "foo:bar")
	}
}
do {
	guard let elements = doc?.getElementsByTagNameNS(namespaceURI: "foo:barz", localName: "a") else {
		return XCTAssert(false)
	}
	XCTAssert(elements.count == 0)
}
do {
	guard let elements = doc?.documentElement?.getElementsByTagNameNS(namespaceURI: "foo:barz", localName: "a") else {
		return XCTAssert(false)
	}
	XCTAssert(elements.count == 0)
}

Get Element By ID

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b id=\"foo\"/><a><b>FOO<b/></b></a></a>\n"
let doc = XDocument(fromSource: docSrc)
XCTAssert(doc?.nodeName == "#document")
guard let element = doc?.getElementById("foo") else {
	return XCTAssert(false)
}
XCTAssert(element.tagName == "b")

XPath

Elements:

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b id=\"foo\"/><a><b>FOO<b/></b></a></a>\n"
guard let doc = XDocument(fromSource: docSrc) else {
	return XCTAssert(false)
}
XCTAssert(doc.nodeName == "#document")
let pathRes = doc.extract(path: "/a/b")
guard case .nodeSet(let set) = pathRes else {
	return XCTAssert(false, "\(pathRes)")
}
for node in set {
	guard let b = node as? XElement else {
		return XCTAssert(false, "\(node)")
	}
	XCTAssert(b.tagName == "b")
}

Attributes:

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b id=\"foo\"/><a><b>FOO<b/></b></a></a>\n"
guard let doc = XDocument(fromSource: docSrc) else {
	return XCTAssert(false)
}
XCTAssert(doc.nodeName == "#document")
let pathRes = doc.extract(path: "/a/b/@id")
guard case .nodeSet(let set) = pathRes else {
	return XCTAssert(false, "\(pathRes)")
}
for node in set {
	guard let b = node as? XAttr else {
		return XCTAssert(false, "\(node)")
	}
	XCTAssert(b.name == "id")
	XCTAssert(b.value == "foo")
}

Text:

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a><b id=\"foo\"/><a><b>FOO<b/></b></a></a>\n"
guard let doc = XDocument(fromSource: docSrc) else {
	return XCTAssert(false)
}
XCTAssert(doc.nodeName == "#document")
let pathRes = doc.extract(path: "/a/a/b/text()")
guard case .nodeSet(let set) = pathRes else {
	return XCTAssert(false, "\(pathRes)")
}
for node in set {
	guard let b = node as? XText else {
		return XCTAssert(false, "\(node)")
	}
	guard let nodeValue = b.nodeValue else {
		return XCTAssert(false, "\(b)")
	}
	XCTAssert(nodeValue == "FOO")
}

Namespaces:

let docSrc = "<?xml version=\"1.0\" encoding=\"UTF-8\"?>\n<a xmlns:foo=\"foo:bar\"><b/><foo:a><b>FOO<b/></b></foo:a></a>\n"
guard let doc = XDocument(fromSource: docSrc) else {
	return XCTAssert(false)
}
let namespaces = [("f", "foo:bar")]
let pathRes = doc.extract(path: "/a/f:a", namespaces: namespaces)
guard case .nodeSet(let set) = pathRes else {
	return XCTAssert(false, "\(pathRes)")
}
for node in set {
	guard let e = node as? XElement else {
		return XCTAssert(false, "\(node)")
	}
	XCTAssert(e.tagName == "a")
	XCTAssert(e.namespaceURI == "foo:bar")
	XCTAssert(e.prefix == "foo")
}

Additional APIs

Beyond the DOM + XPath surface documented above, Sources/PerfectXML also includes:

SAX push-parser (SAX.swift) — event-driven parsing for large documents where building a full DOM tree isn't desirable.
Streaming reader (XMLStream.swift) — an incremental/streaming XML reader.
Codable integration (Codable.swift) — an XMLEncoder/decoding bridge for decoding XML directly into Codable model types.

These aren't yet covered by worked examples in this README; see the source files and Tests/PerfectXMLTests/PerfectXMLTests.swift for usage.

Testing

Run the test suite with:

swift test

See Tests/PerfectXMLTests/PerfectXMLTests.swift for the full suite, including the DOM/XPath examples shown throughout this README.

Name		Name	Last commit message	Last commit date
Sources		Sources
Tests/PerfectXMLTests		Tests/PerfectXMLTests
.gitignore		.gitignore
LICENSE		LICENSE
PADockerfile		PADockerfile
Package.swift		Package.swift
README.md		README.md

Navigation Menu

Uh oh!

Repository files navigation

Perfect-XML

Building

macOS Build Notes

Linux Build Notes

Usage

Parse XML Source

Parse HTML Source

Inspect Node Names

Inspect Text Node

Check Node Type

First & Last Child

Next & Previous Sibling

Element Attributes

Get Elements By Name

Get Element By ID

XPath

Additional APIs

Testing

About

Releases

Packages

Used by

Contributors

Languages

Uh oh!

Latest commit

History

Folders and files

Repository files navigation

Perfect-XML

Building

macOS Build Notes

Linux Build Notes

Usage

Parse XML Source

Parse HTML Source

Inspect Node Names

Inspect Text Node

Check Node Type

First & Last Child

Next & Previous Sibling

Element Attributes

Get Elements By Name

Get Element By ID

XPath

Additional APIs

Testing

About

Topics

Resources

Stars

Watchers

Forks

Releases

Packages

Used by

Contributors

Languages