Overview

libxml2 is the reference XML/HTML parser — used by GNOME, Python's lxml, PHP, and countless production systems. It offers full XML 1.0/1.1, namespaces, DTD and XSD validation, XPath 1.0 queries, and a complete DOM API for building and modifying XML trees.

Parse & Navigate

uses libxml2;

var xml := '<catalog><book id="1"><title>PascalAI</title><price>29.99</price></book></catalog>';

var doc  := ParseString(xml);
var root := RootElement(doc);

WriteLn('Root: ', NodeName(root));   // "catalog"

var books := ChildrenByTag(root, 'book');
for I := 0 to Length(books) - 1 do begin
  var book := books[I];
  WriteLn('id:    ', GetAttr(book, 'id'));
  WriteLn('title: ', NodeText(FirstChild(book, 'title')));
  WriteLn('price: ', NodeText(FirstChild(book, 'price')));
end;

FreeDoc(doc);

XPath Queries

var doc := ParseFile('/data/catalog.xml');

// Select all book titles
var titles := XPath(doc, '//book/title');
for I := 0 to Length(titles) - 1 do
  WriteLn(NodeText(titles[I]));

// Select books with price < 30
var cheap := XPath(doc, '//book[price < 30]');
WriteLn('Cheap books: ', Length(cheap));

// Single string value
var first := XPathString(doc, '//book[1]/title');
WriteLn('First title: ', first);

// Count
var n := XPathNumber(doc, 'count(//book)');
WriteLn('Total books: ', Round(n));

FreeDoc(doc);

Parse HTML

// Parse real-world HTML (lenient — tolerates unclosed tags, etc.)
var doc   := ParseHTMLFile('/web/page.html');
var links := XPath(doc, '//a[@href]');

for I := 0 to Length(links) - 1 do
  WriteLn(GetAttr(links[I], 'href'), ' — ', NodeText(links[I]));

FreeDoc(doc);

Build XML

var doc  := NewDoc('report');
var root := RootElement(doc);
SetAttr(root, 'version', '1.0');

var section := AddChild(root, 'section');
SetAttr(section, 'name', 'results');

var item := AddChild(section, 'item');
SetText(item, 'Passed');
SetAttr(item, 'id', '42');

WriteLn(ToPrettyString(doc));
SaveToFile(doc, '/out/report.xml');
FreeDoc(doc);

Namespaces & Validation

// XPath with namespace prefixes
var nodes := XPathNS(doc, '//soap:Body/m:GetPrice',
  'soap=http://schemas.xmlsoap.org/soap/envelope/' + #10 +
  'm=http://example.com/prices');

// Validate against XSD schema
if ValidateXSD(doc, '/schemas/catalog.xsd') then
  WriteLn('Valid!')
else
  WriteLn('Invalid: ', LastError);

// Check well-formedness
if not IsWellFormed(xmlString) then
  WriteLn('Malformed XML');