CSS Syntax and XPath

CSS syntax

Use using Jint.HtmlParser; and using Jint.HtmlParser.Css; to parse and inspect CSS syntax:

CssStyleSheetSyntax sheet = MarkupParser.ParseCss("@future value; p { color: var(--theme); }");
foreach (CssRuleSyntax rule in sheet.Rules)
{
    Console.WriteLine(sheet.Source.Substring(rule.Span.Start, rule.Span.Length));
}

This prints the two original rules. CssStyleSheetSyntax.Source retains the exact input; rule spans and component spans use UTF-16 offsets into it. Rules, preludes, functions, blocks and token components are immutable syntax objects. Component types also live in Jint.HtmlParser.Css.Syntax.

Method	Result
ParseCss	CssStyleSheetSyntax for a complete stylesheet, with standard recovery
ParseCssRule	One CssRuleSyntax
ParseCssDeclaration	One CssDeclarationSyntax
ParseCssComponentValue	One token, function or block component
ParseCssComponentValues	A CssComponentValueList

All accept CssParseOptions and cancellation. Empty stylesheets succeed. A single-construct method that cannot produce its requested construct raises CssParseException, exposing Code and Offset. An optional ParseDiagnosticCollector retains bounded recoverable errors.

Parsing retains unknown at-rules, unsupported properties, invalid property values and unresolved substitutions as syntax. It does not validate properties or selectors, resolve URLs, load imports, attach style sheets, compute a cascade or expose public mutable CSSOM.

XPath 1.0

NativeXPath.Compile prepares an expression for reuse. Evaluate returns a captured scalar or node-set; Select requires a node-set and returns native identities. Prefixes are resolved through IXmlNamespaceResolver; use XmlNamespaceManager from System.Xml:

var document = MarkupParser.ParseXml("<catalog xmlns='urn:catalog'><item id='one'/></catalog>");
var namespaces = new XmlNamespaceManager(new NameTable());
namespaces.AddNamespace("c", "urn:catalog");
var expression = NativeXPath.Compile("//c:item", namespaces);
IReadOnlyList<object> items = NativeXPath.Select(document, expression);
Console.WriteLine(items.Count);
Console.WriteLine(NativeXPath.Evaluate(document, "count(//c:item)", namespaces).NumberValue);

This prints 1 twice. Unprefixed XPath element tests match the empty namespace, so both HTML's namespace and XML default namespaces require an explicit query prefix. Query prefixes need not equal source prefixes.

NativeXPathResult.ResultType determines which getter to read: NumberValue, BooleanValue, StringValue, or Nodes and FirstNodeStringValue. Reading a getter of another result kind raises InvalidOperationException; invalid expressions or scalar expressions passed to Select raise System.Xml.XPath.XPathException.

Contexts may be a Node, an attached or detached non-XMLNS Attr, or a XPathNamespaceBinding captured from the namespace axis. Results expose those native identities, with no mutable BCL cursor escaping. Node-set membership and captured scalar values are snapshots; selected objects can still be mutated. A namespace context whose URI is no longer in scope is rejected when reused.

Compiled expressions can be reused across documents. Keep trees synchronized during queries and mutation. Cancellation is checked around BCL compilation/evaluation; it cannot interrupt a non-preemptible BCL interval and is not a hard wall-clock deadline.