Crate tagsoup 

TagSoup is a small, fast, fairly forgiving HTML-ish parser with zero required dependencies.

It is built for the boringly useful jobs:

• Parse real-world markup without immediately fainting.
• Walk the resulting tree.
• Query it with a compact CSS-style selector API.
• Pull out text, attributes, and spans.

It is not trying to impersonate a browser engine. It just wants to turn messy markup into something workable, quickly.

Highlights

• Optional serde support, enabled by default.
• Preserves source spans for nodes and parse errors.
• Handles raw-text elements like script and style sensibly.
• Supports query_selector and query_selector_all.
• Supports tree walking with a small visitor API.
• Tries to recover from malformed markup instead of giving up immediately.

Examples

// Parse an HTML tag soup.
let doc = tagsoup::Document::parse("<div><p id=here>Hello, world!</p></div>");

// Check for parsing errors.
assert!(doc.errors.is_empty());

// Query the document for an element using a CSS selector.
let element = doc.query_selector("#here").unwrap();
assert_eq!(element.text_content(), "Hello, world!");

Querying The Tree

let doc = tagsoup::Document::parse(r#"
	<article id="main">
		<p class="lead">Hello</p>
		<p data-kind="feature card">world</p>
	</article>
"#);

assert_eq!(doc.query_selector("#main .lead").unwrap().text_content(), "Hello");
assert_eq!(doc.query_selector_all("[data-kind*=feature]").len(), 1);

Notes

• Whitespace is preserved by default.
• Call Document::trimmed if you want leading and trailing ASCII whitespace removed from text nodes.
• Element::text_content decodes HTML entities, except inside raw-text elements like script and style.
• Invalid selectors currently panic in Document::query_selector and Document::query_selector_all.

This is not a full WHATWG-compliant HTML parser. It is a pragmatic parser for documents that are mostly HTML, occasionally cursed, and still need to be dealt with.

Structs

Attribute

Attribute of an element.

AttributeValue

Value of an attribute.

CommentNode

Comment node in the DOM tree.

DoctypeNode

Doctype node in the DOM tree.

Document

Document represents the entire parsed HTML document fragment.

Element

Element in the DOM tree.

Lexer

TagSoup lexer.

ParseError

Document parse error.

ProcessingInstructionNode

Processing instruction node in the DOM tree.

ResolvedSpan

Span of the information in the parsed source, with line and column information.

SourceSpan

Span of the information in the parsed source.

TextNode

Text node in the DOM tree.

Token

TagSoup Token.

Enums

ElementKind

Kind of element.

Node

Node in the DOM tree.

ParseErrorKind

Document parse error kinds.

TokenKind

TagSoup Token kind.

VisitControl

Visitor control flow for tree traversal.

Functions

normalize_whitespace

Collapses runs of ASCII whitespace into a single space.