HtmlDocumentLoader class final

Loads HTML by reducing it to readable text.

Scripts, styles and comments are dropped entirely; block-level tags become line breaks so paragraphs survive; the rest of the markup is removed and the five predefined entities plus numeric references are decoded.

This is a text extractor, not a parser. It is deliberately dependency-free, which is the right trade for the documentation, help centres and article pages that make up most retrieval corpora. A corpus of application HTML — where the content is assembled by script and the structure carries meaning — wants a real parser upstream, and this loader will happily take its output.

Implemented types

• DocumentLoader

Constructors

HtmlDocumentLoader(List<TextSource> sources)

Creates an HTML loader over sources.

const

Properties

hashCode → int

The hash code for this object.

no setter, inherited

name → String

A short name, used in events and traces.

no setter, override

runtimeType → Type

A representation of the runtime type of the object.

no setter, inherited

sources → List<TextSource>

The HTML texts to load.

final

Methods

load({AgenticContext? context}) → Future<List<RagDocument>>

Produces documents from whatever this loader was constructed with.

override

noSuchMethod(Invocation invocation) → dynamic

Invoked when a nonexistent method or property is accessed.

inherited

toString() → String

A string representation of this object.

inherited

Operators

operator ==(Object other) → bool

The equality operator.

inherited

Static Methods

decodeEntities(String text) → String

Decodes the predefined entities and numeric character references.

extractTitle(String html) → String?

The contents of <title>, or null.

htmlToText(String html) → String

Reduces html to readable text.