document_loader_html_processor 2.0.0-rc1

Changes since 1.0.0-rc1:

• Issue #3575582 by webbywe: Update module for Document Loader 2.x and html_processor 1.0-rc1 compatibility. Extracts all HTML processing services to the standalone html_processor module; this module becomes a bridge plugin. Adds the document_loader_html_processor_url opt-in submodule for URL fetching.

Breaking changes:

• The following classes have been removed. Replace with their html_processor equivalents (see README.md § Breaking migration notice for full mapping): HtmlSanitizerConfigBuilder, HtmlContainerExtractor, HtmlManipulator, RelativeUrlRewriter, AdRemovalHelper, HtmlFetcher, HtmlFetcherInterface, HtmlFetcherFactory, HtmlSanitizerConfigBuilderInterface, HtmlContainerExtractorInterface, HtmlManipulatorInterface, RelativeUrlRewriterInterface.
• HtmlProcessorException renamed to DocumentLoaderHtmlProcessorException.
• HtmlProcessorInput no longer accepts a combined content/url bag. Use HtmlContentInput for in-memory HTML; use the new URL submodule for fetching.
• drush dlhp:test no longer accepts --content-file. Pass HTML via --content.

New:

• HtmlContentInput: in-memory HTML input for document_loader_type:html_content.
• HtmlProcessorTrait (src/Traits/): shared pipeline logic for both loaders.
• document_loader_html_processor_url submodule: opt-in Guzzle-backed URL loader with SSRF scheme allowlist (http/https only), 10 MB bounded streaming read, Content-Length early rejection, redirect protocol pinning, and credential redaction in error messages.
• drush dlhpu:test: Drush test command for the URL submodule.
• Integration test suite using real html_processor services (24 tests).
• URL loader unit test suite (18 tests).