Web Platform · Cluster
HTML and Semantics
Document structure and meaning.
- 01 How the HTML Tokenizer and Tree Builder Recover from Malformed Markup HTML error recovery is specified rather than improvised, which is why every engine builds the same tree from broken markup. The tokenizer states, the insertion modes, and the repair rules behind that guarantee.
- 02 Optional Tags and the Implied End Tag Rules the Parser Enforces Only some tags may be omitted, and only where the tree builder already closes the element. The conformance conditions behind each omission, and the cases where leaving a tag out changes the tree.
- 03 HTML Content Model Categories and What They Actually Constrain Flow, phrasing, sectioning, embedded, interactive, and transparent content describe what may nest inside what. Browsers enforce a few of those rules by repairing the tree and ignore the rest.
- 04 Sectioning Content and the Heading Outline Browsers Never Implemented Nesting sections was meant to derive heading levels automatically. No engine shipped that, the algorithm left the standard, and headings now depend entirely on the level the author writes.
- 05 Constraint Validation and the Form Validity State Machine Required, pattern, min, max, and step produce a set of independent validity flags, not one error. How the browser decides whether a form may submit, and where the API diverges from the interactive path.
- 06 Accessible Name Computation: Which Attribute Wins aria-labelledby, aria-label, label elements, alt text, and element content are consulted in a fixed order. The order, the recursion rules, and the roles that refuse a name entirely.
- 07 The Table Model, Foster Parenting, and Header Association Loose table markup is rebuilt into thead, tbody, and tr by the parser, stray content is moved out, and the header association algorithm decides which cell describes which.
- 08 Template Contents, Slots, and Declarative Shadow DOM A template owns a document fragment that never enters the page, slots decide which light-DOM children render where, and a shadowrootmode attribute can attach a shadow root during parsing.
- 09 Character Encoding Determination Before the Parser Runs A byte order mark, an HTTP charset parameter, and a meta prescan of the first 1024 bytes compete to decide how bytes become characters, and a late declaration can force a reparse.
- 10 How srcset and sizes Select an Image Resource Width and density descriptors describe the candidates, the sizes list converts layout into a pixel width, and the choice among candidates belongs partly to the browser.