wtf is parser?

The riot/parser is a tool that breaks down HTML fragments into structured pieces of information a computer can work with. Instead of treating HTML like a plain text document, it reads through the code and identifies each part, tags, attributes, text, comments, and expressions, so that other programs can understand and manipulate them. At a basic level, you give it a snippet of HTML and it returns an array of objects describing what it found. Each object represents a tag, a piece of text, or a comment, along with details like attribute names and values. The parser is designed to handle a single tag at a time rather than an entire web page. It stops reading once it finds the closing tag of the root element. One notable feature is that it supports expressions embedded in attribute values or text, delimited by curly braces by default, which is useful for templating scenarios where you need to inject dynamic values into markup. This tool is mainly for developers building frameworks or tools that need to process HTML-like templates. For example, someone creating a custom UI framework similar to Riot.js would use this parser to read template files and convert them into a format their rendering engine can use. It is a low-level building block rather than something an end user interacts with directly. The parser takes a deliberately loose approach to HTML parsing. It does not strictly enforce HTML5 specs, for instance, it preserves self-closing tags as-is rather than normalizing them, and it handles attributes in a flexible way, allowing slashes to act as separators. This makes it forgiving for template syntax that might not be perfectly valid HTML but is still meaningful in a framework context. Comments are ignored by default but can be preserved if needed, and the expression delimiters can be customized.