More useful built-in rules

List of built-in rules is rather short. There can be:

• A rule for floating-point number like +34.12e+56
• A rule for some integer number like 123 or 0x24AE.
• A string literal like "qwe\\dfg\"zxc"
• An identifier beginning with a letter or underscore and containing letters, digits and underscore
• ISO datetime

Explicitly writing such rules each time those blocks are needed makes grammars less readable and more error-prone. Also availability of common building blocks discourages being too inventive about grammars (like designing you own escaping scheme for strings) without reason.

So popular and useful rules should be available at hand, either as built-in rules or importable from external crate like pest-common-rules and mentioned in the book.

JSON example in the book should contain additional "final version" of the grammar, where string and number are replaced by STRING_LITERAL and FLOATING_POINT_NUMBER built-ins.

Start with the linked list of built-in rules and the JSON example in the book. Determine which proposed floating-point, integer, string-literal, identifier, and ISO datetime rules are supported and how they should be exposed, then update the JSON example so its string and number rules use the selected built-ins.