Lexical structure & glyphs
The lexer turns a file into tokens before the parser builds a tree. This
page is those tokens: identifiers, numbers, strings, the width markers
(i32, f32, …), and the frontmatter delimiter.
Identifiers are the names you write — score, divide, Span. Numbers
are the integer and floating literals. A "…" string is Unicode text.
Width markers are the bare type tokens for sized numerics. None of these
are keywords. The reserved words on the other family pages are already
known to the parser; everything else that looks like a name is an
identifier.
There is no keyword table here: these productions are the terminal shapes, not the vocabulary of the language. A short program that is mostly those tokens:
main {
let i32 narrow ← 7 ∷ i32
let f32 single ← 1.5 ∷ f32
let string name ← "Marcus"
print name
print narrow
print single
}Return to the grammar overview.
Formal grammar#
The rules for this part of the language are the parser's own definition of it. They are written for models and tools, so this page does not repeat them; the full list is at every production.