5.1 Syntax Description Language

Basically, the syntax of the T3X language is described in a BNF-style form similar to the one accepted by the YACC parser generator. A more detailed description follows.

The T3X grammar is described as a set of rules of the following format:

Name:   Pattern1
        | Pattern2
        | ...
        | PatternN
        ;

It reads 'Name may also be written as Pattern₁ OR Pattern₂ OR ... OR Pattern_N'. Each pattern may consist of names of rules or terminal symbols. Each terminal symbol is enclosed in apostrophes, like '=', 'CONST', or '0x'. An apostrophy may be included in a terminal (symbol) by doubling it. Consequently, a terminal represented by an apostrophy is written ''''.

A terminal symbol is a symbol which represents itself. For example, 'CLASS' represents the literal string CLASS. Non-terminal symbols represent rules.

An example: The rule

BinaryDigit: '0' | '1' ;

is read 'A BinaryDigit may be represented by either the string '0' or the string '1'. A (recursive) rule to define arbitrary-length binary numbers based upon BinaryDigit would look like this:

BinaryNumber:
        BinaryDigit
        | BinaryDigit BinaryNumber
        ;

In this case, a BinaryNumber would be either a single BinaryDigit or a BinaryDigit followed by another BinaryNumber (and therefore more BinaryDigits).

The hash symbol (#) is used to indicate that no white space is allowed between the elements of a pattern. While the above rule would match

1 0 1 1 0 1 0 1

a rule containing the concatenation symbol (#) would match

10110101

Such a rule would be written this way:

BinaryNumber:
        BinaryDigit
        | BinaryDigit # BinaryNumber
        ;

A special rule named <character>, which is not defined inside of the formal grammer, is used to refer to an arbitrary character contained in the character set of the implementation environment.