| TREE-META | |
| Author: | Donald Andrews, Jeff Rulifson |
| Released: | 1968? |
The TREE-META (or Tree Meta, TREEMETA) Translator Writing System is a compiler-compiler system for context-free languages originally developed in the 1960s. Parsing statements of the metalanguage resemble augmented Backus–Naur form with embedded tree-building directives. Unparsing rules include extensive tree-scanning and code-generation constructs.
TREE-META was instrumental in the development of NLS (oN-Line System) and was ported to many systems including the UNIVAC 1108, GE 645, SDS 940, ICL 1906A, PERQ, and UCSD p-System.
This is a complete example of a TREE-META program extracted (and untested) from the more complete (declarations, conditionals, and blocks) example in Appendix 6 of the ICL 1900 TREE-META manual. That document also has a definition of TREE-META in TREE-META in Appendix 3. This program is not just a recognizer, but also outputs the assembly language for the input. It demonstrates one of the key features of TREE-META, which is tree pattern matching. It is used on both the LHS (GET and VAL for example) and the RHS (ADD and SUB).
% This is an ALGOL-style comment delimited by %
% ====================== INPUT PARSE RULES ======================= %
.META PROG
% A program defining driving rule is required. %
% This PROG rule is the driver of the complete program. %
PROG = $STMT ;
% $ is the zero or more operator. %
% PROG (the program) is defined as zero or more STMT (statements). %
STMT = .ID ':=' AEXP :STORE[2]*;
% Parse an assignment statement from the source to the tree. %
% ':=' is a string constant, :STORE creates a STORE node, %
% [2] defines this as having two branches i.e. STORE[ID,AEXP]. %
% * triggers a unparse of the tree, Starting with the last created %
% tree i.e. the STORE[ID,AEXP] which is emitted as output and %
% removed from the tree. %
AEXP = FACTOR $('+' FACTOR :ADD[2] / '-' FACTOR :SUB[2]);
% Here we have the recognizer for arithmetic '+' :ADD and '-' :SUB %
% tree building. Again the [2] creates a 2-branch ADD or SUB tree. %
% Unparsing is deferred until an entire statement has been parsed. %
% ADD[FACTOR,FACTOR] or SUB[FACTOR,FACTOR] %
FACTOR = '-' PRIME :MINUSS[1] / PRIME ;
PRIME = .ID / .NUM / '(' AEXP ')' ?3? ;
% ?3? is a hint for error messages. %
% ===================== OUTPUT UNPARSE RULES ===================== %
STORE[-,-] => GET[*2] 'STORE ' *1 ;
% *1 is the left tree branch. *2 is the right %
% GET[*2] will generate code to load *2. %
% The 'STORE' string will be output %
% followed by left branch *1 a symbol %
% Whatever *2, it will be loaded by GET[*2]. %
GET[.ID] => 'LOAD ' *1 /
[.NUM] => ' LOADI ' *1 /
[MINUSS[.NUM]] => 'LOADN ' *1:*1 /
[-] => *1 ;
% Here an .ID or a .NUM will simply be loaded. A MINUSS node %
% containing a .NUM will have this used, the notation *1:*1 means %
% the first branch (a .NUM) of the first branch (MINUSS). %
% Anything else will be passed on for node recognition %
% The unparse rules deconstruct a tree outputing code. %
ADD[-,-] => SIMP[*2] GET[*1] 'ADD' VAL[*2] /
SIMP[*1] GET[*2] 'ADD' VAL[*1] /
GET[*1] 'STORE T+' < OUT[A] ; A<-A+1 > /
GET[*2] 'ADD T+' < A<-A-1 ; OUT[A] > ;
% Chevrons < > indicate an arithmetic operation, for example to %
% generate an offset A relative to a base address T. %
SUB[-,-] => SIMP[*2] GET[*1] 'SUB' VAL[*2] /
SIMP[*1] GET[*2] 'NEGATE' % 'ADD' VAL[*1] /
GET[*2] 'STORE T+' < OUT[A] ; A<-A+1 > /
GET[*1] 'SUB T+' < A<-A-1 ; OUT[A] > ;
% A percent character in an unparse rule indicates a newline. %
SIMP[.ID] => .EMPTY /
[.NUM] => .EMPTY /
[MINUSS[.NUM]] => .EMPTY;
VAL[.ID] => ' ' *1 /
[.NUM] => 'I ' *1 /
[MINUSS[.NUM]] => 'N ' *1:*1 ;
MINUSS[-] => GET[*1] 'NEGATE' ;
.END