Re: Top-Down Parser Construction Conjectures

sja@vinkku.hut.fi (Sakari Jalovaara) writes:
>[Recursive descent is easily faster than YACC.]


Although some yacc-like tools might do a better job than RD by avoiding
the overhead of function calls. The basic idea is to integrate the parse
table and the interpreter that walks it.


;-D on ( Ingratiating integration ) Pardo


3 references follow:


%A Chris W. Fraser
%A Robert R. Henry
%T Hard-Coding Bottom-Up Code Generation Tables to Save Time and Space
%J Software \- Practice and Experience
%V 21
%N 1
%D January 1991
%P 1-12
%X Code generation by parsing, fast parser.


%A Thomas J. Pennello
%T Very Fast LR Parsing
%J Proceedings of the SIGPLAN 1986 Symposium on Compiler Construction;
SIGPLAN Notices
%V 21
%N 7
%D July 1986
%P 145-151
%X * Partial evaluation of the table interpreter with resepct to each
element of the table
  * Speedup: on a VAX-like machine, 40,000 to 500,000 lines per minute.
On an 80286, 37,000 to 240,000 lines per minute.
  * FSM converted to assembly language.
  * 2-4X increase in table size.


%A Frank G. Pagan
%T Comparative Efficiency of General and Residual Parsers
%J SIGPLAN Notices
%V 25
%N 4
%D April 1990
%P 59-68
%W pardo
%X * Focus on partial evaluation of parsers generating pure code.
  * Hand traslation of Pascal and C.
  * Time (2-10X faster) and size (0.5-1.25 larger).
--

Return to the comp.compilers page.
Search the comp.compilers archives again.