Re: Grammars for future languages

Syntax extension was popular in the late 60's and early 70's. It
usually went together with various kinds of semantic extension, such
as user-defined data types (this was before the power of the Simula 67
approach, popularized by Smalltalk, and now called object orientation,
was widely realized), control structure extensions, etc. Some of the
interpretive systems allowed changing syntax at runtime, and there
were programming environments (e.g. Harvard PDS) built on the idea
that each module could define its own problem-oriented syntax.


Many Lisp-based languages/systems such as Macsyma have an Algol-like
"public" syntax and a Lisp-like internal syntax.


-s


Some relevant bibliography:


On the basis for ELF--an Extensible Language Facility
T.E. Cheatham et al.
Fall Joint Computer Conference 1968, 33:2:937


Extensible Languages (conference)
Carlos Christensen, ed; Christopher J. Shaw, ed
ACM Sigplan Notices 4:8:, 1969


Experience with an Extensible Language
Edgar T. Irons
Commun. ACM 13:1:31


International Symposium on Extensible Languages
Stephen A. Schuman, ed
IBM Centre scientifique de Grenoble FF2.0143, 1971
= ACM Sigplan Notices, 6:12:, 1971


Syntax Extension Using a Run Time Model
Jay Earley
Int. J. Comp. and Info. Sys. 3:3:189 (1974)
[The Irons CACM paper describes IMP which I mentioned a few message ago. The
problem with syntax extension is that it solves the wrong problem, giving
you new and different ways to write code that does the same old thing, and
producing the unpleasant effect that each program in an extensible language is
in effect in a unique language that you'll have completely forgotten the next
time you have to change it. OOP turns this idea on its head, keeping the same
syntax but letting you extend the semantics. -John]
--

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