estwgba@daisy.warwick.ac.uk (J N Petry) (12/15/87)
I am engaged on an M.Sc. project which involves conversion from GWBASIC
to Turbo pascal and vice-versa for IBM pc-compatibles.
My initial research has yielded little information about compilation
into a high-level language rather than assembly code.
I have access to a UNIX system running 4.2BSD and thus intend to use Lex
and Yacc to generate the parser, and KERMIT the results across.
Can anyone recommend any sources of info. on automatic translation or
point out any glaring oversights I have made?
Please send any replies via email and I shall summarize to this group.
Thanks in advance.
James Petry
I.T. pidgeonholes
Department Of Engineering
University Of Warwick
Coventry CV4 7AL
England
[Translating the syntax is the easy part. How, for example, do you
translate Basic's variable length garbage collected strings into Pascal,
or for that matter the PRINT USING statement? In the 1960's there were a
lot of source to source translators in use as people moved from first to
second generation systems, notably from the IBM 7094 to the 360. I tried
IBM's Fortran to PL/I translator, and found its output grotesque because, for
example, even though Fortran and PL/I "format" statements look similar, the
semantics are just different enough to cause enormous trouble. The generic
term is SIFT programs, from the name of the first Fortran translator.
If you look in the back of magazines like Datamation you'll still find ads
from services that translate various dialects of Cobol and even Autocoder
to Cobol. -John]
--
Send compilers articles to ima!compilers or, in a pinch, to Levine@YALE.EDU
Plausible paths are { ihnp4 | decvax | cbosgd | harvard | yale | bbn}!ima
Please send responses to the originator of the message -- I cannot forward
mail accidentally sent back to compilers. Meta-mail to ima!compilers-request