STERRITT%SDEVAX.decnet@ge-crd.arpa (08/03/88)
Hello, we are interested in getting any one of: 1) A YACC (or other, yacc preferred) grammer for a FORTRAN to C compiler 2) A direct C implementation of such a compiler, or 3) A program to do such compilation directly. We have a machine that only has a C compiler, and we have a bunch of large fortran programs that we'd like to run thereon, and avoid hand translation if at all possible. The possibilities above are in order of preferrence; that is, if you have a yacc grammer, we'd like that most of all, and we *REALLY* prefer that it be Public Domain, or minimal cost. PLEASE send me mail DIRECTLY, rather than answering this post with another post, since we don't subscribe to this group. thanks very much, chris sterritt sterritt%sdevax.decnet@ge-crd.arpa on arpanet csterritt on GEnie
wgh@uucsbb.UUCP (Bill Hutchison) (08/08/88)
In article <16715@adm.ARPA>, STERRITT%SDEVAX.decnet@ge-crd.arpa writes: > PLEASE send me mail DIRECTLY, rather than answering this post with > another post, since we don't subscribe to this group. > thanks very much, > chris sterritt > sterritt%sdevax.decnet@ge-crd.arpa on arpanet > csterritt on GEnie I have tried to mail to this gent, but the University of Delaware keeps bouncing my mail back. Does anybody know how I can get mail to him? ----- Unsent message follows ----- Received: by uucsbb.UUCP (5.54/1.14) id AA18233; Wed, 3 Aug 88 16:41:30 EDT Date: Wed, 3 Aug 88 16:41:30 EDT From: wgh (Bill Hutchison) Message-Id: <8808032041.AA18233@uucsbb.UUCP> Apparently-To: burdvax!udel!gatech!rutgers!cmcl2!adm!STERRITT%SDEVAX.decnet@ge-crd.arpa I have been looking into FORTRAN-to-C translation, and I have concluded that it is a bad idea to do directly. The method I recommend is to use teh BSD4.2 or BSD4.3 utility "struct" (written by Brenda Baker of Bell Labs) to translate spaghetti FORTRAN into structured RATFOR, then to write your own translator from RATFOR to C (much easier than FORTRAN to C). A company called Cobalt Blue has FORTRAN-to-C and RATFOR-to-C translators available commercially, and Rapitech in upstate NY has FORTRAN-to-C. Don't wait for this to appear in public domain -- it is a fairly difficult task to do well considering the vast semantic differences between FORTRAN and C. FORTRAN-to-C++ makes a lot more sense (since C++ supports call-by- reference, and you can define what [] means for your own datatypes, so you can redefine how arrays work). -- Bill Hutchison, DP Consultant rutgers!cbmvax!burdvax!uucsbb!wgh Unisys UNIX Portation Center "What one fool can do, another can!" P.O. Box 500, M.S. B121 Ancient Simian Proverb, quoted by Blue Bell, PA 19424 Sylvanus P. Thompson, in _Calculus Made Simple_