Re: Code generation tool

Eric Eide <eeide@cs.utah.edu>
13 Oct 2005 20:40:30 -0400

          From comp.compilers

Related articles
Code generation tool doru-cat@student.matnat.uio.no (Doru-Catalin Togea) (2005-10-05)
Re: Code generation tool snicol@apk.net (Scott Nicol) (2005-10-05)
Re: Code generation tool jatin.bhateja@amdocs.com (Jatin Bhateja) (2005-10-06)
Re: Code generation tool murdesi@yahoo.com (2005-10-13)
Re: Code generation tool eeide@cs.utah.edu (Eric Eide) (2005-10-13)
| List of all articles for this month |

From: Eric Eide <eeide@cs.utah.edu>
Newsgroups: comp.compilers
Date: 13 Oct 2005 20:40:30 -0400
Organization: University of Utah School of Computing
References: 05-10-038
Keywords: code

"Doru-Catalin" == Doru-Catalin Togea <doru-cat@student.matnat.uio.no> writes:


Doru-Catalin> I wonder what tools do the CORBA, ICE and the like
Doru-Catalin> compilers use to translate from their respective IDLs to
Doru-Catalin> other languages, like C++, Java, Python, and so forth.


Since you asked about IDL compilers specifically, you might be interested in
"Flick," the Flexible IDL Compiler Kit from the University of Utah:


<http://www.cs.utah.edu/flux/flick/>


Its intermediate representations are exposed and well documented. The
different phases of the compiler are strongly separated --- into separate
programs, in fact! The phases of the compiler communicate through files.


Caveats and disclaimers: I don't claim that Flick is typical of IDL compilers
in general. In many ways, Flick is atypical. Flick hasn't been updated in a
long time. It has back ends for C and C++ only. I was one of the primary
authors of Flick.


Doru-Catalin> I need a formalized description of what the output
Doru-Catalin> (translation) of parsing a structured text (not XML)
Doru-Catalin> should be.


I'm not sure what you're asking for.


Most translation tools aren't derived from formal models, at least not
end-to-end. Compilers have intermediate representations (IRs), certainly, but
the actual translation from one IR to another is ususally "specified" in an ad
hoc manner. More plainly, the model of the translation --- such as it is ---
is the code, period.


There are counter-examples, of course, and I'm sure that other readers of this
forum will mention some of them :-). But at least, I'm not aware of any
counter-examples that are IDL compilers.


Maybe I'm over-analyzing your question --- maybe you just want to see some
example IRs? If so, check out Flick.


Eric.


--
-------------------------------------------------------------------------------
Eric Eide <eeide@cs.utah.edu> . University of Utah School of Computing
http://www.cs.utah.edu/~eeide/ . +1 (801) 585-5512 voice, +1 (801) 581-5843 FAX


Post a followup to this message

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