Re: OO compiler design ?

steve@gridlock.demon.co.uk (Steve Tracey)
22 Jan 1997 00:14:49 -0500

          From comp.compilers

Related articles
OO compiler design ? jc@chiinc.com (J-C Le Mentec) (1997-01-16)
Re: OO compiler design ? dittmar@berlin.snafu.de (1997-01-17)
Re: OO compiler design ? R.E.Jones@ukc.ac.uk (1997-01-21)
Re: OO compiler design ? steve@gridlock.demon.co.uk (1997-01-22)
Re: OO compiler design ? krotoff@boy.nmd.msu.ru (Alexander Krotoff) (1997-01-22)
Re: OO compiler design ? stephen@lila.york.ac.uk (1997-01-25)
Re: OO compiler design ? trondro@sn.no (Trond Ronde) (1997-01-29)
Re: OO compiler design ? SCHMIDTG@iccgcc.cle.ab.com (1997-02-02)
Re: OO compiler design ? stephen@lila.york.ac.uk (1997-02-07)
Re: OO compiler design ? p.froehlich@link-m.de (1997-02-11)
[1 later articles]
| List of all articles for this month |

From: steve@gridlock.demon.co.uk (Steve Tracey)
Newsgroups: comp.compilers
Date: 22 Jan 1997 00:14:49 -0500
Organization: Gridlock
References: 97-01-136 97-01-149
Keywords: OOP, design

Daniel Dittmar (dittmar@berlin.snafu.de) wrote:
: There is a book named "Object-Oriented Compiler Construction" by Jim
: Holmes (Prentice Hall, ISBN 0-13-630740-X). It describes a compiler
: built in C++.
: --
: Daniel Dittmar dittmar@berlin.snafu.de
: [Haven't seen it, and at $68.00 I guess I'm not likely to unless someone
: wants to send me a review copy. -John]


There is a 'student reprint' of this which is much cheaper (at least
in the UK) - One of those red PH editions.


I wasn't greatly impressed - too much of the front-end is conventional
(YACC) technology. But the discussion of discovering classes is good.
And it's nice to see all the details worked through.


FWIW I experimented with writing a compiler in Jave using recursive
descent augmented with operator precedence for expression parsing - a
class for each type in the *abstract* syntax. It looks like it would
work quite well.


You can use inheritance to support the 'standard' error recovery
mechanism (standard as described in "Algorithms+Data
Structures=Programs").


There is another thread on comp.compilers discussing using operator
precedence expression parsing within a recursive descent compiler.


SteveT
--


Post a followup to this message

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