Re: Writing a Mac Ada compiler

[cons116@twain.oit.umass.edu (Mike White)]

Robert Dewar (dewar@cs.nyu.edu) wrote:
: Mike, at this stage, I am afraid your lack of familiarity with compiler
: technology is showing. Writing a lexical analyzer and parser for Ada is
: by comparison with the rest of the task simple, but even these very
: simple tasks are by no means trivial, especially if you want to get
: any kind of decent error messages.

Very true.  I DO lack familiarity with compiler technology.  I am an
undergrad with NO real-life experience in compiler design. I do,
however, feel that I have an understanding of most of the pieces
which go into a compiler, and was interested in seeing if I can pull
those pieces together and learn something about overall compiler design.

: As for the idea that the only remaining step is generating machine language,
: I am afraid it seems that your cource on compilers gave a rather false
: impression. The hard part of an Ada compiler is the semantic analysis.

True.  I definitely took a simplistic view of what a compiler is.  This
whole thread is based on an interest of mine which I wanted to get some
input on, and I posted some things which I did not think through.

(By the way, I have definitely decided against designing a compiler from
scratch.)

: [more]

: As for porting GCC, this is a task that is probably about one twentieth the
: effort of writing a compiler, but that too expects you to have significant
: experience with GCC and the technology. If you are starting from scratch
: and know little about compilers and processor architectures, you have a lot
: to learn before you can succeed in such a port.

: [more]

I do sincerely appreciate these comments.  I freely admit I am naive
about compilers as my coursework has only provided me with a general 
background in computer science, and not much in compilers.  However, my
coursework _has_ included courses in computer architecture, assembly
language, programming languages, software engineering, and operating
systems.  I have been programming Macs (under system 7) for about two
years now, and have been programming in all sorts of languages on all
sorts of machines for over ten years.  This may or may not be enough
experience to tackle such a chore as we are discussing here.  I am
currently taking a look at the gcc source, analyzing my abilities,
and researching compiler technology.  I am pretty honest with myself
and if I do not see myself able to complete a robust port in a
reasonable amount of time, trust me: I will not take up the project.

mike white