Re: Is it just me or...

genew@mindlink.bc.ca (Gene Wirchenko) writes:


> I would like to know why I can't find a good book on compiler
>writing that isn't either:
> a) "Here's the source for a Very Tiny <language>." or
> b) a jungle of mathematical notation that is not explained in the
>text.


> Currently, I can design and write DFSAs for simple parsing. I
>can construct a parser given a syntax (say in BNF), but I get lost
>whenever I try reading about LL<whatever>, how to implement recursive
>descent, etc.


> Is there any hope that I can find some clear material?


A book I never see people talk about is


A Concurrent Pascal Compiler for Minicomputers
Alfred C. Hartmann
Springer-Verlag 1977
Lecture Notes in Computer Science, 50


I read it when I was younger and learned a lot. In particular, it
does a great job with hand-written scanners and recursive-descent
parsers. Also has interesting ideas about overall design,
concentrating on separation of concerns, along with measurements to
help lend insight.


It does _not_ describe a Pascal compiler; instead, it describes a
compiler for the language "Concurrent Pascal" which was an
experimental language done by Per Brinch Hanson (sort of a precurser
to Edison).


Other than that, my advice is to start writing your own compiler for
some relatively easy language that piques your interest (however, I
wouldn't recommend getting involved with C or C++). Maybe Oberon or
Edison or something. Or design your own. I've done it several times.
Never liked the result, but I learned a lot.


Preston Briggs
[It is, to put it mildly, out of print, which makes it hard to find if you
don't have a good university library nearby. -John]


--

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