Re: What's lacking: a good intermediate form

"Robbert Haarman" <comp.lang.misc@inglorion.net> wrote in message
> On Thu, Mar 05, 2009 at 02:43:00PM -0600, Tony wrote:
>>
>> Note that I said "intermediate form" rather than "intermediate
>> language". "Intermediate representation" (IR) is probably more common
>> than "intermediate form" though. An intermediate language is generated
>> code and a bit farther down the line than the IR I was concerned with,
>> which is probably an AST.
>
> Fair enough. But then again, an intermediate language is just an
> external representation of an intermediate form. How things are
> represented in a running program is something I've always considered to
> be up to the implementer. So what we're left with is the semantics: the
> constructs and what they mean. That is something an intermediate form
> and an intermediate language have in common.


In another post, I categorized all ILs as belonging to the backend of the
compiler and suggested not thinking about a compiler system as only a
frontend and backend but with a middle also. The frontend is the lexer and
parser (and perhaps the language specification if not being general). The
middle is the home of the IR and probably is the AST and associated API to
it. The thing that takes the AST and generates code (be it IL, assembly,
machine code, C...) from it, is the backend. Perhaps if I add that I
visualize "plugging in" the AST into a development environment and working
with it directly, source code view being only one possible representation,
that will distinguish better how I categorize "IR" and "IL" and why.


Take with a grain of salt, I didn't even own "The Dragon Book" until this
year (and I bought a used copy).


Tony

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