HLL syntax & structure suited to rapid compilation?

I understand that there are various methods used to speed up
compilation, such as the way memory is used and the parsing methods
employed.


I'd like to ask about the actual syntax of an HLL and how that might
increase compilation. In x86 assembly, for instance, the mnemonics
relate one-to-one with machine instructions. Compilation would be,
presumably, straight translation.


In languages like c, pascal and basic and in specific compilers
therein there is the effort of translating, intelligently, the
syntactically constructed code a programmer writes.


As anecdotal experience, and from what others say, pascal compilers
have a reputation for speedy compilation, c is middling and basic
relatively slow.


First - does that hold true in a generalised way?


Second - If there are consistent differences between HLL compilation
speeds, why? Is this down to an HLL lowering ambiguity in it's
syntactical rules and rigour in data typing (something pascal seems to
do well), or is that really a secondary matter - and compilation speed
is down to algorithms, memory usage and raw cpu cycles (i.e. the
compiler itself being written to benefit from cpu characteristics)?


Thank you.
[I think the difference in speeds has much more to do with the goals of
the compiler than the language being compiled. -John]

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