Re: GNU-NYU Ada project

[cis.ohio-state.edu!magnus.acs.ohio-state.edu!usenet.ins.cwru.edu!agate!do]

In article <1992Nov24.161852.669@lambda.msfc.nasa.gov> robichau@lambda.msfc.nas
a.gov writes:
>In <61990007@acf3.NYU.EDU> schonber@acf3.NYU.EDU (Ed Schonberg) writes:
>>The figure of 1,000,000 lines per minute on a 16 Mhz PC is correct. It
>>is the the performance of a syntax checker, i.e. the program does not
>>build a parse tree, but does diagnose syntax errors (and recovers from
>>them). This performance is due to several ingredients: 
>
>>a) the program is written in assembly language, with global register 
>>assignments that code  generators are not able to duplicate. 
>
>>b) The program to be parsed is fully in  memory, so the lexical scanner 
>>does not have to check for, nor perform, any I/O. 
>
>This strikes me as slightly curious. Although its performance is
>impressive, this variant of the syntax checker will be helpful only to
>those using PCs.

There seems to be some confusion here. What Ed is referring to is a previous
Ada parser written by Robert Dewar, which was written in assembly for the
PC. The parser part of GNAT is *based upon* this design, but it is *not*
coded in 80x86 assembler, nor PC specific -- it's written in portable Ada.

>-Paul

Mat

| Mathew Lodge                      | "I don't care how many times they go    |
| mjl-b@minster.york.ac.uk          |  up-tiddly-up-up. They're still gits."  |
| Langwith College, Uni of York, UK |  -- Blackadder Goes Forth               |