Version française
Home     About     Download     Resources     Contact us    

This site is updated infrequently. For up-to-date information, please visit the new OCaml website at

Browse thread
[Caml-list] (Release) ocamllex tutorial (ver 0.1)
[ Home ] [ Index: by date | by threads ]
[ Search: ]

[ Message by date: previous | next ] [ Message in thread: previous | next ] [ Thread: previous | next ]
Date: 2004-08-24 (14:55)
From: Nicolas Rougnon-Glasson <nicolas.rougnon-glasson@l...>
Subject: Re: [Caml-list] (Release) ocamllex tutorial (ver 0.1)
another interesting example is the use of several lexer functions (like 
in the ocaml lexer itself...), to analyse properly string literals and 
nested comments.

skaller wrote:

>On Tue, 2004-08-24 at 18:07, SooHyoung Oh wrote:
>>Any comment will be appreciated.
>It looks very good!
>>This tutorial is work-in-progress. The latest version can be found at
>First -- you haven't explained how to use the new argument
>feature. A good example would be to take the non-reentrant
>line counting example and show how to make it re-entrant
>by passing an object containing the counters.
>To justify not using a global variable, you can extend
>the example to allow for recursively counting #include
>files (or something similar :)
>The second thing I'd like to see here is an example
>showing the common technique of lexing an identifier
>and then using a lookup table to see if it is keyword:
>this is done because (a) it is possible and (b) it
>simplifies the lexer which otherwise needs a huge
>number of states.
>The third thing I'd like to see is a more functional example
>where the lexer isn't just executing code for side effects.
>All your examples have the lexer tail calling itself: but
>most real lexing applications do not do this, instead
>you call the lexer repeatedly from a driver loop.
>In particular Ocamlyacc calls Ocamllex to lex one
>token at a time.
>Finally -- since your tutorial is Ocaml specific,
>I'd like to see some information on lexbufs.
>Are they functional? Can I write a backtracking
>functional parser that just 'goes backwards'
>in the input stream using a lexbuf to allow
>re-lexing some input? 
>How do i cope with the fact
>that lexbufs think they're reading named
>files with lines in them .. what if this isn't so?
>This is actually a design fault which is unfortunately
>propagated to Ocamlyacc.

To unsubscribe, mail Archives:
Bug reports: FAQ:
Beginner's list: