[prev in list] [next in list] [prev in thread] [next in thread] 

List:       imap
Subject:    Re: How to parse Incoming Client Requests?
From:       Abhijit Menon-Sen <ams () wiw ! org>
Date:       2004-01-16 19:52:13
Message-ID: 20040117011013.A25560 () lustre ! dyn ! wiw ! org
[Download RAW message or body]

At 2004-01-16 11:33:31 +0100, antonio@stueber.de wrote:
>
> But now I've got recursive Rules and I'm not shure if I can reproduce
> them using constants respectively in Regular Expressions.

Which "recursive rules" are you talking about?

> So those who had the same Problem parsing the incoming requests maybe
> could tell me, if Regular Expressions are a good way, or making to
> much problems.

I'd strongly recommend against trying to parse IMAP requests solely with
regular expressions. To cite just one obvious problem, you need to parse
arbitrarily-nested SEARCH expressions, and a regular expression can't do
that.

> Maybe there's a better way. I'm not shure how to handle the incoming
> messages.

Why not just write a simple recursive-descent parser by hand? You could,
for example, write functions to remove each type of token from the input
string, and then write things like:

    if ( command() == "RENAME" )
        parse_rename();

    ...

    parse_rename( char **src, char **dst )
    {
        space();                // Eat a space after the command name.
        src = astring();        // Eat a mailbox name.
        space();                // ...
        dst = astring();
    }

With suitable (simple) error handling, and translated to the language of
your choice. The grammar is small enough for this to be both easy to do,
and to maintain.

-- ams
[prev in list] [next in list] [prev in thread] [next in thread]