Skip site navigation (1)Skip section navigation (2)
Date:      Mon, 2 May 2011 09:23:43 -0700
From:      Garrett Cooper <yanegomi@gmail.com>
To:        Arnaud Lacombe <lacombar@gmail.com>
Cc:        freebsd-hackers@freebsd.org, =?ISO-8859-1?Q?Martin_M=F6ller?= <moeller.akt@googlemail.com>
Subject:   Re: [LIBC] Modfied Version of sscanf
Message-ID:  <BANLkTimS%2BU7D7yZ8gO7Pw3ZOQC35_c8V-w@mail.gmail.com>
In-Reply-To: <BANLkTinJVG=%2BUq5rc=FdS6WueQSJmDxuaA@mail.gmail.com>
References:  <20110501162925.GB47497@stack.nl> <C9E46C77.389%moeller.akt@googlemail.com> <BANLkTinJVG=%2BUq5rc=FdS6WueQSJmDxuaA@mail.gmail.com>

next in thread | previous in thread | raw e-mail | index | archive | help
On Mon, May 2, 2011 at 8:57 AM, Arnaud Lacombe <lacombar@gmail.com> wrote:
> Hi,
>
> On Mon, May 2, 2011 at 8:13 AM, Martin M=F6ller
> <moeller.akt@googlemail.com> wrote:
>> Hello,
>>
>> Thanks for all the replies.
>> We have so far discovered the following suggetions for the parsing Probl=
em:
>> Using:
>> =A0 =A0o a tokenizer/parser is too much overhead for such a simple task

Explain :).

>> =A0 =A0o strchr, memchr is too low-level and not elegant enough

You forgot strstr ;D!

>> =A0 =A0o strtok would not even parse (tokenize) this simple example

Yes, it could. It just wouldn't be smart to set the delim up as " "
more than once ;D (wrong tool for the job)...

>> =A0 =A0o a regexp library: How would you solve the problem with a regexp=
 lib ?

'^[A-Z]{3,4} (.+) HTTP/[\d\.]+$'

Example:

$ python
>>> import re
>>> r =3D re.compile('^[A-Z]{3,4} (.+) HTTP/[\d\.]+$')
>>> m =3D r.match('GET blah_HTTP/1.1.1.1.1.1_whatever HTTP/1.1')
>>> m.group(1)
'blah_HTTP/1.1.1.1.1.1_whatever

I don't think it's what you wanted earlier -- but you can inspect the
request string further if you wish (and should!).

Oh yeah, and although I used python as a quick and dirty means for
illustrating my example, there's no reason why you can't use C, C++,
Java, perl, ruby, etc to articulate what you need.

Point is, there are a ton of different solutions available, and the
fact that one is homing in on sscanf as the tool to solve the problem
is probably ill put, just like using strtok/strsep wouldn't be the
best choice either for this problem by itself.

Cheers,
-Garrett



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?BANLkTimS%2BU7D7yZ8gO7Pw3ZOQC35_c8V-w>