W3C home > Mailing lists > Public > www-validator@w3.org > October 1999

Re: Infinite loop bug in the validator...

From: <Jukka.Korpela@hut.fi>
Date: Tue, 12 Oct 1999 12:31:51 +0300 (EET DST)
To: www-validator@w3.org
Message-ID: <Pine.OSF.4.10.9910121210130.12673-100000@beta.hut.fi>
On Tue, 12 Oct 1999, Julien PERVILLE wrote:

> Try to validate
> www.premier-ministre.gouv.fr
> the validator starts an infinite loop
- -
>                                      Non SGML character number 0

The document at http://www.premier-ministre.gouv.fr/ as sent by the server
contains a large number on NUL characters (octet/byte where all bits are
zero). Such data is typically ignored by browsers but it a validator
reports it as a "character" which is not allowed in an HTML document.
(Naturally sending NULs is waste of bandwidth.)

Removing the NUL characters helps, though the validation report is still
hard to read, since the lines in the document are _very_ long.

And it's not an infinite loop, just rather long. Each occurrence of the
NUL character is reported separately. 

Perhaps the validator should be prepared to such problems, as a practical
measure against "authoring tools" which produce long strings of NULs.
An ad hoc test perhaps? After getting a NUL, skip all consecutive NULs
after it, just counting them and then reporting something clever, and
giving a polite hint to get a decent authoring tool. :-)

-- 
Yucca, http://www.hut.fi/u/jkorpela/ or http://yucca.hut.fi/yucca.html
Received on Tuesday, 12 October 1999 05:31:54 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:13:52 GMT