W3C home > Mailing lists > Public > www-validator@w3.org > April 2006

Re: problems with W3C::LogValidator

From: olivier Thereaux <ot@w3.org>
Date: Wed, 12 Apr 2006 11:43:23 +0900
Message-Id: <60916296b4280a7111ea88c466b7062a@w3.org>
Cc: www-validator@w3.org
To: Ryan King <ryan@technorati.com>

Hello Ryan,

On Apr 11, 2006, at 4:32, Ryan King wrote:
> I've installed LogValidator via CPAN on a rather vanilla Ubuntu 
> install. I'm using the sample conf file (only changing it to point it 
> at a different log directory).

Thank you very much for sending us you mail, I hope we can fix the 
issues you've been running into.

>> Use of uninitialized value in concatenation (.) or string at 
>> /usr/local/share/perl/5.8.4/W3C/LogValidator/SurveyEngine.pm line 
>> 252.

The SurveyEngine, which got contributed into the LogValidator, is a 
nice idea, but it's not very robust, as it relies a little too much on 
HTML scraping. I fixed this issue and will be sending a new version to 
CPAN soon.

In the meantime, if your goal is only to batch-validate a site from a 
collection logs, you do not need to have it run the SurveyEngine 
module, so I suggest editing your config file and comment out (add a # 
at the beginning of the line) the line:
UseValidationModule W3C::LogValidator::SurveyEngine

Actually, in most cases you will only want to have the following two 
UseValidationModule lines uncommented:
UseValidationModule W3C::LogValidator::HTMLValidator
UseValidationModule W3C::LogValidator::CSSValidator
The other are usually less useful, for a general usage of the 

> Eventually, when the errors stop, I get this:
>> Now using the HTML Validator module...
>> (This may take a long time if you have many files to validate)
> This seemingly hangs forever. I ran it for 30 minutes and it used 0 
> cpu time.

I can imagine two things happening:
- you have a really, really large number of documents in your logs, and 
it takes an awful long time for the logvalidator to process everything 
- it sleeps for one second between requests to the validator, and if 
the validator is overloaded, . The verbose output should show you if 
that's the case. Try setting different values for either 
EntriesPerLogfile, MaxDocuments or MaxInvalid. To speed things up, you 
may also consider installing an instance of the validator on your 
network, and point the logvalidator there, instead of the public 
service at validator.w3.org, which it uses as a default.

- something's timing out. launching the logvalidator in debug mode 
(option -d) should output the requests the logvalidator makes to the 
validation service, and the response headers it gets back. That would 
probably allow you to figure out what's wrong.

If the following steps give you an output you think may be caused by a 
bug in the logvalidator, could you please follow-up on the list, so 
that we can fix the bug(s) ASAP. And if you find out the cause and it's 
not a bug, I'm still curious to know how it could have been avoided.

Received on Wednesday, 12 April 2006 02:43:47 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:48 UTC