«

How do you fetch just the domain name part of a variable in a script? The
variable can be "http://www.domain.com/blahblah/whatever/page.htm" or
"http://sub.domain.com/blahblah/whatever/page.htm".

What I need is to extract just the "domain.com".

[removed non-existant groups, removed off topic AOL group, set followups
to c.l.p.m.]

"Shabam" <> wrote in message
news:3u-dnd1_9JRvQAncRVn-...
> How do you fetch just the domain name part of a variable in a script?
The
> variable can be "http://www.domain.com/blahblah/whatever/page.htm" or
> "http://sub.domain.com/blahblah/whatever/page.htm".
>
> What I need is to extract just the "domain.com".

Try using the Regexp::Common module from CPAN. I seem to recall it has
a method for parsing URIs

Paul Lalli

Look for URI module. IMHO, its a good and simple thing for parsing URLs

use URI;
($domain = URI->new("http://www.domain.com/blahblah/whatever/page.htm")->authority) =~ s/^www\.//i


Regards,
Andrew

Shabam wrote on 12 Ноябрь 2004 16:02:

> How do you fetch just the domain name part of a variable in a script? The
> variable can be "http://www.domain.com/blahblah/whatever/page.htm" or
> "http://sub.domain.com/blahblah/whatever/page.htm".
>
> What I need is to extract just the "domain.com".

--
Andrew

Sorry, did'nt pay attention to sub-domains in your example.
So, IMHO, it depends on your task - if it allows to guess possible
TLD values, then just split domain name into parts and leave just matched
TLD and SLD.

Regards,
Andrew

Ryan Thompson wrote on 12 Ноябрь 2004 17:38:

> [ Cross-post trimmed ]
>
> Shabam wrote to :
>
>> How do you fetch just the domain name part of a variable in a script?
>> The variable can be "http://www.domain.com/blahblah/whatever/page.htm"
>> or "http://sub.domain.com/blahblah/whatever/page.htm".
>>
>> What I need is to extract just the "domain.com".
>
> This is definitely a non-trivial problem. Fortunately, it's been
> partially solved already. I'm involved in the SpamAssassin and SURBL
> projects, where this really became obvious when spammers started
> obfuscating URIs, and using domains from many different TLDs where it
> takes a lot of research to determine where to chop the hostname to get
> the actual registrar domain.
>
> There's much more to it than using a library or regexp.
>
> See get_uri_list() in SpamAssassin 3's PerMsgStatus.pm for one
> "industrial strength" solution to this problem, which still has room for
> improvement.
>
> - Ryan
>

--
Andrew

Shabam wrote:

> How do you fetch just the domain name part of a variable in a script? The
> variable can be "http://www.domain.com/blahblah/whatever/page.htm" or
> "http://sub.domain.com/blahblah/whatever/page.htm".
>
> What I need is to extract just the "domain.com".

The problem is not well defined.

For "http://www.tacp.toshiba.com/" do you want "tacp.toshiba.com" or just
"toshiba.com"? For "http://story.news.yahoo.com", is "news" included or not?
You can't just use the last two components in all cases, such as
"http://www.toyota.co.jp" or "http://www.bbc.co.uk".

-Joe

> The problem is not well defined.
>
> For "http://www.tacp.toshiba.com/" do you want "tacp.toshiba.com" or just
> "toshiba.com"? For "http://story.news.yahoo.com", is "news" included or
not?
> You can't just use the last two components in all cases, such as
> "http://www.toyota.co.jp" or "http://www.bbc.co.uk".

What I would need is just the domain name part. In this case it would be
"toshiba.com" only. No subdomains. My domains will be simple
(com/net/org), so complicated situations like "toyota.co.jp" wouldn't apply.

Shabam wrote:

>>The problem is not well defined.
>>
>>For "http://www.tacp.toshiba.com/" do you want "tacp.toshiba.com" or just
>>"toshiba.com"? For "http://story.news.yahoo.com", is "news" included or
>
> not?
>
>>You can't just use the last two components in all cases, such as
>>"http://www.toyota.co.jp" or "http://www.bbc.co.uk".
>
>
> What I would need is just the domain name part. In this case it would be
> "toshiba.com" only. No subdomains. My domains will be simple
> (com/net/org), so complicated situations like "toyota.co.jp" wouldn't apply.
>
>
I m not an expert, but the following regex will apply:

$url = "http://www.abc.xyz.toy-0-ota.com";
($domain) = ($url =~ /http:\/\/.*\.([0-9a-zA-Z\-]+\.com|net|org)/);
print $domain . "\n";

Sam