regexp

[Jayme Assuncao Casimiro]

I have this piece of html text from Amazon.com

<dt><b><a
href="/exec/obidos/ASIN/0965761762/qid=917872216/sr=1-1/002-1496444-0064804">1
Business, 2 Approaches : How to Succeed in Internet Business by Employing
Real-World Strategies</a></b>
~ <NOBR><font color=#990033>Usually ships in 2-3 days</font></NOBR><dd>
Ron Gielgun / Hardcover / Published 1998
<br>
Our Price: $13.97 ~ <NOBR><font color =#990033>You Save: $5.98
(30%)</font></NOBR>
<br>
<a
href="/exec/obidos/ASIN/0965761762/qid=917872216/sr=1-1/002-1496444-0064804"><i>Read
more about this title...</i></a>
<p>

And I would like to use only one regexp to extract the title, the price,
and the desconunt in percent.

On the above example it would be:
title = 1 Business, 2 Approaches : How to Succeed in Internet Business byEmploying
Real-World Strategies
Price = $13.97
Descount = 30%

I have used:
($title) = $_ =~ m{<a.*?>(.*?)</a>};
($price) = $_ =~ m{.*Our Price:\s(\$?[\d\,.]+)};
($descount) = $_ =~ m{.*You Save:.*?[\d\,.]+.*?([\d\,.]+)};

But I would like to use only one regexp.

Thanks
+---------------------------------------------+
| Jayme Assuncao Casimiro |
| Graduado em Ciência da Computação |
| Estudante de Mestrado em Computação |
| Universidade Federal de Minas Gerais - UFMG |
+---------------------------------------------+

 

[Gunnar Hjalmarsson]

I have used:
($title) = $_ =~ m{<a.*?>(.*?)</a>};
($price) = $_ =~ m{.*Our Price:\s(\$?[\d\,.]+)};
($descount) = $_ =~ m{.*You Save:.*?[\d\,.]+.*?([\d\,.]+)};

But I would like to use only one regexp.

So, what stops you?

($title, $price, $discount) = m{...};
------------------------------------^^^
(to be filles with the regex)

 

[David K. Wall]

I have this piece of html text from Amazon.com
[snip HTML]

And I would like to use only one regexp to extract the title, the price,
and the desconunt in percent.

Don't do that. Use one of the modules designed for parsing HTML. Using REs
to parse HTML is painful and produces easily-broken code.

 

[Gunnar Hjalmarsson]

I have this piece of html text from Amazon.com

[snip HTML]

And I would like to use only one regexp to extract the title, the
price, and the desconunt in percent.

Don't do that. Use one of the modules designed for parsing HTML.
Using REs to parse HTML is painful and produces easily-broken code.

For extracting the first link and two other parts that are not
identified by help of HTML markup? Please, David, there are more
colours in this world than black and white. ;-)

perlfaq9 is less rigid:

http://www.perldoc.com/perl5.8.0/pod/perlfaq9.html#How-do-I-remove-HTML-from-a-string-

http://www.perldoc.com/perl5.8.0/pod/perlfaq9.html#How-do-I-extract-URLs-

 

[David K. Wall]

I have this piece of html text from Amazon.com

[snip HTML]

And I would like to use only one regexp to extract the title, the
price, and the desconunt in percent.

Don't do that. Use one of the modules designed for parsing HTML.
Using REs to parse HTML is painful and produces easily-broken code.

For extracting the first link and two other parts that are not
identified by help of HTML markup? Please, David, there are more
colours in this world than black and white. ;-)

Yeah, you're right. <insert standard excuses>. Thanks for the reality
check.

 

[David K. Wall]

I have this piece of html text from Amazon.com

<dt><b><a
href="/exec/obidos/ASIN/0965761762/qid=917872216/sr=1-1/002-1496444-00648
04">1 Business, 2 Approaches : How to Succeed in Internet Business by
Employing Real-World Strategies</a></b>
~ <NOBR><font color=#990033>Usually ships in 2-3 days</font></NOBR><dd>
Ron Gielgun / Hardcover / Published 1998
<br>
Our Price: $13.97 ~ <NOBR><font color =#990033>You Save: $5.98
(30%)</font></NOBR>
<br>
<a
href="/exec/obidos/ASIN/0965761762/qid=917872216/sr=1-1/002-1496444-00648
04"><i>Read more about this title...</i></a>
<p>

And I would like to use only one regexp to extract the title, the price,
and the desconunt in percent.

I still think you should use one of the HTML parsing modules to get the
otherwise unremarkable piece of HTML, but below is one regex that captures
all three things. Ugly and fragile.

my ($price, $title, $discount);
if ($html =~ m{
<dt>\s*
<b>\s*
<a\s+href\s*=\s*"\S+">
([^<]+) # title
</a>\s*
</b>
.*?
Our\s+Price:\s+
(\S+) # price
.*?
You\s+Save:\s+\S+\s+
\((\S+)\) # discount
}xs )
{
($title, $price, $discount) = ($1, $2, $3);
$title =~ s/\s+/ /g;

print "title: $title\n\n";
print "price: $price\n\n";
print "discount: $discount\n";

}