Need help with some very Practical Extraction

[theapeman]

Hi,

What I'm trying to do seems right up Perl's alley, but I can't get it to work. I'm using the WWW::Mechanize module to retrieve a sprawling HTML document from which I want to extract certain strings and save them. I can get this much to work:

use WWW::Mechanize;
$url = "http://someurl";
my $mechanize = WWW::Mechanize->new(autocheck => 1);
$mechanize->get($url);
my @array_of_data = $mechanize->content;
 

but now I am stuck on how to process that data.

The HTML doc is quite long, and contains numbers that I want to extract, numbers that are always preceeded by a text string that is the same each time, such as:

<a href bla bla bla>bla bla bla<random tag>mydigits=493409 834%bla bla bla<meaningless tag>bla bla</a>

where the string "mydigits=" always preceeds the desired number and is sometimes all lowercase but can occasionally look like "MyDigits=" ; where the number itself may be anywhere from one to 10 digits in length; and where "%" might literally be "%" or any other non-digit character including a space. Moreover, the desired string might appear more than once per line -- assuming Perl doesn't see the HTML doc as just one single long line of text anyway.

What I have tried is many extremely ugly variations on

my $pattern = "[Mm]y[Dd]igits=[0-9]*[^0-9]";
foreach (@array_of_data){
    if ( /$pattern/ ){
    print "$_\n";

but if I don't get an error, all I get is a spew-out of the entire HTML doc instead of what I am hoping for, which would be a printout or file that looks like:

219824
2230239084
04598
98739874
etc., etc.

or better yet, assign the output to an array that looks like:

@desired_array = ( 219824, 2230239084, 04598, 98739874);

I know I must be missing something very fundamental, so if anyone can help steer me away from the major mistakes I'm making, I'd appreciate it. Thanks.

[KevinADC]

as long as the pattern is on the same line this should work:

my @digits = ();
foreach (@array_of_data){
    if ( /mydigits=(\d+)/i ){
       print "found $1 in this line: $_\n";
       push @digits,$1;
    }
}
print "$_\n" for @digits;
 

but can be changed if the pattern is broken over multiple lines.

[theapeman]

Wow, thank you, that is very helpful. Love the i modifier for case insensitivity!

As I suspected, the HTML doc looks like one long line to Perl, so what happens is it finds the first instance, say, 123456789, prints

"found 123456789 in this line:"

followed by what to you and me looks like more than 3000 lines of HTML, then prints:

"123456789"

and then quits. But that is more than I could get it to do before, and this definitely has me pointed in the right direction, so thanks again. :)

[theapeman]

Wow, I just realized what you did with the parentheses and the $1 to extract only the digits. Awesome!

[KevinADC]

See how this works:

use WWW::Mechanize;
$url = "http://someurl";
my $mechanize = WWW::Mechanize->new(autocheck => 1);
$mechanize->get($url);
my $string_of_data = $mechanize->content;
my @digits = $string_of_data =~ m/mydigits=(\d+)/igm;
print "$_\n" for @digits;
 

if that doesn't work, change the 'm' after 'ig' to an 's'