PROS! Stripping!?!? and Listing?

[brianb]

Hey guys,

Here's the deal...

There's a page at AMAZON listing all of he latest DVD releases for the month...basically what I want to do it pull all of the AMAZON AISN numbers from the page.

All of the AMAZON numbers are listed like so:

http://www.amazon.com/exec/obidos/ASIN/<NUMBER GOES HERE>/amazon

So, say there are 50 items on this page, all of the links look exactly the same as above, the only variable being the <NUMBER GOES HERE> part.

I have a script I used in the past to start grabbing at a certain point, then stop grabbing at a certian point. But I need to somehow loop this so that it grabs ALL of these numbers on the page.

Any ideas on how to do so?

Any coding help is appreciated! Thanks in advance!

[diego25]

I think preg_match_all() is a good function to use in this case...

Diego

[brianb]

Ok...hmm...not too sure how to use it. Can you help?

[feldon23]

Damn it took me a long time just to get this far and it's still not quite working yet, but...

<?php

$httpfile = implode ('', file ('http://www.amazon.com/exec/obidos/subst/home/gateway.html'));

$location = 0;
$newlocation = 0;

do {
$location=$newlocation+1;
$location=strpos($httpfile,"obidos/ASIN/",$location)+12;
$newlocation=strpos($httpfile,"/",$location);
echo "ASIN# ".substr($httpfile,$location,$newlocation-$location)."<br>\n";
} while ($location != $newlocation);

?>

[diego25]

That's a lot of code for just grabbing a few numbers 🙂

I was talking more about like this:

// 5 lines for reading webpage into a string, say $str

preg_match_all('#http://www.amazon.com/exec/obidos/ASIN/(\d*)/amazon#',$str, $matches);

Then do a
print_r($matches);

and you'll see your IDs there.

Diego

[brianb]

DAMN!

Still not working...check it out:

(this board screws up the code, so I'm posting the file)

The $url is coming from a form which points to this script....this is so I can retrieve info from any page I want.

So...what's the deal? Any ideas?

[brianb]

Here's the file...

[feldon23]

By the way, I could not get Diego's code to work.

[diego25]

I looked at Amazon's page's source code, and I couldn't find what you were looking for. The closes thing I found were these 2 links:

/exec/obidos/ASIN/0060505389/ref=pd_nfy_gw_anr_a_b/002-6138010-1884040

and

/exec/obidos/ASIN/0684837013/pd_gw_msb2/002-6138010-1884040

But as you can see, it doesn't start with

"http://www.amazon.com"

and end with

"/amazon"

like you said, that's why my regex didn't work. The weird part is, I adjusted my regexp to
'#/exec/obidos/ASIN/(\d*)#'
but it couldn't find the first link....and I can't figure out why 😛 But it does get the second one...

Diego

[feldon23]

What does the # do?

[diego25]

It's just a delimiter....normally, people use '/', i.e. '/regex/', but since I had other '/' inside the regex, I decided to use another delimited so that I didn't have to escape the slashes inside. So, if I had to use '/' as delimited, it would look like:

'/\/exec\/obidos\/ASIN\/(\d*)/'

but that's too many slashes for me 🙂 too hard to read...

Diego

[diego25]

Actually, this regexp works:

'#/exec/obidos/ASIN/(.*?)/#m'

Diego

[brianb]

Finally!

It's spitting out an array now!

Now, the only thing is that it's grabbing some duplicates...PLUS, I wanted to spit out nice and pretty JUST THE AMAZON AISN number...so ARRAY [1] I guess.

How would I throw that into a loop?

Here's what the code is spitting out:

[diego25]

Maybe something like

foreach ($array[1] as $value) {
print $value;
}

Diego

[brianb]

Hey Diego,

Here's my script:

<?PHP

// Open the URL and grab contents
$file = fopen ($url, "r");
$contents = fread ($file, 1000000); // Read around 1 MB
fclose($file);

preg_match_all('#http://www.amazon.com/exec/obidos/ASIN/(.*?)/amazon#',$contents,$matches);

foreach ($array[1] as $value)
{
print $value;
}

?>

Here's the error I get:

Warning: Invalid argument supplied for foreach() in /home/lightso/public_html/dvd/admin/release_date/test.php on line 10

Any ideas on how to spit each ASIN out like so:

B00007G1YQ<br>
B000065464<br>
B00SDG7841<br>

etc etc etc??

[brianb]

Actually...never mind!

I got it....now, I'm getting abunch of duplicates...probably cause of img links and whatnot.

Any ideas on how to get rid of dupes and only print the listing once?

[diego25]

Can you post your current script, so I can see what dupes you're getting?

Diego

[Weedpacket]

Might not the array_unique() function help?