tswhite70, would you check my spider?

[freebits]

Hello,

I was able to create my own spiders which extract music coverart from my own language site.

The thing is I could obtain small coverart but, big one is not easy for me.

Would you check my spider and advise me what I should do to retrieve the bigger picture from my site.

my url is http://www.yes24.com and it's Korean website.

The following is my spider....

url=http://www.yes24.com/searchCenter/searchResult.aspx?keywordAd=&qdomain=%C0%BD%B9%DD&query=%searchstring%

results=<a href='(?<url>/Goods/FTGoodsView.aspx\?.*?)' .*?><b>(?<display>.*?)</a>

src='(?<coverart>http://image.yes24.com/momo/.*?.jpg)'


Thanks in advance,
Freebits

[tswhite70]

Freebits - I took a look at it, tough for me since I don't read Korean, but Babelfish helped me out some.

It looks like the image link you actually want is just to the right of the small image link on the details page in a javascript reference. Looking at the following details page:
http://www.yes24.com/Goods/FTGoodsView. ... 3001001010

The small image link you are getting in your spider is:
http://image.yes24.com/momo/TopCate49/M ... 802649.jpg

The large image link is included in the javascript reference to the right of the small link in the details source and looks like this:
onClick='javascript:ZoomWindow("./FTGoodsFileView.aspx?goodsNo=1969087&CategoryNumber=003001001010&ImgUrl=/TopCate49/MidCate01/4802648.jpg"); return

The section in bold is the url of the large image. I looked at a couple different albums and it appears the large image link is always numerically one less that the small link:

small img: 4802649.jpg
large img: 4802648.jpg

I don't think we can do anything with that though since the final number is different for different albums and we can't tell Xlobby to subtract 1 from the URL . I haven't tested this (I'm not at home today), but I think it may work:

(?<coverart>ImgUrl=.*?)".; return
replace=coverart:ImgURL=:http://image.yes24.com/momo

So we pull the link portion from the javascript reference and then replace the "ImgURL=" with "http://image.yes24.com/momo". I'm not sure what Steven's code will do with the ":" in "http://image", hopefully it won't assume it's another variable in the replace statement.

The other option if the replace doesn't work is to use the full javascript reference as a URL call and then get the link direct from the javascript page, something like:

onClick.'javascript:ZoomWindow."(?<url>.*?)".; return
<span id="LB_CENTER_IMG"><img src='(?<coverart>.*?)' onClick=

EDIT: The second option worked, I've posted the spider in the "Spiders" post so others can find it:
http://www.xlobby.com/forum/viewtopic.php?p=27121#27121

Freebits - let me know if it works for you!

good luck,
tsw

[freebits]

Hello tsw,

What a great work!!! Thanks alot!!!

The hardest party for me was also how to replace the number in image file name. You avoid this problem and did a great job!!! Thank you.

I'm making a Korean verson of Xlobby package with language localization, local spiders, Ant movie catalog with Korean scripts, etc. your help boosts my job alot!!

Regards,
Freebits