Problem using download-multipage-list

[Rodrigo Rech]

Hello, first let me say that i'm really enjoying Web Harvest, good job!

But i'm having one problem with the download-multipage-list function (using the google_images.xml example). If i define more than one word as the search variable, such as:
<var-def name="search">iron maiden</var-def>
the following URLs are "harvested":
1) Downloaded: http://images.google.com/images?q=iron maiden&hl=en&btnG=Search+Images&nojs=1
2) Downloaded: http://images.google.com/images?q=iron+maiden&nojs=1&svnum=10&hl=en&lr=&start=20&sa=N
3) Downloaded: http://images.google.com/images?q=iron%2Bmaiden&nojs=1&svnum=10&hl=en&lr=&start=40&sa=N
4) Downloaded: http://images.google.com/
And then always the same URL as number 4.

Notice that the space character first is codified to + and after to %2B, is this behavior correct? The function only returns the images of URLs 1 and 2.

Thanks for any help!
Rodrigo

[zoombongo]

Problems with google_images example. I'm using webharvest0261.jar

Config file is not harvesting multpie pages. First page is harvested fine

Downloaded: http://images.google.com/images?q=platon&hl=en&btnG=Search+Images&nojs=1, mime type = text/html, length = 25915B.

but the next 4 pages are harvested as:

Downloaded: http://images.google.com/, mime type = text/html, length = 4334B

Any help would be appreciated.

[Vladimir Nikic]

I have tried and realized that in some cases it works and in some not. The problem is in new version of TagSoup dependant library which I added in version 0.26. I have found also some other bugs regarding TagSoup. I'll consider some other library for html clean up.

[zoombongo]

Thank you for your consideration.

[Vladimir Nikic]

Yes, you are right. There is some bug with encoding URLs. I'll fix it as soon as possible.
Thanks for your report.

[Vladimir Nikic]

Bug is fixed now in version 0.26.

[Rodrigo Rech]

Sorry for the delay, i was on vacation.

Thanks a lot, now it's working fine :-)