comp.lang.ruby

Hannes Rammer

11/3/2008 11:33:00 AM

hmmm noone replied... well in case someone hase the same problem.. i
have found the solution



q = "http://www.basketball-bund.net/i...{search_string}"
agent = WWW::Mechanize.new
doc =
agent.get("http://www.basketball-bund.net/index.jsp?Action=100&Verband...)
doc = agent.get(q)

doc = doc.search('body').to_html
#convert iso15 to utf8
doc = Iconv.iconv("UTF-8", "ISO-8859-15", doc).to_s
#make it hpricot
doc = Hpricot(doc)
##end crawling
@q = q
form = doc.search("//form")

it seems that its because of cookcies or something.. that i needed to
reload the page once before i try to do my own search.. thats why i call
the agent.get twice

hope this helps anyone


Hannes Rammer wrote:
> Hi when i get this url
>
> http://www.basketball-bund.net/index.jsp?Action=100&V...
>
> it shows a table with basketball info at page 1
>
> if i want to go to page 2
>
> this is the working url
>
> http://www.basketball-bund.net/index.jsp?Action=100&V...&startrow=10
>
> like it says it starts with the 10th result...
>
> wellif i enter the url into the browser address bar it works fine.. but
> when i look for the html in Hpricot it just returns the first page..
>
> ive found out that if the startrow bit is wrong then it always shows the
> first page.. but itseems to be right as its working in the browser...
>
>
> i got the same problem using
>
> URI.parse
>
>
>
> here is my code
>
>
> q =
> 'http://www.basketball-bund.net/index.jsp?Action=100&V...&viewid=&startrow=10'
>
>
> f = open(q)
> f.rewind
> doc = Hpricot(Iconv.conv('utf-8', f.charset, f.readlines.join("\n")))
> form = doc.search("//form[@name=ligaliste]")
>
>
> can anyone help me pls
>
> thx

--
Posted via http://www.ruby-....