[timoriensis]

Hi,
I'd like to fetch some part of a big file at a URL, but not all the data always. Thus just downloading everything sounds wasting of resources and user's time.

I thought that I could open it, e.g.,
file_in = urllib2.urlopen(file_input_name)

and then use read & seek:
data_in1 = file_in.read(read_some_bytes)
seek to point2 in file_input_name
data_in2 = file_in.read(read_some_bytes_again)
seek to point3 in file_input_name
etc

but after some trying and crying I found discussion at:
http://mail.python.org
"seek() is not a stream operation. It is a random access operation (file-like != stream).... HTTP resources are in essence streams so they implement read() but not seek()... "

Thus sounds good clarifications. But as you guess, what is then the used way to get seek-functionality working in pyS60?
Should I use other methods to open than urllib2.urlopen?

This might be very basic question but I already spend some time with this and did not find anything related that was really useful. Of course I found a lot of other useful and interesting python stuff

[Rafael T.]

Hi timoriensis,

I didn't got you point in this question. Do you want to search a text in the URL?

If yes, you can do this:

url = "http://www.forum.nokia.com"

if url.find(u"nokia") == -1:
    
    print "nokia not found on url"

else:
    print "nokia found on url"

So, this will try to find the text "nokia" on the url variable. If it is -1, the text was not found.



Hope I got it,

Rafael.

[aaaaapo]

Originally Posted by timoriensis

Hi,
I'd like to fetch some part of a big file at a URL, but not all the data always.

If your web server supports HTTP byte range, you can ask server to start serving file from any byte you want
(Range: bytes=nnnnnnn- header). These documents explain something about this:

http://www.lassosoft.com/documentati...dex.lasso?9233
http://en.wikipedia.org/wiki/Byte_serving

I am not sure if you can request multiple chunks at one request, but it may be possible.