[Twisted-web] HTTPPageDownloader closes 'file-like object', file-like object goes bye-bye

[Phil Mayers]

s s wrote:
> Hi,
> 
>     This is my first post to this list.
> 
>     I've been using Twisted for a few (like three) days.  My wife had a 
> baby two days ago, no, really.  I just got the other two kids to bed and 
> thought I'd ask...sorry if this has been answered eleven-thousand times 
> before but I did search the archives first.
> 
>     I'm writing a client app that needs to get the contents of a URL, 
> usually just a web page.
> 
>     I'm trying to use client.py's downloadPage(url, file, 
> contextFactory=None, *args, **kwargs):
> 
>     It claims to use a 'file' which, according to the docstring, can be 
> a file or file-like object.
>     
>     It uses the HTTPDownloader class to perform the actual download.
> 
>     In the pageEnd(): method of HTTPDownloader, the 'file' is closed.
> 
>     Unfortunately, for 'file-like' objects, like StringIOs, this trashes 
> the 'file'; not so useful since I need to work with the result.

Huh, so it does. That's a pretty rubbish feature of StringIO

> 
>     I realize I can subclass and override and such but it seems odd that 
> a function that is documented to use a file-like object uses a class 
> that will destroy something that's not a 'file.'

My guess is that calling "close" is to ensure that data is actually 
written to the file before returning, but I agree it's not helpful.

Since you obviously want the page as a string, I recommend you use 
t.w.client.getPage which returns the result as a deferred(string):

from twisted.internet import reactor
from twisted.web import client

def got(text):
   print "page is", len(text), "bytes"

client.getPage('http://www.google.com').addCallback(got)
reactor.run()



> 
>     I'm sure I'm missing something...or maybe nobody uses this 'cause 
> there's something way better that I haven't discovered yet.
> 
>     Clues?
> 
> Thanks,
> 
> S,
> AKA: Steve Steiner
> 
> 
>     
> 
> 
> 
> 
> _______________________________________________
> Twisted-web mailing list
> Twisted-web at twistedmatrix.com
> http://twistedmatrix.com/cgi-bin/mailman/listinfo/twisted-web