[Tutor] Download file from the web and store it locally.

[Danny Yoo]

On Mon, 22 May 2006, MATATA EMMANUEL wrote:

> import urllib
> import sys
> # Get a file-like object from the web and store it locally.
> url = "http://www.co.rowan.nc.us/enviroservs/downloads.htm"
> f = urllib.urlopen(url)
> # connect to the remote
> try:
>     remote = urllib.urlopen(url)
> except:
>     print "Connot open URL"
>     sys.exit()

Don't do this kind of exception handling: it hides information about the 
exception.

If something bad happens, we want to see something other than "Cannot open 
URL".  So instead, let the exception fall through:

###########################################################
url = "http://www.co.rowan.nc.us/enviroservs/downloads.htm"
f = urllib.urlopen(url)
remote = urllib.urlopen(url)
###########################################################

Or at the very least, modify the exception handler to print out more 
informative error messages other than "An error happened".  We can use the 
'traceback' module in the Standard Library to get us more informative 
error messages.

While you're writing simple programs, I'd recommend avoiding exception 
handlers, just because if something bad happens, we need to preserve all 
the debugging information we can get.



> # Read from the object, storing the page's contents in 's' --> want to
> download parcels.zip.
> s = f.read(Parcels.zip)

What is parcels.zip?  I don't understand what this is trying to do yet.


At this point, 'f' is a file whose contents contains the stuff in

     http://www.co.rowan.nc.us/enviroservs/downloads.htm


I get the feeling you're thinking like a web browser, in the sense that 
you need to first visit the page before downloading the file.

If so, that's not it.  *grin*

If the file you're trying to download has a URL like:

     http://www.co.rowan.nc.us/enviroservs/Parcels.zip

then there is no reason to have Python visit the downloads.htm page first: 
you can go ahead and grab Parcels.zip first, unless the content provider 
has done something weird to their web server.


One other comment:

######
m = urllib.urlretrieve
######

This won't fire off.  In Python, functions that take no arguments still 
need parens to be called.