Download and save a picture - urllib

[MRAB]

mattia wrote:
> You were right, the problem was with the print function, using a normal 
> write everythong works fine.
> 
You should open the output file as binary (it doesn't matter on
Linux/Unix, but is a good idea anyway for portability).

> Il Thu, 10 Sep 2009 18:56:07 +0200, Diez B. Roggisch ha scritto:
> 
>> mattia wrote:
>>
>>> Hi all, in order to download an image. In order to correctly retrieve
>>> the image I need to set the referer and handle cookies.
>>>
>>> opener = urllib.request.build_opener(urllib.request.HTTPRedirectHandler
>>> (), urllib.request.HTTPCookieProcessor())
>>> urllib.request.install_opener(opener) req =
>>> urllib.request.Request("http://myurl/image.jpg")
>>> req.add_header("Referer", "http://myulr/referer.jsp") r =
>>> urllib.request.urlopen(req)
>>> with open("image.jpg", "w" ) as fd:
>>>     print(r.read(), file=fd)
>>>
>>> I'm not able to correctly save the image. In fact it seems that it it
>>> saved in hex format. Any suggestion?
>> How do you come to the conclusion that it's saved as "hex"? It sure
>> isn't - either the request fails because the website doesn't allow it
>> due to missing cookies or similar stuff - or you get the binary data.
>>
>> But you should be aware that in the interpreter, strings are printed out
>> with repr() - which will convert non-printable characters to their
>> hex-representation to prevent encoding/binary-data-on-teriminal-issues.
>>