Re: http_get failing with 404

[Scott Klement <sk@xxxxxxxxxxxxxxxx>]

Hi Mike,

On 8/13/2013 3:05 PM, Mike Krebs wrote:
> I haven't tried it but setting the path like you are and using HTTP_URL_GET doesn't seem like it would work as HTTPAPI would not know what protocol to use (HTTP or HTTPS).

This won't help with Charlie's issue (since he was calling http_url_get 
unnecesarily), but...   In case you're curious -- what HTTPAPI would do 
is try to re-use the protocol, host name, and port of the previous 
request.    So for example, if you did 
http://example.com/somedir/somedoc.ext on your first request (post or 
get, doesn't matter) and then followed up with 
/anotherdir/anotherdoc.ext on the second request, it would go back to 
example.com and use http: to find '/aotherdir/anotherdoc.ext' on the 
same server.

This is done to emulate (as close as I can) the way a browser handles 
"relative links".  (relative urls)

And the reason HTTPAPI tries to handle them is because some sites will 
send a "redirect" code and will redirect to a relative link -- so for 
HTTPAPI to handle those redirects properly, it also needs to handle 
relative links.

One difference, of course, is that browsers handle relative urls based 
on the page that they appeared inside.  HTTPAPI has no notion of 
appearing "inside a page" since it doesn't interpret html, so I just 
made it work based on whatever the previous request was, which so far 
has worked very well for me.

-SK

-----------------------------------------------------------------------
This is the FTPAPI mailing list.  To unsubscribe, please go to:
http://www.scottklement.com/mailman/listinfo/ftpapi
-----------------------------------------------------------------------