adding support for http byte range in http_common.do_http

[guyzmo]

I'm using the same API as tahoe-get.py and I needed to have 'very big files' supported through that API. So I patched do_http() so it has a headers dictionary as parameter in which one can add different options.

I patched also tahoe_get.get() so it supports Partial Content answers (206), and added a 'frange' variable set to None, because I didn't know what CLI option to choose for content retrieval, and did not want to change cli.py. frange shall be a list (or tuple) of two elements :

start = 12
end = 42
frange=[start, end]

[guyzmo]

the actual patch

[zooko]

Dear guyzmo:

Thank you for the patch! To accept it into trunk, it needs a unit test. I'm not sure, but I think right here would be a good place to test this functionality.