-
rohieb authored
This is mostly print -> print(), removing decode("utf8") and encode("utf8") calls because internally strings are now already Unicode (vs. bytes), and changing the shebang to python3 because we're no longer compatible to Python 2. In some cases, make cryptic code better readable in the process. The invented-here version of urlencode is no longer that easy to implement, so replace it with the respective call to urllib. This changes the resulting URL quoting to uppercase, so adapt the test accordingly. Apply relative imports to tests so running nosetests does not depend on the current working directory. Signed-off-by: Roland Hieber <rohieb@rohieb.name>
ca3d9325