This commit improves Python 3 compatibility and fixes two multidict
issues:
1. Headers.items(multi=True) now decodes fields
2. MultiDict.clear(item) has been removed, as Python's
MutableMapping already defines .clear() with different
semantics. This is confusing for everyone who expects a
dict-like object. `.pop("attr", None)` is not fantastic,
but it's the Python way to do it.
This PR improves our handling of HTTP message body encodings:
- The unaltered message body is now accessible as `.raw_content`
- The "content-encoding"-decoded content (i.e. gzip removed) content
is not `.content`, as this is what we want in 99% of the cases.
- `.text` now provides the "content-encoding"-decoded and then
"content-type charset"-decoded message body.
- The decoded values for `.content` and `.text` are cached,
so that repeated access and `x.text = x.text` is cheap.
- The `decoded()` decorator is now deprecated, as we can now just use
`.content`. Similarly `HTTPMessage.get_decoded_content()` is
deprecated.
- Adds default implementations for _kconv and _reduce_values to MultiDict.
Without these, operations fail in really, really non-obvious ways.
- Replace the remaining few instances of ODict
Fixes#1159
- Move more stuff that belongs in netlib.human
- Move some stuff to near the only use
- Zap mitmproxy.utils.timestamp(). I see the rationale, but we used it
interchangeably with time.time() throughout the project. Since time.time()
dominates in the codebase and timestamp() is such low utility, away it goes.
This is just inherently not a determinisitc test. We don't use the log HTTP
interface any more, so it can just go. A more radical "solution" is inbound
shortly.
Fixes#1207
