34 lines
1.6 KiB
Plaintext
34 lines
1.6 KiB
Plaintext
|
|
This is a loose collection of todo items, in case someone else wants to start
|
|
hacking on mitmproxy. Drop me a line (aldo@corte.si) if you want to tackle any
|
|
of these and need some pointers.
|
|
|
|
Features:
|
|
|
|
- Support for content encodings. At the moment, we sidestep this by
|
|
stripping headers that indicate accepted encodings. In theory, this should
|
|
force servers to serve us content uncompressed, but that's not always the
|
|
case. Besides, we should try to modify requests as little as possible, so
|
|
direct support for deflate and gzip is the way to go.
|
|
- Add some "workspace" features to mitmproxy:
|
|
- Flow comments
|
|
- Copying/duplicating flows
|
|
- Ordering by time, size, etc. a-la-mutt (o keyboard shorcut is reserved for this)
|
|
- Post and URL field parsing and editing.
|
|
- We need a built-in graphical editor for these, that knows how to
|
|
break fields up and present them for modificatio individually.
|
|
- No-proxy mode, for when we want to just replay without serving as a proxy ourselves.
|
|
- Upstream proxy support.
|
|
- Strings view for binary responses.
|
|
- Support HTTP Digest authentication through the stickyauth option. We'll
|
|
have to save the server nonce, and recalculate the hashes for each request.
|
|
|
|
- Chunked encoding support for requests (we already support it for responses).
|
|
- Better manipulation of flows in mitmproxy: copying flows, adding
|
|
comments.
|
|
|
|
Bugs:
|
|
|
|
- In some circumstances, long URLs in list view are line-broken oddly.
|
|
- When a bug in mitmproxy causes a stack trace, we hang on exit.
|