Interpreting SSH output

Question 1

When a program displays "static text" or is otherwise exercising control over the whole terminal display, it is actually printing commands to the terminal using hidden control characters (like those shown in your blob). See answer here. Libraries like ncurses provide a portable API for terminal control by getting information about the host terminal from terminfo.

In your project, I suppose you'll have to obtain some kind of dictionary of the host's command sequences and filter or interpret them to display properly when printed to the client's terminal. You might take a look at the colorama project, which translates ANSI command sequences into Win32 calls to properly control a Windows command terminal. But that's for Python, which I know nothing about.

Question 2

You need to write a terminal emulator. Welcome to the weirdness that is Unix terminals. You might want to look at 'man terminfo', 'man termcap' and 'man ncurses' for something of an introduction.

Effectively what a terminal window or console on any Unix machine does, is emulate an old-school serial terminal, which itself was a layer of features on top of an emulation of a teletype machine. So all of those ESC[? sequences are commands to the terminal requesting it to do things. Some of them will be expecting responses to give things like the window size, state of caps lock, etc. The terminfo database records what all the escape sequences are for a given terminal type (you should probably start with the type called 'xterm-color').