|
7 years ago | |
---|---|---|
Types | 7 years ago | |
Views | 7 years ago | |
docfx | 7 years ago | |
docs | 7 years ago | |
ecmadocs | 7 years ago | |
.gitignore | 7 years ago | |
Core.cs | 7 years ago | |
Driver.cs | 7 years ago | |
Event.cs | 7 years ago | |
LICENSE | 7 years ago | |
Makefile | 7 years ago | |
README.md | 7 years ago | |
TODO.md | 7 years ago | |
Terminal.csproj | 7 years ago | |
Terminal.sln | 7 years ago | |
XmlYamlMapping.json | 7 years ago | |
demo.cs | 7 years ago |
This is a simple UI toolkit for .NET.
It is an updated version of gui.cs that I wrote for mono-curses
Go to the API documentation for details.
To run this, you will need a peer checkout of mono-ncurses to this directory. I should make a NuGet package at some point.
The input handling of gui.cs is similar in some ways to Emacs and the Midnight Commander, so you can expect some of the special key combinations to be active.
The key ESC can act as an Alt modifier (or Meta in Emacs parlance), to allow input on terminals that do not have an alt key. So to produce the sequence Alt-F, you can press either Alt-F, or ESC folowed by the key F.
To enter the key ESC, you can either press ESC and wait 100 milliseconds, or you can press ESC twice.
ESC-0, and ESC_1 through ESC-9 have a special meaning, they map to F10, and F1 to F9 respectively.
Currently gui.cs is built on top of curses, but the console driver has
been abstracted, an implementation that uses System.Console
is
possible, but would have to emulate some of the behavior of curses,
namely that operations are performed on the buffer, and the Refresh
call reflects the contents of an internal buffer into the screen and
position the cursor in the last set position at the end.
There are some tasks in the github issues, and some others are being tracked in the TODO.md file.
The original gui.cs was a UI toolkit in a single file and tied to curses. This version tries to be console-agnostic (but currently only has a curses backend, go figure) and instead of having a container/widget model, only uses Views (which can contain subviews) and changes the rendering model to rely on damage regions instead of burderning each view with the details.