Repeat counts

[arpruss]

One of the recent commits broke repeat counts. The cause seems to be that right now io.c translates all number keys to movements. It also doesn't translate keypad left, right, up and down to movements, like the code used to.

[atsb]

Correct, I altered io.c a bit so that it confirms with ASCII and is standards compliant instead of converting chars around.

[arpruss]

Is there a different way to enter a repeat count for a command now, e.g., to wait 9 ticks?

[atsb]

Show me exactly the problem area and I'll get that done.

[arpruss]

The easiest way to see the problem is to put in something like "9h". This should result in moving 9 squares left. But instead it moves diagonally upper-right one square for the "9", and one square left for the "h".

I used to be able to wait for 20 units of time by inputting "20." (this was useful when blind or confused). Now when I do that, the "2" moves me down, the "0" does nothing, and presumably the "." just waits for one unit of time. I can't even do single digit repeats.

The code for repeat number parsing is still there in tok.c, but since io.c translates numerals to movements, the tok.c repeat number parsing code never triggers.

[atsb]

Pushed a fix (at least on my keyboard). Give it a shot, it happened due to me changing a few of the PDCurses/NCurses control sets and I didn't put that into the lex code.

[arpruss]

Pressing 4j goes one square left and down instead of going 4 squares left as it should. That's on my PI (my one working Linux system right now).