potential DECgraphics optimization
Add a comment about something that occurred to me when fixing the ^C-during-DECgraphics-output situation several days ago. I don't think there's any compelling reason to avoid this optimization, but this only describes it without actually implementing it.
This commit is contained in:
@@ -3710,6 +3710,18 @@ g_putch(int in_ch)
|
||||
}
|
||||
(void) putchar((ch ^ 0x80)); /* Strip 8th bit */
|
||||
} else {
|
||||
/*
|
||||
* TODO?
|
||||
* for DECgraphics, we only need to switch back from the
|
||||
* line drawing character set to the normal one if 'ch' is
|
||||
* a lowercase letter or one of a handful of punctuation
|
||||
* characters (the range is contiguous but somewhat odd):
|
||||
* if (GFlag && ch >= 0x5f && ch <= 0x7e).
|
||||
* Leaving it on for other characters might result in only
|
||||
* deferring this graph_off(), but it could possibly skip
|
||||
* both this graph_off() and the need for next graph_on()
|
||||
* depending on whatever follows 'ch'.
|
||||
*/
|
||||
if (GFlag) {
|
||||
graph_off();
|
||||
GFlag = FALSE;
|
||||
|
||||
Reference in New Issue
Block a user