From: Erich W. <ew....@na...> - 2012-07-09 19:26:25
|
Hello Keith, On 07/09/2012 03:05 AM, Keith Amidon wrote: > {-- Sun, 08 Jul 2012 20:01:02 +0200: Erich<ew....@na...> wrote: --} > > Erich> More info below ... > >> I found a minor thing that puzzles me: > >> |C| 11|\ PORTD 7 portpin: PD.7 ( define portD pin #7) > >> |E| 12|\ PD.7 high ( turn portD pin #7 on, i.e. set it high-level) > >> |E=Illegal nested comment > >> make: *** [marker] Error 1 > > Erich> This can be worked around by adding a space before the > Erich> closing ')'. > Erich> ... > Erich> |C| 10|\ Use it this way: > Erich> |C| 11|\ PORTD 7 portpin: PD.7 ( define portD pin #7 ) > Erich> |C| 12|\ PD.7 high ( turn portD pin #7 on, i.e. set it high-level ) > > Ah, yes, I forgot I didn't take care of that. You may notice that the > line starts with |C|. To reduce the amount of data sent to the > microcontroller the script detects lines that are all comments or > whitespace and doesn't send them. > > When I originally wrote this support I was under the mistaken impression > that comments delimited by parenthesis had the following > characteristics: > > 1: the start parenthesis had to be surrounded by whitespace (true I > believe) True. > 2: the end parenthesis had to be surrounded by whitespace (false I > believe) Looking at the code: XT_CSCAN looks for character 'c' (from stack) in the input stream. " ( " is in lparenthesis.asm and looks for the next closing paren ')'. so: False. > 3: these comments could span multiple lines, similar to the way that > C "/* */" comments can (also false I believe) False. cscan will look in the current input buffer. gforth barks as well when trying to break comment over more lines. > > I forgot to go clean this up when I discovered the problem. I'll add > that to my list of things to try to get to when I get a chance to work > on the script. Alas, it doesn't look like that is going to happen this > weekend as I had hoped after all. No sweat! Folks start using your tool in unexpected ways, that's all :-) Cheers, Erich |