LoginLogin
Might make SBS readonly: thread

Syntax Highlighting

Root / Submissions / [.]

MZ952Created:
Download:73XKNXCX
Version:Size:
Whoops meant to upload this like forever ago. Custom syntax highlighter geared towards highlighting SmileBASIC code. I tried to optimize it for highlighting other custom languages, and, well, it can be done, but it's not perfect. I took a few shortcuts that seats it rather firmly at the moment in SB syntax.

Instructions:

Instructions for using the program in your own in the file.
ProgramsGames
MZ952#1
I see a whole boat load of ways this could be improved.
12Me21#2
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
MZ952#3
Replying to:12Me21
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
Already fixed the first, thought I already fixed the last lol thanks Edit: wait, sorry, .E? Like, 
.E1
for example? That's not legal code, or it doesn't seem to be. Stuff like 
1.0E1
.1E1
is though.
MZ952#4
Replying to:12Me21
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
There was also a bug with &H... and &B... when preceded by uh alphanumerics, fixed that too. I'm trying to create a new better version that allows the user to like select which types of tokens are highlighted and whatnot. I also want it to be faster.
12Me21#5
Replying to:12Me21
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
code like this: 
?.END
should be parsed as ?. END rather than ? .E ND (E can't be directly after . in number literals, so it's treated as a separate token) Note that ?1END is not valid in SB though
MZ952#6
Replying to:12Me21
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
All of the code is weird. Yeah, I know it's better to just straight return the raw logical condition. Guess I didn't optimize it further after getting it working.
MZ952#7
Replying to:12Me21
bugs: - ? next alphanumeric chars isn't highlighted - .E is allowed in numbers Weird code:
DEF ISALPHNUM(C$)
 DIM C0=C$=="_"
 DIM C1=C$>="0",C2=C$<="9"
 DIM C3=C$>="A",C4=C$<="Z"
 DIM C5=C$>="a",C6=C$<="z"
 RETURN C0||C1&&C2||C3&&C4||C5&&C6
END
'vs
DEF ISALPHNUM(C$)
 RETURN C$=="_"||C$>="0"&&C$<="9"||C$>="A"&&C$<="Z"||C$>="a"&&C$<="z"
END
'etc
SB is very very good at optimizing comparison chains like this; here's a disassembly of the bytecode generated:
      ==  param(0),"_"
      JF  @RET 
      >=  param(0),"0"
      JNF @30
      <=  param(0),"9"
      JF  @RET
@30:  >=  param(0),"A"
      JNF @4C
      <=  param(0),"Z"
      JF  @RET
@4C:  >=  param(0),"a"
      JNF @RET
      <=  param(0),"z"
@RET: MOVFLAG ->return
      ENDDEF
(JF: jump if flag set (if comparison was true) JNF: jump if flag not set (if compare was false)) vs your function: 
== param(0),"_"
MOVFLAG ->local(0)
>= param(0),"0"
MOVFLAG ->local(1)
<= param(0),"9"
MOVFLAG ->local(2)
>= param(0),"A"
MOVFLAG ->local(1)
<= param(0),"Z"
MOVFLAG ->local(2)
>= param(0),"a"
MOVFLAG ->local(1)
<= param(0),"z"
MOVFLAG ->local(2)
     != local(0),0
     JF @RET
     != local(1),0
     JNF @7E
     != local(2),0
     JF @RET
@7E: != local(3),0
     JNF @96
     != local(4),0
     JF @RET
@96: != local(5),0
     JNF @RET
     != local(6),0
@RET:MOVFLAG ->return
     ENDDEF
Ah I see. I'll probably handle the whole numerics dilemma a bit different in my next iteration. Numbers have .... versatile syntaxes.