This is a multi-part message in MIME format.

------=_NextPart_000_002C_01C7688F.1BEA4B20
Content-Type: text/plain;
	charset="Windows-1252"
Content-Transfer-Encoding: 7bit

On Friday, March 16, 2007 8:23 PM [GMT-3=CET],
William Chops Westfield  wrote:

> > The start bit idea is clever! It allows a character to have
> > up to 7 dits or dahs.
>
> Heh.  The motorola "Black Widow" design contest (for their 8pin
> cpus, essentially) has as one of the "steps" to finish a morse
> code program.  The core that they supply stores the translation
> table as STRINGS:
>
> > > > char *convert(char c){    //Convert ASCII received switch(c){
> > > > case 'a':
> > > > return ".-";
> > > > case 'b':
> > > > return "-...";
> > > > case 'c':
> > > > return "-.-.";
> > > >
> I guess that says something about changes in attention paid to
> code size by "youngins" these days.  And C vs ASM, perhaps.
> Although since the high-end 8pin chip contains 8kbyte of flash
> (more than any other 8pin uC?) I guess it doesn't matter.
>
> > SendMorseCharA
> >        RLF   MorseChar,f  ; shift msb into carry
> >        movf  MorseChar,f  ; test to see if we're done
> >        bz    SendMorseCharZ ; We're done, get out
> >        skpc               ; skip if dit
> >        call  DoDah        ; go send dah - PRESERVE CARRY!
> >        skpnc
> >        call  DoDit        ; go send dit
> >        call  InterCharDelay
> >        goto  SendMorseCharA ; go send next bit, if any
> > SendMorseCharZ
> >        return              ; we're outta here!
>
> Speaking of C compilers, this is a good example of the sort of
> code that I think it would be really difficult for a C compiler
> to generate; It's what I meant by "explicit use of carry."  Does
> anybody want to take a shot at C code/compiler implementation
> that comes close to outputting this?
>
> (And on the third hand, it is perhaps also an example of optimization
> that doesn't need to be done.)
>
> BillW

Bill,

This is my C approach.

//*************************************************
unsigned char MorseChar;
unsigned char i,a,dot,dash;

MorseChar= 0xED;                   // code + stop bit
dot=dash=a=0;                      //testing purpose
                                   //
for(i=0x80;i > 0x00;i >>= 1)       //rotate a 1 used as mask
   {                               //
     a |= i;                       //preparing a mask
     if(MorseChar ^ (MorseChar&a)) //xoring only lsb bits.
       {                           //
        if((MorseChar & i)> 0)     // is a one?
          {                        //
           dot++;                  //send a dot
          }                        //
        else                       //
          {                        //
           dash++;                 //send a dash
          }                        //
       }                           //
     else                          //
       {                           //
         break;                    //end of char!
       }                           //
                                   //
    }                              //
//*************************************************

lookup table (see attached file)
The dots are represented by 1's and the dashes by 0's.
A stop bit was used as Wouter pointed earlier.
The table is ordered by ASCII value minus 34, since 34 is the very first
character value.
So the Table is only 62 bytes long!. I arbitrary used
[ 11111111 91  0xFF
\ 10110100 92  0xB4
to represent Error symbol (it supoused to have no errors)and the "new"
symbol @. The older @ is a combination of AC.
Only 7 bytes of the table are unused.
The C can`t handle the error sign. So that is a particular case.:)

Regards,
Dennis Crawley.





------=_NextPart_000_002C_01C7688F.1BEA4B20
Content-Type: text/plain;
	name="morsetable.txt"
Content-Transfer-Encoding: 7bit
Content-Disposition: attachment;
	filename="morsetable.txt"

" 10110110 34  0xB6
# 00000000 35  0x00
$ 11101101 36  0xED
% 00000000 37  0x00 
& 00000000 38  0x00 
' 10000110 39  0x86
( 01001100 40  0x4C
) 01001010 41  0x4A
* 00000000 42  0x00 
+ 10101100 43  0xAC
, 00110010 44  0x32
- 01111010 45  0x7A
. 10101010 46  0xAA
/ 01101100 47  0x6C
0 00000100 48  0x40 
1 10000100 49  0x84
2 11000100 50  0xC4
3 11100100 51  0xE4
4 11110100 52  0xF4
5 11111100 53  0xFC
6 01111100 54  0x7C
7 00111100 55  0x3C
8 00011100 56  0x1C
9 00001100 57  0xC0 
: 00011110 58  0x1E
; 01010110 59  0x56
< 00000000 60  0x00 
= 01110100 61  0x74
> 00000000 62  0x00 
? 11001110 63  0xCE
@ 10010110 64  0x96
A 10100000 65  0xA0
B 01111000 66  0x78
C 01011000 67  0x58
D 01110000 68  0x70
E 11000000 69  0xC0
F 11011000 70  0xD8
G 00110000 71  0x30
H 11111000 72  0xF8
I 11100000 73  0xE0
J 10001000 74  0x88
K 01010000 75  0x50
L 10111000 76  0xB8
M 00100000 77  0x20
N 01100000 78  0x60
O 00010000 79  0x10
P 10011000 80  0x98
Q 00101000 81  0x28
R 10110000 82  0xB0
S 11110000 83  0xF0
T 01000000 84  0x40
U 11010000 85  0xD0
V 11101000 86  0xE8
W 10010000 87  0x90
X 01101000 88  0x68
Y 01001000 89  0x48
Z 00111000 90  0x38
[ 11111111 91  0xFF
\ 10110100 92  0xB4
] 00000000 93  0x00 
^ 00000000 94  0x00 
_ 11001010 95  0xCA
------=_NextPart_000_002C_01C7688F.1BEA4B20
Content-Type: text/plain; charset="us-ascii"
MIME-Version: 1.0
Content-Transfer-Encoding: 7bit
Content-Disposition: inline

-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist

------=_NextPart_000_002C_01C7688F.1BEA4B20--