|
|
View previous topic :: View next topic |
Author |
Message |
eb7dzp
Joined: 20 Aug 2004 Posts: 2
|
const char *Dia[]={"X","Mon","Tus&q |
Posted: Fri Aug 20, 2004 3:14 am |
|
|
Dear all,
I am trying CCS C compiler. I allways have been working with Hi-tech.
I need to define a constant table of strings for Week days.
Does anybody knows if it is possible in CSS Pic Complier
Code: | const char *Dia[]={"X","Lun","Mar","Mie","Jue","Vie","Sab","Dom"};
const char *Mes[]={"X","Ene","Feb","Mar","Abr","May","Jun","Jul","Ago","Sep","Oct","Nov","Dic"};
printf("F:%2d/%3s/%2d _%3s", month_day,Mes[month],year, Dia[week_day]); |
If not, any other possibility ???
Thanks in advance.
NOTE: Sorry my English
email: eb7dzp@hotmail.com |
|
|
Ttelmah Guest
|
Re: const char *Dia[]={"X","Mon","T |
Posted: Fri Aug 20, 2004 7:31 am |
|
|
eb7dzp wrote: | Dear all,
I am trying CCS C compiler. I allways have been working with Hi-tech.
I need to define a constant table of strings for Week days.
Does anybody knows if it is possible in CSS Pic Complier
Code: | const char *Dia[]={"X","Lun","Mar","Mie","Jue","Vie","Sab","Dom"};
const char *Mes[]={"X","Ene","Feb","Mar","Abr","May","Jun","Jul","Ago","Sep","Oct","Nov","Dic"};
printf("F:%2d/%3s/%2d _%3s", month_day,Mes[month],year, Dia[week_day]); |
If not, any other possibility ???
Thanks in advance.
NOTE: Sorry my English
email: eb7dzp@hotmail.com |
The simple answer, is 'no'.
CCS, does not allow pointers to data in ROM (which a const is).
You have a number of choices:
1) Remove the 'const' declaration, and the data will automatically be copied to RAM on initialisation, and the code will work.
2) Declare the data as a two dimensional array (rather than an array of pointers), and use your own routine to step through the output strings.
3) Take advantage of the CCS shortcut, where if a constant string like this is presented to a routine that accepts single character input, the compiler will automatically 'walk through' the string to send it a character at a time.
It does slightly speed the code, but is an area where CCS, has lagged Hi-Tech for a long time, and been the cause of significant annoyance at times...
Best Wishes |
|
|
treitmey
Joined: 23 Jan 2004 Posts: 1094 Location: Appleton,WI USA
|
|
Posted: Fri Aug 20, 2004 8:33 am |
|
|
I do a simular thing with a 2d array.
Code: |
const char error_txt[16][16] =
{
"USART overflow", // -0-
"TXBuffer ovrflw", // -1-
"RXBuffer ovrflw", // -2-
};
|
Code: | fprintf(STDERR,"ERROR(%U): %s.\r\n",chk,error_txt[chk]);
|
|
|
|
eb7dzp
Joined: 20 Aug 2004 Posts: 2
|
One Solution but not the best. |
Posted: Fri Aug 20, 2004 9:52 am |
|
|
Thanks for your answer.
At the end, I prefered this one:
Code: | const char fecha_hora_txt[21][4] =
{
"X",
"Lun",
"Mar",
"Mie",
"Jue",
"Vie",
"Sab",
"Dom",
"X",
"Ene",
"Feb",
"Mar",
"Abr",
"May",
"Jun",
"Jul",
"Ago",
"Sep",
"Oct",
"Nov",
"Dic"
}; |
I donīt know why CCS, does not allow pointers to data in ROM, itīs a pitty.
So, thanks once again.
Best regards. |
|
|
Ttelmah Guest
|
Re: One Solution but not the best. |
Posted: Fri Aug 20, 2004 11:01 am |
|
|
eb7dzp wrote: | Thanks for your answer.
At the end, I prefered this one:
Code: | const char fecha_hora_txt[21][4] =
{
"X",
"Lun",
"Mar",
"Mie",
"Jue",
"Vie",
"Sab",
"Dom",
"X",
"Ene",
"Feb",
"Mar",
"Abr",
"May",
"Jun",
"Jul",
"Ago",
"Sep",
"Oct",
"Nov",
"Dic"
}; |
I donīt know why CCS, does not allow pointers to data in ROM, itīs a pitty.
So, thanks once again.
Best regards. |
Yes.
It 'makes sense', in two contexts. The first is that with older chips, where this memory cannot be accessed, except by using the 'RETLW' function, a 'pointer', involved a lot of overhead. This disappeared with chips having the table read functions.
The second, is that it implies that a memory 'pointer', has to be large enough, to cover both the ROM space, and the RAM space, implying (for instance), on a 18Fxx20 part, a size of 21bits to cover the possible ROM space, plus another bit to include the RAM space. Conversely, the entire RAM space on the largest chips, can still be covered with only about a 12bit pointer. This has significant implications in speed, and code size.
That being said, there should (to my mind), be an option on the chips supporting the table read operations, similar to the *=16 option, allowing large pointers, and support for ROM access...
Best Wishes |
|
|
|
|
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum
|
Powered by phpBB © 2001, 2005 phpBB Group
|