Value from a single byte

Avex

Maybe i'm overlooking something, but i searched the Netlinx helpfiles and searched this forum, but couldn't find the answer for my 'problem'

From a file i parse a line and in this line is one! character representing an index, this character can be 0-9 and A-Z for calculating i need the value of this character pa. 0=48/ 6=54/A=65/P=80 etc etc
For the numbers i could use ATOI, but for the letters this returns a 0

I can remember from my old ZX81 days that there was something like ASC$(character) even that other brand with the C has a command called byte which returns the value from one byte.
For now i made a stupid lookup table with 36 entrys which do the job, but i'm pretty sure that there must be a more elegant way to do this.

something like this:

INTEGER Byte_Value
CHAR My_Byte[1]

My_Byte = "'A'"
Byte_Value = some_command(My_Byte)

Byte_Value is now 65

Kind Regards

Mark

RogerS

You can just use the integer value directly from the char array:


char my_byte[] = 'A'
integer my_byteAsDec = 0


my_byteAsDec = my_byte[1]

This will give the my_byteAsDec the value of 65

Tested with both numbers and letters


Is this what you were looking for?

Avex

Thanks for the reply, but when i do that i get a compile warning...

WARNING: \Jingle_PC_control.axi(401): C10570: Converting a string to a [INTEGER]
\Studio_Program.axs - 0 error(s), 1 warning(s)

If it works it should be OK, but can it be done without the warning?

Grtzz Mark

RogerS

If you try to convert the whole string to integer you get the warning.


I'm not getting the warning when i convert only 1 char in the string:



my_byteAsDec = my_byte[1] //this is ok. This will only convert the 1 char to integer

my_byteAsDec = my_byte //this will give warning. this tries to convert the whole array.

Avex

I should read better :-)

The [1] behind the byte is what i overlooked, i already did something like :
character = MID_STRING(cLine_Read,24,1) so i presumed this one byte was enough...

Thanks for the answer, it cleared a lot!

Grtzz Mark

DHawthorne

The MID_STRING function returns a string,even if only of length 1. The direct array index only returns the array element at that location. There's a bit more overhead in storing a string, which is why the compiler won't let you do that.

Avex

@DHawthorne

That explains my compiler 'nagging' :-) working with different systems with their small but big influence
differences can sometimes be a real pain in the **s

Many thanks for the answers.

Grtzz Mark

nicholasjames

And don't forget about TYPE_CAST.

Any of the 8 intrinsic data types can be assigned to any other type without the compiler yelling as long as you are not assigning a larger type to a smaller, or assigning a signed type to an unsigned type.

INTEGER = CHAR; // Okay
CHAR = INTEGER; // Not okay

FLOAT = SINTEGER; // Okay
LONG = SINTEGER; // Not Okay

However, you can force NetLinx to ignore this by using TYPE_CAST.

INTEGER = CHAR; // Okay
CHAR = TYPE_CAST(INTEGER); // Okay Now

FLOAT = SINTEGER; // Okay
LONG = TYPE_CAST(SINTEGER); // Okay Now

DHawthorne

Well, it still won't yell at you if it can be converted at all, but you might lose some data. In fact, I learned it as a a method of rounding numbers in my C++ days: add 0.5 to a float, type cast it to an integer, and you got your rounded number.

nicholasjames

DHawthorne wrote: »

Well, it still won't yell at you if it can be converted at all

I'm sorry, but what do you mean "if it can be converted at all"? Unless you're including run-time math. Dividing one integer by another may technically result in a floating point value, but the compiler won't catch it since the state of the variables are unknown.

DHawthorne

nicholasjames wrote: »

I'm sorry, but what do you mean "if it can be converted at all"? Unless you're including run-time math. Dividing one integer by another may technically result in a floating point value, but the compiler won't catch it since the state of the variables are unknown.

I was just hedging my bets. It may never complain for all I know, but neither have I tried type-casting an elaborate structure to a char, for example.