AB Real (F) to ASCII (ST) Conversion

grish05

Member

G

Join Date

Mar 2007

Location

Independence, MO

Posts

6

• Apr 4, 2007

• #1

Has anyone had any experience or written any code to convert a real (F file) to an ASCII (ST file) for a AB SLC/Micro, in this case I am using a 1100. The F register will contain a 6 digit number with a 3 position decmial point (XXX.XXX).



Thanks

Jeff

 

OkiePC

Lifetime Supporting Member

Join Date

Mar 2005

Location

ENE of Nowhere Oklahoma

Posts

11,794

• Apr 4, 2007

• #2

This works in an emulated 5/04...can't be sure it will work in a ML1100 since I don't have one and can't emulate one...

I get each half of the float into two integers, convert them to strings, and concatenate the strings together including one with a manually inserted period (ST9:2).

 

Last edited: Apr 4, 2007

grish05

Member

OP

G

Join Date

Mar 2007

Location

Independence, MO

Posts

6

• Apr 4, 2007

• #3

Thanks for the response,
I have some similar code but when I have a decimal less than .100 the conversion is off by a digit. For example when X.063 is entered (1/16” = .063), the conversion puts the decimal at .630 instead of .063.

Will your code convert over in this situation OK?

Thanks, Jeff

 

kamenges

Member

K

Join Date

Nov 2002

Location

Green Bay, WI

Posts

4,332

• Apr 4, 2007

• #4

OkiePC's code will suffer from the same <0.1 issue you have. To the best of my knowledge there is no way to make the conversion pad with zeros. You need to insert the leading zeros on your own. How dificult that will be will depend on how many places after the decimal point you need to display.


Keith

 

OkiePC

Lifetime Supporting Member

Join Date

Mar 2005

Location

ENE of Nowhere Oklahoma

Posts

11,794

• Apr 5, 2007

• #5

OOPS!, You're right, I forgot all about leading zeroes in the converted integer...hmmm...

 

OkiePC

Lifetime Supporting Member

Join Date

Mar 2005

Location

ENE of Nowhere Oklahoma

Posts

11,794

• Apr 5, 2007

• #6

Here's a slightly improved version...note that the float must be less than 32767 and that the output string is to 4 places past the decimal point, but that could be changed:

 

kamenges

Member

K

Join Date

Nov 2002

Location

Green Bay, WI

Posts

4,332

• Apr 5, 2007

• #7

Good one, OkiePC. That will do it.

But we now see why we all hate to mess with string conversions in most plcs. They just weren't intended to do that very smoothly. This would be a one-line deal in C.

Keith

 

jshiepe

Member

Join Date

Sep 2003

Location

Strongsville, OH

Posts

229

• Apr 5, 2007

• #8

I ran across this same issue while working with one of AB's InView displays that I wanted to use to display some statistical data from a SLC 5/03. The InView doesn't support decimal variables, so I had to do something very similar for my floating point values. This was several years ago, so I hope they've rectified this by now.

 

geniusintraining

Lifetime Supporting Member + Moderator

Join Date

Jun 2005

Location

SC

Posts

8,277

• Apr 5, 2007

• #9

OkiePC said:

Here's a slightly improved version...

Click to expand...

Very nice Paul!!!

 

OkiePC

Lifetime Supporting Member

Join Date

Mar 2005

Location

ENE of Nowhere Oklahoma

Posts

11,794

• Apr 6, 2007

• #10

Yeah, it would be "much improved" if it handled longer floats...

If the ML1100 supports long integers and the AIC instruction does too then it should be simple to make that little nugget of code even better...

 

grish05

Member

OP

G

Join Date

Mar 2007

Location

Independence, MO

Posts

6

• Apr 6, 2007

• #11

Thanks, I did not know about the AEX command nor would I thought to use it in that way. I am looking forward to putting that in next week.

One thought for your first line of code, if you use a ABS - absolute value command and set the destination for your integer file you don't have to worry about the rounding up. Plus it is a little cleaner.

Thank you, Jeff

 

OkiePC

Lifetime Supporting Member

Join Date

Mar 2005

Location

ENE of Nowhere Oklahoma

Posts

11,794

• Apr 6, 2007

• #12

Ah, thanks, I forgot that A/B had an ABS instruction.

 

mellis

Member

M

Join Date

Nov 2004

Location

Cincinnati, Ohio

Posts

829

• Apr 6, 2007

• #13

Here's my version...

It's a bit messy (I'd prefer to reorder the string file), but it works pretty well.

It does handle input up to 999999.999, after that you start to get roundoff in the float.

It also handles negative numbers.


A surprisingly ugly little problem, didn't expect it to be this difficult. I'd definitely like to hear from anyone that has a more elegant solution.

One last thing... What's up with the discussion around the ABS instruction? Looks to me like it rounds off just like a MOV.

 

chavak

Member

Join Date

Jul 2002

Posts

750

• Apr 6, 2007

• #14

mellis said:

Here's my version...

One last thing... What's up with the discussion around the ABS instruction? Looks to me like it rounds off just like a MOV.

Click to expand...

ABS outputs the absolute value, say if the input is a negative value, it turns the output to positive. If the input is positive value, ABS just output it as it is.

____________