String to Floating point?

Sniper King

Tried doing a search on this and I cant find it.· Any help here.






bytemove(@str,string("31.2345"), 7)

I want X [url=mailto:x=@str]= @str[/url]· where x is Long

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.



Michael King
Application Engineer
R&D
Digital Technology Group

Mike Green

There's no ready-made generalized string to float conversion routine. If you have constants at compile time, you can write

X := 31.2345

and the compiler will do the conversion for you

Sniper King

The is a constanly changing value. GPS

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.



Michael King
Application Engineer
R&D
Digital Technology Group

John Abshier

Here is some pseudo code off the top of my head
string :+ "31.2345\0" ' extracted from gps read
integer := 0
i := 0
repeat until string == "."
integer := 10 * integer + string - "0"
i ++
fraction := 0
i := 0
repeat until string == $0
fraction:= 10 * fraction + string - "0"
i ++
fpResult := FAdd(FFloat(integer), FDiv(FFloat(fraction), ###))
### is either a constant or calculated from i
I did notice that the number of decimals from my gps was different from the number of decimals in the example in the NMEA documentation I was using.

I wish the Spin complier would let me define variables as float and translate C := A + B to C := FAdd(A,B) for me.

John Abshier

StefanL38

Hello Michael,

in the Extended_FD_Serial-object

found by a search on the word fraction in the obex

obex.parallax.com/objects/search/?q=fraction

there are two PUB-Methods to extract the digits before and after the decimaldot

PUB IntOfString (stringptr): Value  | negFlag
{{ Returns the integer or whole portion of a string with a decimal point.
   serial.RxString(@myStr))
   x := serial.IntOfString(@myStr)
   "-123.456" will return -123
}}
    repeat strsize(stringptr)
        if  byte[noparse][[/noparse]stringptr] == "-"
            negFlag := true
        if (byte[noparse][[/noparse]stringptr] => ("0")) and (byte[noparse][[/noparse]stringptr] =< ("9"))
            value := value * 10 + (byte[noparse][[/noparse]stringptr]-"0")
        if byte[noparse][[/noparse]stringptr] == "."
           if NegFlag ==  true
               value := value * -1
               !negFlag
           quit
        stringptr++
    if NegFlag ==  true
       value := value * -1
       !negFlag 

PUB FracOfString (stringptr): Value  | decFlag
{{ Returns the fractional or decimal portion of a string with a decimal point.
   serial.RxString(@myStr))
   x := serial.FracOfString(@myStr)
   "-123.456" will return 456
}}
    repeat strsize(stringptr)
        if  byte[noparse][[/noparse]stringptr] == "."
            decFlag := true
        if decFlag == True
           if (byte[noparse][[/noparse]stringptr] => ("0")) and (byte[noparse][[/noparse]stringptr] =< ("9"))
              value := value * 10 + (byte[noparse][[/noparse]stringptr]-"0")
        stringptr++

you can use this as a base for your needs

best regards

Stefan

John Abshier

The Extended_FDSerial.spin that is distributed with Propeller Tool Ver 1.2 is 1.0. You need to go to the object exchange to get version 1.1 that thas the IntOfString and FracOfString. I wish I knew how to manage my objects and how to easily check to see if any objects have been updated.

John Abshier

StefanL38

hello John,

well you can sort the obex-table on each columtitle by clicking on the colummtitle

I made a shortcut to the configuration: all objects sorted by published-date descending

obex.parallax.com/objects/?o=0&ot=dsc

I use this for my regularly check about new uploaded objects in the obex.

best regards

Stefan

Sniper King

I spent all night but I got the thing to work. Simple but it works...

I know, I know I should just pass a string pointer but when I did that it only passed the string up to the decimal point. So this worked and i stayed with it.

Oh yeah, I am finding some weird byte array issues that i fix by moving their declaration lines in front of other declaration. it doesn't affect the variables that I moved in front of.... Weird huh?

[noparse][[/noparse]code]

PUB Str2Float :c | a· ,i , B· ,k
a:=1
c:=0
·
a:=fp.ffloat(a)
···

· c:=fp.ffloat(c)
· repeat i from 0 to strsize(@buf)-1
····
···
····· if buf[noparse][[/noparse]i]=="."
········
······ quit
··· if buf[noparse][[/noparse]i]<>"-"··
···· a:=fp.fmul(a,10.0)···
··
·a:=fp.fdiv(a,10.0)
·· k:=0
· repeat i from 0 to strsize(@buf)
··· b:=0.0
·· if buf[noparse][[/noparse]i]=="-"
····· k:=1
···
·· if buf[noparse][[/noparse]i]=="1"
··· b:=fp.fmul(1.0,a)
·· if buf[noparse][[/noparse]i]=="2"
··· b:=fp.fmul(2.0,a)
·· if buf[noparse][[/noparse]i]=="3"
··· b:=fp.fmul(3.0,a)
·· if buf[noparse][[/noparse]i]=="4"
··· b:=fp.fmul(4.0,a)
·· if buf[noparse][[/noparse]i]=="5"
··· b:=fp.fmul(5.0,a)
·· if buf[noparse][[/noparse]i]=="6"
··· b:=fp.fmul(6.0,a)
·· if buf[noparse][[/noparse]i]=="7"
··· b:=fp.fmul(7.0,a)
·· if buf[noparse][[/noparse]i]=="8"
··· b:=fp.fmul(8.0,a)
·· if buf[noparse][[/noparse]i]=="9"
··· b:=fp.fmul(9.0,a)
····
·· if buf[noparse][[/noparse]i]<>"." and buf[noparse][[/noparse]i]<>"-"
······ a:=fp.fdiv(a,10.0)
·· C:=fp.fadd(c,b)
··
···
·if k==1
······ c:=fp.fneg(c)
·
[noparse][[/noparse]code/]

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
·- Was it pin 11 or 26?· Hmmm....··I think the smell of smoke tells the whole story.· Must be 26.



Michael King
Application Engineer
R&D
Digital Technology Group