François Van Emelen writes: > Thank you for your 'ADD$' function, but it doesn't solve my problem. > I'll rephrase my question with a concrete example: > A DBAS-file is open. One of the fields is of type 3 (floating point) > The user wants to update that field. > That could be done with something like: > 120 input new_value$ > Suppose the value assigned to new_value$ exceeds 9 digits > (new_value$='12345678.55') or another value with more than 9 digits > As long as '12345678.55' is manipulated with a text variable everything > will be ok, of course. > But that large number has to be stuffed into the floating point field. > set#channel, x,new_value$ , x=field number > New_value$ is 'save' to the floating point field in its internal format > (I suppose) > Let us now (or later on, of course) retrieve the content of that field: > a=fetch(#channel,x): 'a$' could be used instead of 'a' > print a (or a$) will display '1.234568E7' which is not the same as the > value assigned to new_value$. > > And here comes my original question again: > > Is there a way to get the original value.In other words is it possible > to act upon 'a' (or 'a$') to 'extract' '12345678.55' from '1.234568E7' ?
Use a string variable in DBAS instead of a float and use the sort of routines I described to do the calculations. Didnt someone write an IEEE fp (64 or 80 bit) emulation package for the QL? Even if there is one, youd probably have to store the values in a string unless DBAS happens to support the type. Per
