Bug in the function SigDig

[Wiljea]

I discover the following bug about the function SigDig to compute the number of significant digits for number greater than "0" having trailing "0":
7 SigDig = 1 OK
70 SigDig = 1 OK
700 SigDig = 1 OK
7000 SigDig = 3 NOT OK, should be 1
70000 SigDig = 3 NOT OK, should be 1
700000 SigDig = 3 NOT OK, should still be 1
...

[c3d]

Good catch, @Wiljea. This issue only happens when an integer to decimal conversion happens. In other words, 7000. SigDig returns 1 as intended.

There is a bug in that integer to decimal conversion, which can be reduced to the following case:

7000 TODECIMAL BYTES
@ Produces #AF02 0428h 6

7000. BYTES
@ Produces #AF01 0428h 5

In the first case, we produce a decimal number where the trailing zeroes are not correctly stripped. The internal representation otherwise remains correct, so the rest of the computations will not be affected, but it throws SigDig off.

[Wiljea]

But it seems that, in some case, SigDig doesn't produce the right result even with a decimal number:
70000. SigDig => 1 OK
70000 →NUM SigDig => 3 NOT OK even if after →NUM, we have 70000. on the stack.