[Patch, Fortran] PR32555 re-allow zero to precede the P edit descriptor

[Jerry DeLisle]

Tobias Burnus wrote:
> :ADDPATCH fortran:
> 
> My patch for rejecting the "i0" edit decriptor in READ caused a
> regression in the NIST test suite by fixing format_lex() such that it
> can also return FMT_ZERO.
> 
> "k P" seems to be the only place where a preceding number is allowed to
> be zero (or negative).
> 
> Build (yesterday's svn version)
> and regression tested (x86-64/Linux). Ok for the trunk?
> 
> Tobias
> 
OK, I confirmed this fixes the NIST failures I was seeing.

Thanks,

Jerry