[Date Prev][Date Next][Thread Prev][Thread Next]
- Subject: Re: string.sub for suffix not usable with varying length (including zero)
- From: Dirk Laurie <dirk.laurie@...>
- Date: Sat, 16 Feb 2013 23:10:11 +0200
2013/2/16 Egor Skriptunoff <email@example.com>:
> i=0 should be interpreted as "after last symbol" for suffixes' sake
> j=0 should be interpreted as "before first symbol" for prefixes' sake
> This quirk is more desirable, all special cases: sub(-0,0), sub(-0),
> sub(1,0) becomes predictable.
> IMHO, sub(0) should be the same as sub(0,-1) - both should return empty string.
> I never needed sub(0,j) for prefixes.
> Can anybody give an example from real life code
> where it was more convenient having ('abcd'):sub(0,4)=='abcd' ?
I have never needed str:sub(0) or str:sub(0,j) for any reason whatsoever.
I stick to string indices in the ranges -#str to -1 or 1 to #str. Even
though str:sub has fallbacks otherwise, I don't use them. If that is
a habit, errors can easily be found this way:
local strsub = string.sub
string.sub = function(str,i,j)
j=j or n
if not ((i>0 and i<=n) or (i<0 and i>=-n)) then
error("Bad parameter #2 to string.sub: out of range") end
if not ((j>0 and j<=n) or (j<0 and j>=-n)) then
error("Bad parameter #3 to string.sub: out of range") end