Re: reliably test integer size for portable bit-fiddling? (~0)

Replies are listed 'Best First'.
Re^2: reliably test integer size for portable bit-fiddling? (~0) by davido (Cardinal) on Dec 09, 2014 at 00:50 UTC
Beautiful use of the Maths. ;) Dave	[reply]
Re^3: reliably test integer size for portable bit-fiddling? (~0) by Anonymous Monk on Dec 09, 2014 at 00:53 UTC
cough `$ perl -E " say log(~0)/log(2) " 31.9999999996641 $ perl -E " say int 0.999+log(~0)/log(2) " 32` [download]	[reply] [d/l]
Re^4: reliably test integer size for portable bit-fiddling? (~0) by tye (Sage) on Dec 10, 2014 at 15:34 UTC
`$ perl -le "print log(~0+1)/log(2)" 32` [download] or, to be safe `$ perl -e 'printf "%.0f\n", log(~0+1)/log(2)' 32` [download] - tye	[reply] [d/l] [select]
Re^2: reliably test integer size for portable bit-fiddling? (~0) by wollmers (Scribe) on Dec 09, 2014 at 08:46 UTC
`log(~0)/log(2)` [download] Hmm, I tried a similar idea (get the index of the highest bit), but it failed. Now I remember: `helmut@w3:~$ perl -e 'print log(-1)/log(2),"\n";' Can't take log of -1 at -e line 1.` [download] Thx, this is extremely fast and IMHO reliable, because it tests the capabilty directly. Helmut "wollmers" Wollmersdorfer	[reply] [d/l] [select]
Re^2: reliably test integer size for portable bit-fiddling? (~0) by wrog (Friar) on Dec 14, 2014 at 08:47 UTC
I figure for base2 stuff `length` is better than `log` (no need to worry about rounding/transcendental crap), e.g., `4*length sprintf'%x',~0` [download]	[reply] [d/l] [select]
Re^3: reliably test integer size for portable bit-fiddling? (~0) by BrowserUk (Patriarch) on Dec 14, 2014 at 09:46 UTC
Avoid the multiply: `say length sprintf "%b", ~0;; 64` [download] With the rise and rise of 'Social' network sites: 'Computers are making people easier to use everyday' Examine what is said, not who speaks -- Silence betokens consent -- Love the truth but pardon error. "Science is about questioning the status quo. Questioning authority". In the absence of evidence, opinion is indistinguishable from prejudice.	[reply] [d/l]