-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
On Sun, Feb 18, 2018 at 04:55:28PM +0100, Ionel Mugurel Ciobîcă wrote:
>
> Anyone care to explain what exactly means the -g option of sort? The
> fine manual only says "general numerical", but I doubt that is true,
> because -g (and all other options I have tried, -n, -M, -h, -V) will
> all put Roman numeral 9 in between 4 and 5. See here:
>
> # echo "III\nII\nI\nV\nIV\nVII\nVI\nVIII\nX\nIX" | sort -g | nl
>
> What I expect is to put 9 in between 8 and 10.
The info documentation has more (alternative, you can look
that up on the web, as stated in the man page itself).
Extracted from the info:
‘-g’
‘--general-numeric-sort’
‘--sort=general-numeric’
Sort numerically, converting a prefix of each line to a long
double-precision floating point number. *Note Floating point::.
Do not report overflow, underflow, or conversion errors. Use
the following collating sequence:
• Lines that do not start with numbers (all considered to be
equal).
• NaNs (“Not a Number” values, in IEEE floating point
arithmetic) in a consistent but machine-dependent order.
• Minus infinity.
• Finite numbers in ascending numeric order (with -0 and +0
equal).
• Plus infinity.
Use this option only if there is no alternative; it is much slower
than ‘--numeric-sort’ (‘-n’) and it can lose information when
converting to floating point.
So '-g' basically means (decimal representation of) float, plus a
couple of NaNs. No roman numerals, alas...
[...]
> How do I sort in a pipe those roman numerals? I have written two bash
> scripts roman_to_arab.sh and arab_to_roman.sh, but I do not know how
> to adapt it to use it in pipes. Also, it may be too cumbersome to make
> the conversion to arab digits, sort with -n and then convert back into
> roman numerals...
I fear sort is out of its smarts on that. There are libraries for
different languages to do this, e.g. Perl's Roman.pm (in Debian
package libperl-roman).
> Anyone has encounter this issue? Any ideas how to sort out this sort
> issue? Of course, the easier will be if, indeed, the sort -g would
> work as expected, e.g. if "_general_ numeric" will not be particular
> to exclude Roman numerals...
I guess your idea of "general" is just too general to be practical :)
Cheers
- -- tomás
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.12 (GNU/Linux)
iEYEARECAAYFAlqJrS0ACgkQBcgs9XrR2kZo4ACcDkY4H1RzyWYaQnQF7E/PfLN9
AbsAmgPSPyn7r5kWyTH7CFOir/OMPAwo
=SXF9
-----END PGP SIGNATURE-----
