Re: OT: Gawk match() and numbers in scientific notation

On May 10, 8:10=A0pm, Hermann Peifer <pei...@[EMAIL PROTECTED]
> wrote:
>
> $ cat collating_chars.sh
> #!/bin/bash
> # collate sequence of characters which belong
> # to a given character range or class
> # based on Ed's and Steffen's code
> # seems to work for Unicode locales
> #
> # usage:
> # collating_chars.sh [A-Z]
> # collating_chars.sh [[:upper:]]
>
> gawk 'BEGIN {
>
> =A0 =A0 =A0for (i=3D0;i<=3D32767;i++) {
>
> =A0 =A0 =A0 =A0 =A0num =3D sprintf("%X", i)
> =A0 =A0 =A0 =A0 =A0l =A0 =3D length(num)
>
> =A0 =A0 =A0 =A0 =A0# construct a format that is
> =A0 =A0 =A0 =A0 =A0# understood by /usr/bin/printf
>
> =A0 =A0 =A0 =A0 =A0if (i < 16)
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0num =3D "\\\\x0" num
> =A0 =A0 =A0 =A0 =A0else if (i < 128)
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0num =3D "\\\\x" num
> =A0 =A0 =A0 =A0 =A0else if (l =3D=3D 2)
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0num =3D "\\\\u00" num
> =A0 =A0 =A0 =A0 =A0else if (l =3D=3D 3)
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0num =3D "\\\\u0" num
> =A0 =A0 =A0 =A0 =A0else
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0num =3D "\\\\u" num
>
> =A0 =A0 =A0 =A0 =A0# exclude C1 control chars
> =A0 =A0 =A0 =A0 =A0# as printf doesnt like them
>
> =A0 =A0 =A0 =A0 =A0if (i < 128 || i > 159)
> =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0 =A0print num
> =A0 =A0 =A0}
>
> }' |
>
> # Print Unicode chars with /usr/bin/printf
> while read num ; do /usr/bin/printf "$num\n" ; done =A0| sort |
>
> # Collate characters which are matched by the given range or class
> gawk -v re=3D$1 '$1 ~ re { s =3D s $1 } END { print s }'


For those who might be interested: here another variation of the
script. Reading an additional local file is perhaps not the smartest
solution, but guarantees that all relevant Unicode code points are
covered. (I promise this to be my last OT posting in this thread ;-)

Hermann

$ cat collate_rechars.sh
#!/bin/bash
# Collate a sorted list of chars that belong to a range or class
# Based on Ed's and Steffen's code, works only for Unicode locales
#
# Usage:
# LC_ALL=3Den_GB.UTF-8 collate_rechars.sh [A-Z]
# LC_ALL=3Dda_DK.UTF-8 collate_rechars.sh [[:upper:]]
#
# For this script you need to have a local copy of
# http://www.unicode.org/Public/UNIDATA/UnicodeData.txt

gawk -F";" '

# Construct a format for chars in ASCII range
#
NR < 129 { print "\\\\x" substr($1,3,2) ; next }

# Construct a format for other code points in UnicodeData.txt
# Exclude C1 control chars and some other code points
# as usr/bin/printf re****ts errors for them
#
NR > 160 && $1 !~ /^(D800|DB7F|DB80|DBFF|DC00|DFFF)$/ {
        printf "\\\\U%08s\n", $1 }' UnicodeData.txt |

# Print chars with /usr/bin/printf and sort them
#
while read f ; do /usr/bin/printf "$f\n" ; done | sort |

# Collate chars for the given character range or class
#
gawk -v re=3D"$1" '$1 ~ re { s =3D s $1 } END { print re "=3D" s }'