Interval expressions should count characters, not UTF-8 bytes

[paulmiller]

Try putting "≤" (UTF-8 encoding 0xE2 0x89 0xA4) in a file:

$ echo "≤≤≤" > test.txt
$ xxd test.txt
00000000: e289 a4e2 89a4 e289 a40a                 ..........
$ grep -E '^.{3}$' test.txt
≤≤≤
$ grep -E '^.{9}$' test.txt
$ ack '^.{3}$' test.txt
$ack '^.{9}$' test.txt
≤≤≤

grep counts each "≤" as 1, whereas ack counts each as 3. Thus grep's . matches 3 times, but ack's . matches 9 times.

$ ack --version
ack v3.4.0 (standard build)
Running under Perl v5.32.0 at /usr/bin/perl

$ grep --version
grep (GNU grep) 3.4