#1147602 sort with LC_ALL=C.UTF-8 instead of LC_ALL=C.

Package:
wamerican
Source:
wamerican
Submitter:
Akkana Peck
Date:
2026-09-13 17:53:01 UTC
Severity:
normal
#1147602#5
Date:
2026-09-13 15:52:36 UTC
From:
To:
Dear Maintainer,

The wamerican word list in /usr/share/dict/words is not alphabetical, though man american-english says it's "an ASCII file which contains an alphabetic list of words, one per line."

Apparently the sorter ignores all accented characters and sorts according to the remaining characters in the word.

For instance, here's a section from the Cs:

claiming
claim's
claims
Clair
éclair
Claire
Claire's
Clairol
Clairol's
Clair's
éclair's
éclairs
clairvoyance
clairvoyance's
clairvoyant
clairvoyant's
clairvoyants
clam

Note all those éclair words mixed into words that otherwise begin with C.

(The man page saying it's ASCII is also incorrect, since it contains accented characters which are not ASCII.)

On my system,
/usr/share/dict/words is a symlink to /etc/dictionaries-common/words
which is a symlink to /usr/share/dict/american-english
#1147602#8
Date:
2026-09-13 17:41:20 UTC
From:
To:
On Sun, 13 Sep 2026, Akkana Peck wrote:n

That's correct; it's sorted with LC_ALL=C sort -d currently.

There's an argument that since it's UTF-8, it should actually be
LC_ALL=C.UTF-8 sort -d.

I'll go ahead and do that the next time I make a release of scowl.