Hi,

spelling.sh has false alarms like this:

debian/aide.conf.d/10_aide_distribution:14: parm ==> param, pram, parma
(is it really nitpicking about my choice of variable names? srsly?)
debian/aide.conf.d/31_aide_hald:1: hald ==> held, hold, half, hall
debian/aide.conf.d/31_aide_libvirt:44: passt ==> past, passed
debian/aide.conf.d/31_aide_locales:1: fo ==> of, for, to, do, go
debian/aide.conf.d/31_aide_locales:1: te ==> the, be, we, to
debian/aide.conf.d/31_aide_munin:62: MIS ==> MISS, MIST
debian/aide.conf.d/31_aide_udev:20: nd ==> and, 2nd
(yes, the hardward abstraction layer daemon is actually called hald, fo
and te are valid language codes and commonly found in paths to manpages,
munin actually uses a file called MIS, and udev actually uses nd as a
directory name).

From what I found in the sparse docs of codespell, you can write a
.codespellrc which tells codespell to ignore all those words, (which
will stop the typo "fo" for "of" to be detected if I do so).

Is it possible to tell codespell to just ignore "fo" in the file
debian/aide.conf.d/31_aide_locales? If so, I need an example about how
to do this.

Greetings
Marc

^

Hi!

You are right to point out that the man page at
https://manpages.debian.org/unstable/codespell/codespell.1.en.html is
lacking in this regard.

Seem the readme at
https://github.com/codespell-project/codespell#using-a-config-file is
much better.

This issue should perhaps be filed upstream to Codespell to improve
their man page?