I noticed that kate thinks that a particular word is spelled correctly, even though is it misspelled. When I check the word using aspell or enchant utilities, they both correctly identify it as being misspelled. How can I tell which spell checker utility or dictionary kate/KDE is using and why it doesn't identify the misspelling?


First find the PID of the kate process, e.g. ps aux | grep kate gives

user     23838  0.9  1.1 862548 97028 ?        Sl   12:17   0:05 kate -b /path/to/open/file
user     24710  0.0  0.0  12796   968 pts/0    S+   12:27   0:00 grep --color=auto kate

The kate process is the first line (check the command at the end of the line) and so it's PID is 23838 (the first number).

Now run lsof | grep <PID> | grep spell. I get (abridged)

kate      23838         jlippuner  mem       REG                8,1    110400    2884298 /var/lib/aspell/en-wo_accents-only.rws
kate      23838         jlippuner  mem       REG                8,1   2412496    2884269 /var/lib/aspell/en-common.rws
kate      23838         jlippuner  mem       REG                8,1    424224    3420495 /usr/lib/x86_64-linux-gnu/libhunspell-1.4.so.0.0.0
kate      23838         jlippuner  mem       REG                8,1     52128     657021 /usr/lib/x86_64-linux-gnu/qt5/plugins/kf5/sonnet/hunspell.so
kate      23838         jlippuner  mem       REG                8,1     77328     657020 /usr/lib/x86_64-linux-gnu/qt5/plugins/kf5/sonnet/hspell.so
kate      23838         jlippuner  mem       REG                8,1    714800    3408001 /usr/lib/x86_64-linux-gnu/libaspell.so.15.2.0
kate      23838         jlippuner  mem       REG                8,1     19088     657009 /usr/lib/x86_64-linux-gnu/qt5/plugins/kf5/sonnet/aspell.so
kate      23838         jlippuner  mem       REG                8,1     91248    2884348 /var/lib/aspell/en_US-wo_accents-only.rws

So in my case, it looks like kate loaded both the hunspell and aspell libraries (libhunspell-1.4.so.0.0.0 and libaspell.so.15.2.0) and it is using the aspell dictionaries (/var/lib/aspell/en-wo_accents-only.rws and /var/lib/aspell/en-common.rws)

| improve this answer | |

So I realise this is an old question, but I seem to keep revisiting this situation and maybe you are too.

KDE uses "Sonnet" these days as the spelling framework. Apparently it can use Hunspell underneath (and also Aspell, which is no longer maintained?)

Anyway, I needed to figure this out myself again today, and wanted to leave a note for posterity... the Sonnet setting in the KDE spell checking (Regional Settings, Spell Check) will cause a silent failure if your particular regional variant is not fully installed - in my case Australian English (en_AU).

Try starting KMail (for example) from a shell and notice:

sonnet.core: Missing trigrams for languages: QSet("en_GB", "en_CA", "en_AU")

So my distro did not automatically pull in my en_AU dictionaries, and setting that region for spelling did not complain about or change that situation.

This may be a source of your problem - also try disabling automatic language detection until you have your dictionary issues sorted.

| improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.