https://bugs.freedesktop.org/show_bug.cgi?id=35960

--- Comment #14 from Owen Genat <[email protected]> ---
I think this report relates to how the dictionary files store their information
and how this information is interpreted. It may actually be an enhancement
request. Here are the grep results from some of the dictionaries in v4.1.3.2:

$ grep -e "'" /opt/libreoffice4.1/share/extensions/dict-en/en_AU.dic | wc -l
1410
$ grep -e "’" /opt/libreoffice4.1/share/extensions/dict-en/en_AU.dic | wc -l
0
$ grep -e "'" /opt/libreoffice4.1/share/extensions/dict-en/en_GB.dic | wc -l
925
$ grep -e "’" /opt/libreoffice4.1/share/extensions/dict-en/en_GB.dic | wc -l
0
$ grep -e "'" /opt/libreoffice4.1/share/extensions/dict-en/en_US.dic | wc -l
390
$ grep -e "’" /opt/libreoffice4.1/share/extensions/dict-en/en_US.dic | wc -l
0

My goodness, us Australian's really do like to shorten everything, don't we?
Even the main system dictionaries here under Crunchbang 11 x86_64 show the same
pattern:

$ grep -e "'" /usr/share/dict/american-english | wc -l
26226
$ grep -e "’" /usr/share/dict/american-english | wc -l
0
$ grep -e "'" /usr/share/dict/british-english | wc -l
26245
$ grep -e "’" /usr/share/dict/british-english | wc -l
0

Presumably this is because, as reported in comment #9, the auto-correction of
U+0027 contextually to U+2019 is handled by the word processor. To add to what
that comment states:

$ cat .config/libreoffice/4/user/wordbook/standard.dic
OOoUserDict1
lang: <none>
type: positive
---

No entries. Now I enter the text as indicated in comment #9 and add the term to
the dictionary:

$ cat .config/libreoffice/4/user/wordbook/standard.dic
OOoUserDict1
lang: <none>
type: positive
---
O'Briaaan

Spell check is as indicated in comment #9. Now I edit the dictionary to
manually alter the character from U+0027 to U+2019:

$ cat .config/libreoffice/4/user/wordbook/standard.dic
OOoUserDict1
lang: <none>
type: positive
---
O’Briaaan

The result is that the spell check is no different. All four instances are
marked as correct. It is possible that this bug may actually be an enhancement
request to get the dictionary facility of LO to detect these differences i.e.,
mark U+0027 as incorrect and U+2019 as correct. Whether that is even possible
would require a developer to comment on. In any case it would seem a huge
request.

-- 
You are receiving this mail because:
You are the assignee for the bug.
_______________________________________________
Libreoffice-bugs mailing list
[email protected]
http://lists.freedesktop.org/mailman/listinfo/libreoffice-bugs

Reply via email to