Back to LanguageTool Homepage - Privacy - Imprint

To language mantainers: advanced typography in LT messages

We have created a method that applies automatically advanced typography (typographical apostrophes, smart quotes, ellipsis, non-breaking spaces…) to LanguageTool rule messages. That means that messages can be written with the characters ', ", ... by the rule creator, but they will be shown to the users with typographical characters ( “ ” ‘ ’ « » ... ).
Maintainers just need to tell me which characters to use for each language (single and double opening and closing quotation marks), and any other special need (i.e. white spaces in French, etc.), and I will do the setup. It is already done for English, German, French, Spanish and Catalan.

I guess Microsoft Office and LibreOffice uses “ ” for Portuguese.

Done here: https://github.com/languagetool-org/languagetool/commit/9de02c94c9c0f5a8785684b034aaf3c557cb5c1f

Thank you!

Should this not be in Slack?

By the way, does this include apostrophes inside words? If so, doesn’t that impact the spell checking of those words?

Anyway, I will have a look at what is the official advice for Dutch.

Apparently there is no official advice, but this seems to be common use:

Jan zei: “Fatima zei ‘Hallo’ tegen me.”

The default for a ’ within a word seems to be: m’n.

(I know this auto typography improvement is in Word and LibreOffice, but personally, I dislike any automatic change of what I am typing. But anyway, that’s just me.)

Can you inform me when this has been implemented for Dutch, so I can see what the consequences are in the rules and speller?

The default behavior now (when advanced typography is enabled) is to convert apostrophes inside words (') to typographical apostrophes (). Let us know what are the preferences for Dutch, and provide examples so that we can test it.

This is just for the LanguageTool messages shown to the users. There is no change in the language rules, dictionary, etc.

Okay, I edited my post to show what I could find. I will see what it does.

Done here: https://github.com/languagetool-org/languagetool/commit/a641ad5b6813175c06a9a24345665173fdea8a0a

There is no change in rules and suggestions, only in rule messages.

Suggestions can have also advanced typography, but now it is only enabled in Catalan (as explained here: https://github.com/languagetool-org/languagetool/issues/3390).

These characters are usually used in Russian: ( « » ).

I think steps 1, 2 and 3 are problemless in Dutch too. But let’s wait a few days to not change too much at once.

It seems the advanced typography is on now for Dutch. Great.
But there is a rule that was affected:
https://internal1.languagetool.org/regression-tests/via-http/2020-08-28/nl/result_java_WHITESPACE_RULE.html

The superfluous spaces reported there, are not superfluous at all. These are all falses.

How did that come? Can it be corrected?

Also affected seems to be:
https://internal1.languagetool.org/regression-tests/via-http/2020-08-28/nl/result_java_SENTENCE_WHITESPACE.html

No idea. I can’t see any relation with advanced typography, which only changes messages.

It was probably a priority thing.

In Ukrainian «» are used as main ones and „“ as a backup.
There’s no single quote quotes, as single quote (’, U+2019, or U+02BC) can be part of the word.

Then, I am not sure what to do. Could you provide examples of actual rule messages and the desired output?
The suggestions inside messages (<suggestion></suggestion>), which now are shown as "..." will become «...». And apostrophes will become typographical (пов'язане -> пов’язане).
'пов'язане' would become ‘пов’язане’. Or do you prefer „пов’язане“? But if it is not used, then it doesn’t matter.

Oh, ok, let’s change quotes then and leave apostrophe as is (a direct one - 0x27 - is the most common). The other two are not as common and have problems. I may change it later at some time. Thanks.