Words with apostrophe in find and replace

[GColl]

Lo 25.8.1.1 x86 64
Win 10

I tagged this in writer ss thats where i noticed it but it may well be wider.

In find and replace dialogue it seems a word such as doesn't which has an apostrophe is not treated as a united word, insomuch that a global replace of t with something else will replace the t in doesn't (etc.) with the replace string even when Whole Words only is checked. OTOH, when the cursor is placed inside such a word and a style applied, the style does affect the whole word (as i believe it should). Is there a reason for this difference of interpretation (or indeed a setting somewhere to select between the two? Or is it perhaps a bug?

Thanks
/Gary

[ajlittoz]

This may be a side effect of Whole words only being ticked: it references an internal “definition” of a word. It is likely that this “definition” only takes “letters” (a subset of Unicode characters valid in the current language). For English, this is A-Za-z.

[karolus]

@ajlittoz

For English, this is A-Za-z

IMO: A-Za-z0-9_

@GColl I’ve no problem to find »doesn't« with regex: \w+'t\b

[Lupp]

@ajlittoz

It is likely that this “definition” only takes “letters” ...

Under my en (UK) / en (Canada) setting the apostrophe as well as the “typographic” RIGHT SINGLE QUOTATION MARK are treated as word characters when searching for Whole words only.
Also note: With Regular expressions enabled the Whole words only option is not supported. There the mentioned characters are included with \W (non-word).

(LibO V 25.8.2 under Win10).