-
Suggestion
-
Resolution: Unresolved
NOTE: This suggestion is for Confluence Cloud. Using Confluence Server? See the corresponding suggestion.
We have an issue here with confluence that prevent us to adopt it fully in our company .
In the French language, we use accents, such as é,à,é,etc.. and we need Confluence to ignore those accents for searchs, like google does.
For instance the search text: "majore", should give us pages with "majoré" or "majore" or "majorè"
It would be very useful if we could configure the search feature to be able to ignore accents.
Languages that should ignore accents for searches are:
- English (for loanwords, and also cover other European languages not supported in Confluence indexing)
- French
- Spanish
- Catalan (not supported in Confluence indexing)
- Portuguese (currently only Brazilian Portuguese is supported in Confluence indexing)
- Czech
- is related to
-
- Closed
- relates to
-
- Gathering Interest
-
- Gathering Interest
- mentioned in
-
Page Failed to load
[AI-611] Allow search feature to ignore accent
This is basic necessity feature. Also a feature that influences which project management SW to pay for.
Hi,
we just received email that our maintenance expired. Why to renew it?
For us is important to improve daily things like search without accent. We do not need another improvements.
I am apologize but this is nice example how Attlasian take care about their users.
David
David Petrík
added a comment - - edited Hi,
we just received email that our maintenance expired. Why to renew it?
For us is important to improve daily things like search without accent. We do not need another improvements.
I am apologize but this is nice example how Attlasian take care about their users.
David
How much money would Atlassian earn by improving its product as requested?
Probably none.
Hence this basic request will probably never be implemented.
Christophe Carpentier
added a comment - Since this utterly annoying issue hasn't moved in 6 years, I feel free to add a friendly +1.
Accents being part of my language, this makes searching for pages virtually unusable. Plus, people sometimes get accents wrong here, and you know you're looking for a page, you know what word to look for, and you can't get it because someone messed up a é for an è. That's truly unnerving.
Christophe Carpentier
added a comment - Since this utterly annoying issue hasn't moved in 6 years, I feel free to add a friendly +1.
Accents being part of my language, this makes searching for pages virtually unusable. Plus, people sometimes get accents wrong here, and you know you're looking for a page, you know what word to look for, and you can't get it because someone messed up a é for an è. That's truly unnerving. So to point out an example for a user search. Assuming there exists a user with the full name René Niño, then this user is not found by rene or nino. And that's another issue beside finding pages. This involves of course any other indexed content entity as well.
I agree with Paul and Lars-Erik.
And why the status of this issue is still set to NEEDS VERIFICATION ?
Trust us, it's a really annoying issue.
This is also necessary to search for users! This is a real issue in finding a certain person among 10 thousands. Please fix this.
This is related to how text is normalised before being indexed in Lucene. Currently, only the GermanAnalyzer removes umlauts, but they face a different issue as a result. Once that's solved, we could change all European languages that accents can be dropped to use the ASCIIFoldingFilter during normalisation.
What has to happen is a very simple process. Apache StringUtils normalises and then removes diacritics.
StringUtils.stripAccents(body);