Tags starting with upper case special characters are not imported properly

[marph91]

Operating system

Linux

Joplin version

3.1.17

Desktop version info

No response

Current behaviour

Unzip and import the following folder with the "MD + Front Matter (Directory)" import. It contains two empty notes with tag Ökonomie.

example_tag_umlaut.zip

Result: The tag ökonomie is created and assigned to the first note. The second note gets imported, but without tag. Instead, there is an error message and the import stops:

References:

• Problem with Windows CLI marph91/jimmy#6 (comment)
  • Điện doesn't work
  • điện works
  • Dien works
• https://discourse.joplinapp.org/t/jimmy-a-joplin-import-tool/38503/13
  • Ökonomie doesn't work
  • ökonomie works
  • Oekonomie works

Expected behaviour

Import both notes including tags.

Logs

No response

[JackGruber]

Possible the same problem with the SQLite function LOWER as in #5102

[richardsprague]

when I import the enclosed file with a single tag, joplin explodes the tag into multiple tags of one letter each. Oddly, it imports correctly when the imported note contains more than one tag.
Notes 241019 Saturday.md

[marph91]

when I import the enclosed file with a single tag, joplin explodes the tag into multiple tags of one letter each. Oddly, it imports correctly when the imported note contains more than one tag.

The tags field expects a list. I guess if you specify a single tag only, it's handled as a list of characters (each a separate tag). Imo. this is expected behavior and not related to the original issue.

NB: I didn't try, but

tags: [notes]

or

tags:
  - notes

should work in your case.

[pedr]

It seems like this is limitation of sqlite, we can't have a "case insensitive" search for all Unicode characters without adding a extension to SQLite.

I found that this extension could be used:
https://github.com/nalgeon/sqlean/blob/main/docs/text.md

In the image text_lower is the function added by the extension.