After an evaluation, GNOME has moved from Bugzilla to GitLab. Learn more about GitLab.
No new issues can be reported in GNOME Bugzilla anymore.
To report an issue in a GNOME project, go to GNOME GitLab.
Do not go to GNOME Gitlab for: Bluefish, Doxygen, GnuCash, GStreamer, java-gnome, LDTP, NetworkManager, Tomboy.
Bug 343906 - No unicode support for bookmark completion
No unicode support for bookmark completion
Product: epiphany
Classification: Core
Component: [obsolete] URL bar
git master
Other Linux
: Normal normal
: ---
Assigned To: Epiphany Maintainers
Epiphany Maintainers
Depends on:
Reported: 2006-06-05 12:07 UTC by Wouter Bolsterlee (uws)
Modified: 2008-08-14 22:53 UTC
See Also:
GNOME target: ---
GNOME version: ---

Description Wouter Bolsterlee (uws) 2006-06-05 12:07:56 UTC
The bookmark completion in the location entry has no i18n support. The tokenization logic iterates over chars right now, while g_utf8_next_char should be used.

See bug #339167 for more information about the keyword completion logic.
Comment 1 Diego Escalante Urrelo (not reading bugmail) 2006-08-15 23:31:28 UTC

In Ephy 2.14 I can type 'El ú' and it knows that:
El Útero de Marita (a blog)
is different from:
El Utero de Marita (same blog, different name for testing)

If I type 'El u' it will offer both bookmarks. I think this is technically broken as you say Wouter, but it's a nice feature (at least i see it like that) for non english speaking people.
In the case of spanish, people tend to write without punctuation signs like acutes so most users would bookmark 'El Útero de Marita' (since that's the title of the page being bookmarked) and when they go to the location bar they will expect to see that bookmark as an option if they type 'El u'.
I would say this is not a bug from the usability point of view.

Opinions from non english speakers will be appreciated :)

I don't know if I was clear.
Comment 2 Wouter Bolsterlee (uws) 2006-08-16 08:11:39 UTC
So, basically, what you're saying is that if a user enters only ascii characters, unicode bookmark names should be roughly 'converted' (if possible) to ascii as well before matching?
Comment 3 Diego Escalante Urrelo (not reading bugmail) 2006-08-16 08:26:46 UTC
Yes, that's what happens right now.
Comment 4 Diego Escalante Urrelo (not reading bugmail) 2006-08-16 08:27:07 UTC
And it's ok that way, by the way.
Comment 5 Diego Escalante Urrelo (not reading bugmail) 2006-10-03 10:39:10 UTC
Actually this was requested as a feature in bug #328162, I think this should be closed.
What do you say Wouter?
Comment 6 Wouter Bolsterlee (uws) 2007-03-21 13:53:39 UTC
I doubt the title is really converted to ascii before matching right now... :)
Comment 7 Diego Escalante Urrelo (not reading bugmail) 2008-07-09 14:55:40 UTC
< wouter at guadec> It works, but no unicode normalization is done, just bytewise comparisons... covers like 99.99% of all uses cases though
Comment 8 Diego Escalante Urrelo (not reading bugmail) 2008-08-14 22:52:10 UTC
Fixed, part of #517960.