Unicode group name gets broken

[dolugen]

When I create a group with a long name in cyrillic (unicode), very rarely, the name gets broken, like not-a-working-group-name%d. So I can't access the group clicking its link. Notice the end of the name.

You can try recreating the bug with the name "Уран бичлэг сонирхогчид", any other name I tried was fine ("Amateur Calligraphers" in Mongolian).

[apeatling]

Milestone 1.0.4 deleted

[apeatling]

Need i18n help on this one.

[erosaung]

Myanmar Unicode is also got error

[boonebgorges]

The problem seems to happen when the slug gets run through the chain sanitize_title() -> sanitize_title_with_dashes() -> utf8_uri_encode(). The process of uri_encoding the non-Latin characters makes the slug very long. sanitize_title_with_dashes() limits the length of the converted slug to 200 characters, but the 'slug' column in the BP groups table only allowed for 100 characters. As a result, some of the uri-encoded characters were getting cut off and made unreadable.

Allowing the slug column to accept 200 characters fixes the problem.

[boonebgorges]

(In [3677]) Allow group slugs to be up to 200 characters to account for uri-encoding of non-latin alphabets. Fixes #880