Post slug contains Vietnamese characters

[anonymized_15729769]

In WP 5.0 + Gutenberg Editor,

If my post like "Some characters: Những ký tự tiếng việt" (contains ":" character), then the post slug is auto set to "some-characters-những-ký-tự-tiếng-việt", I think it should be "some-characters-nhung-ky-tu-tieng-viet"

[Joen]

I can confirm that if you add the suggested phrasing, the previewed permalink does not appear to be sanitized. However the published version has a sanitized slug, and if you reload the editor you'll see the accurate slug represented.

Screenshots: ​https://cloudup.com/cDhLpMIBITM

The enhancement we could make is to do this sanitization _before_ publish also, so it would match the before and after. I would suggest this could be done in a future release. Thanks for the ticket!

[lonelyvegan]

Agreed that we should handle this in an upcoming patch 👍🏻

[SergeyBiryukov]

To clarify a bit more, it looks like neither sanitize_title() nor wp_unique_post_slug() runs until publishing.

А post titled Hello, world! shows hello,-world! in the previewed permalink, see 45478.png​.

After publishing, the actual slug was hello-world-2, because it was a second post with that title.

[SergeyBiryukov]

Replying to SergeyBiryukov:

А post titled Hello, world! shows hello,-world! in the previewed permalink, see 45478.png​.

• If I edit the slug to hello,-world!!! and click "Save" on the permalink input, it stays as entered. At this point, it's not sanitized and not saved to database yet.
• If I click "Save Draft" then, it changes to hello-world. sanitize_title() appears to run now, for some reason only if you manually edit slug. wp_unique_post_slug() does not run at this point. In the database, post_name now has this temporary value.
• If I publish the post, it changes to hello-world-2. sanitize_title() and wp_unique_post_slug() finally run without depending on manual slug edits, and post_name is now correct in the database.

Would be great to have some consistency here :)

[SergeyBiryukov]

#45588 was marked as a duplicate.

[audrasjb]

Hello,

5.0.3 is going to be released in a couple of weeks.

It doesn't appear this ticket can be handled in the next couple of weeks (still needs a patch, unit-tests and actual testing). Let's address it in 5.1 which is coming in February. Feel free to ask for changing the milestone if you think this issue can be quickly resolved.

Cheers,

Jb

[pento]

This is an issue with how the block editor generates the post slug for previewing. The slug is generated client side, which doesn't have an equivalent of sanitize_title().

I'm going to close this issue in favour of ​GB12907, as it needs to be handled in the block editor JS.

[ocean90]

#46088 was marked as a duplicate.

[SergeyBiryukov]

#47532 was marked as a duplicate.

[SergeyBiryukov]

#50246 was marked as a duplicate.

[SergeyBiryukov]

Looks like ​GB12907 is now fixed, but only the initial issue about punctuation, not the ones about sanitizing special characters in other languages: ​GB16695 or ​GB18177, which were both closed as duplicates.

Per ​https://github.com/WordPress/gutenberg/pull/21007:

This fixes the original issue from ​GB12907 regarding punctuation in the slug. It doesn't resolve the related issues referenced there that were opened for characters in other languages (such as German and Danish) not being converted properly. If we merge this and close ​GB12907, we may want to reopen one of those that was closed as a duplicate in order to keep a more accurate record.

So either ​GB16695 or ​GB18177 should be reopened.

[SergeyBiryukov]

#54492 was marked as a duplicate.