Published page URI & national characters?

Support
1

Hi guys.
I’m hoping this is a matter of my setup ‘suffers’ from misconfiguration - otherwise I’d have call it a bug for…
a topic with national chars in the title gets a URL with those chars fine, no problems there, but the page published off such topic shows a URL mangled horribly.
I cannot use, put correct chars there manually for Discourse says that slug contains incorrect chars.

2

does changing this site setting help?

IMG_0614
IMG_06141107×398 66.9 KB

4

I’ve had encoded - which I believed was a must for non-english lang, so I did not try anything else, yet.
nope… does not resolve the issue, I would (imagine everybody) prefer to keep my national chars. Looks increasing more like a bug, no?

Any @devel here? Should I move it into bug?

5

If you’d like to make a bug report rather than ask the community for support, there’s a good bug reporting guide here - Writing a Good Bug Report

We would need some repro steps, and screenshots would also be of value.

1 Like
6

@JammyDodger There is not much to it - deploy, change lang (PL in my case), set encode - should be very easy to reproduce.
Can create titles (off which URLs/slugs are created I presume) with national chars – cannot publish such a page with those national chars.
I was hoping somebody/devel can reproduce before I file a bug

7

Have you got a topic title I could copy and paste to try it out?

1 Like
8

Jeśli nie sałata to nie żaba

9

Or second example:

Jäällä öljyä vanhassa Åbossa.

10

I’ve had a test run:

  • Set default locale to polski (or suomi)
  • Set slug generation method to ‘encoded’
  • Enabled enable page publishing

  • Created topic with title ‘Jeśli nie sałata to nie żaba’ (or ‘Jäällä öljyä vanhassa Åbossa’)
  • Topic slug generated /t/je%C5%9Bli-nie-sa%C5%82ata-to-nie-%C5%BCaba/758 (or /t/j%C3%A4%C3%A4ll%C3%A4-%C3%B6ljy%C3%A4-vanhassa-%C3%A5bossa/759)

  • Attempt to publish topic with the encoded slug je%C5%9Bli-nie-sa%C5%82ata-to-nie-%C5%BCaba as the URL
  • Can’t publish as URL contains invalid characters
  • Removing encoding allows page to be published

Alternatively…

It doesn’t seem like encoded slugs are supported for page publishing.

What’s expected here? I checked out both the Polish and Finnish government websites to see what their URLs are like, and they don’t encode theirs to account for this.

I’ve also had a search, and I don’t think this has come up as an issue before, which may make this feature request territory rather than a bug if they’ve never been supported.

1 Like
11

Don’t mind the governments - those are extremely rarely the brightest among us, though many would hope they were.
As for the rest of us – yes, I’m surprised nobody reported that ever before – many if not most, are sheeple or lazy or both.

What expected - I’d only worry, focus here on the one thing which matters, certainly should matter to solid devel/programmers — which is logic & consistency. If one part of system/code does(or doesn’t) do something, then another bit(s) which relays, depends on the first one, should… obviously. So on so forth.
Bug is - published pages brake with national characters in titles and they should not. Discourse should be able to publish pages with national characters in URLs/slugs.

12

You can reproduce this far more easily.

Just try to publish a page and add a % sign to the slug. It will immediately display the message about their being an invalid character.

1 Like
13

Have you got a URL of what you would expect to see that you can give as an example?

I think the idea behind page publishing is to have a readable URL, so an encoded one would seem to go against that?

1 Like
14

No they don’t, in Finland anyway - and AFAIK same thing applies thru all Nordic, including almost or totally every country in the world.

Just because that way internet works because of historical reasons and english (and its very limited alphabet system :sweat_smile:). But it is always bad idea using those characters in urls and filenames on linuxes.

I don’t expect anything but letters another than a…z shall be cleaned, or coded. We have to remember too the fact that quite few ever see actual urls.

So… my role here was only offering some lego blocks more to play with, and I’ve had never any issues with urls. Neither my users.

3 Likes
16

I don’t understand - I filed a bug report.

18

Cross-posting your issue is unnecessary. If you could provide the extra details here that would be sufficient. :+1: