Turn off character emoji :short_code: conversion of characters?

amapanda_ᚐᚋᚐᚅᚇᚐ · 2022-09-14T08:32:20.835Z

Currently Discourse converts certain unicode characters into a plain ASCII version of the character (called shortcodes), i.e. (one character U+00270A) is converted to :fist: (i.e. 6 characters). Sometimes I see the character, sometimes I see the ASCII text like this. (e.g. my current discourse profile)

Can this text conversion be turned off please? It’s also done in the emails as well.

Is there any other automatic text conversion that’s enabled here?

Yes this is me being pedantic about text characters! I’m sure you’ll tell me to go away

nukeador · 2022-09-14T09:42:57.694Z

As far as I’ve been able to find, this is a discourse feature to allow easily pick an emoji when you don’t have an emoji selector on your keyboard

Discourse Meta – 9 Mar 15

Auto-replace emoji like ☺ with Discourse codes

Over on TDWTF, I’ve been using a bot to automatically replace Unicode emoji like 🙂 with Discourse emoji codes like 😄; it’s useful when posting from mobile. @codinghorror suggests it could be made a Discourse feature; it never occurred to me before,...

Reading time: 3 mins 🕑 Likes: 51 ❤

That’s why depending on how the poster created the emoji, you will see the unicode (probably someone with a keyboard emoji picker) or the shortcode (using the discourse emoji picker).

How is this conversion affecting you negatively? On emails they should be converted into the unicode one.

amapanda_ᚐᚋᚐᚅᚇᚐ · 2022-09-14T10:20:54.000Z

No they aren’t. (see attached screen shot)

Screenshot from 2022-09-14 12-19-56.png

nukeador · 2022-09-14T10:44:48.604Z

I’ve found this, it seems they disabled the conversion in emails to avoid being flagged as spam by spamassassin?

Discourse Meta – 19 Sep 14

Suppress images for short emails

To prevent Discourse emails being detected as spam, we should suppress images from short emails (having text length 2800 Bytes or less). Currently short emails with images are detected as spam by Apache SpamAssassin, due to HTML_IMAGE_ONLY_24 rule. ...

Reading time: 1 mins 🕑 Likes: 20 ❤

amapanda_ᚐᚋᚐᚅᚇᚐ · 2022-09-14T12:36:22.000Z

I’ve found this, it seems they disabled the conversion in emails to
avoid being flagged as spam by spamassassin?
https://meta.discourse.org/t/suppress-images-for-short-emails/20235

That’s mostly talking about removing images from emails to reduce spam likelyhood. But I don’t want images removed. II just want the textual characters I enter to be saved and not mandled… Is that possible to change?

nukeador · 2022-09-14T12:40:12.930Z

amapanda_ᚐᚋᚐᚅᚇᚐ:

II just want the textual characters I enter to be saved and not mandled… Is that possible to change?

I haven’t found a way to do that. This might be something to ask upstream.

/cc @Firefishy

Firefishy · 2022-09-14T12:59:25.693Z

We already have the “Strip images from emails having size less than 2800 Bytes” option enabled (default).

Please note we do not have the “bandwidth” to support custom patches on discourse. Please push these sorts of requests upstream if you need new features. Check out https://meta.discourse.org/

amapanda_ᚐᚋᚐᚅᚇᚐ · 2022-09-15T09:44:35.293Z

I don’t want images in emails, I just want the text I type in to be shown.

But it looks like Discourse will mangle my text. alas.

Minh_Nguyen · 2022-09-15T15:25:14.990Z

HTML character entity references look like the shortcodes when rendered but are preserved in the source. For example, ✊ becomes .

SomeoneElse · 2022-09-18T19:57:10.682Z

Oddly this happens even if you explicitly overtype the incorrect “correction” with what you wanted to type (I just did that here).

Edit: and for completeness there 2 ascii (remember them?) characters were converted to 14 in this example.