🩹 dev/core#4149 Change CKEditor 4 default to not encode HTML entities by default which breaks Smarty syntax and prevents use of crmDate function

[agileware-justin]

Change CKEditor 4 default to not encode HTML entities by default which breaks Smarty syntax and prevents use of crmDate function. Looks like this has been a reported problem since 2015, https://civicrm.stackexchange.com/questions/7212/escaping-quotes-in-ckeditor-breaks-smarty-syntax/7213#7213

Overview

User editable Message Templates, the Contact Action: Send an Email and the Contact Action: Print Merge/Document cannot use any Smarty Tokens with the crmDate function because CKEditor 4 unnecessarily HTML encodes single quotes (') and double quotes (").

So if you use a Smarty Token like:

{contribution.receive_date|crmDate:"%E%f %B %Y"}

{contribution.receive_date|crmDate:'%E%f %B %Y'}

When CKEditor 4 parses the HTML it converts this into:

{contribution.receive_date|crmDate:"%E%f %B %Y"}

{contribution.receive_date|crmDate:'%E%f %B %Y'}

Effectively rendering the crmDate function unusable in these situations.

https://civicrm.stackexchange.com/questions/7212/escaping-quotes-in-ckeditor-breaks-smarty-syntax/7213#7213
https://lab.civicrm.org/dev/core/-/issues/4149

Before

As per the description above.

After

These Smarty tokens are rendered correctly.

{contribution.receive_date|crmDate:"%E%f %B %Y"}

{contribution.receive_date|crmDate:'%E%f %B %Y'}

Technical Details

Comments

Also noting that the CKEditor configuration file is written to a "persistent" location: [civicrm.files]/persist/crm-ckeditor.js and really should be stored in a "dynamic" location: [civicrm.files]/persist/contribute/dyn/crm-ckeditor.js

Agileware Ref: CIVICRM-2103

[civibot]

(Standard links)

• If you are reviewing this pull-request, you may wish to consult the test sites and the Review Standards (long template, short template).

[demeritcowboy]

I think this makes sense. I was a little worried about how it might affect activities but it still encodes < and >. The docs aren't too clear but I think it still encodes these nbsp,gt,lt,amp which is good.