Fix token regression - re-add support for empty tokens

[eileenmcnaughton]

Overview

Fix token regression - re-add support for empty tokens since this was originally being accessed from replaceGreetingTokens which is no longer called by tokenCompatSubscriber

Before

{contact.first_name}{ }{contact.last_name} leaves in { }

After

{contact.first_name}{ }{contact.last_name} removes { }

Technical Details

I originally thought the behaviour was to remove those tokens when there was a string either side - the tests show that not to be working for replaceGreetingTokens - this syncs the behaviour of the 2 rather than trying to improve it.

Comments

@totten @colemanw note this test specifically covers the 2 variants of prefix_id - but won't pass without #21705 merged

[civibot]

(Standard links)

• If this is your first pull-request for CiviCRM, please browse CONTRIBUTING.md for information about the development and testing processes.
• If you are reviewing this pull-request, you may wish to consult the test sites and the Review Standards (long template, short template).

[colemanw]

retest this please

[eileenmcnaughton]

@totten @colemanw we can merge this as is for parity but I think the before & after are not working as well as I thought - there is still a double space when middlename is empty. I suspect it's clever enough for some cases but not others

[totten]

OK, so your hesitation is that this produces a double-space? Which would be invisible on most HTML outputs - but visible on text outputs?

I think if this produces the same output on 5.42, then I wouldn't worry about it.

[totten]

I copied+trimmed the test to try it on 5.41. Same result with two spaces (for both TokenProcessor and replaceGreetingTokens()).

[eileenmcnaughton]

thanks @totten

[totten]

jenkins, test this please

[totten]

I had some issues where testOddTokens didn't pass when I ran the branch locally, but I think that was because of the relative-order-of-operations with the prior PR. (They started passing when I did a local rebase.) So if testOddTokens is good on the next run, then merge it.