Entity Name vs. Entity Number

[johnjamesjacoby]

In our strings, we currently use ’ and ’ interchangeably. They result in the same right quote mark (’) so this isn't a bug, nor is it grammatically incorrect. It may however be a potential bottleneck for individuals translating our heavy slang and contraction use, which appears to be in the hundreds of strings.

I'd like to suggest we do one of two things:

• Switch completely to ’ as it's easier to grok than ’
• Switch completely to ’ unless we have a complementary ‘ usage. (Note that we only currently have 1 ‘ and it's incorrectly used in a contraction.)
• Remove our contraction usages completely. This results in a subtle tone change and removes some of WordPress's Texan personality, but also makes internationalization easier and potentially more inviting as a result.

There are likely other entities worth discussing, so I titled this ticket intentionally broad and highlighted one of the more obvious usages. Definitely feel free to retitle and modify this ticket for maximum traction, y’all.

[SergeyBiryukov]

Related: #7098, #8714, #9030, #13341, #14225, #17362.

[GaryJ]

​Polyglot markup requires numbered entities and the W3C Recommendation is that ​numbered entities SHOULD use the hexadecimal form when it exists.

Is WP UTF-8 safe yet? Could we use the real ’ in the code?

Even with numbered or named entities, do translators ensure those entities are used in contractions in the translated strings? Does GlotPress do any sort of conversion? How about plugin authors? I feel some more answers are needed beyond just whether to change the source, to see what effect (good or bad) it will have for translators.

[SergeyBiryukov]

Replying to GaryJ:

Is WP UTF-8 safe yet? Could we use the real ’ in the code?

There's a precedent in [38359].

Even with numbered or named entities, do translators ensure those entities are used in contractions in the translated strings?

I think entities were originally used to make sure the code is ASCII-only, and prevent code editors from introducing issues like the one just fixed in [38517] :) That said, I guess most editors should be UTF-8 safe now.

Translators can use actual characters instead of entities, that should not cause any issues.

Does GlotPress do any sort of conversion?

Not that I know of.