Add a dummy language for to test if all texts are translatable
When replacing hard coded English strings with i18n helpers, there is currently no way to check that all texts were caught.
Pseudo-localization helps verify that all strings have been externalized and can help verify that designs accommodate languages that have longer words and phrases than English.
Proposal
Add a dummy language for development which replaces all translatable texts with psuedo-localized strings.
- accent each character (yields a visibly different but still legible string)
- add tilde (~) to make all words 35% longer (simulates longer words in other languages to identify overflow problems)
We will need to handle inline HTML, placeholders, and emoji so they are not mangled with accents. (e.g em
should not be mangled in string like "this is important")
Links
Localizing Slack: https://i736c61636bo656e67696e656572696e67z.oszar.com/localizing-slack-680c4bc7f45a
Edited by James Ramsay (ex-GitLab)