Source string check for strings with max-length: The source string is too long for other languages to fit within the limit.

[emmapeel2]

Describe the problem

See: https://github.com/orgs/WeblateOrg/discussions/18353

Many languages need more characters than English, and often the original string is too close to the string limit for most languages to translate.

We have some strings on our project that create lots of frustration because translators cannot fit all the meaning on the space allowed.

Is hard for translators to leave stuff out! Is contra natura!

And many developers are native English speakers blissfully unaware of such matters unless you tell them.

https://ititranslates.com/size-matters-text-length-when-translating-from-english/

Solution brainstorm

So, if a string is allowed 100 characters with max-length, an alert or failing check for mantainers should appear (similar to the "Multiple failing checks" check) when the English string is more than 70 characters. Something like: 'The original string is too long to fit on $max-length when translated to other languages'.

Or maybe we could make (and mantain!) a list of the languages that have more characters than English, and check that the component is translating to those languages before running the check?

Describe alternatives you have considered

No response

Screenshots

No response

[nijel]

Generally I like this idea, but we need to be conservative in this to avoid making the check too annoying.

The factor seems to also differ based on the length of the string; see https://www.w3.org/International/articles/article-text-size.en.html.

PS: Some language data are also at https://www.scriptis.com/wp-content/uploads/2021/03/expansion-chart-1.pdf.

[nijel]

Reading through the documents, I think 30% is too strict; something like 15-20% would be better for a source string check. We would rather not make it too annoying, as people will then just ignore the finding.