1491867 - Enclose accented pseudolocale strings in markers

Status: RESOLVED FIXED

(Core :: Internationalization, defect, P3)

Description

[Francesco Lodolo [:flod]]

Every pseudolocale string should have an opening and closing marker (e.g. brackets). 

This allows to identify truncation easily: if the string is ending with an ellipsis and not a marker, that's a truncation. 

If the opening marker is missing, the string is not displayed for some reasons.

Comment 1

[Francesco Lodolo [:flod]]

Attachment: test.png

I took a look at the code, and it's really simple to add opening and closing markers, so I guess it mostly comes down to a product decision.
https://searchfox.org/mozilla-central/rev/21588b2a9824e0758fe11d10065e2c01ea9f32be/intl/l10n/L10nRegistry.jsm#341

With a caveat: placeables will be enclosed in markers as well, but I personally don't see a problem in that.

On a side note, not sure why tests use .include and not equal() (it makes totally sense for BiDi)
https://searchfox.org/mozilla-central/rev/21588b2a9824e0758fe11d10065e2c01ea9f32be/intl/l10n/test/test_pseudo.js#69

Comment 2

[Staś Małolepszy :stas]

Looking at the screenshot, adding markers in this way would be misleading, I think?

    [Connect with a ][Firefox Account]

This suggests that the string is created via concatenation, whereas in reality it's properly constructed via interpolation in FTL:

    Connect with a { -fxaccount-brand-name }

IIUC, the goal of adding the markers is to detect cases were this string would be built as:

    <span data-l10n-id="connect-string" /><span data-l10n-id="fxaccount-name" />

To that end, the implementation should only add them around the result of `bundle.format()`. The interpolation is encapsulated in the format() method, and would have no effect on the position of the markers.

Comment 3

[Francesco Lodolo [:flod]]

(In reply to Staś Małolepszy :stas from comment #2)
> IIUC, the goal of adding the markers is to detect cases were this string
> would be built as:

This is something I didn't think of, but I'm also more interested in detecting truncations, especially if we move to small screens.