Translation Memory can now match strings across projects even when they use different placeholder formats or HTML entity encodings.
For example, I have %{count} dogs (Rails), I have %1$s dogs (Android), and I have %@ dogs (iOS) now all match each other. When a match is found, placeholders are automatically adapted to the correct format for the target project.
This applies to:
- TM suggestions in the translation editor (shown as “Normalized TM”)
- Translate using TM batch operation
- Plural strings — each plural form is adapted individually
Supported placeholder formats include %{name}, {{name}}, {name}, %1$s, %d, %@, {0}, and XLIFF inline tags. HTML entities like &, <, and > are also decoded for matching.