A parser prompt gives examples for canonical digits but says nothing about surrounding wh…

A parser prompt gives examples for canonical digits but says nothing about surrounding whitespace. What should happen before implementation?

1. Mark surrounding whitespace as unresolved and obtain a decision to trim, preserve, or reject it
2. Trim it because most users probably expect that behavior
3. Reject it because strict parsing is always safer
4. Let the generated tests choose whichever behavior the implementation already has