We have one failing test case, bracketsCrossTags: selection is expanded to "long in brackets" while expected (by test) is "(long in brackets)". Actually both variants are accaptable, but it should be consistent with other cases (parentheses, square brackets, curly braces, all variants of quotes). I'd rather expect surroundings not included into selection (or included when they are explicitly selected by user, but this contradicts to our concept of coding only word index). Current implementation has different behavior for different surroundings, which is certainly not good.
We have one failing test case,
bracketsCrossTags: selection is expanded to "long in brackets" while expected (by test) is "(long in brackets)". Actually both variants are accaptable, but it should be consistent with other cases (parentheses, square brackets, curly braces, all variants of quotes). I'd rather expect surroundings not included into selection (or included when they are explicitly selected by user, but this contradicts to our concept of coding only word index). Current implementation has different behavior for different surroundings, which is certainly not good.