The tokens currently detect a piece of a formula and also a type of a formula, e.g. it detects prefix through exclamation mark at the end of a token (e.g. SomeName!) and that way it is differentiated from a name token that doesn't have the exclamation mark at the end.
It works in most cases, but there are likely some edge cases, where name and other types are aliased and tokens are not what we expect.
The only case I have seen so far is a 3D reference, where A5:Sheet5!A1:A5 is recognized as a 3D reference instead of A1_CELL RANGE SHEET_PREFIX A1_AREA and I made a workaround for it, but there likely are others.
Instead, move to ident based parser, where token doesn't necessary mean specific type and it is handled in the parser and token is responsible only for tokens.
Instead, use IDENT token for sheet name or a name and separate exclamation mark.
The tokens currently detect a piece of a formula and also a type of a formula, e.g. it detects prefix through exclamation mark at the end of a token (e.g. SomeName!) and that way it is differentiated from a name token that doesn't have the exclamation mark at the end.
It works in most cases, but there are likely some edge cases, where name and other types are aliased and tokens are not what we expect.
The only case I have seen so far is a 3D reference, where
A5:Sheet5!A1:A5is recognized as a 3D reference instead ofA1_CELL RANGE SHEET_PREFIX A1_AREAand I made a workaround for it, but there likely are others.Instead, move to ident based parser, where token doesn't necessary mean specific type and it is handled in the parser and token is responsible only for tokens.
Instead, use
IDENTtoken for sheet name or a name and separate exclamation mark.