This grammar currently interprets f(+) as (infix_op (identifier) (oplus) ... instead of (bound_op (identifier) (plus)).
It's uncertain which interpretation should actually be correct; the resolution of this issue is blocked until such time as a decision is made elsewhere. Tree-sitter currently chooses the infix interpretation because it will tokenize f as an identifier, then choose to tokenize (+) as oplus instead of function application, due to rules either 2, 3, or 4 in tree-sitters conflicting tokens rules.
This behavior is exhibited for the infix operators (+), (-), (/), (\X), and (.) (uncertain about last two).
Similar to https://github.com/tlaplus/tlaplus/issues/625. The grammar's current behavior matches SANY's current behavior.
This grammar currently interprets
f(+)as(infix_op (identifier) (oplus) ...instead of(bound_op (identifier) (plus)).It's uncertain which interpretation should actually be correct; the resolution of this issue is blocked until such time as a decision is made elsewhere. Tree-sitter currently chooses the infix interpretation because it will tokenize
fas an identifier, then choose to tokenize(+)as oplus instead of function application, due to rules either 2, 3, or 4 in tree-sitters conflicting tokens rules.This behavior is exhibited for the infix operators
(+),(-),(/),(\X), and(.)(uncertain about last two).