ID token used for parsing—not part of the lexer.

[keleshev]

In parsing chapter token ID is extensively used, however, it is not produced by the lexer. This leads to syntax errors while parsing {}-style objects.