Context
PR #12038 fixes the tokenization of && and adds proper whitespace handling in ConditionParser.tokenize().
Problem
The remaining arithmetic operators (-, *, /) are still not recognized as delimiter characters in the tokenizer. This means expressions without spaces around these operators (e.g., 5-3, 5*3, 10/2) are tokenized as a single token and fail to parse.
Notes
- needs special care since it can appear in property names with hyphens (e.g., my-property) and as a unary negation operator* and / are straightforward to add as delimiters- These operators are rarely used in profile activation conditions, so this is low priority
- The fix should follow the same pattern used for
&& and || tokenization
Related: #11882, #12038
Claude Code on behalf of Guillaume Nodet
Context
PR #12038 fixes the tokenization of
&&and adds proper whitespace handling inConditionParser.tokenize().Problem
The remaining arithmetic operators (
-,*,/) are still not recognized as delimiter characters in the tokenizer. This means expressions without spaces around these operators (e.g.,5-3,5*3,10/2) are tokenized as a single token and fail to parse.Notes
-needs special care since it can appear in property names with hyphens (e.g.,my-property) and as a unary negation operator*and/are straightforward to add as delimiters&&and||tokenizationRelated: #11882, #12038
Claude Code on behalf of Guillaume Nodet