# Difference between revisions of "Proposals of Changes to the Mathematical Language Specification"

From Event-B

Line 4: | Line 4: | ||

* Matthias: The Paragraph "Arithmetic Operators" on page 19 says that "the exponentiation operator has the least priority". That would mean 1+1^2 is parsed as (1+1)^2 and evaluates to 4, which is not the case in Rodin. Giving the exponentiation operator least priority is also uncommon. I therefore propose to change the paragraph to say that the exponentiation operator has highest priority among all arithmetic operators. | * Matthias: The Paragraph "Arithmetic Operators" on page 19 says that "the exponentiation operator has the least priority". That would mean 1+1^2 is parsed as (1+1)^2 and evaluates to 4, which is not the case in Rodin. Giving the exponentiation operator least priority is also uncommon. I therefore propose to change the paragraph to say that the exponentiation operator has highest priority among all arithmetic operators. | ||

+ | |||

+ | [[Category:Design proposal]] |

## Latest revision as of 13:53, 26 March 2010

# Change Proposals to Version 1.0

- Matthias: Section 2.3 says that integer literals are unsigned, i.e., nonnegative. In Arithmetic Rewrite Rules, it is however assumed that integer literals may also be negative. I therefore propose to say in Section 2.3 that integer literals can be positive, zero, or negative, but that negative literals such as -1 are parsed to unary minus followed by 1.

- Matthias: The Paragraph "Arithmetic Operators" on page 19 says that "the exponentiation operator has the least priority". That would mean 1+1^2 is parsed as (1+1)^2 and evaluates to 4, which is not the case in Rodin. Giving the exponentiation operator least priority is also uncommon. I therefore propose to change the paragraph to say that the exponentiation operator has highest priority among all arithmetic operators.