Define missing definitions for variables #775
Conversation
During publication of v45, I found that we were missing a number of terms that we actively use in the specification. Among these are: - variable - external variable - local variable This PR defines those terms.
aphillips
added
syntax
aphillips
requested review from
stasm,
catamorphism,
eemeli,
echeran,
gibson042,
mradbourne and
mihnita
spec/syntax.md
Outdated
| @@ -97,6 +97,14 @@ Attempting to parse a _message_ that is not _valid_ will result in a _Data Model | |||
|
|
|||
| A **_<dfn>message</dfn>_** is the complete template for a specific message formatting request. | |||
|
|
|||
|
|
|||
| A **_<dfn>variable</dfn>_** is a data value bound to a _name_. | |||
There was a problem hiding this comment.
I don't think "name" has been defined -- or is it meant to refer to name as defined in the grammar? But if so, then variable doesn't need to be defined here either, since it's in the grammar.
Also, this conflicts with the later statement that: "A local-declaration binds a variable to the resolved value of an expression." Maybe that sentence should be changed to say "name" instead of "variable"?
There was a problem hiding this comment.
I don't think "name" has been defined
name has a definition here
variable needs to be defined because we (try to) link it many times in the spec.
Maybe that sentence should be changed to say "name" instead of "variable"?
Good point. Perhaps by saying something like:
A
local-declarationcreates a variable by binding a name to the resolve value of an expression.
Responding to @catamorphism's comment. This change adopts the already-problematic phrase "resolved value", which is begging for us to define it.
During publication of v45, I found that we were missing a number of terms that we actively use in the specification. Among these are:
This PR defines those terms.