delayedAssign: Delay Evaluation

Description

delayedAssign creates a promise to evaluate the given expression if its value is requested. This provides direct access to the lazy evaluation mechanism used by R for the evaluation of (interpreted) functions.

Usage

delayedAssign(x, value, eval.env = parent.frame(1), assign.env = parent.frame(1))

Arguments

a variable name (given as a quoted string in the function call)

value

an expression to be assigned to x

eval.env

an environment in which to evaluate value

assign.env

an environment in which to assign x

Value

This function is invoked for its side effect, which is assigning a promise to evaluate value to the variable x.

Details

Both eval.env and assign.env default to the currently active environment.

The expression assigned to a promise by delayedAssign will not be evaluated until it is eventually ‘forced’. This happens when the variable is first accessed.

When the promise is eventually forced, it is evaluated within the environment specified by eval.env (whose contents may have changed in the meantime). After that, the value is fixed and the expression will not be evaluated again.

Examples

Run this code

msg <- "old"
delayedAssign("x", msg)
substitute(x) # shows only 'x', as it is in the global env.
msg <- "new!"
x # new!

delayedAssign("x", {
    for(i in 1:3)
        cat("yippee!\n")
    10
})

x^2 #- yippee
x^2 #- simple number

ne <- new.env()
delayedAssign("x", pi + 2, assign.env = ne)
## See the promise {without "forcing" (i.e. evaluating) it}:
substitute(x, ne) #  'pi + 2'


### Promises in an environment [for advanced users]:  ---------------------

e <- (function(x, y = 1, z) environment())(cos, "y", {cat(" HO!\n"); pi+2})
## How can we look at all promises in an env (w/o forcing them)?
gete <- function(e_)
   lapply(lapply(ls(e_), as.name),
          function(n) eval(substitute(substitute(X, e_), list(X=n))))

(exps <- gete(e))
sapply(exps, typeof)

(le <- as.list(e)) # evaluates ("force"s) the promises
stopifnot(identical(unname(le), lapply(exps, eval))) # and another "Ho!"