jonkeane commented on a change in pull request #8947:
URL: https://github.com/apache/arrow/pull/8947#discussion_r549861998
##########
File path: r/R/expression.R
##########
@@ -173,84 +205,84 @@ Expression <- R6Class("Expression", inherit = ArrowObject,
ToString = function() dataset___expr__ToString(self)
)
)
-
+Expression$create <- function(function_name,
+ ...,
+ args = list(...),
+ options = empty_named_list()) {
+ assert_that(is.string(function_name))
+ dataset___expr__call(function_name, args, options)
+}
Expression$field_ref <- function(name) {
- assert_is(name, "character")
- assert_that(length(name) == 1)
+ assert_that(is.string(name))
dataset___expr__field_ref(name)
}
Expression$scalar <- function(x) {
dataset___expr__scalar(Scalar$create(x))
}
-Expression$compare <- function(OP, e1, e2) {
- comp_func <- comparison_function_map[[OP]]
- if (is.null(comp_func)) {
- stop(OP, " is not a supported comparison function", call. = FALSE)
- }
- comp_func(e1, e2)
-}
-comparison_function_map <- list(
- "==" = dataset___expr__equal,
- "!=" = dataset___expr__not_equal,
- ">" = dataset___expr__greater,
- ">=" = dataset___expr__greater_equal,
- "<" = dataset___expr__less,
- "<=" = dataset___expr__less_equal
-)
-Expression$in_ <- function(x, set) {
- dataset___expr__in(x, Array$create(set))
-}
-Expression$and <- function(e1, e2) {
- dataset___expr__and(e1, e2)
-}
-Expression$or <- function(e1, e2) {
- dataset___expr__or(e1, e2)
-}
-Expression$not <- function(e1) {
- dataset___expr__not(e1)
-}
-Expression$is_valid <- function(e1) {
- dataset___expr__is_valid(e1)
+build_dataset_expression <- function(.Generic, e1, e2, ...) {
+ if (.Generic %in% names(.unary_function_map)) {
+ expr <- Expression$create(.unary_function_map[[.Generic]], e1)
+ } else if (.Generic == "%in%") {
+ # Special-case %in%, which is different from the Array function name
+ expr <- Expression$create("is_in", e1,
+ options = list(
+ value_set = Array$create(e2),
+ skip_nulls = TRUE
+ )
+ )
+ } else {
+ if (!inherits(e1, "Expression")) {
+ e1 <- Expression$scalar(e1)
+ }
+ if (!inherits(e2, "Expression")) {
+ e2 <- Expression$scalar(e2)
+ }
+
+ # In Arrow, "divide" is one function, which does integer division on
+ # integer inputs and floating-point division on floats
+ if (.Generic == "/") {
+ # TODO: omg so many ways it's wrong to assume these types
+ e1 <- Expression$create("cast", e1, options = list(to_type = float64()))
+ e2 <- Expression$create("cast", e2, options = list(to_type = float64()))
+ } else if (.Generic == "%/%") {
+ e1 <- Expression$create("cast", e1, options = list(to_type = float64()))
+ e2 <- Expression$create("cast", e2, options = list(to_type = float64()))
+ return(Expression$create("cast",
Expression$create(.binary_function_map[[.Generic]], e1, e2, ...), options =
list(to_type = int32(), allow_float_truncate = TRUE)))
+ } else if (.Generic == "%%") {
+ # {e1 - e2 * ( e1 %/% e2 )}
+ # TODO: there has to be a way to use the form ^^^ instead of this.
Review comment:
Ah that works fine (though casting all of the operands to be right here
doesn't quite work (the previous hack didn't work in this case, but I failed to
have a test).
With array expressions, I get "cannot add bindings to a locked environment"
(I'm pushing a commit that will have this in a comment in a sec)
----------------------------------------------------------------
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:
[email protected]
