[Rd] Parser oddity with <- and =

[Mikael Jagan]

The `<-` operator has higher precedence than the `=` operator,
so it makes sense to me that `a <- b = c` would be parsed as

call("=", call("<-", quote(a), quote(b)), quote(c))

For `a <- b = c` to be parsed instead as

call("<-", quote(a), call("=", quote(b), quote(c)))

R would need to assign equal precedence to `<-` and `=`, but
I'm guessing that `=` _needs_ to have lower precedence than `<-`
for calls with named arguments to be parsed in a sensible way.

Mikael

> Sorry, I wrote in a sloppy way.  The parsing I see is just what you saw.
>    It's the evaluation of that expression that tries to call `<-<-`:
> 
>   > a <- 1
>   > a <- b = c
> Error in a <- b = c : could not find function "<-<-"
> 
> This happens because R is trying to make an assignment using = with a
> LHS that is the function call a <- b.
> 
> Duncan Murdoch
> 
> 
> On 04/02/2022 3:28 p.m., Bill Dunlap wrote:
>> In R-4.1.2 and R-devel from two weeks ago I do not get the `<-<-`:
>>
>>       > str.language(parse(text = "a <- b = c"))
>>      expression: structure(expression(a <- b = c), sr ...
>>         language: a <- b = c
>>           symbol: =
>>           language: a <- b
>>             symbol: <-
>>             symbol: a
>>             symbol: b
>>           symbol: c
>>       > identical(parse(text = "a <- b = c")[[1]], parse(text = "`=`(
>>      `<-`(a, b), c)")[[1]])
>>      [1] TRUE
>>
>> str.language() is a rudimentary parse tree displayer:
>>
>>      str.language <- function(expr, name = "", indent = 0)
>>      {
>>           trim... <- function(string, width.cutoff) {
>>               if (nchar(string) > width.cutoff) {
>>                   string <- sprintf("%.*s ...", width.cutoff-4, string)
>>               }
>>               string
>>           }
>>           cat(sep="", rep("  ", indent), typeof(expr), ": ",
>>               if(length(name)==1 && nzchar(name)) { paste0(name, " = ") },
>>               trim...(deparse1(expr, width.cutoff=40), width.cutoff=40),
>>               "\n")
>>           if (is.function(expr)) {
>>               str.language(formals(expr), name="[formals]", indent =
>>      indent + 1)
>>               str.language(body(expr), name="[body]", indent = indent + 1)
>>           } else if (is.recursive(expr)) {
>>               expr <- as.list(expr)
>>               nms <- names(expr)
>>               for (i in seq_along(expr)) {
>>                   str.language(expr[[i]], name=nms[[i]], indent = indent + 1)
>>               }
>>           }
>>           invisible(expr)
>>      }
>>
>>
>> -Bill
>>
>> On Fri, Feb 4, 2022 at 9:34 AM Duncan Murdoch <murdoch.duncan using gmail.com
>> <mailto:murdoch.duncan using gmail.com>> wrote:
>>
>>      Here's an odd parse:
>>
>>           a <- b = 1
>>
>>      This appears to be parsed as
>>
>>           `<-<-`(a, b, 1)
>>
>>      instead of being equivalent to
>>
>>           a <- b <- 1
>>
>>      I wonder if that's intentional?
>>
>>      (This showed up at https://stackoverflow.com/q/70989067/2554330
>>      <https://stackoverflow.com/q/70989067/2554330>, where
>>      it caused a lot of confusion.  I think the original intent was that `a`
>>      would be a macro holding `b = 1`, but I'm not sure of that.)
>>
>>      Duncan Murdoch
>>
>>      ______________________________________________
>>      R-devel using r-project.org <mailto:R-devel using r-project.org> mailing list
>>      https://stat.ethz.ch/mailman/listinfo/r-devel
>>      <https://stat.ethz.ch/mailman/listinfo/r-devel>
>>