target_name order shouldn't matter

[aappling-usgs]

It seems I can only use target_name in a command if it's the first argument. I would like to be able to use that variable anywhere in a command.

this remake file creates an i.txt file as expected:

target_default: i.txt
targets:
  i.txt :
    command: file.create(target_name, showWarnings=FALSE)

> make()
<  MAKE > i.txt
[ BUILD ] i.txt |  file.create("i.txt", showWarnings = FALSE)

but this one gives an error:

target_default: i.txt
targets:
  i.txt :
    command: file.create(showWarnings=FALSE, target_name)

> make()
<  MAKE > i.txt
[ BUILD ] i.txt |  file.create(showWarnings = "i.txt", target_name)
Error in file.create(showWarnings = "i.txt", target_name) : 
  object 'target_name' not found

(and also my argument showWarnings=FALSE has been altered in the BUILD call)

[jordansread]

Note that this is possible when your function supports named arguments. In the example above the args to file.create are ..., showWarnings = TRUE. I agree that it would be nice that target_name could be found when unnamed, it does work when you can tie it to an argument name:

target_default: i.txt
targets:
  i.txt:
    command: file.create(target_name, showWarnings=FALSE)

  i2.txt:
    command: file.copy(from = 'i.txt', to = target_name)

works with

remake::make('i2.txt')
<  MAKE > i2.txt
[    OK ] i.txt
[ BUILD ] i2.txt |  file.copy(from = "i.txt", to = "i2.txt")
[  READ ]        |  # loading packages