This forum is a read-only archive. Posting, replying and signing in are disabled.

Using Global get Ex on Noop call giving error when deploying app

jjEvans

I’m running the counter applications in the docs, and I made a small change: I’d like to check that the count value exists in state.

Because count is initialized to 0, I figured I could call global get ex with the current application ID and use the Maybe Value to do the check. However, I get the following error:

File "/usr/local/lib/python3.9/site-packages/pyteal/compiler/compiler.py", line 215, in compileTeal
    localSlotAssignments = assignScratchSlotsToSubroutines(
  File "/usr/local/lib/python3.9/site-packages/pyteal/compiler/scratchslots.py", line 95, in assignScratchSlotsToSubroutines
    raise TealInternalError(msg) from errors[0]
pyteal.TealInternalError: Encountered 1 error when assigning slots to subroutine

Here is the code: The only change between this and the counter (add, deduct) example in the official algorand docs is the first line , and the third line in the And block.

   myStatus = App.globalGetEx(Txn.applications[0], Bytes("Count")).hasValue()
   handle_noop = Cond(
       [And(
            Global.group_size() == Int(1),
            Txn.application_args[0] == Bytes("Add"),
            myStatus > Int(0)
        ),add],
        [And(
            Global.group_size() == Int(1),
            Txn.application_args[0] == Bytes("Deduct"),
        ), deduct]
   )
Ben

Hi jjEvans,

This trips up a lot of folks. It has to do with the AppGlobalGetEx returning a MaybeValue which needs to be included in a Seq (or some other expression) prior to calling hasValue

You may try something like

   myStatus = App.globalGetEx(0, Bytes("Count"))
   handle_noop = Seq(
    myStatus,
    Cond(
       [And(
            Global.group_size() == Int(1),
            Txn.application_args[0] == Bytes("Add"),
            myStatus.hasValue()
        ),add],
        [And(
            Global.group_size() == Int(1),
            Txn.application_args[0] == Bytes("Deduct"),
        ), deduct]
   )
)

This is a strange way to offer this functionality so I hope will be tweaked in upcoming versions of PyTEAL.

Ben