AFAIK, there's no Lua anywhere in AutoKey. AutoKey is entirely written in Python 3.x - which is also the language our scripts are written in. All we do is add our API so they can perform keyboard, mouse, and some window oriented actions. Subprocess is a Python module. Maybe Lua has a similar facility with the same name. I have never looked into Lua.

Oh, wait! Yeah, no Lua. I'm thinking of the scripts I have to write to get devilspie2 to work… Crossed my wires there.

I use devilspie2 to position the things I have auto-start when I log in, such as the position of all comm apps on the laptop display, and all service request research on another display, etc. I use xwininfo in a YAD script and take the values from that to determine everything, and the Lua part is baffling to me because it's such a different scripting language. Why did they use that and not Bash, Python, Perl, etc.?

I'm a big fan of yad. I have a small bash project that uses it for its GUI interface. I'll have a look at devilspie2. New to me.

I use xdotool for such things. I have a moderately large autostart script that compensates for some of the shortcomings of KDE's session restore. Maybe we can compare notes outside of this project.

I write a lot of bash and it does get pretty tiring trying to convince it that a string is just a string, not something that "needs" further expansion.

Join us over on Gitter where the developers and advanced users hang out. I have an empty room there for off-topic discussions.

(We also have a user support forum for tier one support.)

Okay, final answer, before I dive into the newer version where this has been fixed:

#!/bin/bash

CMD=$1
$CMD $2 $3
echo

And that seems to work. I tried several iterations, but in each case (through habit mostly) I was using "exec", and since that essentially starts a new instance of bash, it caused the script error, thus defeating my attempts at a solution I thought was going to work.

Now that I'm directly executing the command ($CMD) without exec, it's working. No more call last scripting error.

You told me not to use exec, or at least implied it could be my problem. It was.

And I'm now on Gitter, which I'd never heard of before it was mentioned in this thread. --Thank you!

One small nit that my bash OCD won't let go of: quoting your variables tends to avoid a lot of problems.

#!/bin/bash
"$1" "$2" "$3" ; echo

See you over on Gitter!

HOW can i fix this error

I love beating a dead horse. LOL

i is an input argument to the function so using it as a local index variable will destroy its original value in the function. I'm not sure of how Python handles parameters, but if it's call by reference (relatively unlikely), it will also destroy the value of i outside of the function.

However, the code reuses i in just this way even though i should be a loop invariant.

This code appears to be written by someone just learning their first programming language who doesn't yet have the skills to keep all the moving pieces in their proper places and roles.

The exception is thrown by

uroll = i[1]

Since Python uses 0 based indexes, that means that i has 0 or 1 characters in it. Among other things, this implies that i does not have the contents in it that the program expects it to have. I would be wrong about this if the for loop forms it's own scope so the iterator i is not the same variable as the function argument i. If this is the case, then the variables should be renamed to avoid such confusions.

I think the OP's code is expecting a number (the list index) and getting a string (the list element) instead. Since uroll is the second element in the OP's list, the code can't get to it. It manages to get to uid, which is the first element in the OP's list, because it was fed data, so it can't be empty. It can't go past that one element, though, because it wasn't given numbers. At least I think that's what's going on. I'm still very much a Python n00b.

If I'm right, then my suggested replacement line of code should fix that issue. If I'm right about the other issue, then that one will suddenly show up once the first one is fixed. Otherwise, all should be well.

Meanwhile, since it would be awfully messy for us to try to run that code up there without the necessary data, I created a couple of snippets below that can be run from a .py file. There are two ways that I know of so far to get at a list's indices and elements. I used range() above and have included it again here along with enumerate() as another way of doing the same thing. You can see that you've got the choice, either way, of accessing the index or the element, which you don't have with for i in fruits: which only gives you access to each element:

# You can use enumerate to access a list's indices or elements:
print("\nENUMERATE EXAMPLE")
fruits = ["apple", "banana", "coconut"]
for i, e in enumerate(fruits):
	# print the index:
	print(i)
	# print the element:
	print(e)

# Or you can use a range in combination with the length of a list to access a list's indices or elements:
print("\nRANGE EXAMPLE")
fruits = ["apple", "banana", "coconut"]
for i in range(len(fruits)):
	# print the index:
	print(i)
	# print the element:
	print(fruits[i])

Aside from the likelihood that this post was probably just posted in the wrong project, the elephant in the room is

def mark_attendance(self, i, r, n):

and 

        if ((i in myDataList) or (r in myDataList) or (n in myDataList)):
            pass
            # speak_va("thank you your attendance has been added")
        elif ((i not in myDataList) and (r not in myDataList) and (n not in myDataList)):

The variable i is a function parameter, so why is it being reused as a loop index? Doesn't that mess things up. If not, it's just bad programming style.

Please explain to me: how is the condition of the elif ever true?

The more I look at this code the more wrong it looks.

All the other variables have meaningful names. Why not the function parameters - which are arguably the least known because they come from a larger external scope?

name_list is built, but never used.

              for  i in myDataList:
               uid = i[0]
               uroll = i[1]
               uname = i[2]
               udate = i[3]

is completely fubar. On each iteration (possibly, a lot of them), those four variables keep getting overwritten by the next value without ever being used. If you just want the values from the last element in MyDataList, that's a terrible way to get them.

We don't have all of the code. For example, we never actually get to see the function call that runs the main function that's holding all of that code. As far as reusing parameters as variables, that's probably considered sloppy code, but it's allowed. Check this out:

print("\nEXAMPLE 1")
def example(i):

	fruits = ["apple"]

	print(i) # green

	if fruits:
		print(i) #green

	for x in fruits:
		print(x) # apple
		print(i) # green

	for i in fruits:
		print(i) # apple

	print(i) # apple

example("green")

print("\nEXAMPLE 2")
def example(i):

	fruits = ["apple"]

	print(i) # green

	if fruits:
		print(i) # green
		i = "red"

	for x in fruits:
		print(x) # apple
		print(i) # red

	if fruits:
		print(i) # red

	for i in fruits:
		print(i) # apple

	print(i) # apple

example("green")

You can see that in these examples, the value we gave the parameter in the function call holds the trump card until/unless another object creates a variable with the same name and gives it a different value. At that point, all other references to it will be the new value until/unless that happens again. So, I guess I'm going to have to think of Python objects as Godzilla from now on.

Please explain to me: how is the condition of the elif ever true?

We don't have the contents of MyDataList, but presumably, if what's in it doesn't match what's fed to it by this script, then the elif would be true, I suppose.

The more I look at this code the more wrong it looks.

You'd probably have nightmares if you took a look at some of the stuff I've coded. It works, but it's not pretty. I'm certain that I can create in a room full of code what others can create in a shoe-box.

All the other variables have meaningful names. Why not the function parameters - which are arguably the least known because they come from a larger external scope?

No idea. It could be a convention or an oversight.

name_list is built, but never used.

I get the impression this is a work-in-progress and has hit a wall, so we're seeing it in an unfinished state.

              for  i in myDataList:
               uid = i[0]
               uroll = i[1]
               uname = i[2]
               udate = i[3]

is completely fubar. On each iteration (possibly, a lot of them), those four variables keep getting overwritten by the next value without ever being used. If you just want the values from the last element in MyDataList, that's a terrible way to get them.

I totally missed this. You're absolutely right. They will always be those four values no matter how many iterations are run. I suspect it was the OP's way to try to get the indices of the list elements.