PureBasic Forums - English

Hello,
I'm experiencing a return value issue with the GetGadgetItemText() function from an ExplorerTreeGadget(), and I'm wondering if this is a bug.
GetGadgetText() returns the full path of the selected item, while GetGadgetItemText() returns the root directory of a file, but a different root directory for a folder.
I find this very strange!???
A little test code to see the problem. Have I misinterpreted it?
(Test performed with PB 6.21 32-bit, Windows 11 Professional)

I dont know, but I think that you cannot do this?
SelectElement is the position.
The Constants are not "compatible" with the position. I added a comment with what the compiler see..... You can use FileSize() to detect the File or Directory Type.

Hi Axoltl,

In my documentation ExplorerTreeGadget() is explained:
GetGadgetState() : Lets you know if the selected item is a directory or a file. <<
GetGadgetItemText() : Returns the full path of the specified item.
GetGadgetText() : Returns the full path of the selected directory or file.

This can be a big problem if you want to use GetGadgetItemText(#Gadget, Item [, Column]) to delete a file or directory. Thinking you got the full path to your file, you got a root directory.
It can be funny

Jacobus wrote: Thu Jul 17, 2025 5:28 pm In my documentation ExplorerTreeGadget() is explained:
GetGadgetState() : Lets you know if the selected item is a directory or a file. <<

That is correct, but you are using the return value as 'position' for GetGadgetItemText.
It's not returning an index, just the type of the selected item.

Jacobus wrote: Thu Jul 17, 2025 5:28 pm ExplorerTreeGadget() and GetGadgetItemText() >> is this a bug?

I wonder whether it is possible to ask a simple coding question without using the word “bug”.

Little John wrote: Thu Jul 17, 2025 6:48 pm

Jacobus wrote: Thu Jul 17, 2025 5:28 pm ExplorerTreeGadget() and GetGadgetItemText() >> is this a bug?

I wonder whether it is possible to ask a simple coding question without using the word “bug”.

bug was removed from the title.

Sirius-2337 wrote: Thu Jul 17, 2025 6:07 pm

Jacobus wrote: Thu Jul 17, 2025 5:28 pm In my documentation ExplorerTreeGadget() is explained:
GetGadgetState() : Lets you know if the selected item is a directory or a file. <<

That is correct, but you are using the return value as 'position' for GetGadgetItemText.
It's not returning an index, just the type of the selected item.

Ok, but in these conditions how do you get the position of the element in question?

Okay, I read you code and not the help for that command.
You are right. The GetGadgetState() is different for this Gadget.

HELP-Text:
The following functions can be used to control the gadget:

- GetGadgetText(): Get the full path of the currently selected directory/file.
- SetGadgetText(): Set the currently selected file/directory.
- GetGadgetState(): Check if the selected item is a file or a directory.
- GetGadgetItemText(): Returns the full path of the specified item.

So there seems to be no PB way to identify the selected element. Which would be needed for the parameter Position in the GetGadgetItemText()
Furthermore, GetGadgetItemText() would only return the same result as GetGadgetText().
From my point of view, the help should probably be corrected by removing the forth function.

Axolotl wrote: Fri Jul 18, 2025 11:03 am So there seems to be no PB way to identify the selected element. Which would be needed for the parameter Position in the GetGadgetItemText()
Furthermore, GetGadgetItemText() would only return the same result as GetGadgetText().
From my point of view, the help should probably be corrected by removing the forth function.

I came to the same conclusion. That was also why I wanted to test both functions and see the difference in the results. But the function is incomplete if we can't have the position argument.

I'd say GetGadgetState() should work like for all the other gadgets and return the selected index.
It feels strange, when it here returns the kind of selection.
Anyway, no bug, more a feature request.
Unfortunately it would break existing codes.
Not mine though, I never use ExplorerTreeGadgets

HeX0R wrote: Fri Jul 18, 2025 5:26 pm I'd say GetGadgetState() should work like for all the other gadgets and return the selected index.
It feels strange, when it here returns the kind of selection.
Anyway, no bug, more a feature request.
Unfortunately it would break existing codes.
Not mine though, I never use ExplorerTreeGadgets

See GetGadgetState Help. Is not always the index ...

Depending on the gadget type, it is different.

I was talking about those Gadgets which allow more entries.
Here I see exactly one, where things are different.
Especially if you compare ExplorerList and ExplorerTree gadget.

HeX0R wrote: Fri Jul 18, 2025 8:39 pm I was talking about those Gadgets which allow more entries.
Here I see exactly one, where things are different.
Especially if you compare ExplorerList and ExplorerTree gadget.

This is exactly what is confusing. Especially since GetGadgetState() in ExplorerTreeGadget() must obtain the index of the selection to be able to tell if it is a file or a folder. At least, it seems so to me. In this case, it would indeed be desirable for a function to be able to return the index of a file in an ExplorerTreeGadget() to fully exploit it, without torturing oneself to write a routine that gives the index.

The index changes every time a tree is opened or closed.
Thus, the index does not help.

mk-soft wrote: Fri Jul 18, 2025 11:57 pmThe index changes every time a tree is opened or closed.

But if you run the code from the first post and DON'T open anything, and just simply right-click "C:\" and don't do anything else, the debug output shows both "C:\" and "D:\" for that selected item. (Assuming you have a D: drive, like I do).