This might qualify as the most bizarre issue I have run across.
I have a command line utility which returns results via an errorlevel value (and stdout, but that's not relevant here). I use that utility in a complex script and at times I saw that the script generated unexpected results which I could not explain. I was finally able to determine how to trigger the behavior and came up with a simple .BTM which doesn't rely on any external utility to demonstrate what I'm seeing.
It appears that when the Errorlevel is set to 4095, and AFAIK no other value, subsequent Do Loops exit after only a single iteration.
Here's the .BTM which demonstrates the issue:
And here's the output it generates:
Note that in the Second Test both Do Loops seem to end after a single iteration.
As noted in the .BTM comments, this is not specific to version 37. TCC versions 34-37 have this behavior. TCC version 24 does not.
Now that I know what is happening, it's a simple enough workaround to reset the Errorlevel after checking it. And I may be the only person in the universe who encountered this, so I don't expect addressing this to be a high priority.
I have a command line utility which returns results via an errorlevel value (and stdout, but that's not relevant here). I use that utility in a complex script and at times I saw that the script generated unexpected results which I could not explain. I was finally able to determine how to trigger the behavior and came up with a simple .BTM which doesn't rely on any external utility to demonstrate what I'm seeing.
It appears that when the Errorlevel is set to 4095, and AFAIK no other value, subsequent Do Loops exit after only a single iteration.
Here's the .BTM which demonstrates the issue:
Code:
@echo off
setlocal
REM --------- Problem occurs if you select the first 12 items displayed by MULTIPLECHOICEBOX
REM --------- - This will return an Errorlevel set to 4095 (twelve 1's in base 2 = 111111111111 = 4095 in base 10)
REM --------- - For some reason this value, and only this value, causes subsequent DO loops to exit after 1 iteration
REM --------- - So after checking the return code, resetting Errorlevel to 0 makes things work correctly
REM ---------
REM --------- TCC Versions 34-37 Have this issue
REM --------- TCC Version 24 Does NOT have this issue
set ItemList=item1;item2;item3
echo.
echo ------------------ First Test ------------------------
REM SETERROR 4095
SETERROR 0
DO 3
echo.
echo *** Outer Loop: %_Do_Loop
set ItemList*
Do Item in /T";" /L %ItemList
echo Processing %Item
Enddo
enddo
echo.
echo ------------------ Second Test ------------------------
SETERROR 4095
DO 3
echo.
echo *** Outer Loop: %_Do_Loop
set ItemList*
Do Item in /T";" /L %ItemList
echo Processing %Item
Enddo
enddo
quit
And here's the output it generates:
Code:
TCC 37.00.18 x64 Windows 11 [Version 10.0.26200.9974]
Copyright 2026 JP Software Inc. All Rights Reserved
Your evaluation period expires in 23 days.
You can buy Take Command, TCC, and CMDebug at https://jpsoft.com
[C:\Program Files\JPSoft\TCMD37]c:\batfile\TestEL4095.btm
------------------ First Test ------------------------
*** Outer Loop: 1
ItemList=item1;item2;item3
Processing item1
Processing item2
Processing item3
*** Outer Loop: 2
ItemList=item1;item2;item3
Processing item1
Processing item2
Processing item3
*** Outer Loop: 3
ItemList=item1;item2;item3
Processing item1
Processing item2
Processing item3
------------------ Second Test ------------------------
*** Outer Loop: 1
ItemList=item1;item2;item3
Processing item1
[C:\Program Files\JPSoft\TCMD37]
Note that in the Second Test both Do Loops seem to end after a single iteration.
As noted in the .BTM comments, this is not specific to version 37. TCC versions 34-37 have this behavior. TCC version 24 does not.
Now that I know what is happening, it's a simple enough workaround to reset the Errorlevel after checking it. And I may be the only person in the universe who encountered this, so I don't expect addressing this to be a high priority.