| Key: |
QB-1824
|
| Type: |
Improvement
|
| Status: |
Resolved
|
| Resolution: |
Fixed
|
| Priority: |
Major
|
| Assignee: |
Unassigned
|
| Reporter: |
productivity
|
| Votes: |
0
|
| Watchers: |
0
|
|
If you were logged in you would be able to see more operations.
|
|
|
|
Original Estimate:
|
Unknown
|
Remaining Estimate:
|
Unknown
|
Time Spent:
|
Unknown
|
|
Currently, the behavior of the "Shell/Batch command" is different when executing a single command line (which AFAIK, uses the Java ProcessBuilder) than when executing multiple line commands (which actually creates a temporary batch or shell file that is interpreted directly by the console/terminal). I believe that in both cases, the behavior should be the same and they should both create a temporary batch/shell file for interpretation by the shell.
See http://forum.pmease.com/viewtopic.php?f=1&t=2876 for more details and for an example of the kind of problem this issue can bring.
|
|
Description
|
Currently, the behavior of the "Shell/Batch command" is different when executing a single command line (which AFAIK, uses the Java ProcessBuilder) than when executing multiple line commands (which actually creates a temporary batch or shell file that is interpreted directly by the console/terminal). I believe that in both cases, the behavior should be the same and they should both create a temporary batch/shell file for interpretation by the shell.
See http://forum.pmease.com/viewtopic.php?f=1&t=2876 for more details and for an example of the kind of problem this issue can bring. |
Show » |
|
QuickBuild should create a temporary batch file/shell script even for a single line command
Now, encoding of an shell script is undefined for me and I could not even discover what encoding they now have.
In examples:
I have single line shell script which is run on Windows agent
curl http://hostname/path/to/mantis -d <variable1> -d <variable2>
where <variable1> is a string on Russian (in my case; Windows-1251 encoding)
I've run it on 5.0.41 (and I've just checked it on 5.0.42) - it works as it was expected.
After upgrade to 5.1.0 <variable1> now have unknown and undiscoverable encoding, so I even cannot specify its encoding.
P.S. I've checked multiline script on 5.0.41/42 - it's "broken", too
P.P.S. Unfortunately, I've discovered this problem after several months after upgrade to 5.1.x, so downgrade is not an option