JobScheduler can monitor file creation, file changes and timestamp changes with a (standalone) job. This job can then start a job chain or job group.
An example configuration for a "file-watcher":
<job name = "scheduler_file_notification_1">
<script language="shell">
<![CDATA[
set
echo %SCHEDULER_JOB_NAME%
echo %SCHEDULER_TASK_TRIGGER_FILES%
rem del %SCHEDULER_TASK_TRIGGER_FILES%
]]>
</script>
<start_when_directory_changed directory = "c:/temp" regex = "^file[1|2]\.xml$"/>
</job>
This job will watch the folder c:/temp on a Windows system. The regular expression is used to define which file(s) JS has to watch. In this example the regex means that every event involving the files file1.xml and/or file2.xml will fire the execution of the embedded script (which in this case will echo values for JS-Environment Variables to stdout).
Every change of files in the folder c:/temp which is matched by the filter ^file1\.xml$ starts a task for the job.
The SCHEDULER_TASK_TRIGGER_FILES environment variable contains all the filtered files in the c:/temp folder at the time where the task is started. These files are held in the form of a semicolon separated list. This list doesn't just contain the file which was changed.
Note that this behaviour could be critical when two files (e.g. file1.xml and file2.xml) are changed at the same time.
Two tasks will then be started for the job.
In the first task the environment variable SCHEDULER_TASK_TRIGGER_FILES has the value c:/temp/file1.xml;c:/temp/file2.xml.
In the second task the environment variable SCHEDULER_TASK_TRIGGER_FILES will have a value that depends on what the first task has done with the files.
If, for example, the first task deletes the files then SCHEDULER_TASK_TRIGGER_FILES will be empty for the second task.
For further information we recommend the Directory Monitoring with File Orders.