You can execute “in batch” (without any human interaction) any Anatella-graph.
Let’s assume that the Anatella executables (“Anatella.exe” or “AnatellaConsole.exe”) are inside a directory inside your “PATH” environmental variable (this is the default setting for a standard Anatella installation).
To run an Anatella-Graph “in batch”, type:
Anatella.exe –e "<file_to_execute.anatella>"
Anatella.exe –r "<file_to_execute.anatella>"
The 2 parameters (“-e” and the <Anatella filename>) must be the first 2 parameters on the command line. The “-e” and “-r” parameters are equivalent (“e” stands for “execute” and “r” stands for “run”): A simple log-window will appear (the Anatella graph is thus not visible) (you can hide the log window with the “–s” option) and the graph is executed. The “AnatellaConsole.exe” executable will never create any new log-window: it will use the current shell (or console) window to display all the trace messages.
When a program terminates, it always returns back to the operating system (i.e. to MS-Windows) a number. This number is named the “error level” of the program. By convention, an “error level” that is non-zero indicates that there was an error during the execution of the program. Anatella extends slightly the convention: More precisely:
•Error Level=0: Everything went ok.
•Error Level=1: There were some warnings during the execution of the Anatella Graph.
•Otherwise: There were some errors during the execution of the Anatella Graph.
•Any external software (such as curl, scp, cmd, etc.)
•Any Anatella graph (because you only need to run “Anatella.exe” with the proper command-line parameters).
Using the ProcessRunner class, you can access the “error level” of all the programs that have been launched. For example: it’s very easy to implement a very complex logic based on the success (“error level”=0) or the failure (“error level”>1) of the execution of your Anatella graphs. For a simple example of usage of the ProcessRunner class, see the RunProcesses Action (see section 5.20.3.)
The ParallelRun Action (see section 5.3.3.) also makes extensive use of the “error level” to detect errors or warnings during graph execution.