See more than one level deep for the tree visual of JCL about mapa HOT 3 CLOSED

Is there some way to get the parser to output an outtree that resolves the FORMATPGM tree, and any programs IT calls as well?

Sort of.

If you run your COBOL source code through the CallTree.jar in the mapa/cobol directory, you will get an output file detailing information about each COBOL program. You might also need the DB2zSQL.jar file.

The intended use of these delimited output files is they can be processed in any way you like. You could load them into a database and process them from there with SQL or you could just read the files and process those to produce whatever sort of output meets your needs.

So, you could process the JCL with JCLParser.jar, then process the COBOL with CallTree.jar, then put the two outputs together in a nice report or tree structure represented in a GUI with code you write.

from mapa.

Ah this is great thank you! We may try to add this functionality from the csv files output or potentially intercepting the intermediary AST from ANTLR.

Re: the docs around -include / -includeList:

 -include <arg>       name of a single path in which to locate cataloged
                      procs and JCL INCLUDEs, takes precedence over the
                      includeList option
 -includeList <arg>   name of a comma-delimited file containing a list of
                      paths in which to locate cataloged procs and JCL
                      INCLUDEs and the PDS(E)s to which they correspond

These are required flags, so I'm wondering what the purpose of providing these are -- our initial assumption when reading these docs were that it would "resolve" program files and unfurl them under the call site of procs or programs in the csv / tsv output, but that doesn't appear to be happening at first glance / from your response. What does providing these flags do?

Thanks again for your prompt replies!

from mapa.

Those options are for specifying a directory or directories for resolving cataloged procedures and JCL INCLUDE statements. The former are the code run via an EXEC PROC= statement, the latter resolve to JCL fragments which IBM calls "include groups". My suggestion would be to map a directory to each PROCnn DD in the JES2 cataloged procedure and to each library mentioned in a JCLLIB statement in a job being analyzed.

from mapa.