Recovering Lost Source Code

The Problem

Before ITMS offers a modernisation service project, a complete in-depth analysis must be performed on all of the customer’s code.

At that stage ITMS discovers very often discrepancies between the productive (binary) version of a program and its source code, e.g.:

  • the load module exists, but the source code got lost.
  • the source code version differs from the version which is in production, i.e. a compile and link does not produce the productive load module.
  • the compiler which was used to create the binary file, does no longer exist

This fact will cause difficulties and risks for the modernisation project.

In any case this situation means an unacceptable risk for the IT operation - though yet undiscovered.

The Solution

The framework accepts the following binary file types:

  • z/OS and z/VSE load modules and object decks
  • VM/CMS modules
  • Created by compilers or mainframe assemblers

The binary code is converted in several steps, where each iteration produces a functionally identical binary file.

The resulting source code can be enriched by information which is obtained through

  • data names contained in descriptions of input and output files
  • comparative compilations
  • customer specifications

Lost source code happens
more often than you might expect.
ITMS services recover that code.

ITMS advanced source code processing
technologies, create source code files
which are functionally equivalent and maintainable.