F
FreeChemistry
Hi
I guess this is a bit of a strange one but I can't think of a better
place to start.
I have just started a fixed term contract that requires me to
investigate a large enterprise system written by (or more accurately
managed by) a large, well known company that has had a rather
important role in the development of Java. The documentation consists
of javadoc comments in the code and a rudimentary UML model for one
small part of the system. I reckon a first estimate of the number of
lines of source code (including comments) to be in the region of
250,000, 500,000 lines of code in over 1000 source files (yep, some of
these classes have over 2K lines in them). I have found out how the
state is persisted (a combination of JDBC access, Castor and EJB 2.0
CMP) Now I need to figure out how the business works. I have access to
the front end and the struts config stuff but I can't really exercise
the system as it's in the wild and being used. There are some very
strange things going on in the code (Like VOs that access the database
and DAOs that extend the VOs !!!!)
I would be most interested in any techniques used by
developers/programmers in a similar situation. I have the full source
available and I'm using the Studio 4 IDE that comes with appserver7
(I've also got Netbeans). I cant find the Ant build file although I
know the developers are using Ant so I'm unable to build or deploy the
application at the moment althought I hope to fix that soon.
Cheers
FreeChemist
I guess this is a bit of a strange one but I can't think of a better
place to start.
I have just started a fixed term contract that requires me to
investigate a large enterprise system written by (or more accurately
managed by) a large, well known company that has had a rather
important role in the development of Java. The documentation consists
of javadoc comments in the code and a rudimentary UML model for one
small part of the system. I reckon a first estimate of the number of
lines of source code (including comments) to be in the region of
250,000, 500,000 lines of code in over 1000 source files (yep, some of
these classes have over 2K lines in them). I have found out how the
state is persisted (a combination of JDBC access, Castor and EJB 2.0
CMP) Now I need to figure out how the business works. I have access to
the front end and the struts config stuff but I can't really exercise
the system as it's in the wild and being used. There are some very
strange things going on in the code (Like VOs that access the database
and DAOs that extend the VOs !!!!)
I would be most interested in any techniques used by
developers/programmers in a similar situation. I have the full source
available and I'm using the Studio 4 IDE that comes with appserver7
(I've also got Netbeans). I cant find the Ant build file although I
know the developers are using Ant so I'm unable to build or deploy the
application at the moment althought I hope to fix that soon.
Cheers
FreeChemist