The differentiation between different file types
enables COBOL Translation Toolkit to organize the processing of the files in a
better order. If all your programs use the same extension, you will need to
modify them to use different extension. Note that ALL PROGRAM FILES REQUIRE AN
EXTENTION.
Programs translated by COBOL Translation Toolkit from
COBOL to DELPHI
or JAVA.
There are 4 stages to migration of each COBOL program
module: Pre-Processor:
Minimal Syntax changes and preparation of COBOL programs.
Parsing:
Semantic parsing of programs - generates a logical model of program structure
and data-definitions of each program.
Analysis: Further Analysis of program modules and data-definitions will produce
an intermediate model of both data and operation logic programs and data-files.
Code Generation: The process of translation of the intermediate data/logical
models into a selected target program.
There are many steps within each of the above stages.
The user interface is organized in such a way to help you going through these
processes.
There are two options on parsing.
Parse of data-definitions is useful if you may need to examine and modify the
data structures of you programs - prior to the main processing. It effectively
cuts out the processing of the PROCEDURE divisions, hence saving time in
parsing time required. Note that the date-time conversions will NOT be picked
up by the system - however, you may manually change the data type of a variable
using the functions provided. The main benefit of parsing the
data-definitions first is that you can see the effect of REDEFINITIONS of
variables, and how redefined map on to each other, and where necessary
remove/change the redefinitions if they are not absolutely essential. Such
changes must be saved, and the subsequent re-parsing of the programs will pick
these changes up.
DO NOT ATTEMPT TO FURTHER PROCESS (ANALYSE) the partly parsed programs. You
will need to start complete parse before any analysis can be done by the
system.
You can modify the REDEFINEs on a field/group to
Ignore Redefines or Delayed Redefines. By changing the Redefines to Ignored
Redefines the system essentially creates different variable/memory spaces for
each of the involved variables, as opposed to the REDEFINES statement, which
essentially creates a SINGLE memory space/variable for the involved fields.
The system will make no attempt to consolidate the variables whose definition
has been changed to Ignored Redefines.
Delayed Redefined is similar to above, but the system will attempt to
consolidate the variables after each operation on the variables.
For example, the variables in following COBOL code
03 var-1 pic x(3) .
03 var-2 REDEFINES var-1 pic x(2) comp
....
move 'abc' to var-1
move 'xyz' to var-2
will be translated into only the 1 variable in the
target language:
String var_1;
...
var_1 = 'abc';
var_1 = 'xyz';
By changing the redefinition to Ignored Redefines, the
target code will become:
String var_1;
String var_2;
...
var_1 = 'abc';
var_2 = 'xyz';
By changing the redefinition to Delayed Redefines, the target code will become:
String var_1;
String var_2;
...
var_1 = 'abc';
var_2 = var_1;
var_2 = 'xyz';
var_1 = var_2;
Note that the system does not guarantee that the
variables will always be equated on Delayed Redefines. Particularly, the
equating processmay be skipped on operations involving sub-groups of the
Redefined item, for example in translation of the following COBOL statements:
02 group-1 .
03 group-2 .
05 var-1 pic x(3).
02 var-2 redefines var-1 pic x(3)
....
MOVE 'abc' TO VAR-1 .
If the Redefinition on Var-2 is modified to Delayed
Redefines, then the system may just generate var_1 = abc';and var_2 = var1; may not be generated.
This is useful when an FD is used in more than 1 COBOL
module/program.
The first time you parse and analyze a COBOL program, which utilizes this
particular FD, the COBOL Translation Toolkit will generate its own internal
representation of the FD. Saving your work at this stage will improve the
system performance in dealing with other COBOL modules also utilizing this FD.
Saving your work will also allows COBOL Translation Toolkit to remember any
modifications you may have made to the COBOL Translation Toolkit’s
representation of the FD using Expanded
FD Screen. None of the migrated programs access the data migrated
from COBOL WS/FD directly. All access is achieved through a set of objects
automatically created by COBOL Translation Toolkit. This structure eases
optimization of the memory access routines (for WS wrappers) or the database
access routines (for FD wrappers) by allowing changes to be made to the objects
representing the WS section or the FD, and none of the migrated programs will
require modifications.