home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Monster Media 1994 #1
/
monster.zip
/
monster
/
PROG_GEN
/
PEO.ZIP
/
WHATSNEW.DOC
< prev
Wrap
Text File
|
1994-01-30
|
10KB
|
233 lines
WHAT'S NEW WITH THIS VERSION OF PEO
Since we first offered our custom analysis tool, PEO, we've
made significant enhancements that have greatly increased
the efficacy of PEO. The enhancements
* make it easy to document the results of the analysis
* help the user keep track of the locations along the data
transformation paths and logic flow paths
* "filter" data element names for specified character strings
* recognize CICS and IMS DL/I language elements
* display the hiearchy of subprograms.
The most significant enhancement is Code Lifter PLUS. Now you
can copy code from the COBOL program into a buffer area and add
notes and comments. This is an easy way to build documentation
about the results of the program analysis.
The other important advances are:
PEO now automatically returns the cursor to the proper location
on the PEO ELEMENT DISPLAY SCREEN when you return from the PEO
SOURCE CODE DISPLAY SCREEN. This eliminates the need for the
user to re-position the cursor.
The parser (SEO) can produce a file of selected data element
names based upon specified strings of characters in the data
element name. The "filter" can be helpful to guide the user to
those data elements which likely contain dates, for example.
CICS commands that contain the INTO, FROM, COMMAREA, and RIDFLD
clauses are recognized and related to the elements referenced
by the commands.
IMS DL/I calls ("CBLTDLI") are recognized and related to the
elements referenced by the call.
The parser automatically produces a file showing the hierarchy
of subprograms. This can be helpful to view the overall structure
of a program. The subprograms are listed in the order in which
they were encountered as the parser reads sequencially through
the main program and subprograms.
PEO / Code Lifter PLUS
PEO / Code Lifter PLUS gives the user the ability to lift text
from the source code and copy it into a buffer. You can edit the
code and add your own notes and comments. It's a fast, powerful
tool to quickly document the things you learn about the program.
After you have finished you can save the buffer into a file.
It's easy to use:
On the PEO ELEMENT DISPLAY SCREEN:
Press F-8. The TARGET element will be copied to the
buffer and identified as the TARGET ENTITY.
On the PEO SOURCE CODE DISPLAY SCREEN:
Just position the cursor on the line of code you want
to copy into the buffer and press F-8. The cursor will
automatically advance to the next line.
To enter the editor simply press F-9. After you've
finished you can exit back to the PEO screens by
pressing CTL+W to keep the changes; or, you can press
ESC to exit the editor without saving the changes. The
buffer is not written to a file until you exit PEO.
When you exit from PEO you'll be asked if you want to
save the buffer in a file. If you do you simply type
in a name without an extension. PEO automatically adds
.cap as the extension.
It's reccommended that you do not attempt to do all analysis
required and document the analysis in a single file, especially
if it's a large project. Instead, break the project up into
small fragments and name the files you save in a manner that
will suggest some hiearchy or structure. Later, you can assemble
the information from all the files into a final document.
DATA NAME FILTER.
This feature provides the capability to tell the parser to
select and write to a file all the data element names which
contain specified strings of characters.
The specified strings are entered into a file called,
seofiltr.fil. This file must reside in the application
identifier sub directory of MACHY. There must be only one
string per record and the string should start in column one.
The file must be an ASCII text file. You can edit the file to
suit your needs. If the file is missing or empty, the parser
simply bypasses the filter logic. Here is a sample file
containing strings of characters which pertain to date fields.
E.g.,
DATE
DTE
DT
MDY
DMY
YMD
YY
MM
DD
The file that contains the selected data element names is
called progid.dte; where progid is the name of the COBOL
program or the name you typed into the parser when you
combined programs for parsing. The file is written to the
EXS sub directory of the application identifier.
CICS commands and IMS DL/I
CICS commands that contain the INTO, FROM, COMMAREA, and
RIDFLD clauses are recognized and related to the elements
referenced by the commands. The relations are defined in
terms of data entering the program or leaving the program,
i.e., a READ or WRITE, including group read and group write.
IMS DL/I calls ("CBLTDLI") are recognized and related to the
elements referenced by the call. This is accomplished by a
skeleton COBOL program that contains and entry point, "CBLTDLI".
The arguments associated with the entry point are the parameters
normally passed to DL/I. The skeleton program is included with
PEO install diskette and resides in the MACHY directory.
SUBPROGRAM HIEARCHY
The parser writes the called entry point and the name of the
COBOL source file containing the entry point to a file called
progid.hih; where progid is the name of the COBOL program or
the name you typed into the parser when you combined programs
for parsing. The file is written to the EXS sub directory of
the application identifier. The names are indented to suggest
the hiearcical relation of the nested subprograms. Here is an
example:
000 MACHYM20
001 M20MAT
002 XMATSE
003 MINMAX
003 MAT006
004 MAT002
004 MAT010
005 MAT011
006 MAT012
007 MAT020
007 MAT004A (MAT004)
006 MAT004C (MAT004)
006 MAT004B (MAT004)
006 MAT004D (MAT004)
005 MAT004A (MAT004)
005 MAT004B (MAT004)
005 MAT011
004 MAT030
A NOTE ABOUT THE SORT
The sort that the parser uses is actually a series of merges.
The sortin file is split into strings. The strings are merged
and split into other strings that are then merged and split
again until all the records are in ascending sequence. Each
string is displayed as a dot (.) in order to show that something
is happening and the computer has not "hung up". Large files
will fill the screen with dots and may cause some concern if
you did not know what is happening.
A NOTE ABOUT THE COBOL SOURCE CODE
The COBOL source code must be in 80-column card image format.
That is, the source code must be in columns eight through
seventy-two. The CR/LF can terminate the record in any column.
The sequence columns, columns one through six must be physically
present and the indicator column (column 7) must be present.
The current version of the parser edits for this format.
BUG FIXES
The data definition clause, PIC X VALUE '''', is now processed
correctly. Note: the clause will appear on the PEO SOURCE CODE
DISPLAY screen as: PIC X VALUE '#'.
The COPY ... REPLACING == ... == BY == ... == logic has been
corrected. Before the fix, the REPLACING clause,
==OPERAND1== BY ==OPERAND2==
was not processed correctly because the parser expected a space
to preceed the operand within the == delimiter.
The logic has been changed to process the RENAMES clause
correctly.
The first source statement that is not a comment must be one
of the following:
IDENTIFICATION DIVISION
ID DIVISION
PROGRAM-ID
AUTHOR
REMARKS
INSTALLATION
DATE-WRITTEN
DATE-COMPILED
SECURITY
ENVIRONMENT DIVISION
CONFIGURATION SECTION
SOURCE-COMPUTER
OBJECT-COMPUTER
SPECIAL-NAMES
INPUT-OUTPUT SECTION
FILE-CONTROL
SELECT file-name ASSIGN [TO] ....
I-O-CONTROL
DATA DIVISION
FILE SECTION
WORKING-STORAGE SECTION
LOCAL-STORAGE SECTION
LINKAGE SECTION
COMMUNICATION SECTION
REPORT SECTION
SCREEN SECTION
PROCEDURE DIVISION