Release notes: Version 20.08.01 of BMC Compuware Topaz Workbench

Git integration improvements

For a Git project in Topaz, we have removed the pop-up dialog boxes to make sure that during Git to ISPW mapping, file download runs in the background without interruption.

Sandbox improvements

We have made the following Sandbox improvements:

• You can now use Transfer Task for Sandbox Codelines to transfer your work to a Release. (Promotion and Deploy will be enabled later.)
• ISPW views now include the Code line path.

DevOps improvements

We have made the following improvements:

• We have added a new REST API to update fields in an Assignment or Release. Hence, you no longer must use ISPF or Topaz to update the field such as Description.
• The CLI accepts a Certificate in addition to the User ID and Password to authenticate the TSO connection.

Line, Record and Page numbers 

The Line number and Record number are now available during File-AID Edit of a z/OS file. These values are useful in determining the current position within a file. The Page number is now displayed at the top of the editor for the z/OS file.

Line and Page Navigation 

Direct navigation to a specific line or page is now available for Edit of z/OS files. You can navigate to this option in following ways:

• Go To Line/Page icon on the Toolbar displays a dialog box where you can enter the line or page number
• Enter the value for the page to be browsed and click GO
• Go To Line/Page navigation from the right-click options which are displayed for each line

Data Source Statistics

Data Source Statistics are available from the Toolbar for Edit of z/OS files. The statistics now includes the total number of pages needed to display the entire file.

File-AID Services Ports

File-AID Services (FAS) can now use the same port for Topaz Data Privacy client communications and Data Privacy Execution. This will eliminate the requirement to have two different ports for FAS.

ISPW/Eclipse

ISPW Distributed Deploy is no longer supported.

SonarLint Analyze Changed Files option is not functional for ISPW controlled files.

For Tasks generated from a Sandbox Codeline, the Parts View shows an incorrect data set name.

Topaz Enterprise Data

The potential for loss of precision exists on FLOAT and DOUBLE data types between the mainframe and the File-AID Data Editor user interface. For more information, see Chapter 9. Floating-Point Overview and Support Instructions in IBM’s z/Architecture Principles of Operation (SA22-7832-09).

Picture clauses for COMP and COMP-4 (binary data) field types are not used when displaying the decimal numeric value of these fields. Any implied decimals (V) or implied scaling (P) in the Picture are ignored and not used to display the numeric value of the field. The value displayed is a positive or negative integer.

When saving data to a locked Oracle table, the timeout error does not return after the set timeout period. It does return after the lock has been released.

Topaz Workbench

After installing multiple Topaz components into an Eclipse-based IDE, you are asked to restart. During that restart, however, the Eclipse IDE sometimes fails to rewire the bundle dependencies. Symptoms can include UI components that do not function properly and views that do not display. To correct this, manually close the Eclipse IDE, then start it back up again with the -clean option. This might be necessary to do only once.

The topaz.ini file sets the osgi.instance.area.default to the @user.home location, which is typically mapped to a local drive or absolute path. If the user has their @user.home set to a UNC network path, the toolbars and menus within Topaz Workbench might not be displayed.

To correct the problem, edit the topaz.ini file (default location is the Topaz Workbench installation’s eclipse directory), replacing the @user.home location with a local drive or absolute path. If Topaz Workbench is installed via p2repo into an Eclipse instance, you must edit the eclipse.ini file instead.

Navigating Topaz COBOL Editor search results do not place the cursor at the exact position of the match in the Editor if the match exists within a non-editable region. This pertains to the Sequence Number or Identity areas within the Editor where either have been set via preferences to be non-editable. The cursor in this case is placed at the beginning of the statement or in the general vicinity of the match in the editable region.

When exporting an image from Program Analysis or Visualizer if the View document after export check box is selected, *.png is selected from the Save as type list. The default application associated with .png files is Microsoft Paint and the exported image cannot be opened automatically. Browse to the export location and open the file manually.

Under Open JDK, when the Team Profile Import process finds an invalid profile, it fails to display the full error message.

When analyzing a project, for example after adding a large copybook concatenation PDS, Program Analysis can take a long time or hang Topaz Workbench.

When installing Topaz Workbench plug-ins into IBM Aqua, the following issues might occur:

• Deleting a PDS member in Host Explorer and then trying to open the same member from the Project Explorer might cause Aqua to hang.
• Creating a new folder in a Compuware Online Project and then creating a new file in that folder causes Aqua to hang.
• The shortcut keys for Toggle Comment (Ctrl+/) do not work in the JCL editor.

If you open a file in Host Explorer, perform Search Text > File within it, close the Editor, and then click on the search results, the selected file fails to open. This happens because when the Editor is closed, the file is no longer present in the Topaz Workbench local directory. Reopening the file through Host Explorer re-syncs the search results with the file. The issue affects all Topaz Editors.

Cut (Ctrl+x) is not supported in the Topaz 3270 Emulator.

Topaz Workbench

Error markers now appear for a local COBOL program.

Assembler comments beginning with a symbol are now colored correctly.

Direct conversion of an Object value to Boolean during Import of a Team Profile throws a ClassCastException.

When auto-tagging is on and user deletes a line, the tag in col. 73-80 will be properly deleted too.

Updated the Japanese message when there are mismatching shift out/in characters in source file.

PL/I grammar syntax error caused StackOverFlow for DEACTIVATE and ACTIVATE in Exec statement.

When reopening the Find/Replace dialog, the checkbox for "Find all occurrences" will be set to the previously used state.

The user will now be able to copy both ASCII and binary files from Windows to a VB PDS with record length up to max PDS record length. 

PL/I parser error: invalid statement on dcl with init. Added hex_literal "G X"  in the grammar file to handle the init statement.

NullpointerException in the PL/I preprocessor. Added fix that allows the use of an empty body on parenthesis. 

PL/I Parser error: mismatched input 'e' expecting … Updated grammar for float literal to support lower case letters.

Code Coverage source matching failed in Jenkins when the source is downloaded using the Native Endevor APIs. Source matching was change to ignore all leading and trailing white spaces for each line of text in the source.

An infinite loop generated due to a StackOverflowError during parsing of the Outline view for PL/I source. 

Handled the Null pointer exception in the case when procedure has a null value in the RETURN statement. The PL/I program should pass value in return statement of procedure as per language grammar rule.

Jenkins plugin and Topaz Workbench CLI were unable to pass the new x509 certificate parameter correctly.

Made internal changes to the Cobol incremental parser to better recover when there are syntax errors that break the ability to parse subsequent code successfully. Previously there were occasional problems with syntax errors being marked in the code after the cause had been fixed.

Added condition to check whether the variable to be added in Content Assist dialog is anonymous. If its anonymous, then avoid adding in variable Collections to be displayed in content assist popup.

Hovering over a copybook name will now show the copybook, even if there are blank lines or comments between the COPY statement and the copybook name.

Fixed the alignment of the Text editor preference page.

Fixed alignment of the Assembler preference page.

PL/1 parser grammar corrected so that an ON Condition with a Label does not trigger a parser error.

Updated the syntax rule in the PL/I parser for assignment expression to fix the extraneous input ''1'B' error.

The notifications from the Eclipse framework that indicate when a document is edited were not timely, and inconsistent about how much of the change was reported at once.  A timely and reliable way of detecting changes was determined, and the Cobol incremental parsing was updated to give a much better user experience. 

A 2-second delay was also added before incremental parse begins - parsing now only starts 2 seconds after the last change is made. This improves the user experience since parsing while the user is typing is not useful.

LINKAGE SECTION updates outline view when placed in a comment. There was a missing filter for comment lines in the parser.

Formerly we were storing information directly from the parser for later use, which contains for more data than we need. This was potentially causing hangs during edit. Changed it to save off only the data we need and discard the original excess parse data.

Fixed PL/I code folding for comment blocks

The Cobol editor incrementally parses small changes as the user edits a Cobol file in order to update the outline view and display syntax errors.  In some cases this resulted in runaway memory usage that significantly slowed down the parsing.

Changed Cobol parser rule to not require a period following a PROGRAM-ID statement due to errors being generated for a nested program.

PL/I Outline view is sorted by the order that items are defined.

The put data formats (A,B,C,F, etc.) are now treated as variables.

Warning dialog message for the Denumber function was updated to reflect the changes in the preference pages.

The cursor no longer jumps to the beginning of the file when saving in the PL/I editor.

Error markers disappear from the vertical ruler and duplicate errors in the Problems view.

Improved code to honor the Tab Width / Tabulations value configured in Preferences.

Adjusted the spacing of the Cobol Editor general and Area Preference pages. 

Hovering on subsequent include statements on the same line does not display copybook contents. 

Appropriate text is now highlighted in the editor when clicking on an item in the outline view for a PL/I source.

Open declaration stopped working correctly when adding some code inside of a procedure that had a variable declared outside of the procedure. 

Fixed junit failure due to different time zone.

Opening .data files from Project Explorer was not using the correct editor.

Handle multiple files in an Open File statement in a PL/I program.

Handle PL/I expressions with negative variables (-K) similar to signed numbers in expression operand.  

Add support for the PL/I preprocessor built-in statement %REPLACE.

The Cobol Denumber hyperlink was going to the wrong location. It now points to the correct preference page.  

Search view will now refresh if the file is dirty, or the file was recently modified.

NullPointException occurs in the Compuware log when opening Cobol file from an offline project. 

PL/I preprocessor for built-in function SUBSTR now generates error messages for offsets that are out of range and returns a null string per the language spec.

Improved parser logic that determines if Cobol variables are undeclared.

In a "EXEC SQL INCLUDE ...." statement, blank lines and comments are allowed between "EXEC SQL" and "INCLUDE"

Corrected the behavior for toggling comments in JCL where lines that did not start with "//" were included.

Performing a Find/Replace function will now report the user's TSO id to zAdviser when possible. Windows id will still be provided for instances where TSO id cannot be obtained, eg: local resource. 

SVN files don't have an eclipse workspace resource associated with them. The "Find All" option in the Topaz search dialog requires such a resource, otherwise the Search view is empty. Disabled the "Find All" option when dealing with SVN files.

Cobol parsing conditions were based on variable name only, creating issues with duplicate Titles in the Outline view that went to the same code line. Improved handling of that condition and corrected the highlighting behavior. 

Cobol parser will no longer generate unresolved variable messages for records defined in the File Section.

Fixed Runtime Visualizer chart to display multiple tables that are part of the same SQL statement.

In the Outline view, nested programs will be displayed under the program that defines the program.

Error markers lost for PL/I programs in Project Explorer. The PL/I editor maintains various internal data structures, and one of them was being recreated each time it was needed rather than reusing the existing one. This was throwing off its ability to keep track of the current state. 

The PL/I block comment shortcut/keybinding now works consistently.

Search / Find All Occurrences now highlights the correct positions for COBOL when the Identity Area columns are locked.

When saving/closing a file with unmatched shfitout-shiftin combinations, give the user the option to save or not instead of refusing to save.

Improvements to colorizing of the PL/I %do skip; to %end; block(s) of code.

Beginning and ending brackets are now updated in response to changes made in the COBOL Editor.  IF loop marker positions were being incorrectly displayed.

JCL containing binary data can no longer be submitted from the editor.

Updated the PL/I parser to allow NULL as a pseudo variable/built-in procedure.

Key binding for open definition in PL/I Editor should now work using F3 on an include. 

Rexx variable list was missing in Outline view after reopening the view. 

Corrected ability to edit or remove a JDBC database connection entered with invalid data.

Removed invalid record formats that were available when creating a PDS or Library.

Updated the online CLI documentation.

The PL/I preprocessor now handles generated %include statements and prompts for copybooks to be downloaded.

Updated the PL/I outline parser to download Includes that use Allowed, and other non-standard Include statement formats.

Topaz Connect: Add Change Request Field to the Topaz Connect Endover Pakage Options.

Updated copybook download process for COBOL programs to use our new COBOL parser.

Create or Save of a new PDS member in an online project causes an error to be displayed in the Console. If the new member is in a copybook PDS, then "Open Definition" from the other source calling it doesn’t work. Added code to use createLink for online projects instead of using an internal control file/folder.

Program Analysis: when user clicks on a variable in the Data Flow view, the correct node in the Logic Flow view will be highlighted.

The "Show Compile Diagnostics" action was updated so the Problems view is always displayed after taking this action.

Topaz Enterprise Data

Fixed a problem where Compare results were not generated when a compare using the Match Record number was as the key (1 to 1, and others) or Match Records using sort field. The problem occurred when 1 file had a record length shorter than the COBOL layout specified.

Double-byte characters are preserved in Data Editor request names in Request Management file.

Adjusted the document for the Spec Generation utility to include correct parameters.

Workaround added for DB2 JDBC driver issue to correctly calculate length of GRAPHIC and VARGRAPHIC columns defined with CODEUNITS16 specifier.

Fixed a problem where a host alias was contained within another host definition.

Fixed a problem parsing a specification name that contained the META keyword.

Added error handling for migration problem that prevents the Repository Management Utility to hang.

Removed the unnecessary creation of ScrolledForm in row mode.

Fixed an issue where the repository list is not being updated on Manage Data Privacy Repositories editor when user connects/disconnects a repository.

ISPW Eclipse

Topaz load does not find the proper assignment when there are multiple assignments with same description, such as when one is a Sandbox and one is not a Sandbox assignment.

Trying to add a renamed task using the Add Task dialog box does not have any entries in the Check out from level field leading to a level not specified error.

In the Add Task dialog box, selecting the Add from history action shows a highlighted dark text for the Check out from level field making the text difficult to read.

In the VScode extension, the generate action loads a task into a normal assignment if already loaded, instead of a Sandbox assignment.

The build action error messages do not show in the Compuware log.

An inactive task in the Task view does not have the View > Parts action enabled.

When refreshing an ISPW project, you may experience poor performance if you are selecting many files.

Topaz for Total Test

Fixed the CICS CALL type descriptions.

SQL statements with continued lines have been reduced correctly.

When the Context file is stored in Topaz for Total Test, this fix forces UTF-8 as the code page (requires PTF CXSE80A).

Allow scenarios to be opened when there are no structures or interfaces present.

Corrected the error "getAttributes - rc(8) reason(008C0028)".

Corrected the misspelling of "Environments".

Corrected the error message in NVT.

Updated code to allow “Read Records”, “Read Dataset” and "Write Dataset" to use smartcard.

Fixed several error messages.

CICS recording will now generate a context file and include the transaction load library.

Unable to record sub-program if it does not contain any CICS statements and is called from a CICS main program.

End of recording pop-up might not be displayed.

Removed duplicate entry for "Total Test Scenario" from the Project Explorer> New context menu in the Total Test perspective.

Opening context file may not ensure UTF-8.  The fix is to force UTF-8 when the context file is read.

Subprogram stubs were not properly recorded in a persistence scenario.

Test cases might show failures when using RRDS input only files.

Some tests terminate before all data is recorded.

Some tests fail with recently introduced persistence changes.

Do not allow duplicate unit test scenarios in total test suite.

Engine version is updated.

Unable to record the program stub.