Release Notes - 4.7 Versions

January 3, 2023 (v.4.7.25)

• AS400 SDK - Fixed issue with NumberFormatException for code-page values such as CP037

December 16, 2022 (v.4.7.24)

Improvements

• JCL Connector - Added pre processing and post processing placeholders to the submitted JCL to allow additional actions to be performed on the mainframe side (decrypting records for example)

Fixes

• Kafka serialization/deserialization now uses the Original-name attribute as the field's labels when analyzing JSON request and responses. If Original-name is not provided, it will use the field name.

• IBM MQ API project - Fixed issue with messages occasionally returning to queue in NO-COMMIT mode when service would stop and multiple consumer queues defined.

• File connector - You can now override the configured file name from the code to something other than what was configured in the application.yml

• AS400 SDK - Fixed Carriage return serialization/deserialization.

Known Issues:

• AS400 - Getting NumberFormatException for code-page values such as CP037 in application.yml

November 9, 2022 (v.4.7.23)

Improvements

• File Connector :

  • The Swift producer now supports producing Swift header and footer blocks, in addition to the data block which contains the actual payment.

• File Consumer (API)

  • Users can now specify a retry limit and interval for moving a file to the success/failure folder. This allows separating the retry logic between cases where the invocation of the backend failed, and cases where the SFTP is not available for some reason.

  • Consuming CSV files now supports complex types with nested collections. The project can be configured to support such cases.

  • Polling was improved and now will not stop the consumer if polling failed due to connectivity issue. It will try again after a specified period of time.

  • The file pattern parameter of the File consumer was changed to retrieve files by Regular expression instead of only specific file name.

• CICS Adapter Tracing - Users can now configure their CICS SDK project to receive Trace information from the OpenLegacy adapter running on the mainframe. This will make tracing problems easier.

Fixes

• Fixed validation issue caused by spaces which were not being trimmed when running a mock with a trail file in Screens projects.

• Fixed Issue with Trimming redundant spaces from JSON field values by using the String-format configuration. Spaces now trim correctly.

• IBM-MQ Consumer (API) - can now set the Correlation ID as a byte-array to support non-alphanumeric characters

• Fixed issue with File Consumer (API) opening two threads . Now a single thread is opened as required.

• Fixed issue where Data-Queue consumer would stop because of AS/400 connection exception

August 29, 2022 (v.4.7.22)

Improvements

• UI Improvement - List fields are now displayed showing Collection Type.

• API Project Wizard has slightly changed for IBM-MQ, Kafka and AS/400 Data-queue.

  • Default error handling behavior is now selected in the project wizard

  • Users will be able to override the selected behavior on the service level and choose another. This will allow more flexibility per service.

• Support for Cobol CHAR type when parsing Cobol/copybook files.

• Mainframe-CICS , user can now log outbound/ inbound HTTP communication in debug level.

• File connector :

  • CSV Entity parser now supports a collection of primitive fields

Fixes

• Row gap property for screen tables works properly now. Users can define table items across multiple screen rows.

July 12, 2022 (v.4.7.21)

Improvements

• Additional improvements to Temenos T24 TAFJ connector performance.

July 4, 2022 (v.4.7.20)

Improvements

• Improved Temenos T24 TAFJ connector performance.

June 23, 2022 (v.4.7.19)

Improvements

• Kafka Consumer - It is now possible to override Error Handling logic using a bean.

• AS400 DataQueue - It is now possible to override Error Handling logic using a bean.

Fixes

• Temenos T24 (TAFJ) now properly handles XML tags in responses.

June 13, 2022 (v.4.7.18)

Improvements

• New feature - CORBA Endpoint
  Users can now create a CORBA API project that generates a light weight server to receive CORBA messages , map them to an associated SDK project and execute entity operations , then return the results to the CORBA client that called the server.
  The Server service can be generated from an IDL file that will define the service input and output models.

• The Temenos T24 Fetcher was enhanced to support fetching from SOLACE.

• We now expose Hikari metrics in OpenLegacy microservices, so customers can examine their connection pools

• ISO-8583

  • Added support for RAW and ASCII channels for ISO-8583 connector (SDK)

  • Added support for RAW and ASCII channels for ISO-8583 Server (API)

• Users can now optionally implement a bean to dynamically construct user names for SSO connection.

Fixes

• Fetching Oracle 19 (and above) stored procedures that contain a list of simple types as an input parameter are now parsed and executed correctly.

• COBOL Parser now properly handles fields that are defined as “USAGE IS POINTER”.

• IBM MQ API project now begins consuming messages from the queue after the application finished loading properly.

• The SFTP file connector now properly saves the file to the configured path.

• The IBM MQ connector, when set to PRODUCE & CONSUME mode, can now retrieve responses by Correlation ID properly.

May 9, 2022 (v.4.7.17)

Improvements

• NEW Connector - CORBA

  • You can now execute CORBA programs and generate entities based on CORBA IDL files.

  • When generating a model it is important to have all referenced IDL files in the same directory of the main IDL file being parsed.

• MQ Consumer (API) - Consumed message MESSAGE_ID has been added as a property in addition to the Correlation ID.

• CSV File SDK supports writing multiple lines to a file simultaneously

• CSV Files API supports reading multiple lines from the file and processing them as a collection or reading and processing one line at a time

Fixes

• Serialization / Deserialization issue of unnamed ZBAPI in SAP was fixed.

• Kafka consumer no longer creates unrequired producers.

• File Connector (CSV) no longer generates quotation marks around values by default. A flag was added to specify whether they are required. Flag value is False by default.

• File connector - date values are no longer generated as a timestamp, when a date format has been provided.

• MSSQL Stored Procedure Fetcher no longer generates fields with a negative length when encountering a varchar (max) definition in the stored procedure.

March 30, 2022 (v.4.7.16)

Improvements

• Improved Tuxedo JCA log further, now includes field names and values.

• Improved Buffer serialization (affects Mainframe and AS/400) to allow fields to return null values. This is a project level configuration and will effect all fields. However, you can also exclude certain field types from this behavior (strings for example).

March 21, 2022 (v.4.7.15)

Improvements

• Improved Tuxedo JCA debug capabilities to enable viewing Input and Output in the logs.

Fixes

• De-serializing XML to Entity no longer produces redundant objects.

• Fixed issue in Oracle (v19 and above) Stored Procedure fetcher. It now generates list of fields properly.

• AS/400 Data Queue, message key encoding issue fixed.

March 9, 2022 (v.4.7.14)

Improvements

• Added the ability to retrieve additional screen metadata from the legacy system by demand. This improves our ability to reflect screen behavior in our Angular application:

  • Enable/disable fields

  • Reflect screen colors

  • Better generation of tables

• Improved Tuxedo JCA TPESVCFAIL(11) error handling to be more informative.

• 3rd Party library upgrades :

Fixes

• Fixed generated MQ consumer Service to implement the correct consumer class.

• File Consumer Processor was fixed. Will not process the same file more than once.

• CTG connector - Fixed issue of timeout on doAction not overriding timeout set on YML.

• Fixed Tuxedo JCA Code Generation issue when dealing with a List of Characters.

February 22, 2022 (v.4.7.13)

Improvements

• It is now possible to configure the treatment of space values in fields. By default, they will be considered as nulls, however it is now possible to configure them to be treated as regular spaces.

February 20, 2022 (v.4.7.12)

Improvements

• It is now possible to configure AS/400 Data Queue consumer error handling strategy , so that upon error. the message will be:

  • ignored

  • re-processed until successful, or

  • Moved to a dead letter queue

• IBM MQ Connector - Added trace logs indicating connection status

January 24, 2022 (v.4.7.11)

Fixes

• Model is now generated correctly when request body is referenced (#ref) in Swagger specification.

January 20, 2022 (v.4.7.10)

Improvements

• It is now possible to generate a Legacy HTTP API Project. This is a lightweight server that:

  • Gets a Buffer from the legacy system (Mainframe or AS/400) via HTTP

  • Transforms the Request into JSON

  • Maps the input to any desired SDK and executes it

  • The SDK response (JSON) is then transformed into a Buffer and sent back to the legacy system
    This will allow Mainframe programs to trigger external processes such as pushing messages to Kafka.

• File connector (Producer) - Added Support for producing CSV file

  • It is possible to define:

    • File Delimiter (comma by default)

    • Customize Header/Footer - Add placeholder for field names in headers or create your own by defining the desired text

• File Consumer - Added support for consuming CSV files.

  • Select CSV as file type

  • Select the Consumed file CSV delimiter

  • Set the number of header lines in the consumed CSV file (we skip those in runtime)

  • Set the number of footer lines in the consumed CSV file (we skip those in runtime)

  • Parse an OL-Formatted-CSV file that describes the entity.

• AS/400 PCML parser can now automatically calculate the length of a field when we know the length of the containing structure and this is the only field within the structure that doesn’t have a length property defined.

• Enhanced Kafka SDK logging capabilities with some additional logs.

• 3rd Party library upgrades :

Fixes

• Log4J vulnerabilities

• MQ Consumer can now reconnect after the consumer is restarted.

• MQ consumer now closes the connection properly in the case of connection timeout.

December 30, 2021 (v.4.7.9)

Improvements

• Core was updated to protect against Log4J vulnerabilities

• REST SDKs can now be configured to support HTTP2

• Users can now instruct MQ Consumers how to handle errors. The options are:

  • Ignore message

  • Reprocess message until successful

  • Send to Dead Letter Queue

• Data Queue Consumer can now read queued messages sequentially: reading a message from the queue, processing it and only when done, removing it from the queue.

Fixes

• The Mapping tab now works correctly with Contract First projects

• Reconnection time can now be configured, even if the session pool is greater than 1

November 25, 2021 (v.4.7.8)

Improvements

• Users can now delegate the RPC Session connection to enable implementing external logic to the underlying OL connector.

• MQ Consumer :

  • Users can now configure the Reconnect Time to allow reprocessing a message a number of times before failing it, according to the error handling strategy that was selected.

  • The Correlation ID of a consumed message can now be passed to the processing logic and from there to the MQ output queue.

• MQ Connector:

  • Users can now set a timeout for getting a connection from the pool, so the message request can be aborted quickly if there are connectivity issues.

  • Users can now create MQ SDKs that will support consuming messages only. This allows REST APIs to be created that get messages from a queue.

• Tuxedo JCAs can now be executed asynchronously to improve performance.

Fixes

• SAP Serialization no longer throws NPE

• When using MQ Consumer in Sync + Transactional mode, the processMessageTimeout in the MQ consumer now behaves correctly.

October 27, 2021 (v.4.7.7)

Improvements

• Consumed or Failed files can now be written to multiple specified directories. This allows compliance with different failure handling strategies

• Data Queue SDKs now support Consume Only mode to allow a REST API to retrieve messages from a queue

• 3rd Party library upgrades

Fixes

• Service names now generate properly from Screen Trail files

• AS/400 Timeout mechanism has been adjusted to prevent errors

• MQ Consumer - Fixed an issue with setting the pool size.

October 12, 2021 (v.4.7.6)

Improvements

• Improved Sybase connector handling of fields with default values.

October 06, 2021 (v4.7.5)

New Features

• New API project type : Legacy-HTTP
  New functionality to listen for an HTTP request made by a Mainframe or AS/400 (basically sending buffers) , transform it to JSON , execute an SDK operation (with the input we got from the legacy) and return the response, if needed, to the legacy system (in the form of a buffer).

• CICS-resilience:

  • A new flag was added that can, when sending messages to the mainframe:

    • For  each request to the MF, will also return a header that specifies if the region we are interacting with is in a Closing state so we can close the socket on our side to allow the region to close.

    • In addition, we can have the adapter return a header with the information about the CICS region that answered, so we can build a list of all the CICS regions we interacted with.

  • A second flag will disable/enable CICS Resilience feature that takes a region list (either manually created or built using the aforementioned flag) and when a CICS region changes state from down to up, we can rotate our HTTP connections to balance the pool against all active CICS regions.

• The Emulation Type setting for an AS/400 SDK project can now be set using a drop down selection, and not only by configuring the YML file (See Advanced Options in the AS/400 screens SDK project wizard)

• The timeout value for a ‘Network is Unavailable’ scenario is now configurable and not hard-coded to 130 seconds

• Whether to use the ‘Depend On’ mechanism in runtime can now be set with a flag on the SDK level (can be overridden on the Entity level).
  The default will be set to TRUE, meaning that collections with DEPENDING ON definition will retrieve the number of items set in their respective counter field. When set to FALSE, the counter field will be dismissed and the number of occurrences set in the entity (OCCURS) will be returned.

Fixes

• The Add Reference Project command no longer adds dependencies in the wrong place in the POM file

• AS/400 data queue runtime now processes correctly using auto-generated code page value

August 27, 2021 (v4.7.4)

New Features

• Users can now generate services (APIs) directly from Screen entity metadata ('Trail' files).
  Right clicking a trail file in a Screens SDK project will allow the user to select the “Generate Model and Service” option. This will:

  • Identify any unidentified screen in the trail file

  • Allow the user to select an API project from the workspace

  • Automatically generate a service based on the trail file scenario.

Improvements

• 3rd Party Upgrades:

  • spring boot 2.5.3 → 2.5.4

  • spring data commons 2.5.3 → 2.5.4

  • spring security 5.5.1 → 5.5.2

  • kaml 0.34.0 → 0.35.2

  • slf4j 1.7.31 → 1.7.32

  • kotlin logging 2.0.8 → 2.0.11

  • logback 1.2.3 → 1.2.5

  • jaxb 2.3.4 → 2.3.5

  • jaxws rt 2.3.4 → 2.3.5

  • jackson 2.12.3 → 2.12.4

  • kotlin compile testing 1.4.2 → 1.4.3

  • testcontainers 1.15.3 → 1.16.0

  • mockito 3.11.2 → 3.12.1

  • hikaricp 4.0.3 → 5.0.0

  • mysql connector java 8.0.25 → 8.0.26

  • projectreactor 3.4.7 → 3.4.9

  • active jms pool 5.16.2 → 5.16.3

  • activemq artemis jms client 2.17.0 → 2.18.0

  • aws sdk v2 2.16.88 → 2.17.25

  • aws sdk v1 1.12.10 → 1.12.53

  • swagger ui 3.50.0 → 3.51.2

  • springdoc 1.5.9 → 1.5.10

  • maven 3.8.1 → 3.8.2

  • prowidesoftware SRU2020-9.1.5 → SRU2020-9.1.6

  • hazekcast 4.2 → 4.2.2

  • redisson 3.16.0 → 3.16.1

  • ktlint 0.41.0 → 0.42.1

  • artifactory gradle plugin 4.21.0 → 4.24.16

• A property was added to the consumers infrastructure , which allows a consumer to read more than the default 5 messages , according to the value that was set in the property

• A separate connection idle timeout was added to the connection pool to ensure that a connection in the pool will close before the mainframe closes the session due to inactivity. This will help us avoid cases where we are trying to use a connection whose session was already closed by the mainframe.

• Users can now configure message logging for ISO 8583 requests and responses in DEBUG log level.

• Users can now close File Consumers gracefully, and control the Close process.
  
  

Fixes

• ISO 8583 TCP/IP server processes messages correctly after redundant annotation was removed

• Username and Password fields names are correctly generated to prevent duplication errors when two password fields appear on screen with similar (but not identical) names

• Users can now choose to significantly improve XML performance by selecting to skip XML formatting.

• @RpcDateField annotation is now correctly generated for the RpcEntity after generating BAPI entity for SAP project

• Cobol Parser successfully generates a model from programs that are chain-referencing multiple copybooks.

August 09, 2021 (v4.7.3)

Improvements

• Upgrades

  • The core version was upgraded to v4.7.32

  • Spring Boot was upgraded to v2.5.3

  • Kotlin was upgraded to v1.5.21

  • Vertx was upgraded to 4.1.2

• The ISO8583 Message consumer can now receive message headers separately from the message itself.

• Users can now view ISO 8583 message requests and responses in the log by setting the log level to DEBUG in the application.yml

• Kafka consumers can now be closed gracefully.

Fixes

• Annotation processing performance for ISO 8583 has been improved

• DB2 fetcher can now return metadata for stored procedures in the SYSPROC schema

• Temenos T24

  • Temenos enquiry fetcher was enhanced to retrieve report column information.

  • Service models can now be generated with List<Date> type fields

  • Fields in Temenos SDKS that include underscores can now be correctly mapped.

  • '@transaction_info” : object has been removed from request logs

• MQ Consumer now shuts down gracefully when the API uses Ribbon

• The CICS Header TransactionOption property now propagates correctly

• Generating model from xsd now supports large XSD files

July 22, 2021 (v4.7.2)

New Features

• A new Microfocus connector has been added. This connector supports Visual COBOL and allows non-native connection, such as CICS applications that use the OL CICS Adapter for calls

  • The New SDK Project wizard now has a Microfocus backend selection option

  • RPC entity operations can now be performed against a Microfocus backend

• Users can now create new TCP SDK projects and configure common properties directly from the New SDK Project wizard. This would provide a skeleton project for a pure TCP implementation

Improvements

• Temenos Error responses are now more informative

• Users can now rotate HTTP connections to allow CICS regions to be shut down gracefully, without impacting in-flight requests

• Core upgrades:

  • Kotlin coroutines 1.5.0 → 1.5.1, Kotlin serialization 1.2.1 → 1.2.2

  • Vertx upgraded to v4.1.1

  • Spring Boot 2.5.1 -> 2.5.2, Kotlin 1.5.10 -> 1.5.20

• Connectors were aligned to v4.7

• An SDK-level flag was added to select not to trim RPC response string fields automatically

• Redis was added as a caching option

Fixes

• Temenos fixes:

  • Local Reference Fields are now supported

  • The names of List fields are now fetched correctly (without the XX prefix)

  • Temenos applications are now fetched by latest version

• ‘Copy’ statements in copybooks are now resolved properly

• The CicsHttpRequestCustomizer now propagates Action Properties correctly

• Requests with Bulk Entities are no longer sent to a hardcoded address

• Empty List Fields do not cause exceptions

• ISO-8583 Header is not duplicated in the request

• IBM-MQ consumer now now tries to reconnect to MQ when a connection is lost

July 6, 2021 (v4.7.1)

Important Note

All installations of this version require a license, whether a new installation or an upgrade from 4.6.X and below or an upgrade from v4.7.0

Improvements

• The Oracle stored procedures parser was updated to support generating entities with Output directional definitions that are correct for Oracle Stored procedures

• Users can now select whether to trim raw data in string fields or leave them. When choosing not to trim string fields , string field values will return with their additional white space

Fixes

• COBOL fields with +/- (positive/negative) signs are now generated correctly

Breaking Changes

• Upgrading to this version requires a new license file

June 15, 2021 (v4.7.0)

New Infrastructure for Upcoming Features

• We laid the groundwork to enable users to run SDK projects in Record or Replay mode with a mock backend. This will enable tests to be executed against a recorded session , thus lowering the hits to the backend system.

• DB connectors such as ORACLE, MSSQL and Sybase can now execute entities based on prepared statements , rather than only stored procedures. We’ll be publishing a how-to guide very soon.

Improvements

• Vertx was upgraded to v4.1.0

• You can now set a base path for AS/400 programs in the application.yml , in runtime it will be concatenated to the RPC entity’s execution-path , so as to easily switch between environments

• Cache information is now collected from the annotations on the entity into a cache definition so it will be available in runtime

• File connector SDK projects can now save files to an SFTP file location, so produced files can be securely placed in a remote location

• MQ Connector

  • The MQ connector now calculates the output data length dynamically when using the mqcih header

  • When receiving a response from the MQ, the mqcih header will now be removed automatically, to avoid manual work on the entity

  • When receiving a response from the MQ, the mqiih header will now be removed automatically, to avoid manual work on the entity

Fixes

• REST SDK now parses response JSONs correctly if the error models have the same structure as the response model

• The CICS connector now passes HDR properties correctly

• H3270 Terminal Screen projects now handle Double Bytes correctly

Breaking Changes

• Due to new infrastructure changes , POM dependencies will need to be changed in SDK/API projects.
  Follow this guide to upgrade your projects: https://openlegacy.atlassian.net/wiki/spaces/OV45/pages/1985937413