3.17.0 Launch with Computed Columns, Audit Columns, Sample Matching, Reactive Transactions and Kotlin Coroutine Help – Java, SQL and jOOQ.

This launch contiues the work from earlier releases round extra refined SQL transformation capabilities, together with:

• Consumer facet computed columns for each learn and write operations
• Audit columns
• Sample matching SQL transformations
• Extra implicit JOIN capabilities

Consumer facet computed columns

A floor breaking new core function out there in all business distributions is
the brand new shopper facet computed columns function, constructing on prime of jOOQ 3.16’s
business help for readonly columns and server facet computed columns.

Not all RDBMS help computed columns (e.g. utilizing the usual SQL syntax
GENERATED ALWAYS AS), and in the event that they do, they may not help them in each STORED (computed on write) and VIRTUAL (computed on learn) variants. jOOQ can now emulate each options on the shopper facet, by remodeling your SQL queries:

• STORED impacts INSERT, UPDATE, DELETE, and MERGE
• VIRTUAL impacts SELECT and the RETURNING clause of DML statements. To make use of those, mix them with the brand new artificial column technology function.

Not like their server facet counterparts, these shopper facet options can produce arbitrary expressions, together with:

• Implicit joins
• Scalar subqueries
• MULTISET subqueries
• Way more

Consider this as “views” written in jOOQ, on a per-column foundation. An expecially helpful function mixture is to mix these computed columns with the brand new visibility modifier that permits for conserving computed columns (or the underlying base columns) non-public and thus invisible to consumer code.

Extra about this function right here

Audit columns

A particular case of STORED shopper facet computed columns are audit columns, whose most simple implementation comes within the type of:

• CREATED_AT
• CREATED_BY
• MODIFIED_AT
• MODIFIED_BY

Different approaches to auditing exist, together with tender deletion, further meta information, (bi)temporal versioning, however these columns are among the many hottest approaches, making this business solely comfort function very helpful to loads of clients.

Extra about this function right here

Java 17 baseline for the jOOQ Open Supply Version

Java 17 has been the most recent LTS, and it consists of loads of actually cool options, together with:

• sealed varieties (important for sample matching)
• data
• instanceof sample matching
• textual content blocks
• swap expressions

jOOQ 3.16’s experimental new Question Object Mannequin (QOM) API experiments with sealed varieties, which shall be adopted extra typically as soon as the QOM API is finalized.

To get broader consumer suggestions on these enhancements, in addition to to embrace Java’s new LTS replace cadence, we’ve determined to make Java 17 the baseline for the jOOQ 3.17 Open Supply Version, persevering with our Java 8 and 11 help within the business jOOQ distributions.

The next older jOOQ releases will proceed to obtain upgrades for some time:

• jOOQ 3.14: The final launch with Java 8 help within the jOOQ Open Supply
  Version and Java 6 help within the jOOQ Enterprise Version
• jOOQ 3.15 and three.16: The final releases with Java 11 help within the jOOQ Open
  Supply Version and Java 8 help within the business editions.

PostgreSQL information sort help

The jooq-postgres-extensions module, which contained help for the HSTORE sort, now has much more help for PostgreSQL particular information varieties, together with array kinds of every of:

• CIDR
• CITEXT
• LTREE
• HSTORE
• INET
• RANGE (together with all of the specialisations for INT4, INT8, and so on.)

With a purpose to revenue from these information varieties, simply add the org.jooq:jooq-postgres-extensions module to your code technology and runtime dependencies, and the categories are generated routinely.

Implicit JOIN enhancements

On this launch, we experimented with just a few new implicit JOIN options, together with help for implicit JOIN in DML statements. The present implementation produces correlated subqueries the place JOIN isn’t supported in DML statements.

We’ve additionally experimented with making a “comfort syntax” for different generally used correlated subqueries, equivalent to EXISTS(...) subqueries or MULTISET(...) subqueries. The experiment has been very fascinating. The prototype, nonetheless, was rejected. See the discussions right here:

Future jOOQ variations will implement the specified comfort within the type of extra implicit JOIN performance, providing the function additionally as an implicit to-many JOIN.

A leftover from the prototype is the truth that now you can extra simply challenge expressions apart from basic Subject<T> in your SELECT clause, specifically:

• Desk<R> now extends SelectField<R>
• Situation now extends Subject<Boolean>

This implies you may write a question like this:

Consequence<Record3<CustomerRecord, AddressRecord, Boolean>> end result =
ctx.choose(

      // Mission a CustomerRecord immediately
      CUSTOMER,

      // Mission an AddressRecord from an implicit JOIN
      CUSTOMER.handle(),

      // Mission a boolean expression, as a substitute of wrapping it with discipline()
      exists(
        selectOne()
        .from(PAYMENT)
        .the place(PAYMENT.CUSTOMER_ID.eq(CUSTOMER.CUSTOMER_ID))
      )
   .from(CUSTOMER)
   .fetch();

Sample matching SQL Transformations

SQL transformations have been a strategic function set to latest jOOQ releases, providing further compatibility between SQL dialects to business clients, equivalent to, for instance:

• Reworking Oracle’s ROWNUM into equal window features or LIMIT clauses.
• Turning desk lists together with Oracle’s (+) operator into ANSI JOIN syntax.

This launch ships with a brand new business solely function that immediately transforms the brand new Question Object Mannequin (QOM)’s expression tree previous to rendering. It does so by making use of sample matching to the expression tree. Some assorted examples embody:

• LTRIM(RTRIM(x)) into TRIM(x)
• x != a AND x != b into x NOT IN (a, b)
• x IN (a, b, c) AND x IN (b, c, d) into x IN (b, c)
• NOT (NOT (x = 1)) into x = 1
• NOT (x = 1) into x != 1

And way more. The first use-cases for this performance are:

• SQL linting, e.g. as a part of an ExecuteListener
• SQL auto cleanup, together with in a ParsingConnection
• Dialect migration, when upgrading database variations, or shifting between dialects
• Patching particular SQL options

For extra details about the function, see right here

Word that this function can also be out there without spending a dime on-line at https://www.jooq.org/translate

Reactive and kotlin coroutine help

A number of minor enhancements have been applied. A couple of extra vital ones
embody:

• R2DBC 0.9.1.RELEASE is now supported
• A brand new reactive transaction API has been added, which gives the identical nested
  transaction semantics as the present blocking transaction API, see additionally:
  https://weblog.jooq.org/nested-transactions-in-jooq/
• jOOQ’s reactive streams bindings through the Writer SPI are actually
  bridged routinely to kotlin coroutines within the new
  org.jooq:jooq-kotlin-coroutines module utilizing the standard utilites
  org.jetbrains.kotlinx:kotlinx-coroutines-core and
  org.jetbrains.kotlinx:kotlinx-coroutines-reactor
• The org.jooq:jooq-kotlin extensions module now has further
  extension features for extra MULTISET and different nesting associated
  comfort.
• The whole blocking execution API is now annotated with
  org.jetbrains.annotations.Blocking to assist reactive jOOQ customers
  keep away from by accident blocking on a question, when utilizing IntelliJ. As well as, we
  now annotate experimental and inside API with the ApiStatus
  annotation from the identical package deal.

Full launch notes right here