lacanoid/pgddl

Release for postgres 16 New `lite` option to move defaults and constraints into `create table` section.

Objects from `pg_amop` and `pg_amproc` now handled by `create_operator_class()`. New option `nodrop` to omit drop statements alltogether. Bug fixes with respect to partitions, where too many constraint/indexes were dumped.

Extensions are now supported. New options: `nodcl`, `noowner`, `nogrants`, `noalter`. Bug fixes.

New option `ie` will add `IF EXISTS` to a bunch of places. New option `ine` will add `IF NOT EXISTS` to a bunch of places. New option `drop` will add drop statements in the beginning of the script. One can pass options as text array to `ddlx_create` and `ddlx_script` and other functions. Like: ``` mydb=# select ddlx_create('my_table'::regclass , '{ine}') ``` Now one can get individual parts of object definition. Try: ``` mydb=# select * from ddlx_definitions('my_table'::regclass) ``` Some internal factorization to make this more sensible.

Tests on Postgres 14 no longer fail. ddlx_apropos() function now uses POSIX regular expressions.

A bug in generated create event trigger was fixed. Objects from extensions are now excluded from script unless 'ext' option is specified.

Improved compatibility with PostgreSQL 13. All tests pass now.

Support for GENERATED columns

This version has been tested to work with PostgreSQL 12 and test have been adjusted to pass. Publications and subscriptions are now supported. New ddlx_apropos() function lets you conveniently search for stuff. Bugs have been fixed.

This version improves compatibility with PostgreSQL version 12

Another release, which brings a bunch of improvements and bugfixes. See CHANGELOG for more details.

Another spring release! This version adds support for a number of features, including partitions, policies (row level security), clustering, disabled triggers, statistics and operator classes.

New fall release! No naming changes this time :) The extension now installs and runs on different PostgreSQL versions, tested from version 9.1 on. This version adds support for many new objects, including databases, languages, transforms, rules and tablespaces. Few bugs have been fixed. Regression tests were extended and improved.

New spring release! This time, the extension name itself has been changed to `ddlx`. Also, the API has been changed to use just `ddlx_` for function prefixes. Hopefully this is final and makes things more consistent and shorter to type. Now you can just `select ddlx_script('tablename')`! Support has been added for a bunch of objects. There is now support for all `reg*` types, including text search. Also for more SQL standard stuff, like foreign data wrappers, servers, user mappings, casts, collations and conversions. Bugs were fixed. Regression tests were extended and improved.

API has been renamed to use `pg_ddlx` prefix, as in DDL eXtractor! This keeps the API neatly in one place in functions browser. Public API has been extended with `pg_ddlx_create(oid)` and `pg_ddlx_drop(oid)` functions. Added support for operators. Initial support for dependancies.

New release supports extracting DDL for roles (users and groups). Also, range types are supported, as are collations on domains. This further enhances support for types, making it more complete.

Since this is apparently being used by some people, it might make sense to make a release, so that I can see how this works. Also that the work can continue on the next version. Some of the stuff on the roadmap has been implemented recently and stupid bugs fixed. Enjoy, and please report all issues in the issue tracker. Thanks!