Log message:
databases/ruby-sequel: update to 5.93.0

5.93.0 (2025-06-01)

* Add support to Dataset#explain for all PostgreSQL EXPLAIN options
  (jeremyevans)

* Make Dataset#{explain,analyze} work with the pg_auto_parameterize
  extension (jeremyevans)

* Support multilevel qualified identifiers in Database#create_table
  (jeremyevans)

* Add Sequel::OLD and Sequel::NEW constants, for use in RETURNING statements
  on PostgreSQL 18+ (jeremyevans)

* Fix SQL::QualifiedIdentifier handling of SQL::Identifier wrapping
  LiteralString values (jeremyevans)

* Fix literalization of SQL::Identifier wrapping LiteralString values
  (jeremyevans)

* Add Dataset#prepare_sql_type for setting the SQL type to use for
  prepare/call (jeremyevans)

Log message:
databases/ruby-sequel: update to 5.92.0

5.91.0 (2025-04-01)

* Recognize datetime(N) types as :datetime again (broke in 5.32)
  (jeremyevans) (#2293)

* Fix possible concurrency issue in eager loading due to Method#hash
  iterating over a hash that can be inserted into concurrently (jeremyevans)
  (#2292)

* Add pg_auto_validate_enums plugin for automatically validating enum values
  on PostgreSQL (jeremyevans)

* Make pg_auto_constraint_validations cache file creation deterministic by
  sorting subhashes (jeremyevans)

* Support treat_string_list_as_untyped_array Database option in
  pg_auto_parameterize_in_array extension (jeremyevans)

* Support allow_by_default option in forbid_lazy_load extension (brandur)
  (#2288)

5.92.0 (2025-05-01)

* Use a qualified primary key for Model.paged_each and for the
  paged_operations plugin when using joined datasets (jeremyevans, davekaro)
  (#2299, #2300)

* Add Model.qualified_primary_key for returning a qualified version of the
  model's primary key (jeremyevans)

* Support sql_literal_allow_caching? method on objects that respond to
  sql_literal_append/sql_literal (jeremyevans)

* Support set_column_not_null as a reversible migration method (jeremyevans)

Log message:
databases/ruby-sequel: update to 5.90.0

5.90.0 (2025-03-01)

* Avoid one query per enum when loading pg_enum extension when pg_array
  extension is already loaded (nick96) (#2278, #2279)

* Automatically load the sql_comments Database extension when loading the
  sql_comments model plugin (jeremyevans) (#2276)

* Support :pg_auto_parameterize_min_array_size Database option in
  pg_auto_parameterize_in_array extension to control minimum array size to
  handle (jeremyevans)

* Add pg_eager_any_typed_array plugin, automatically transform eager loads
  to use = ANY(array_expr::type[]) instead of IN (value_list) (jeremyevans)

* Support :eager_loading_predicate_transform association option
  (jeremyevans)

* Support explicitly specifying :use_placeholder_loader association option
  (jeremyevans)

* Make SQL::BooleanExpression.from_value_pairs handle
  Dataset::PlaceholderLiteralizer::Argument with existing transformer
  correctly (jeremyevans)

* Make connection_validator extension not swallow exceptions raised by
  Database#valid_connection? (jeremyevans)

* Do not call Model.name while holding a mutex when setting temporary names
  for anonymous modules (jeremyevans) (#2273)

Log message:
databases/ruby-sequel: update to 5.89.0

5.89.0 (2025-02-01)

* Make connection_validator extension handle case where
  Database#valid_connection? raises an exception (jeremyevans)

* Support :compare_connections_by_identity Database option to work around
  bugs in other libraries or the ruby implementation (jeremyevans)

* Make mysql2 adapter handle invalid statement handles when closing prepared
  statements (jeremyevans)

* Add query_blocker extension, for blocking queries inside a block
  (jeremyevans)

* Support alter_table add_primary_key/add_unique_constraint :using_index
  option on PostgreSQL 9.1+ (jeremyevans)

* Add temporary names for internally defined anonymous classes and modules
  on Ruby 3.3+ (jeremyevans)

Log message:
databases/ruby-sequel: update to 5.88.0

5.88.0 (2025-01-01)

* Add subset_static_cache plugin for statically caching subsets of a model
  class (jeremyevans)

* Allow class-level dataset methods to be overridable and call super to get
  the default behavior (jeremyevans)

* Support column aliases with data types on PostgreSQL, useful for selecting
  from functions returning records (jeremyevans)

* Break ties in timestamp migrator version handling using lexicographic sort
  of rest of migration filename (jeremyevans)

* Fix strict_unused_block warnings when running specs on Ruby 3.4
  (jeremyevans)

Log message:
databases/ruby-sequel: update to 5.87.0

5.87.0 (2024-12-01)

* Add pg_schema_caching extension, for reloading OIDs for custom types when
  loading cached schema (jeremyevans)

* Make Database#schema hashes include :comment field on MySQL and PostgreSQL
  (Bahanix) (#2248, #2249)

* Add inspect_pk plugin to make it easier to retrieve model instance based
  on inspect output (jeremyevans)

* Treat all Trilogy errors with error code 1205 as
  Sequel::DatabaseLockTimeout in the trilogy adapter (jeremyevans)

* Allow Data{base,set}#extension to not require files if the extension is
  already registered (jeremyevans) (#2246)

* Do not limit imports to 40 rows at a time if using pg_auto_parameterize
  extension with the no_auto_parameterize method (davekaro) (#2242)

Log message:
databases/ruby-sequel: update to 5.86.0

5.86.0 (2024-11-01)

* Support the :disable_dqs Database option in the sqlite adapter to disable
  treating double quoted values as strings (jeremyevans) (#2233)

* Have the subset_conditions plugin support where_all and where_any methods
  to combine existing subsets with AND or OR (jeremyevans)

* Have the subset_conditions plugin add *_conditions methods for exclude
  method calls in addition to where and subset (jeremyevans)

* Use Ruby module naming instead of Java package naming for access to Java
  libraries in the jdbc adapters (kalenp) (#2235, #2236)

* Make schema_dumper extension format options similar to Hash#inspect on
  Ruby 3.4+ (jeremyevans)

* Avoid deprecation warnings on Ruby 3.4.0-preview2 (jeremyevans)

* Handle FROM tables that are SQL::DelayedEvaluation instances when trying
  to get primary key values after INSERT on PostgreSQL (tomasmiguez) (#2230,
  #2232)

Log message:
databases/ruby-sequel: update to 5.85.0

5.85.0 (2025-10-01)

* Support json_table on PostgreSQL 17+ in the pg_json_ops extension
  (jeremyevans)

* Make Dataset#get and #first without argument not create intermediate
  datasets if receiver uses raw SQL (jeremyevans)

* Add dataset_run extension, for building SQL using datasets, and running
  with Database#run (jeremyevans)

* Switch default connection pool to timed_queue on Ruby 3.2+ (jeremyevans)