Between v3 and v4 in node-redis the client options changed for a collection of socket related
paramaters. In the caching documentation the v3 format for used for `host` and `socket`. When
following the documented example the redis client would default to `localhost` even when a
different value for `host` was specified.
Closes: #9987
For previous examples typescript error occurs in createQueryBuilder that says “Argument of type 'string' is not assignable to parameter of type 'QueryRunner'.ts(2345)”.
If you check the type of createQueryBuilder, the string cannot come as the first argument.
The types of createQueryBuilder are as follows.
myDataSource.createQueryBuilder<ObjectLiteral>(entityClass: EntityTarget<ObjectLiteral>, alias: string, queryRunner?: QueryRunner | undefined):
* added transaction retry logic in cockroachdb
* added option to control max transaction retries;
added delay before transaction retry;
updated docs;
* fixes in retry logic
* enable storing queries after retrying transaction
* fix: add support for mongodb driver v5
This new fix allow support to mongodb driver v5
Closes: #7907
* refactor: remove callback from MongoDriver connect
* fix: check for propertyName in transform
* fix: add support for mongodb driver v5
This new fix allow support to mongodb driver v5
Closes: #7907
* feat: mongodb 5.2.0 typings
* fix: instanceof ObjectId check
Instanceof check now no longer references to just the
type but the loaded class ref instead.
* test: fix test name to mongodb v5
---------
Co-authored-by: Matheus Melo Antiquera <matheusantiquera@finchsolucoes.com.br>
Co-authored-by: tgrassl <grassl.timon@gmail.com>
Co-authored-by: Matheus Melo Antiquera <matheus.melo.a@hotmail.com>
* Added a section on debugging query builder
It covers using getQuery() and getQueryAndParameters() which are both useful when trying to figure out why your query isn't doing what you might expect it should do.
* forgot await
* fix typo
* documented getCount()
getCount() is currently only mentioned in the documentation once, and you're never shown how to use it. This patch adds a little documentation for getCount()
* clarification
* feat: Add support for SQL aggregate functions SUM, AVG, MIN, and MAX to the Repository API
* rename field name to make tests work in oracle
* fix the comments
* update the docs
* escape column name
* address PR comment
* format the code
* feature: adds support for enum type (fixes#9068)
* temporarily ran package to test on different repo
* playing around - bumped version
* Revert "playing around - bumped version"
This reverts commit 7df4adb3e698419c174c2daee88614f8dafdbb6c.
* Revert "temporarily ran package to test on different repo"
This reverts commit 48f394e8eb32c22fe757010b446c85740bf80b5f.
* feat: add support for geometry data type
* feature: properly hydrate enum array values
* feature: adds support for geography and geometry for cockroachdb
* bugfix: fixes issue with primary generated columns being invalid column type (fixes#8532)
* Revert "bugfix: fixes issue with primary generated columns being invalid column type (fixes#8532)"
This reverts commit e00cdb090638d34668e3e10acd5f8267fe3bd028.
* bugfix: type casts to string when using ANY
* feature: cast geometry/geography to geojson
* feature: added references to srid
* bugfix: prevent error if trying to close already closed connection
* feature: added cockrachodb as part of postgres family
* feature: ensures support for spatial columns for cockroachdb
* feature: adds support for UPSERT for CockroachDB (fixes#9199)
* minor: added TODO; unsure how to achieve this
* feature: adds support for time travelling queries for cockroachdb
* bugfix: only run time travel query on SELECT statements
* refactor: changed UsertType from 'upsert' to 'primary-key' since this is more logical
* feature: added posibility to set timeTravelQuery to false, instead of the parameter function; help for disabling time travel queries during tests
* feature: allow timeTravelQueries in find* queries
* bugfix: when using timetravel queries with joinAttributes it now prevents error 'AS OF SYSTEM TIME must be in top level' error
* lint
* minor fix
* fixed failing test
* implemented ENUM type;
added tests;
* fixed failing tests
* fixed failing test
* fixed spatial types synchronization;
implemented spatial indices;
added tests for spatial columns;
* refactored Time Travel Query functionality;
removed TTQ from find options;
added tests for TTQ;
* added docs for Time Travel Queries
* minor changes
* added GeoJSON types;
other minor fixes;
* updated docs
* updated docs
Co-authored-by: Dmitry Zotov <dmzt08@gmail.com>
* feat: allow for partial index conditions for on conflict statments in postgres
* fix: change variable to match postgres documentation and convert from array to string type
* fix: generalize use of index predicate. add documentation and expand test cases
* lint fix
Co-authored-by: Umed Khudoiberdiev <pleerock.me@gmail.com>
Co-authored-by: Alex Messer <dmzt08@gmail.com>
* fix: disable transaction support for cordova driver
the cordova-sqlite-storage plugin does not support transactions. fc4133c
introduced an exception when starting a transactions but did not disable
the transactionSupport in general leading to errors when using the
`save` and `remove` method of the `EntityPersistExecutor`. With this PR
the `EntityPersistExecutor` will respect the drivers `transactionSupport`
option.
* docs: add note on transaction limitations for cordova driver
* Improve comprehension of update repository API
Because the previous doc example was `await repository.update({ firstName: "Timber" }, { firstName: "Rizzrak" })`, I got confused and reversed the **WHERE** condition and the actual **change**.
This change is just a suggestion to avoid that kind of confusion
* Apply the same doc suggestion change to the entity manager update API
* feat: Added new indices attribute to View
* feat: Added view indices methods
Such as dropViewIndex, addViewIndices, addViewIndex
Added "View" type in some parameters of methods
* feat: Added view indices support when creating new indices and dropping old indices
* ref: Renamed "table" to "view" in log when dropping view index
* feat: changed order of schema sync operations
To create a new view index, a view has to be created first.
* feat: removed unreachable code
A view object don't have its indices when creation. The indices are added to the view through the createViewIndex method.
* feat: Added view when returning TableIndex
* feat: Added view paths as argument in getViews on log method
* feat: Created createViewIndexSql
This method reuses code from createIndexSql, but eliminates the isSpatial part, because a viewColumn doesn't support this attribute.
* fix: Added missing columns const to createViewIndexSql
* feat: Removed isSpatial attribute when returning TableIndex
* feat: Added unit tests
* fix: Dropped current index to leave unique index on indices array
There was a bug that when asserting the unique index, it would compare with the previous index, even when explicitly selecting the unique index in the indices array.
* ref: lint files
* feat: added "postgres" in enabledDrivers attribute
This is to enable only PostgreSQL for the tests
* feat: added doc for materialized view indices
* ref: lint files
* feat: Added new method to create mat. view indices
This new method goes after creating the views. Aditionally, the views are now created at the end (as it was before)
* ref: prettify files
* feat: revamped tests
Replaced previous unit tests with more significant ones
