slick.jdbc

Specialized JDBC ResultConverter for non-Option values.

Set parameters on a new Connection. This is used by DataSourceJdbcDataSource.

Slick profile for IBM DB2 UDB.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.relational.RelationalCapabilities.reverse: This String function is not available in DB2.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if generated keys should be returned. Otherwise the operation is performed natively on the server side.

• slick.jdbc.JdbcCapabilities.booleanMetaData: DB2 doesn't have booleans, so Slick maps to SMALLINT instead. Other JDBC drivers like MySQL map TINYINT(1) back to a Scala Boolean. DB2 maps SMALLINT to an Integer and that's how it shows up in the JDBC meta data, thus the original type is lost.

• slick.jdbc.JdbcCapabilities.supportsByte: DB2 does not have a BYTE type.

Note: The DB2 JDBC driver has problems with quoted identifiers. Columns which are returned from inserts must not require quoted names (in particular, they must not contain lower-case characters or be equal to a reserved word), otherwise a bug in the DB2 JDBC driver triggers a SQL Error -206 (SQLState 42703).

A JdbcDataSource for a DataSource

A DataSource that wraps the DriverManager API. It can be configured as a Java Bean and used both stand-alone and as a source for a connection pool. This implementation is design specifically to handle a non-JDBC Database URL in the format defined by the libpq standard.

Specialized JDBC ResultConverter for handling non-Option values with a default. A (possibly specialized) function for the default value is used to translate SQL NULL values.

Slick profile for Derby/JavaDB.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.relational.RelationalCapabilities.functionDatabase: Functions.database is not available in Derby. Slick will return an empty string instead.

• slick.relational.RelationalCapabilities.replace, slick.relational.RelationalCapabilities.reverse: These String functions are not available in Derby.

• slick.relational.RelationalCapabilities.pagingNested: See DERBY-5911.

• slick.jdbc.JdbcCapabilities.returnInsertOther: When returning columns from an INSERT operation, only a single column may be specified which must be the table's AutoInc column.

• slick.sql.SqlCapabilities.sequenceCurr: Sequence.curr to get the current value of a sequence is not supported by Derby. Trying to generate SQL code which uses this feature throws a SlickException.

• slick.sql.SqlCapabilities.sequenceCycle: Sequence cycling is supported but does not conform to SQL:2008 semantics. Derby cycles back to the START value instead of MINVALUE or MAXVALUE.

• slick.relational.RelationalCapabilities.zip: Ordered sub-queries and window functions with orderings are currently not supported by Derby. These are required by zip and zipWithIndex. Trying to generate SQL code which uses this feature causes the DB to throw an exception. We do not prevent these queries from being generated because we expect future Derby versions to support them with the standard SQL:2003 syntax (see <a href="http://wiki.apache.org/db-derby/OLAPRowNumber" target="_parent"

  http://wiki.apache.org/db-derby/OLAPRowNumber).

• slick.relational.RelationalCapabilities.joinFull: Full outer joins are emulated because there is not native support for them.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side because there is no native support for them (but there is work in progress: see DERBY-3155).

• slick.jdbc.JdbcCapabilities.booleanMetaData: Derby <= 10.6 doesn't have booleans, so Slick maps to SMALLINT instead. Other jdbc drivers like MySQL map TINYINT(1) back to a Scala Boolean. Derby maps SMALLINT to an Integer and that's how it shows up in the jdbc meta data, thus the original type is lost.

• slick.jdbc.JdbcCapabilities.supportsByte: Derby doesn't have a corresponding type for Byte. SMALLINT is used instead and mapped to Short in the Slick model.

• slick.relational.RelationalCapabilities.repeat: There's not builtin string function repeat in Derby. <a href="https://db.apache.org/derby/docs/10.10/ref/rrefsqlj29026.html" target="_parent"

  https://db.apache.org/derby/docs/10.10/ref/rrefsqlj29026.html

• slick.jdbc.JdbcCapabilities.returnMultipleInsertKey: Derby's driver is unable to properly return the generated key for insert statements with multiple values. (DERBY-3609)

A DataSource that wraps the DriverManager API. It can be configured as a Java Bean and used both stand-alone and as a source for a connection pool.

Basic conversions for extracting values from PositionedResults.

GetResult for tuple types.

Slick profile for H2.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.relational.RelationalCapabilities.reverse: This String function is not available in H2.

• slick.sql.SqlCapabilities.sequenceMin, slick.sql.SqlCapabilities.sequenceMax, slick.sql.SqlCapabilities.sequenceCycle: H2 does not support MINVALUE, MAXVALUE and CYCLE

• slick.jdbc.JdbcCapabilities.returnInsertOther: When returning columns from an INSERT operation, only a single column may be specified which must be the table's AutoInc column.

• slick.relational.RelationalCapabilities.joinFull: Full outer joins are emulated because there is not native support for them.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if the data to insert contains an AutoInc fields. Otherwise the operation is performed natively on the server side.

Slick profile for HyperSQL (starting with version 2.0).

This profile implements the slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.sql.SqlCapabilities.sequenceCurr: Sequence.curr to get the current value of a sequence is not supported by Hsqldb. Trying to generate SQL code which uses this feature throws a SlickException.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if generated keys should be returned. Otherwise the operation is performed natively on the server side.

Base trait for all statement invokers of result element type R.

Specialized JDBC ResultConverter for handling isDefined checks for Option values.

A JDBC-based database back-end that is used by slick.jdbc.JdbcProfile.

Capabilities for slick.jdbc.JdbcProfile.

A JdbcDataSource provides a way to create a Connection object for a database. It is similar to a javax.sql.DataSource but simpler. Unlike JdbcBackend.JdbcDatabaseDef it is not a part of the backend cake. This trait defines the SPI for 3rd-party connection pool support.

Create a JdbcDataSource from a Config object and an optional JDBC Driver. This is used with the "connectionPool" configuration option in JdbcBackend.DatabaseFactoryDef.forConfig.

JDBC profile component which contains the mapping compiler and insert compiler

Build a Slick model from introspecting the JDBC metadata.

In most cases you are better off transforming the generated model instead of overriding functionality here. It is only useful if you need easy access to the JDBC metadata in order to influence how the model is generated. A good use case would be interpreting column types or default values that Slick doesn't understand out of the box. If you just want to remove or hard code some default values, transform the resulting model instead.

The tight coupling can easily lead to source code incompatibilities in future versions. Avoid hooking in here if you don't have to.

Abstract profile for accessing SQL databases via JDBC.

A JdbcType object represents a Scala type that can be used as a column type in the database. Implicit JdbcTypes for the standard types are provided by the profile.

A wrapper for java.sql.PreparedStatement that logs statements, parameters and benchmark results to the appropriate JdbcBackend loggers.

A wrapper for java.sql.Statement that logs statements and benchmark results to the appropriate JdbcBackend loggers.

Slick profile for MySQL.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.jdbc.JdbcCapabilities.returnInsertOther: When returning columns from an INSERT operation, only a single column may be specified which must be the table's AutoInc column.

• slick.sql.SqlCapabilities.sequenceLimited: Non-cyclic sequence may not have an upper limit.

• slick.relational.RelationalCapabilities.joinFull: Full outer joins are emulated because there is not native support for them.

• slick.jdbc.JdbcCapabilities.nullableNoDefault: Nullable columns always have NULL as a default according to the SQL standard. Consequently MySQL treats no specifying a default value just as specifying NULL and reports NULL as the default value. Some other dbms treat queries with no default as NULL default, but distinguish NULL from no default value in the meta data.

Sequences are supported through an emulation which requires the schema to be created by Slick. You can also use an existing schema with your own sequence emulation if you provide for each sequence ''s'' a pair of functions s_nextval and s_currval.

The default type for strings of unlimited length is "TEXT", falling back to "VARCHAR(254)" if a Default or PrimaryKey column option is set. This can be changed by overriding slick.jdbc.MySQLProfile.defaultStringType in application.conf.

Note: Slick 3.2 also checks the old config path "slick.driver.MySQL" that was used by Slick 3.0 and 3.1, and logs a warning if anything is found there. Values from the old path are not used anymore. This deprecation warning will be removed in a future version.

Specialized JDBC ResultConverter for handling values of type Option[T]. Boxing is avoided when the result is None.

Slick profile for Oracle.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.relational.RelationalCapabilities.foreignKeyActions: Foreign key actions ''Cascade'', ''SetNull'' and ''NoAction'' are directly supported for ''onDelete''. ''Restrict'' and ''SetDefault'' are ignored (i.e. equals to ''NoAction''). No ''onUpdate'' actions are supported but specifying ''Cascade'' adds the option INITIALLY DEFERRED to the foreign key constraint, thus allowing you to perform the cascading update manually before committing the current transaction. Other ''onUpdate'' actions are ignored.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if generated keys should be returned. Otherwise the operation is performed natively on the server side.

• slick.jdbc.JdbcCapabilities.booleanMetaData: Oracle doesn't have booleans, so Slick maps to CHAR instead and that's how it appears in model and generated code.

• slick.jdbc.JdbcCapabilities.distinguishesIntTypes: Oracle doesn't distinguish integer types and Slick uses NUMBER, which is always mapped back to BigDecimal in model and generated code.

• slick.jdbc.JdbcCapabilities.supportsByte: Oracle does not have a BYTE type.

• slick.jdbc.JdbcCapabilities.returnMultipleInsertKey: Oracle returns the last generated key only.

• slick.jdbc.JdbcCapabilities.insertMultipleRowsWithSingleStatement: Oracle doesn't support this feature directly. There are several alternative ways, but the library doesn't support them, so far.

Note: The Oracle JDBC driver has problems with quoted identifiers. Columns which are returned from inserts must not require quoted names (in particular, they must not contain lower-case characters or be equal to a reserved word), otherwise a bug in the Oracle JDBC driver triggers an ORA-00904 error. The same issue arises when trying to update such a column in a mutable result set.

Updating Blob values in updatable result sets is not supported.

A wrapper for a JDBC PreparedStatement which allows inceremental setting of parameters without having to sepcify the column index each time.

A database result positioned at a row and column.

Slick profile for PostgreSQL.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the server side with a single JDBC statement executing multiple server-side statements in a transaction. This is faster than a client-side emulation but may still fail due to concurrent updates. InsertOrUpdate operations with returning are emulated on the client side.

• slick.jdbc.JdbcCapabilities.nullableNoDefault: Nullable columns always have NULL as a default according to the SQL standard. Consequently Postgres treats no specifying a default value just as specifying NULL and reports NULL as the default value. Some other dbms treat queries with no default as NULL default, but distinguish NULL from no default value in the meta data.

• slick.jdbc.JdbcCapabilities.supportsByte: Postgres doesn't have a corresponding type for Byte. SMALLINT is used instead and mapped to Short in the Slick model.

Notes:

• slick.relational.RelationalCapabilities.typeBlob: The default implementation of the Blob type uses the database type lo and the stored procedure lo_manage, both of which are provided by the "lo" extension in PostgreSQL.

Ensure that every expression in a GroupBy's "by" clause contains a reference to a proper source field. If this is not the case, wrap the source in a Subquery boundary.

Represents a result set concurrency mode.

Represents a result set holdability mode .

An invoker which calls a function to retrieve a ResultSet. This can be used for reading information from a java.sql.DatabaseMetaData object which has many methods that return ResultSets.

For convenience, if the function returns null, this is treated like an empty ResultSet.

Represents a result set type.

Slick profile for Microsoft SQL Server.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.jdbc.JdbcCapabilities.returnInsertOther: When returning columns from an INSERT operation, only a single column may be specified which must be the table's AutoInc column.

• slick.sql.SqlCapabilities.sequence: Sequences are not supported because SQLServer does not have this feature.

• slick.jdbc.JdbcCapabilities.forceInsert: Inserting explicit values into AutoInc columns with ''forceInsert'' operations is not supported.

• slick.jdbc.JdbcCapabilities.createModel: Reading the database schema is not supported.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if generated keys should be returned. Otherwise the operation is performed natively on the server side.

• slick.jdbc.JdbcCapabilities.supportsByte: SQL Server's TINYINT is unsigned. It doesn't have a signed Byte-like type. Slick maps Byte to SMALLINT instead and that's how it shows up in model and code-generation.

The default type for strings of unlimited length is "VARCHAR(MAX)", falling back to "VARCHAR(254)" if a PrimaryKey column option is set. This can be changed by overriding slick.jdbc.SQLServerProfile.defaultStringType in application.conf.

Slick profile for SQLite.

This profile implements slick.jdbc.JdbcProfile ''without'' the following capabilities:

• slick.relational.RelationalCapabilities.functionDatabase, slick.relational.RelationalCapabilities.functionUser: Functions.user and Functions.database are not available in SQLite. Slick will return empty strings for both.

• slick.relational.RelationalCapabilities.joinFull, slick.relational.RelationalCapabilities.joinRight: Right and full outer joins are emulated because there is not native support for them.

• slick.jdbc.JdbcCapabilities.mutable: SQLite does not allow mutation of result sets. All cursors are read-only.

• slick.sql.SqlCapabilities.sequence: Sequences are not supported by SQLite.

• slick.jdbc.JdbcCapabilities.returnInsertOther: When returning columns from an INSERT operation, only a single column may be specified which must be the table's AutoInc column.

• slick.relational.RelationalCapabilities.typeBigDecimal: SQLite does not support a decimal type.

• slick.relational.RelationalCapabilities.typeBlob: Blobs are not supported by the SQLite JDBC driver (but binary data in the form of Array[Byte] is).

• slick.relational.RelationalCapabilities.zip: Row numbers (required by zip and zipWithIndex) are not supported. Trying to generate SQL code which uses this feature throws a SlickException.

• slick.jdbc.JdbcCapabilities.insertOrUpdate: InsertOrUpdate operations are emulated on the client side if the data to insert contains an AutoInc field. Otherwise the operation is performed natively on the server side.

• slick.jdbc.JdbcCapabilities.defaultValueMetaData: The stable xerial sqlite-jdbc driver 3.7.2 does not return default values for columns in the DatabaseMetaData. Consequently they also do not appear in Slick's model. This has been fixed in sqlite-jdbc, but the only released version that contains the fix is milestone 3.7.15-M1. You can use it instead of the stable 3.7.2 in order to get default values with SQLite. Also see https://code.google.com/p/sqlite-jdbc/issues/detail?id=27

• slick.jdbc.JdbcCapabilities.booleanMetaData: SQLite doesn't have booleans, so Slick maps to INTEGER instead. Other jdbc drivers like MySQL map TINYINT(1) back to a Scala Boolean. SQLite maps INTEGER to an Integer and that's how it shows up in the jdbc meta data, thus the original type is lost.

• slick.jdbc.JdbcCapabilities.distinguishesIntTypes: SQLite does not distinguish integer types and maps them all to Int in the meta data.

• slick.jdbc.JdbcCapabilities.supportsByte: SQLite does not distinguish integer types and maps them all to Int in the meta data.

• slick.jdbc.JdbcCapabilities.returnMultipleInsertKey: SQLite returns the last generated key only.

Basic conversions for setting parameters in PositionedParameters.

SetParameter for tuple types.

A non-streaming Action that wraps a JDBC call.

Factory methods for JdbcResultConverters which are manually specialized on the underlying JdbcType. A generic implementation of this factory still provides allocation free call paths but performs almost 100% slower in the fast path benchmark.

An invoker which executes an SQL statement through JDBC.

A streaming Action that wraps an Invoker. It is used for the Lifted Embedding, Direct Embedding, Plain SQL queries, and JDBC metadata.

Converts between TIMESTAMPTZ and java.time times and back. Oracle jar not on path at compile time (but must be a run time) Use reflection to get access to TIMESTAMPTZ class

Represents a transaction isolation level.