HHH-3404 support for regular expressions in HQL LIKE operator

documentation/src/main/asciidoc/querylanguage/Expressions.adoc

@@ -1495,7 +1495,7 @@ Their syntax is defined by:
 include::{extrasdir}/predicate_like_bnf.txt[]
 ----
 
-The expression on the right is a pattern, where:
+The expression on the right is usually a SQL-style pattern, where:
 
 * `_` matches any single character,
 * `%` matches any number of characters, and
@@ -1509,6 +1509,14 @@ from Book where title not like '% for Dummies'
 
 The optional `escape` character allows a pattern to include a literal `_` or `%` character.
 
+Alternatively, the `regexp` keyword specifies that the pattern should be interpreted as a regular expression:
+
+[[like-regexp-predicate-example]]
+[source, hql]
+----
+from Book where title not like regexp '.+ for Dummies'
+----
+
 As you can guess, `not like` and `not ilike` are the enemies of `like` and `ilike`, and evaluate to the exact opposite boolean values.
 
 [[in-predicate]]

documentation/src/main/asciidoc/querylanguage/extras/predicate_like_bnf.txt

@@ -1 +1 @@
-expression "NOT"? ("LIKE" | "ILIKE") expression ("ESCAPE" character)?
+expression "NOT"? ("LIKE" | "ILIKE") REGEXP? expression ("ESCAPE" character)?

hibernate-core/src/main/antlr/org/hibernate/grammars/hql/HqlLexer.g4

@@ -292,6 +292,7 @@ POSITION			: [pP] [oO] [sS] [iI] [tT] [iI] [oO] [nN];
 PRECEDING			: [pP] [rR] [eE] [cC] [eE] [dD] [iI] [nN] [gG];
 QUARTER				: [qQ] [uU] [aA] [rR] [tT] [eE] [rR];
 RANGE				: [rR] [aA] [nN] [gG] [eE];
+REGEXP              : [rR] [eE] [gG] [eE] [xX] [pP];
 RESPECT				: [rR] [eE] [sS] [pP] [eE] [cC] [tT];
 RETURNING			: [rR] [eE] [tT] [uU] [rR] [nN] [iI] [nN] [gG];
 RIGHT				: [rR] [iI] [gG] [hH] [tT];

hibernate-core/src/main/antlr/org/hibernate/grammars/hql/HqlParser.g4

@@ -669,7 +669,7 @@ predicate
 	| expression NOT? MEMBER OF? path											# MemberOfPredicate
 	| expression NOT? IN inList													# InPredicate
 	| expression NOT? BETWEEN expression AND expression							# BetweenPredicate
-	| expression NOT? (LIKE | ILIKE) expression likeEscape?						# LikePredicate
+	| expression NOT? (LIKE | ILIKE) REGEXP? expression likeEscape?				# LikePredicate
 	| expression NOT? CONTAINS expression										# ContainsPredicate
 	| expression NOT? INCLUDES expression										# IncludesPredicate
 	| expression NOT? INTERSECTS expression										# IntersectsPredicate
@@ -1998,6 +1998,7 @@ xmltableDefaultClause
 	| PRECEDING
 	| QUARTER
 	| RANGE
+	| REGEXP
 	| RESPECT
 	| RETURNING
 //	| RIGHT

hibernate-core/src/main/java/org/hibernate/dialect/Dialect.java

@@ -149,7 +149,6 @@
 import org.hibernate.tool.schema.spi.SchemaManagementTool;
 import org.hibernate.tool.schema.spi.TableMigrator;
 import org.hibernate.type.BasicType;
-import org.hibernate.type.BasicTypeRegistry;
 import org.hibernate.type.SqlTypes;
 import org.hibernate.type.StandardBasicTypes;
 import org.hibernate.type.descriptor.WrapperOptions;
@@ -235,7 +234,6 @@
 import static org.hibernate.type.SqlTypes.BIGINT;
 import static org.hibernate.type.SqlTypes.BINARY;
 import static org.hibernate.type.SqlTypes.BLOB;
-import static org.hibernate.type.SqlTypes.BOOLEAN;
 import static org.hibernate.type.SqlTypes.CHAR;
 import static org.hibernate.type.SqlTypes.CLOB;
 import static org.hibernate.type.SqlTypes.DATE;
@@ -434,7 +432,7 @@ protected void initDefaultProperties() {
 	protected void registerColumnTypes(TypeContributions typeContributions, ServiceRegistry serviceRegistry) {
 		final DdlTypeRegistry ddlTypeRegistry = typeContributions.getTypeConfiguration().getDdlTypeRegistry();
 
-		ddlTypeRegistry.addDescriptor( simpleSqlType( BOOLEAN ) );
+		ddlTypeRegistry.addDescriptor( simpleSqlType( SqlTypes.BOOLEAN ) );
 
 		ddlTypeRegistry.addDescriptor( simpleSqlType( TINYINT ) );
 		ddlTypeRegistry.addDescriptor( simpleSqlType( SMALLINT ) );
@@ -575,7 +573,7 @@ protected String columnType(int sqlTypeCode) {
 		return switch (sqlTypeCode) {
 			case ROWID -> "rowid";
 
-			case BOOLEAN -> "boolean";
+			case SqlTypes.BOOLEAN -> "boolean";
 
 			case TINYINT -> "tinyint";
 			case SMALLINT -> "smallint";
@@ -1103,8 +1101,8 @@ public int ordinal() {
 	 * functions with the same names.
 	 */
 	public void initializeFunctionRegistry(FunctionContributions functionContributions) {
-		final TypeConfiguration typeConfiguration = functionContributions.getTypeConfiguration();
-		final BasicTypeRegistry basicTypeRegistry = typeConfiguration.getBasicTypeRegistry();
+		final var typeConfiguration = functionContributions.getTypeConfiguration();
+		final var basicTypeRegistry = typeConfiguration.getBasicTypeRegistry();
 		final BasicType<Date> timestampType = basicTypeRegistry.resolve( StandardBasicTypes.TIMESTAMP );
 		final BasicType<Date> dateType = basicTypeRegistry.resolve( StandardBasicTypes.DATE );
 		final BasicType<Date> timeType = basicTypeRegistry.resolve( StandardBasicTypes.TIME );
@@ -1114,6 +1112,7 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		final BasicType<LocalTime> localTimeType = basicTypeRegistry.resolve( StandardBasicTypes.LOCAL_TIME );
 		final BasicType<LocalDate> localDateType = basicTypeRegistry.resolve( StandardBasicTypes.LOCAL_DATE );
 
+		final var functionRegistry = functionContributions.getFunctionRegistry();
 		final var functionFactory = new CommonFunctionFactory( functionContributions );
 
 		//standard aggregate functions count(), sum(), max(), min(), avg(),
@@ -1196,20 +1195,20 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		//only some databases support the ANSI SQL-style position() function, so
 		//define it here as an alias for locate()
 
-		functionContributions.getFunctionRegistry().register( "position",
+		functionRegistry.register( "position",
 				new LocatePositionEmulation( typeConfiguration ) );
 
 		//very few databases support ANSI-style overlay() function, so emulate
 		//it here in terms of either insert() or concat()/substring()
 
-		functionContributions.getFunctionRegistry().register( "overlay",
+		functionRegistry.register( "overlay",
 				new InsertSubstringOverlayEmulation( typeConfiguration, false ) );
 
 		//ANSI SQL trim() function is supported on almost all of the databases
 		//we care about, but on some it must be emulated using ltrim(), rtrim(),
 		//and replace()
 
-		functionContributions.getFunctionRegistry().register( "trim",
+		functionRegistry.register( "trim",
 				new TrimFunction( this, typeConfiguration ) );
 
 		//ANSI SQL cast() function is supported on the databases we care most
@@ -1220,7 +1219,7 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		// - casts to and from Boolean, and
 		// - casting Double or Float to String.
 
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"cast",
 				new CastFunction(
 						this,
@@ -1239,7 +1238,7 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		//additional non-standard temporal field types, which must be emulated in
 		//a very dialect-specific way
 
-		functionContributions.getFunctionRegistry().register( "extract",
+		functionRegistry.register( "extract",
 				new ExtractFunction( this, typeConfiguration ) );
 
 		//comparison functions supported on most databases, emulated on others
@@ -1250,7 +1249,7 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		//two-argument synonym for coalesce() supported on most but not every
 		//database, so define it here as an alias for coalesce(arg1,arg2)
 
-		functionContributions.getFunctionRegistry().register( "ifnull",
+		functionRegistry.register( "ifnull",
 				new CoalesceIfnullEmulation() );
 
 		//rpad() and pad() are supported on almost every database, and emulated
@@ -1261,23 +1260,23 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 
 		//pad() is a function we've designed to look like ANSI trim()
 
-		functionContributions.getFunctionRegistry().register( "pad",
+		functionRegistry.register( "pad",
 				new LpadRpadPadEmulation( typeConfiguration ) );
 
 		//legacy Hibernate convenience function for casting to string, defined
 		//here as an alias for cast(arg as String)
 
-		functionContributions.getFunctionRegistry().register( "str",
+		functionRegistry.register( "str",
 				new CastStrEmulation( typeConfiguration ) );
 
 		// Function to convert enum mapped as Ordinal to their ordinal value
 
-		functionContributions.getFunctionRegistry().register( "ordinal",
+		functionRegistry.register( "ordinal",
 				new OrdinalFunction( typeConfiguration ) );
 
 		// Function to convert enum mapped as String to their string value
 
-		functionContributions.getFunctionRegistry().register( "string",
+		functionRegistry.register( "string",
 				new StringFunction( typeConfiguration ) );
 
 		//format() function for datetimes, emulated on many databases using the
@@ -1290,89 +1289,91 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		//since there is a great variety of different ways to emulate them
 		//by default, we don't allow plain parameters for the timestamp argument as most database don't support this
 		functionFactory.timestampaddAndDiff( this, SqlAstNodeRenderingMode.NO_PLAIN_PARAMETER );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "dateadd", "timestampadd" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "datediff", "timestampdiff" );
+		functionRegistry.registerAlternateKey( "dateadd", "timestampadd" );
+		functionRegistry.registerAlternateKey( "datediff", "timestampdiff" );
 
 		//ANSI SQL (and JPA) current date/time/timestamp functions, supported
 		//natively on almost every database, delegated back to the Dialect
 
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"current_date",
 				new CurrentFunction(
 						"current_date",
 						currentDate(),
 						dateType
 				)
 		);
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"current_time",
 				new CurrentFunction(
 						"current_time",
 						currentTime(),
 						timeType
 				)
 		);
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"current_timestamp",
 				new CurrentFunction(
 						"current_timestamp",
 						currentTimestamp(),
 						timestampType
 				)
 		);
-		functionContributions.getFunctionRegistry().registerAlternateKey( "current date", "current_date" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "current time", "current_time" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "current timestamp", "current_timestamp" );
+		functionRegistry.registerAlternateKey( "current date", "current_date" );
+		functionRegistry.registerAlternateKey( "current time", "current_time" );
+		functionRegistry.registerAlternateKey( "current timestamp", "current_timestamp" );
 		//HQL current instant/date/time/datetime functions, delegated back to the Dialect
 
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"local_date",
 				new CurrentFunction(
 						"local_date",
 						currentDate(),
 						localDateType
 				)
 		);
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"local_time",
 				new CurrentFunction(
 						"local_time",
 						currentLocalTime(),
 						localTimeType
 				)
 		);
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"local_datetime",
 				new CurrentFunction(
 						"local_datetime",
 						currentLocalTimestamp(),
 						localDateTimeType
 				)
 		);
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"offset_datetime",
 				new CurrentFunction(
 						"offset_datetime",
 						currentTimestampWithTimeZone(),
 						offsetDateTimeType
 				)
 		);
-		functionContributions.getFunctionRegistry().registerAlternateKey( "local date", "local_date" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "local time", "local_time" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "local datetime", "local_datetime" );
-		functionContributions.getFunctionRegistry().registerAlternateKey( "offset datetime", "offset_datetime" );
+		functionRegistry.registerAlternateKey( "local date", "local_date" );
+		functionRegistry.registerAlternateKey( "local time", "local_time" );
+		functionRegistry.registerAlternateKey( "local datetime", "local_datetime" );
+		functionRegistry.registerAlternateKey( "offset datetime", "offset_datetime" );
 
-		functionContributions.getFunctionRegistry().register(
+		functionRegistry.register(
 				"instant",
 				new CurrentFunction(
 						"instant",
 						currentTimestampWithTimeZone(),
 						instantType
 				)
 		);
-		functionContributions.getFunctionRegistry().registerAlternateKey( "current_instant", "instant" ); //deprecated legacy!
+		functionRegistry.registerAlternateKey( "current_instant", "instant" ); //deprecated legacy!
 
-		functionContributions.getFunctionRegistry().register( "sql", new SqlFunction() );
+		functionRegistry.register( "sql", new SqlFunction() );
+
+		functionFactory.regexpLike();
 	}
 
 	/**

hibernate-core/src/main/java/org/hibernate/dialect/MariaDBDialect.java

@@ -45,7 +45,6 @@
 import org.hibernate.tool.schema.extract.internal.SequenceInformationExtractorMariaDBDatabaseImpl;
 import org.hibernate.tool.schema.extract.spi.SequenceInformationExtractor;
 import org.hibernate.type.SqlTypes;
-import org.hibernate.type.StandardBasicTypes;
 import org.hibernate.type.descriptor.jdbc.JdbcType;
 import org.hibernate.type.descriptor.jdbc.VarcharUUIDJdbcType;
 import org.hibernate.type.descriptor.jdbc.spi.JdbcTypeRegistry;
@@ -59,11 +58,11 @@
 
 import static org.hibernate.exception.spi.TemplatedViolatedConstraintNameExtractor.extractUsingTemplate;
 import static org.hibernate.internal.util.JdbcExceptionHelper.extractSqlState;
-import static org.hibernate.query.sqm.produce.function.FunctionParameterType.NUMERIC;
 import static org.hibernate.type.SqlTypes.GEOMETRY;
 import static org.hibernate.type.SqlTypes.OTHER;
 import static org.hibernate.type.SqlTypes.UUID;
 import static org.hibernate.type.SqlTypes.VARBINARY;
+import static org.hibernate.type.StandardBasicTypes.BOOLEAN;
 
 /**
  * A {@linkplain Dialect SQL dialect} for MariaDB 10.6 and above.
@@ -121,15 +120,20 @@ public NationalizationSupport getNationalizationSupport() {
 	public void initializeFunctionRegistry(FunctionContributions functionContributions) {
 		super.initializeFunctionRegistry( functionContributions );
 
+		final var functionRegistry = functionContributions.getFunctionRegistry();
 		final var commonFunctionFactory = new CommonFunctionFactory( functionContributions );
+		final var basicTypeRegistry = functionContributions.getTypeConfiguration().getBasicTypeRegistry();
 
 		commonFunctionFactory.windowFunctions();
 		commonFunctionFactory.hypotheticalOrderedSetAggregates_windowEmulation();
-		functionContributions.getFunctionRegistry().registerNamed(
+		commonFunctionFactory.inverseDistributionOrderedSetAggregates_windowEmulation();
+		commonFunctionFactory.median_medianOver();
+
+		commonFunctionFactory.regexpLike_regexp();
+
+		functionRegistry.registerNamed(
 				"json_valid",
-				functionContributions.getTypeConfiguration()
-						.getBasicTypeRegistry()
-						.resolve( StandardBasicTypes.BOOLEAN )
+				basicTypeRegistry.resolve( BOOLEAN )
 		);
 		commonFunctionFactory.jsonValue_mariadb();
 		commonFunctionFactory.jsonArray_mariadb();
@@ -139,13 +143,6 @@ public void initializeFunctionRegistry(FunctionContributions functionContributio
 		commonFunctionFactory.jsonArrayAppend_mariadb();
 		commonFunctionFactory.unnest_emulated();
 		commonFunctionFactory.jsonTable_mysql();
-
-		commonFunctionFactory.inverseDistributionOrderedSetAggregates_windowEmulation();
-		functionContributions.getFunctionRegistry().patternDescriptorBuilder( "median", "median(?1) over ()" )
-				.setInvariantType( functionContributions.getTypeConfiguration().getBasicTypeRegistry().resolve( StandardBasicTypes.DOUBLE ) )
-				.setExactArgumentCount( 1 )
-				.setParameterTypes(NUMERIC)
-				.register();
 	}
 
 	@Override
@@ -332,11 +329,6 @@ public IdentifierHelper buildIdentifierHelper(IdentifierHelperBuilder builder, D
 		return super.buildIdentifierHelper( builder, metadata );
 	}
 
-	@Override
-	public String getDual() {
-		return "dual";
-	}
-
 	public ViolatedConstraintNameExtractor getViolatedConstraintNameExtractor() {
 		return EXTRACTOR;
 	}