5
Rules

This chapter explains the concepts related to rules.

This chapter contains these topics:

• The Components of a Rule
• Rule Set Evaluation
• Database Objects and Privileges Related to Rules

  See Also: Chapter 6, "How Rules Are Used In Streams" Chapter 15, "Managing Rules and Rule-Based Transformations" Chapter 24, "Rule-Based Application Example"

The Components of a Rule

A rule is a database object that enables a client to perform an action when an event occurs and a condition is satisfied. Rules are evaluated by a rules engine, which is a built-in part of Oracle. Both user-created applications and Oracle features, such as Streams, can be clients of the rules engine.

A rule consists of the following components:

• Rule Condition
• Rule Evaluation Context (optional)
• Rule Action Context (optional)

Each rule is specified as a condition that is similar to the condition in the WHERE clause of a SQL query. You can group related rules together into rule sets. A single rule can be in one rule set, multiple rule sets, or no rule sets.

Rule Condition

A rule condition combines one or more expressions and operators and returns a Boolean value, which is a value of TRUE, FALSE, or NULL (unknown). An expression is a combination of one or more values and operators that evaluate to a value. A value can be data in a table, data in variables, or data returned by a SQL function or a PL/SQL function. For example, the following condition consists of two expressions (department_id and 30) and an operator (=):

department_id = 30

This logical condition evaluates to TRUE for a given row when the department_id column is 30. Here, the value is data in the department_id column of a table.

A single rule condition may include more than one condition combined with the AND, OR, and NOT conditional operators to form compound conditions. For example, consider the following compound condition:

department_id = 30 OR job_title = 'Programmer' 

This rule condition contains two conditions joined by the OR conditional operator. If either condition evaluates to TRUE, then the rule condition evaluates to TRUE. If the conditional operator were AND instead of OR, then both conditions would have to evaluate to TRUE for the entire rule condition to evaluate to TRUE.

Variables in Rule Conditions

Rule conditions may contain variables. When you use variables in rule conditions, precede each variable with a colon (:). The following is an example of a variable used in a rule condition:

:x = 55

Variables enable you to refer to data that is not stored in a table. A variable may also improve performance by replacing a commonly occurring expression. Performance may improve because, instead of evaluating the same expression multiple times, the variable is evaluated once.

A rule condition may also contain an evaluation of a call to a subprogram. These conditions are evaluated in the same way as other conditions. That is, they evaluate to a value of TRUE, FALSE, or unknown. The following is an example of a condition that contains a call to a simple function named is_manager that determines whether an employee is a manager:

is_manager(employee_id) = 'Y'

Here, the value of employee_id is determined by data in a table where employee_id is a column.

You can use user-defined types for variables. Therefore, variables can have attributes. When a variable has attributes, each attribute contains partial data for variable. In rule conditions, you specify attributes using dot notation. For example, the following condition evaluates to TRUE if the value of attribute z in variable y is 9:

:y.z = 9

Simple Rule Conditions

A simple rule condition is a condition that has either of the following forms:

• simple_rule_expression operator constant
• constant operator simple_rule_expression

In a simple rule condition, a simple_rule_expression is one of the following:

• Table column
• Variable
• Variable attribute
• Method result where the method takes no arguments and the method result can be returned by the variable method function, so that the expression is either a numerical or character type

For table columns, variables, and variable attributes, all numeric (NUMBER, FLOAT, DOUBLE, INTEGER) and character (CHAR, VARCHAR2) types are supported. Use of other types of expressions results in non-simple rule conditions.

In a simple rule condition, an operator is one of the following:

• <=
• <
• =
• >
• >=

Use of other operators results in non-simple rule conditions.

A constant is a fixed value. A constant can be:

• A number, such as 12 or 5.4
• A character, such as x or $
• A character string, such as "this is a string"

Therefore, the following conditions are simple rule conditions:

• tab1.col = 5
• :v1 > 'aaa'
• :v2.a1 < 10.01
• :v3.m() = 10

Rules with simple rule conditions are called simple rules. You can combine two or more simple rule conditions with the conditional operators AND and OR for a rule, and the rule remains simple. However, using the NOT conditional operator in a rule's condition causes the rule to be non-simple. For example, rules with the following conditions are simple rules:

• tab1.col = 5 AND :v1 > 'aaa'
• tab1.col = 5 OR :v1 > 'aaa'

Simple rules are important for the following reasons:

• Simple rules are indexed by the rules engine internally.
• Simple rules can be evaluated without executing SQL.
• Simple rules can be evaluated with partial data.

When a client uses DBMS_RULE.EVALUATE to evaluate an event, the client can specify that only simple rules should be evaluated by specifying true for the simple_rules_only parameter.

Rule Evaluation Context

A rule evaluation context is a database object that defines external data that can be referenced in rule conditions. The external data can exist as variables, table data, or both. The following analogy may be helpful: If the rule condition were the WHERE clause in a SQL query, then the external data in the rule's evaluation context would be the information referenced in the FROM clause of the query. That is, the expressions in the rule condition should reference the tables, table aliases, and variables in the evaluation context to make a valid WHERE clause.

A rule evaluation context provides the necessary information for interpreting and evaluating the rule conditions that reference external data. For example, if a rule refers to a variable, then the information in the rule evaluation context must contain the variable type. Or, if a rule refers to a table alias, then the information in the evaluation context must define the table alias.

The objects referenced by a rule are determined by the rule evaluation context associated with it. The rule owner must have the necessary privileges to access these objects, such as SELECT privilege on tables, EXECUTE privilege on types, and so on. The rule condition is resolved in the schema that owns the evaluation context.

For example, consider a rule evaluation context named hr_evaluation_context that contains the following information:

• Table alias dep corresponds to the hr.departments table.
• Variables loc_id1 and loc_id2 are both of type NUMBER.

The hr_evaluation_context rule evaluation context provides the necessary information for evaluating the following rule condition:

dep.location_id IN (:loc_id1, :loc_id2)

In this case, the rule condition evaluates to TRUE for a row in the hr.departments table if that row has a value in the location_id column that corresponds to either of the values passed in by the loc_id1 or loc_id2 variables. The rule cannot be interpreted or evaluated properly without the information in the hr_evaluation_context rule evaluation context. Also, notice that dot notation is used to specify the column location_id in the dep table alias.

Explicit and Implicit Variables

The value of a variable referenced in a rule condition may be explicitly specified when the rule is evaluated, or the value of a variable may be implicitly available given the event.

Explicit variables are supplied by the caller at evaluation time. These values are specified by the variable_values parameter when the DBMS_RULE.EVALUATE procedure is run.

Implicit variables are not given a value at evaluation time. The value of an implicit variable is obtained by calling the variable value evaluation function. You define this function when you specify the variable_types list during the creation of an evaluation context using the DBMS_RULE_ADM.CREATE_EVALUATION_CONTEXT procedure. If the value for an implicit variable is specified during evaluation, then the specified value overrides the value returned by the variable value evaluation function.

Specifically, the variable_types list is of type SYS.RE$VARIABLE_TYPE_LIST, which is a list of variables of type SYS.RE$VARIABLE_TYPE. Within each instance of SYS.RE$VARIABLE_TYPE in the list, the function used to determine the value of an implicit variable is specified as the variable_value_function attribute.

Whether variables are explicit or implicit is the choice of the designer of the application using the rules engine. The following are reasons for using an implicit variable:

• The caller of the DBMS_RULE.EVALUATE procedure does not need to know anything about the variable, which may reduce the complexity of the application using the rules engine. For example, a variable may call a function that returns a value based on the data being evaluated.
• The caller may not have execute privileges on the variable value evaluation function.
• The caller of the DBMS_RULE.EVALUATE procedure does not know the variable value based on the event, which may improve security if the variable value contains confidential information.
• The variable may be used infrequently, and the variable's value always can be derived if necessary. Making such variables implicit means that the caller of the DBMS_RULE.EVALUATE procedure does not need to specify many uncommon variables.

For example, in the following rule condition, the values of variable x and variable y could be specified explicitly, but the value of the variable max could be returned by running the max function:

:x = 4 AND :y < :max

Alternatively, variable x and y could be implicit variables, and variable max could be an explicit variable. As you can see, there is no syntactic difference between explicit and implicit variables in the rule condition. You can determine whether a variable is explicit or implicit by querying the DBA_EVALUATION_CONTEXT_VARS data dictionary view. For explicit variables, the VARIABLE_VALUE_FUNCTION field is NULL. For implicit variables, this field contains the name of the function called by the implicit variable.

Evaluation Context Association with Rule Sets and Rules

A single rule evaluation context can be associated with multiple rules or rule sets. The following list describes which evaluation context is used when a rule is evaluated:

• If an evaluation context is associated with a rule, then it is used for the rule whenever the rule is evaluated, and any evaluation context associated with the rule set being evaluated is ignored.
• If a rule does not have an evaluation context, but an evaluation context was specified for the rule when it was added to a rule set using the ADD_RULE procedure in the DBMS_RULE_ADM package, then the evaluation context specified in the ADD_RULE procedure is used for the rule when the rule set is evaluated.
• If no rule evaluation context is associated with a rule and none was specified by the ADD_RULE procedure, then the evaluation context of the rule set is used for the rule when the rule set is evaluated.

  Note: If a rule does not have an evaluation context, and you try to add it to a rule set that does not have an evaluation context, then an error is raised, unless you specify an evaluation context when you run the ADD_RULE procedure.

Evaluation Function

You have the option of creating an evaluation function to be run with a rule evaluation context. You may choose to use an evaluation function for the following reasons:

• You want to bypass the rules engine and instead evaluate events using the evaluation function.
• You want to filter events so that some events are evaluated by the evaluation function and other events are evaluated by the rules engine.

You can associate the function with the rule evaluation context by specifying the function name for the evaluation_function parameter when you create the rule evaluation context with the CREATE_EVALUATION_CONTEXT procedure in the DBMS_RULE_ADM package. Then, the rules engine invokes the evaluation function during the evaluation of any rule set that uses the evaluation context. The function must have each parameter in the DBMS_RULE.EVALUATE procedure, and the type of each parameter must be same as the type of the corresponding parameter in the DBMS_RULE.EVALUATE procedure, but the names of the parameters may be different.

An evaluation function has the following return values:

• DBMS_RULE_ADM.EVALUATION_SUCCESS: The user specified evaluation function completed the rule set evaluation successfully. The rules engine returns the results of the evaluation obtained by the evaluation function to the rules engine client using the DBMS_RULE.EVALUATE procedure.
• DBMS_RULE_ADM.EVALUATION_CONTINUE: The rules engine evaluates the rule set as if there were no evaluation function. The evaluation function is not used, and any results returned by the evaluation function are ignored.
• DBMS_RULE_ADM.EVALUATION_FAILURE: The user specified evaluation function failed. Rule set evaluation stops, and the rules engine returns the results of the evaluation obtained by the evaluation function to the rules engine client using the DBMS_RULE.EVALUATE procedure.

If you always want to bypass the rules engine, then the evaluation function should return either EVALUATION_SUCCESS or EVALUATION_FAILURE. However, if you want to filter events so that some events are evaluated by the evaluation function and other events are evaluated by the rules engine, then the evaluation function may return all three return values, and it returns EVALUATION_CONTINUE when the rules engine should be used for evaluation.

If you specify an evaluation function for an evaluation context, then the evaluation function is run during evaluation when the evaluation context is used by a rule set or rule.

Rule Action Context

A rule action context contains optional information associated with a rule that is interpreted by the client of the rules engine when the rule is evaluated for an event. The client of the rules engine can be a user-created application or an internal feature of Oracle, such as Streams. Each rule has only one action context. The information in an action context is of type SYS.RE$NV_LIST, which is a type that contains an array of name-value pairs.

The rule action context information provides a context for the action taken by a client of the rules engine when a rule evaluates to TRUE. The rules engine does not interpret the action context. Instead, it returns the action context information when a rule evaluates to TRUE. Then, a client of the rules engine can interpret the action context information.

For example, suppose an event is defined as the addition of a new employee to a company. If the employee information is stored in the hr.employees table, then the event occurs whenever a row is inserted into this table. The company wants to specify that a number of actions are taken when a new employee is added, but the actions depend on which department the employee joins. One of these actions is that the employee is registered for a course relating to the department.

In this scenario, the company can create a rule for each department with an appropriate action context. Here, an action context returned when a rule evaluates to TRUE specifies the number of a course that an employee should take. Here are the rule conditions and the action contexts for three departments:

Rule Name	Rule Condition	Action Context Name-Value Pair
rule_dep_10	department_id = 10	course_number, 1057
rule_dep_20	department_id = 20	course_number, 1215
rule_dep_30	department_id = 30	NULL

These action contexts return the following instructions to the client application:

• The action context for the rule_dep_10 rule instructs the client application to enroll the new employee in course number 1057.
• The action context for the rule_dep_20 rule instructs the client application to enroll the new employee in course number 1215.
• The NULL action context for the rule_dep_30 rule instructs the client application not to enroll the new employee any course.

Each action context can contain zero or more name-value pairs. If an action context contains more than one name-value pair, then each name in the list must be unique. In this example, the client application to which the rules engine returns the action context registers the new employee in the course with the returned course number. The client application does not register the employee for a course if a NULL action context is returned or if the action context does not contain a course number.

If multiple clients use the same rule, or if you want an action context to return more than one name-value pair, then you can list more than one name-value pair in an action context. For example, suppose the company also adds a new employee to a department electronic mailing list. In this case, the action context for the rule_dep_10 rule might contain two name-value pairs:

Name	Value
course_number	1057
dist_list	admin_list

The following are considerations for names in name-value pairs:

• If different applications use the same action context, then avoid naming conflicts by using different names or prefixes of names.
• Do not use $ and # in names to avoid conflicts with Oracle-supplied action context names.

Streams uses action contexts for rule-based transformations and, when subset rules are specified, for internal transformations that may be required on LCRs containing UPDATE operations.

You can add a name-value pair to an action context using the ADD_PAIR member procedure of the RE$NV_LIST type. You can remove a name-value pair from an action context using the REMOVE_PAIR member procedure of the RE$NV_LIST type. If you want to modify an existing name-value pair in an action context, then you should first remove it using the REMOVE_PAIR member procedure and then add an appropriate name-value pair using the ADD_PAIR member procedure.

Rule Set Evaluation

The rules engine evaluates rule sets based on events. An event is an occurrence that is defined by the client of the rules engine. The client initiates evaluation of an event by calling the DBMS_RULE.EVALUATE procedure. The information specified by the client when it calls the DBMS_RULE.EVALUATE procedure includes the following:

• The name of the rule set that contains the rules to use to evaluate the event
• The evaluation context to use for evaluation. Only rules that use the specified evaluation context are evaluated.
• Table values and variable values. The table values contain rowids that refer to the data in table rows, and the variable values contain the data for explicit variables. Values specified for implicit variables override the values that might be obtained using a variable value evaluation function. If a specified variable has attributes, then the client can send a value for the entire variable, or the client can send values for any number of the variable's attributes. However, clients cannot specify attribute values if the value of the entire variable is specified.
• An optional event context. An event context is a varray of type SYS.RE$NV_LIST that contains name-value pairs that contain information about the event. This optional information is not directly used or interpreted by the rules engine. Instead, it is passed to client callbacks, such as an evaluation function, a variable value evaluation function (for implicit variables), and a variable method function.

The client can also send other information about the event and about how to evaluate the event using the DBMS_RULE.EVALUATE procedure. For example, the caller may specify if evaluation must stop as soon as the first TRUE rule or the first MAYBE rule (if there are no TRUE rules) is found.

The rules engine uses the rules in the specified rule set to evaluate the event. Then, the rules engine returns the results to the client. The rules engine returns rules using the two OUT parameters in the EVALUATE procedure: true_rules and maybe_rules. That is, the true_rules parameter returns rules that evaluate to TRUE, and, optionally, the maybe_rules parameter returns rules that may evaluate to TRUE given more information.

Rule Set Evaluation Process

Figure 5-1 shows the rule set evaluation process:

1. A client-defined event occurs.
2. The client sends the event to the rules engine by running the DBMS_RULE.EVALUATE procedure.
3. The rules engine evaluates the event based on rules in the rule set and the relevant evaluation context. The client specifies both the rule set and the evaluation context in the call to the DBMS_RULE.EVALUATE procedure. Only rules that are in the specified rule set and use the specified evaluation context are used for evaluation.
4. The rules engine obtains the results of the evaluation. Each rule evaluates to either TRUE, FALSE, or NULL (unknown).
5. The rules engine returns rules that evaluated to TRUE to the client. Each returned rule is returned with its entire action context, which may contain information or may be NULL.
6. The client performs actions based on the results returned by the rules engine. The rules engine does not perform actions based rule evaluations.

Figure 5-1 Rule Set Evaluation

Text description of the illustration strms017.gif

Partial Evaluation

Partial evaluation occurs when the DBMS_RULE.EVALUATE procedure is run without data for all the tables and variables in the specified evaluation context. During partial evaluation, some rules may reference columns, variables, or attributes that are unavailable, while some other rules may reference only available data.

For example, consider a scenario where only the following data is available during evaluation:

• Column tab1.col=7
• Attribute v1.a1='ABC'

The following rules are used for evaluation:

• Rule R1 has the following condition:

  (tab1.col = 5)
  
  

• Rule R2 has the following condition:

  (:v1.a2 > 'aaa')
  
  

• Rule R3 has the following condition:

  (:v1.a1 = 'ABC') OR (:v2 = 5)
  
  

• Rule R4 has the following condition:

  (:v1.a1 = UPPER('abc'))
  
  

Given this scenario, R1 and R4 reference available data, R2 references unavailable data, and R3 references available data and unavailable data.

Partial evaluation always evaluates only simple conditions within a rule. If the rule condition has parts which are not simple, then the rule may or may not be evaluated completely, depending on the extent to which data is available. If a rule is not completely evaluated, then it can be returned as a MAYBE rule.

For example, given the rules in the previous scenario, R1 and the first part of R3 are evaluated, but R2 and R4 are not evaluated. The following results are returned to the client:

• R1 evaluates to FALSE, and so is not returned.
• R2 is returned as MAYBE because information about attribute v1.a2 is not available.
• R3 is returned as TRUE because R3 is a simple rule and the value of v1.a1 matches the first part of the rule condition.
• R4 is returned as MAYBE because the rule condition is not simple. The client must supply the value of variable v1 for this rule to evaluate to TRUE or FALSE.

  See Also: "Simple Rule Conditions"

Database Objects and Privileges Related to Rules

You can create the following types of database objects directly using the DBMS_RULE_ADM package:

• Evaluation contexts
• Rules
• Rule sets

You can create rules and rule sets indirectly using the DBMS_STREAMS_ADM package. You control the privileges for these database objects using the following procedures in the DBMS_RULE_ADM package:

• GRANT_OBJECT_PRIVILEGE
• GRANT_SYSTEM_PRIVILEGE
• REVOKE_OBJECT_PRIVILEGE
• REVOKE_SYSTEM_PRIVILEGE

To allow a user to create rule sets, rules, and evaluation contexts in the user's own schema, grant the user the following system privileges:

• CREATE_RULE_SET_OBJ
• CREATE_RULE_OBJ
• CREATE_EVALUATION_CONTEXT_OBJ

These privileges, and the privileges discussed in the following sections, can be granted to the user directly or through a role.

Privileges for Creating Database Objects Related to Rules

To create an evaluation context, rule, or rule set in a schema, a user must meet at least one of the following conditions:

• The schema must be the user's own schema, and the user must be granted the create system privilege for the type of database object being created. For example, to create a rule set in the user's own schema, a user must be granted the CREATE_RULE_SET_OBJ system privilege.
• The user must be granted the create any system privilege for the type of database object being created. For example, to create an evaluation context in any schema, a user must be granted the CREATE_ANY_EVALUATION_CONTEXT system privilege.

  Note: When creating a rule with an evaluation context, the rule owner must have privileges on all objects accessed by the evaluation context.

Privileges for Altering Database Objects Related to Rules

To alter an evaluation context, rule, or rule set, a user must meet at least one of the following conditions:

• The user must own the database object.
• The user must be granted the alter object privilege for the database object if it is in another user's schema. For example, to alter a rule set in another user's schema, a user must be granted the ALTER_ON_RULE_SET object privilege on the rule set.
• The user must be granted the alter any system privilege for the database object. For example, to alter a rule in any schema, a user must be granted the ALTER_ANY_RULE system privilege.

Privileges for Dropping Database Objects Related to Rules

To drop an evaluation context, rule, or rule set, a user must meet at least one of the following conditions:

• The user must own the database object.
• The user must be granted the drop any system privilege for the database object. For example, to drop a rule set in any schema, a user must be granted the DROP_ANY_RULE_SET system privilege.

Privileges for Placing Rules in a Rule Set

This section describes the privileges required to place a rule in a rule set.

The user must meet at least one of the following conditions for the rule:

• The user must own the rule.
• The user must be granted the execute object privilege on the rule if the rule is in another user's schema. For example, to place a rule named depts in the hr schema in a rule set, a user must be granted the EXECUTE_ON_RULE privilege for the hr.depts rule.
• The user must be granted the execute any system privilege for rules. For example, to place any rule in a rule set, a user must be granted the EXECUTE_ANY_RULE system privilege.

The user also must meet at least one of the following conditions for the rule set:

• The user must own the rule set.
• The user must be granted the alter object privilege on the rule set if the rule set is in another user's schema. For example, to place a rule in the human_resources rule set in the hr schema, a user must be granted the ALTER_ON_RULE_SET privilege for the hr.human_resources rule set.
• The user must be granted the alter any system privilege for rule sets. For example, to place a rule in any rule set, a user must be granted the ALTER_ANY_RULE_SET system privilege.

Privileges for Evaluating a Rule Set

To evaluate a rule set, a user must meet at least one of the following conditions:

• The user must own the rule set.
• The user must be granted the execute object privilege on the rule set if it is in another user's schema. For example, to evaluate a rule set named human_resources in the hr schema, a user must be granted the EXECUTE_ON_RULE_SET privilege for the hr.human_resources rule set.
• The user must be granted the execute any system privilege for rule sets. For example, to evaluate any rule set, a user must be granted the EXECUTE_ANY_RULE_SET system privilege.

Granting EXECUTE object privilege on a rule set requires that the grantor have the EXECUTE privilege specified WITH GRANT OPTION on all rules currently in the rule set.

Privileges for Using an Evaluation Context

To use an evaluation context, a user must meet at least one of the following conditions for the evaluation context:

• The user must own the evaluation context.
• The user must be granted the EXECUTE_ON_EVALUATION_CONTEXT privilege on the evaluation context, if it is in another user's schema.
• The user must be granted the EXECUTE_ANY_EVALUATION_CONTEXT system privilege for evaluation contexts.