Decision Table Configuration

Introduction

A decision table is the core component of a DMN decision. It defines a set of rules that map input conditions to output values. Each row in the table represents a rule, and when the input conditions of a row match, the corresponding output values are produced.

This page covers the detailed configuration of decision table columns and rules within the DMN Visual Editor.

Decision Table Structure

A decision table is composed of the following elements:

Decision table structure

  • Input columns (1): Define the input variables and their conditions

  • Output columns (2): Define the output variables produced by the decision

  • Rules (rows) (3): Each row is a rule mapping input conditions to output values

  • Hit policy indicator (4): Shows the active hit policy (always F for FIRST)

  • Annotation column (5): Optional free-text description for each rule

Hit Policy

The hit policy defines how multiple matching rules are handled. In the Axelor DMN editor, the hit policy is enforced as FIRST (displayed as F in the table header).

With the FIRST hit policy, when rules overlap (multiple rules match the same input), the first matching rule in order is chosen. The remaining matching rules are ignored.

The Axelor DMN editor enforces the FIRST hit policy. Other standard DMN hit policies (Unique, Any, Collect, Priority, Rule Order, Output Order) are not available through the editor UI.
To control which rule takes priority, arrange your rules from most specific (top) to most general (bottom).

Input Columns

Each input column defines a condition that incoming data must match for a rule to fire.

Input Column Properties

Click on an input column header to open its properties panel:

Input column properties panel
Property Description

Label

The display label for the column header. Used in Excel export/import headers.

Expression

The input expression that defines what value to evaluate. Can be entered manually or selected via the field selector.

Expression Language

The language used to evaluate the expression. See Expression Languages for available options.

Input Variable

An optional variable name to reference this input in other expressions within the same rule.

Type

The data type of the input. Determines the type-specific editor for rule cells. See Data Types.

Default Value

A default value used when no input is provided.

Field Selector

The expression field provides a field selector dialog to help you build expressions based on model fields:

Field selector dialog

  1. Select the model to browse from the model dropdown

  2. Navigate through the field hierarchy to select the desired field

  3. Sub-fields of relational fields can be accessed using dot notation (e.g., partner.name)

  4. The selected field path is automatically set as the input expression

When an expression text is set (non-empty) on an input column via the field selector, the expression language is automatically switched to Groovy.

Output Columns

Each output column defines a value that is produced when a rule fires.

Output Column Properties

Click on an output column header to open its properties panel:

Output column properties panel
Property Description

Label

The display label for the column header. Used in Excel export/import headers.

Expression

The output expression. Can be sourced from context (a model field) or model (the entire model reference).

Output Name

The variable name used to reference this output in the decision result. This name is used when mapping outputs to process variables.

Type

The data type of the output. See Data Types.

Default Value

A default value when no rule matches or the output cell is empty.

Expression Source

Output expressions can be configured with two source modes:

  • Context: The output value is a field from the associated data model. Use the field selector to choose the target field.

  • Model: The output value references the entire model object.

Data Types

Decision table columns (both input and output) support the following data types:

Type Description

string

Text values. Rule cells support "match one of" and "match none of" editors.

boolean

True/false values. Rule cells display a yes/no selector.

integer

Whole number values. Rule cells support exact match, comparison operators, and ranges.

long

Large whole number values. Same editors as integer.

double

Decimal number values. Same editors as integer.

date

Date values. Rule cells support "exactly", "before", "after", and "between" operators.

datetime

Date and time values. Same editors as date.

Rule Editing

Each cell in a rule row can be edited by clicking on it. The editor displayed depends on the data type of the column.

String Rules

String rule editor

String rule cells provide the following match modes:

  • Match one of: The rule matches if the input equals any of the specified values (comma-separated list)

  • Match none of: The rule matches if the input does NOT equal any of the specified values

Boolean Rules

Boolean rule cells display a simple Yes/No selector.

Number Rules (integer, long, double)

Number rule editor

Number rule cells support:

  • Exact value: Matches a specific number

  • Comparison: Less than, less than or equal, greater than, greater than or equal

  • Range: Between two values (inclusive or exclusive bounds)

Date Rules (date, datetime)

Date rule editor

Date rule cells support:

  • Exactly: Matches a specific date

  • Before: Matches dates before the specified date

  • After: Matches dates after the specified date

  • Between: Matches dates within a specified range

Expression Languages

The DMN editor supports the following expression languages for input expressions and output definitions:

Language Description

FEEL

Friendly Enough Expression Language — the default DMN standard expression language. Suitable for most use cases.

Groovy

Apache Groovy scripting language. Automatically selected when a model field expression is used.

JUEL

Java Unified Expression Language. Uses ${…​} syntax.

JavaScript

JavaScript expressions evaluated on the JVM.

Python

Python expressions evaluated on the JVM (Jython).

JRuby

Ruby expressions evaluated on the JVM.
FEEL is the default expression language for new columns. When you use the field selector dialog to set an expression, the language automatically switches to Groovy.

Adding and Removing Columns

  • To add an input column: Right-click on the table header area and select "Add Input Column"

  • To add an output column: Right-click on the table header area and select "Add Output Column"

  • To remove a column: Right-click on the column header and select "Remove"

Adding and Removing Rules

  • To add a rule: Click the "+" button at the bottom of the table, or right-click a row and select "Add Rule Below"

  • To remove a rule: Right-click a row and select "Remove Rule"

  • Rules can be reordered by drag-and-drop

Rule order matters with the FIRST hit policy. The first matching rule takes precedence, so place more specific rules before general ones.