Version 4.6 release notes

Navigation:  Reference chapter >

Version 4.6 release notes

Previous pageReturn to chapter overviewNext page

 

General

•CopyTarget operations with invalid paths are now caught by the repository validator - this was previously allowing code generation to occur, causing compilation errors.

•Changed the background colour of selected but not focused nodes so that selected variable nodes are more visible when not focused.

•Fixed an issue preventing users from changing their own usernames.

•Fixed an issue with copying onto repeating variable nodes that was causing compilation errors.

Inline Debugger

•Users can now use decisions, decision groups, and node operations to focus on their respective contexts in the debug input document tab via the right click context menu or F4 shortcut.

•The document history list box now shows the timestamp for each entry added from the debugging process.

•Document sets can now be loaded from the document history list box when a particular entry in the history was used with at least one other schema during rules execution. This means that users can go back to a particular input document set state in one go rather than having to go to each individual input document schema tab to reload every document. Document sets can also be saved/removed.

 

Test Executive

•Support for JSON added.

•Test executive console windows changed to use a fixed width font, with the ability to increase/decrease the font size via mouse wheel scrolling while holding Ctrl down.

•Test executive now has a "Logging Level" dropdown under the "Advanced Settings" tab to allow users to change the logging level that the Decision Server will use.

•

Document Builder

•Support for JSON added.

•The "Show where used" feature is now also available from document builder nodes.

•Added a context menu item to allow users to clear the values of the selected node and its descendants.

•Added a context menu item to allow users to remove the child nodes of the selected node.

•Results in the XPath browser can now be copied onto the clipboard.

•The XPath browser now adds the current xpath to its dropdown list when the dropdown loses focus, and when its xpath value is set from the document builder's "Show in XPath browser" menus.

•The XPath browser results column widths are now adjustable.

•Fixed an issue that can cause an exception when right clicking in the document builder when no node is selected.

•Fixed an issue where pasting a repeating node structure in the debug input document tab after executing the inline-debugger would cause an exception.

Decision Model Pallet

•Fixed an issue with the "Create Processes from Schema Node" feature where it would incorrectly assign the "Create if Necessary" action type to repeating non-output nodes, causing problems with code generation and compilation down the line.

•The "Show Schema(s)" toolbar button in the Decision Model pallet now has a dropdown to allow users to select a specific schema to show.

•Added a new toolbar button "Generate Default Formula(s) for selected item" to allow users to generate default formulas for the selected cell and its descendants (alternative to existing "Generate Default Formula(s)" that targets the entire Decision Model).

•Using action types (e.g. AlwaysCreate, CreateIfNecessary etc.) as decision model names no longer causes generated decision model code to fail compilation.

•The action type indicator for "Create if necessary" and "Create" on decisions and decision groups now have a slightly darker yellow so that it is more obvious.

•Indicator decorations on decisions and decision groups have been moved from above to below the cell.

•The indicator decoration for "Iterate collection in reverse order" has been updated.

•Results from executing the decision model are now shown above each decision and decision group cell (toggling "Show execution timings" on the toolbar replaces the execution results). The background colors of cells are also updated to indicate whether that decision or decision group was executed.

•Decision/decision group cells that have a breakpoint set on them now have a red selection highlight instead of the cell color itself changing.

•Decision/decision group cells that are currently the location of an active debugging breakpoint now have a yellow selection highlight instead of the cell color itself changing.

•Added a new toolbar button "Reset test inputs" to reset all debug input documents for a decision model's schemas to their respective default empty documents.

•Editing controls (e.g. Decision/decision group name editing, label editing) now scale with zoom factors greater than 100%.

•Fixed an issue with tooltips not appearing at the correct location when font scaling of over 100% is used on a monitor.

•The "reverse order iteration" option is now allowed with the "use previous context" option.

•Inline debugger performance has been improved.

Formula/UDO Pallet

•Fixed an issue where changing the target column of a table lookup operation would result in an exception if the "Default if not matching rows" branch does not have an operation.

•The 'NodeLiteralEnum' operator is no longer context dependent and can be used from any location in the node tree.

•Fixed an issue where node lookup aliases were not being resolved correctly.

•The comment indicator has been moved to the bottom right of cells, and color changed to be more consistent with the Decision Model Pallet.

•Node lookup cells now expand to the right instead of below, and the expand/collapse box has been moved to the right.

•Fixed an issue with child operations of summation operations appearing slightly opaque when the summation operation has a label.

•Fixed an issue with labels not correctly extending past the summation operation itself when collapsed.

•Fixed an issue where the data type dialog would not show at the center of the main form.

•Editing controls (e.g. Literal editing, variable name editing) now scale with zoom factors greater than 100%.

•Added a new "Copy Variable Name" context menu item when right clicking on operations that have variable names set.

•A new "a" decoration below cells will now be shown when operations have additional optional arguments that can be specified.

•Fixed an issue with tooltips not appearing at the correct location when font scaling of over 100% is used on a monitor.

•Pressing 'p' is now a shortcut for expanding the operators palette to show whichever operator group was last selected.

•Pressing 'o' or 'Ctrl + Space' now brings up an operator search box at the cursor location that allows users to find the operator that they want to add to their formula/UDO. The search box only contains the same list of operators available from the operators palette.

•Operation pallet cells now display their mode/table name/target column/xpath as colored decorations inside the cell itself.

•Double clicking on an operation pallet cell's variable name puts the variable name in edit mode.

•Add context menu items for users to add the 'If' operation, 'abort' operations, and node lookups based on the context and data type of the selected (empty) cell.

Operators

•[Financial] New operator "NPV": Calculates the net present value given a set of values and discount rate.

•[Financial] New operator "IRR": Calculates the internal rate of return for a series of cash flows.

•[Arithmetic] New operator "StandardDeviation": Calculates the standard deviation from a given set of values (as a population or sample selectable via the operation's mode).

•[Arithmetic] Updated operator "Multiply": Now has an optional "Default when empty" argument to allow users to specify a value to return when there are no values to multiply.

•[Arithmetic] Updated operator "AverageOperation": Now has an optional "Default when empty" argument to allow users to specify a value to return when there are no values to average.

•[Arithmetic] Updated operator "MaximumOperation": Now has an optional "Default when empty" argument to allow users to specify a value to return when there are no values to search.

•[Arithmetic] Updated operator "MinimumOperation": Now has an optional "Default when empty" argument to allow users to specify a value to return when there are no values to search.

•[String (text) manipulation] New operator "IsEmptyString": Tests if a given string value is empty or just whitespace (depending on the selected mode).

•[String (text) manipulation] New operator "RightSubstring": Obtains a specified number of characters starting from the ride side of a string.

•[String (text) manipulation] New operator "Encase": Surrounds text with a specified string, using the opposite order on the right side and opposite symbols if possible (e.g. '[' and ']', '{' and '}', '(' and ')', '<' and '>').

•[String (text) manipulation] New operator "Reverse": Reverses the character order of the given string.

•[String (text) manipulation] Updated operator "IndexOf": Now has an optional "start index" argument to allow users to specify where in the string they would like to start the search from.

•[String (text) manipulation] Updated operator "Concatenate": Now has an optional "delimiter" argument to allow insertion of a string in between each string to concatenate (like the ConcatenateNodes operator).

•[String (text) manipulation] Updated operators "RegxMatch" and "RegxReplace": Now have "case sensitive" and "ignore case" modes.

•[Miscellaneous] New operator "DecisionName": Returns the name of the currently executing decision or decision group.

•[Node List] New operator "NodeValues": Allows users to return an array of values from a repeating node collection - this is accessed by dragging a repeating node collection onto the pallet like other node list operators.

•The following operators now accept an array of values returned from the new "NodeValues" operator (along with their default results when provided with an empty array or invalid values):

▪Add (default: 0)

▪Multiply (default: optional user specified value, else 1)

▪Subtract (default value subtracted: 0)

▪AnyOperationTrue (default: false - "at least one item needs to return true")

▪AllOperationsTrue (default: true - "zero or more items all return true")

▪CountTrueOperations (default: 0)

▪MaximumOperation (default: optional user specified value, else throws DecisionExecutionException)

▪MinimumOperation (default: optional user specified value, else throws DecisionExecutionException)

▪AverageOperation (default: optional user specified value, else throws DecisionExecutionException)

▪IsMemberOfStringSet (default: false)

▪IsMemberOfNumericSet (default: false)

▪Concatenate (default: empty string)

▪NPV (default: user specified value)

▪IRR (default: user specified value)

▪StandardDeviation (default: user specified value)

▪If

▪Select

▪PreviousContext

▪ExternalContext

Tables

•Copies of tables from other repositories along with their associated table contents can now be imported via the new "Table Import Wizard". This is a new feature that allows users to specify tables that can be shared across other repositories.

•Excel spreadsheets can now be imported via the new "Spreadsheet Import Wizard".

•Fixed an issue where the progress dialog would not show at the center of the main form when opening formula versions or decision models.

•The table form now has a new "Shared" checkbox to allow the user to indicate that a table can be shared by decision models across multiple repositories.

•The inline-tester and test executive will automatically detect if a shared table is being used, generating and compiling the code for use in testing as long as the shared table's source repository is available on the current repository server instance.

•The release form and production release settings editor now lists tables as entities that can be released/deployed.

•There is now a new "Table Import Wizard" to facilitate the import of shared tables - all shared tables and their associated table contents can only be edited from the source repository.

Decision Server Runtime

•Support added for standalone tables that can be shared by decision models across multiple repositories.

•C# Decision Server API updates:

o[Added] IDecisionServer.GetTable(string tableID)

o[Added] IDecisionServer.GetTableInfo(string tableID)

o[Added] IDecisionServer.GetTableInfoList()

o[Added] ITableSet.Table

o[Removed] ITable.DecisionModel (getting the Decision Model from the Table this was unnecessary)

•Java Decision Server API updates:

o[Added] DecisionServer.getTable(String tableID)

o[Added] DecisionServer.getTableInfo(String tableID)

o[Added] DecisionServer.getTableInfoList()

o[Removed] Table.getDecisionModel() (getting the Decision Model from the Table this was unnecessary)

•Support for JSON added - Newtonsoft (https://www.newtonsoft.com/json) for C# and Jackson (https://github.com/FasterXML/jackson) for Java). As such, the Decision Server now supports multiple DOMs in a single call to execute rules. This means that IDecisionExecutionArgs and IDocumentProvider interfaces now use the IRulesDocument interface to represent the actual document being used in rules execution (instead of the actual document directly in the past, i.e. XmlDocument).

oAll constructors of DecisionExecutionArgs have been marked as obsolete/deprecated and will be removed in the future. Going forward, the preferred way to create instances of DecisionExecutionArgs will be via static factory methods. DecisionExecutionArgs.Create overloads accept IRulesDocuments while DecisionExecutionArgs.CreateWithObjects accept documents themselves (e.g. System.Xml.XmlDocument, org.w3c.dom.Document) that get converted to their respective IRulesDocuments implementations.

o[C#] The default implementations of IRulesDocument are XmlRulesDocument and JsonRulesDocument which each take System.Xml.XmlDocument and Newtonsoft.Json.Linq.JToken respectively in one of their constructor overloads.

o[Java] The default implementations of IRulesDocument are XmlRulesDocument, JdomRulesDocument, and JsonRulesDocument which each take org.w3c.dom.Document, org.jdom.Document, and com.fasterxml.jackson.databind.JsonNode respectively in one of their constructor overloads.

•Updated the way decisions and decision groups iterate over repeating nodes so that performance for the C# XML DOM is improved when iterating in reverse order.

•Decision Groups that iterate over a repeating node collection will now cache its node collection using a key specified in its referenced formula. This works similar to implicit caching for the PositionToNode and AnyNodeTrue operators where it attempts to detect the use of StringCompare or Equals operators that specify one of the node's child attributes/elements as its key. Node collections to be cached also need to be non-output nodes (i.e. readonly).

•Fixed an issue causing the Copy/CopyTarget operation not creating a new instance of an ancestor repeating node when the AlwaysCreate mode is specified.

•Fixed an issue with the IsValidNode operator and typed repeating nodes where it would include every single instance of that repeating node collection as part of its determination of whether the node is valid. This meant that as long as a single node in the collection was deemed valid, IsValidNode would return true, even if other nodes were empty or contained invalid data.

•Fixed an issue causing a null reference exception when using node lookups under the "Default when empty" branch of a cached PositionToNode operation.

Local and Production releases

•Decision model settings used during release now use the settings as defined in the decision model pallet by default (as long as the release settings for that decision model have not previously been set before), making it an explicit decision now to override these settings.

•New path variables added: <Name>, <NameNoSpaces>, <ID> - these refer to the name and ID of the entity being released.

•Path variables are no longer case sensitive, e.g. using <NaMe> will yield the same result as <Name>.

•Textboxes for all build settings that accept path variables now highlight supported path variables and provide an auto-complete dropdown (with tooltips about what each path variable does) when the user types the '<' character to start a path variable.

 

Decision Server JSON behaviour differences to XML

•The default parsing for Newtonsoft JSON always attempts to parse string values as a DateTime object if it sees it in the ISO8601 format - this means that the Decision Server is always going to get a JSON document that has pre-parsed DateTime values regardless of whether the value is supposed to be DateTime typed or not. This behaviour however can be overridden by the calling application when it deserializes the JSON to use with the call to Decision Server.

•Date (without time) typed nodes are updated in the JSON DOM as string values regardless of whether it was deserialized as a DateTime object.

•Properties/arrays that contain typed values other than string can no longer have its output controlled by Decision Model settings. This is entirely at the mercy of the calling application when it performs serialization. This is especially relevant to DateTime typed values as all original time zone information will no longer be available.

•Attributes are not directly supported by JSON, Decision Server will attempt to mimic this by giving JSON objects properties with the attribute name prefixed by the '@' character.

•Typed elements with attributes are not directly supported by JSON, Decision Server will attempt to mimic this by giving the JSON object a "#text" property.

For example: <foo bar="baz">goo</foo> will be represented as { "foo": { "@bar": "baz", "#text": "goo" } }.

•JSON does not support namespaces. Decision Server will only attempt to respect namespace prefixes originally defined in the source schema without looking at the actual URIs.

For example: <ns1:foo>bar</ns1:foo> and  <ns2:foo>bar</ns2:foo>  will be represented as { "ns1:foo": "bar" } and  { "ns2:foo": "bar" } respectively even if the prefixes ns1 and ns2 have the same namespace URI.