The Formula Working Group are pleased to begin public review of a set of additional modules and functions for the formula specifications.  See for details.

  Most of the following features have been implemented to some degree.  The main points are:

1.         Messaging: Both conformant formula processors have deployed mechanisms to generate human readable messages based on assertion results.  Stakeholders have various facets of this feature in deployed production, the draft standard details the ability to embed XPath expressions in the messages. The implementation experience has helped in evolving to a pair of specifications, one for the general mechanism of messages, which can embed XPath expressions in the message text, so that this mechanism is usable by non-formula XBRL consumers, and the other spec for the specific arc roles and variables visibility for messages for assertion validation results.

2.         Aspect cover filter:  Variable set evaluations with terms of unlike units, periods, and other facets depend on filters to support aspect intermixing, with a feature of implicit filtering that ignores covered aspects.  The aspect cover filter makes explicit and uniform the declaration of aspects to be covered in the implicit filtering process, and will improve the readability and maintenance of formulas that used existing 'trick' expressions to achieve this currently.

3.         Custom function implementation:  Custom functions have been supported in arbitrary programming languages, such as Java, but not in a manner that allows the custom functions to be implemented within formula linkbases, be portable between vendor implementations, and be maintainable in the linkbases.  This feature supports multi-step XPath expressions to implement custom functions, and should reduce the clutter of existing products that resulted to repeatedly copying expression code or were unable to implement, without Java, some mathematical functions (like present value calculations and such).

4.         Concept relation filters and associated DTS functions:  The dimension filter provided a means of filtering based on dimensional relationships in the DTS, this filter and companion functions extends that to all kinds of relationships.  It allows calculation processing implemented by formula, movement analysis driven by patterns of relationship structure, and reduction of difficult formula maintenance where one repeated pattern of expression can be driven by linkbase relationships and coded once, instead of repeated for each summation or movement formula.  Also, supportive functions are provided to obtain available link roles, labels, role and arcrole definitions.

5.         Instances, Tuples, and Chaining are supported by a set of related specifications for arc roles and techniques.  There has been steady demand for these features by stakeholders, and a prototype implementation to prove its feasibility.