Tracing fixed

A small issue prevented tracing to work properly since the path parameter was changed to an internal field.

Version 1.1.1

• Fix an issue when returning a date or an array from a rule.

Version 1.1.0

• Deprecated InfernalEngine#defRule and replaced it by InfernalEngine#def
• Deprecated InfernalEngine#undefRule and replaced it by InfernalEngine#undef
• Fixed some documentation sentences

1.0.1 - assertAll

• Added the InfernalEngine#assertAll method.
• Added the InfernalEngine#fact static method to create facts to be consumed by assertAll.
• Improved code documentation.

1.0.0 - First stable release

• Complete rewrite of the engine with ECMAScript 2016.
• Method names have been changed to align with other inference engines (namely by CLIPS)
• Added the possibility to retract facts and undefine rules with wildcards.
• Improved tracing.

Add annotations and fact return parameter to `next`

• Add parameters to rules to reference same context facts.
• Add annotations to rules references to reference relative or absolute facts.
• Change the nextcallback to allow a second parameter expecting a rule_path / value mapping object.

Fixed dependency vulnerabilities

The project itself has no dependencies but mocha (a dev dependency) does. Updated the mocha reference to the latest version and rebuilt the package-lock.json to get rid of alerts on the github page.

This version is exactly the same as v0.17.0

After thinking about it, I should have released a 0.17.1 version instead. Too late, my bad 🤷‍♂️

Fixed agenda execution issue

• Fixed the agenda execution issue where only the last rule was executed by adding a local valiable to the async loop.
• Fixed the trace output event ordering to match the actual order of steps that are taken during inference.
• Changed the trace property "rule" to "fromRule" which is clearer, especially two years after my last commit to this project...
• Improved documentation!

return next() paradigm

This release contains a documentation updates and updated test code. The done callback have been renamed next and instead of calling next and returning on separate lines, I changed code and documentation to return next() on the same line instead.

Improved engine behavior

• It is now possible to load a sub model without clearing the engine state
• The inference callback function now receives the fact changed during inference
• The trace function now tells when a rule is added to the agenda.