Description Valentina Popescu 2007-10-10 12:42:15 EDT

The current SML validation implementation offers only the capability of reporting errors as a document or set of documents is validated.

As a next step, the implementation should provide a set of API's for accessing the underlying data model as well as for running the evaluation in separate steps.

The scenarios we should be able to cover :
- be able to validate in memory objects as well as file system documents.
A parse()throws SMLExceptions API that takes as input a Path/String ( system file path )or InputStream
- be able to collect the parsing/validation errors and use them in custom scenarios ( the default approach if nothing is specified is to show the errors to the default console output; the user can overwrite this option and decide to not show the errors or show them in a different format or output)
- run only a subset of the validation steps. 
The user may want to run a schematron rule every time some data changes, without having to run the entire model validation. Need an API runSchematron(rule, Element(s) ) so that I can run a specific rule against a set of elements.
- API which allows binding a schematron rule with a set of documents, similar with what the IF ruleBindings element does.