Validation Logging

The validation logging framework superficially mimics Python’s logging framework, however instead of formatting and outputting validation logging messages as soon as they are generated, they are stored in an object that provides list functionality, the ability to collate repeated results, and to serialise itself as plain text, CSV or JSON.

The ValidationLogger is a list of ValidationResult objects.

Each ValidationResult has:

• A status, being GOOD, INFO, WARN, ERROR or SKIP

• A code reflecting what it relates to - these are an enum at ValidationCode

• A location str to identify where in the document the result applies

• A message str to describe in more detail the result of the check

To decide whether or not the validation passed or failed a relevant subset of checks, the ValidationPassChecker module offers “pass checkers” that include appropriate check codes, for example the XmlPassChecker has a _check_codes list that contains all the appropriate ValidationCode values for which check failures would indicate that the input is not valid XML, whereas the TtmlPassChecker has TTML-specific check failures.

Typically a ConstraintSet would check each appropriate ValidationPassChecker that is relevant for its purpose, in its static summarise method, so that the results can be used to establish where the error lies, for example the input document might be valid XML but contain TTML errors.