David Randall Stokes
David Randall Stokes
@athewsey Seems reasonable... I'll try to get that done for/before the next release.
Note: This also involves universally converting the docstrings markup from its current Epydoc-inspired format.
`Path` objects for loading schemata were introduced as part of #103, but there are other places it could be implemented.
`children` could easily be changed to a list of IDs; being a dict doesn't really improve things. It would simplify the process of replacing an `Element` subclass.
It looks like the non-apropos `IOError` exceptions are primarily in `ebmlite.decoding`.
Clarification: * "Ignore Voids" would be an argument for `Schema.load()` and affect the resulting Document. * Void elements would not show up when either the Document or Master elements are...