[svd-parser] Better validation of unknown tokens

Today I was playing with svd-parser on hand-written svd files.
I made typo in derivedFrom (double 'r') but it went without errors.
After that I tried with this sample:

<!-- This is the sample of minimal, valid SVD file -->
<device schemaVersion="1.3" xmlns:xs="http://www.w3.org/2001/XMLSchema-instance" xs:noNamespaceSchemaLocation="CMSIS-SVD.xsd">
  <name>Minimal SVD</name>
  <version>0.1</version>
  <description>This is syntetic example of minimal valid SVD file</description>
  <addressUnitBits>8</addressUnitBits>
  <width>32</width>
  <unknownEmptyToken></unknownEmptyToken>
  <peripherals>
    <peripheral>
        <unknownEmptyToken></unknownEmptyToken>
        <name> Bunny </name>
        <baseAddress>0x21370000</baseAddress>
    </peripheral>
  </peripherals>
</device>

But again - no errors.

Start at the svd-parser parsing and validation entry points and reproduce the issue with the malformed derivedFrom attribute and the supplied unknownEmptyToken sample. Trace how unrecognized XML tokens are handled; done means these inputs produce validation errors while the minimal valid SVD remains accepted.