Fix NodeDescriptorImporterBase.fromString to handle non ascii chars in xml file

[WolfgangHG]

This fixes #119

Arqullian uses this code to parse "arquillian.xml": https://github.com/arquillian/arquillian-core/blob/main/config/impl-base/src/main/java/org/jboss/arquillian/config/impl/extension/ClasspathConfigurationPlaceholderResolver.java
It exports a descriptor to a string, replaces some placeholders and reimports the descriptor from this string.
This fails if "arquillian.xml" contains non-ascii chars (though the xml files defines "UTF-8" encoding).

This pull request adds an override fromString to NodeDescriptorImporterBase. In this method, I create a java.io.StringReader, which is passed to the new method NodeImporter.importAsNode(Reader), implemented in XmlDomNodeImporterImpl. Here, a org.xml.sax.InputSource is created using this Reader, and the DocumentBuilder parses this source.

The NodeDescriptorImporterBase base class DescriptorImporterBase still contains the previous implementation of fromString that calls fromStream(new ByteArrayInputStream(string.getBytes())). I would have preferred to add a fromReader to the base class too, but this would have meant adding a method to the interface org.jboss.shrinkwrap.descriptor.api.DescriptorImporter.

Things to improve - please give me feedback:

• NodeDescriptorImporterBase: the existing fromStream and my new method fromString contain similar code. The part after the comment "// Create the end-user view" could be extracted to a helper method.
• same applies to XmlDomNodeImporterImpl.importAsNode(Reader).
• I didn't add a test. Actually, I don't have an idea how to implement one - probably I would have to do something similar to the arquillian code and load the arquillian descriptor from a file, convert to a string and re-parse it from this string? But currently, the test suite does not run for me - probably because I use Java 1.8 and it requires older versions? I saw that there is a pull request to switch to Java 1.8 - maybe this one fixes the test suite.

I tested it in my sample project and will attach a ut8 "arquillian.xml" and a "windows-1252" file (both containing umlauts in a comment) to #119. Both files did not show the failure from the initial bugreport.