0.0.6-beta1

[clarkis117]

Mime-Detective 0.0.6-beta1 includes numerous changes and improvements.

• Analyzer Abstraction (IFileAnalyzer) for allowing extensibility
• Static extension method extensibility through the static MimeAnalyzer class
• Various improvements and additions to the underlying file header definitions
• Significantly faster file header matching algorithms
• Seekable Streams are now reset to position 0 by default for extension methods that accept streams
• Secondary Analyzer for MS Document Type matching the MSDoc header (aka MS_Office)
• More test coverage

Tries (prefix trees) and Analyzers This release now includes 3 different file header matching implementations:

• ArrayBasedTrie -- Fastest implementation by far -- Consumes the most amount of memory
• DictionaryBasedTrie -- Significantly slower than ArrayBasedTrie -- Significantly faster than LinearCountingAnalyzer -- Consumes significantly less memory than ArrayBasedTrie -- This is the default
• LinearCountingAnalyzer -- A simple linear Algorithm, iterates through a list -- Significantly slower than all other implementations -- Consumes the least memory

Default header matching algorithm is now the DictionaryBasedTrie, constructed from MimeType.Types, and can be manipulated via the static MimeAnalzyer.PrimaryAnalzyer property. The Linear Algorithm now has the same behavior as tries. It will try to find the highest completely matching definition.

Resolves #12 Resolves #9 Resolves #11