Convert Java library to C# (open-source)

C# lacks a cross-platform image metadata parser, and we need to fix that. The excellent Metadata-extractor project is written in Java and is probably the easiest to automatically convert to C#. [login to view URL] It is less than 20K lines of code according to [login to view URL] In fact, there is already an old (2011) port of Metadata-extractor to C#, which can be found here and used for reference: [login to view URL] We have created a GitHub repository for the result here: [login to view URL] You must push your changes directly to the repository or create pull requests. The older port can be found in the 'renaud' branch. Your changes should go into the 'new-conversion' branch. The newly converted project should 1. Retain all documentation, comments, and tests from the java equivalent. License files and headers *MUST* be retained. 2. Use idiomatic, concise C# when possible, and the appropriate .NET types ([login to view URL], [login to view URL], etc) that correspond to the Java types. No external dependencies should be required. 3. Use XUnit or NUnit instead of JUnit, and include and pass all tests from the original project, both locally and on AppVeyor. 4. Support automatic documentation generation with Doxygen or docu as part of the [login to view URL] script/ 5. Reflect the work done in the 'master' branch of metadata-extractor, as the official releases are far behind. You can 1. Use the existing sample-images repository at [login to view URL] 2. Use existing code portions from the 'renaud' branch as long as you add attribution in the header of each file and a note in the main LICENSE file. An automated conversion is appreciated; if used, it is important that 1. Automated conversions are committed separately from manual corrects, and tagged as such. This allows us to re-run the conversion and reapply (rebase) the manual edits on top of the updated base source code. 2. The conversion process is fully documented (preferably scripted!), with details of versions and options used. This project will be released under the Apache 2.0 license, and you will be recognized for your work (if you so choose). This is a significant resume booster.