Tailgunner

allright i will change it to byte[] for the previews and back to magick for the textures and make it load lazy so it only croaks when somebody accesses textures and the magick lib isn't there. that way you can keep the parser slick and when somebody needs the texture stuff /(s)?he/ can add the magick lib. i'll do installers with and without. btw, tony if you want to continue on, please feel free to merge my changes. I refactored quite a bit and it's quite more friendly for unit tests imho. edit: oh, and because i added one unit test already (along with the test project).

Mmh. Allright, I will think about that.

those textures are png format anyway. turning the byte array into a usable image is as simple as private static BitmapImage LoadImage( byte[] imageData ) { return LoadImage( new System.IO.MemoryStream( imageData ) ); } private static BitmapImage LoadImage( Stream stream ) { try { var i = new BitmapImage(); i.BeginInit(); i.StreamSource = stream; i.EndInit(); i.Freeze(); return i; } catch { return null; } }

Works fine for me. There are no references to the Magick library anymore, so they hardly can throw an error. I'm pretty sure something has been mixed up during install. Try re-installing while making sure the extractor isn't running. Meh, I packed outdated files into the installer. That is now corrected. Also: corrected link above.

And some more tinkering. I removed the dependency of the ImageMagick Library, it was 10 MB fat! Now uses byte[]. new installer: https://github.com/clausthaler-devel/crp-parser/blob/master/Installer/CRP-Extractor-0.2-Setup.exe?raw=true

And some more tinkering. While above tinkerings focus the developer sense of smugness, the following changes will appeal to the common user. The code now contains a windows app, as well as a (now working) command line tool to extract crp files. With both you can chose the output directory. Also some minor improvements to the code. Oh and I added a fincy fancy belly whistely installer for the clicky people. screenie:

I have tinkered around with the code a bit. My fork mainly introduces a AssetInfo object that gets returned by parseFile and bundles the parsed data. So you can say var info = new CrpDeserializer(path).parseFile(); var name = info.metadata["name"]; var preview = info.images[1]; etc. Here's the link: https://github.com/clausthaler-devel/crp-parser/tree/master/CRPParser