I have a package (System.IO.Packaging.Package) hosted on a webserver (URL) and, if I feed the package URL in web browser I can download the package.
My requirement is to extract a single file (stream associated to a file) from the package without downloading the entire package.
System.IO.Packaging has classes to allow manipulating packages. These classes work well if the package is located on a SMB share, but fails to extract file if the package is hosted on a web-server.
It fails when I call PackagePart.GetStream() method, giving an exception which says that the stream is write-only and hence a read operation cannot be performed. I have tried using overloads of Package.Open() and PackagePart.GetStream() method passing read/write
permissions but still it does not work.
If I understand the stack trace correctly, somehow the stream instance in the MS.Internal.IO.Zip.ZipIOLocalFileBlock._blockManager does not have permissions propagated from the PackagePart.GetStream() call.
Is this a bug? Did I do something wrong? Is there a workaround? I have been struggling for over two weeks now to get a solution for this problem and there is no ample documentation around usage of these classes.
PS: I have already tried some popular Zip libraries (SharpZipLib, DontNetZip and .NET 4.5 ZipArchive class) they all fail to realize my requirement.
Thank you.
