BT

Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ

Topics

Choose your language

InfoQ Homepage News Support for Zip Files Still Lacking In .NET 3.0

Support for Zip Files Still Lacking In .NET 3.0

Bookmarks
The ability to use file compression like the venerable ZIP format is very important to many developers. For those developers using.NET, that means dropping to command shell or using a third-party component. With .NET 3.0, there is built-in support for ZIP files, though the implementation is somewhat questionable.

The first step to create a ZipPackage object. One would expect this to work just like any other object.

   Dim zipFile As New ZipPackage("C:\Temp\test.zip", FileMode.Create)
ZipPackage zipFile = new ZipPackage("C:\\Temp\\test.zip", FileMode.Create);

Unfortunately that isn't the case. Instead, one has to use a factory method in the base class Package.

   Dim zipFile As ZipPackage = Package.Open("C:\Temp\test.zip", FileMode.Create)
ZipPackage zipFile = (ZipPackage)Package.Open("C:\\Temp\\test.zip", FileMode.Create );

Though ZipPackage is the default return type for Package.Open, there is no way to actually specify that in any of the overloads making it somewhat difficult to create your own implementations.

Moving on, adding files to the zip package in the normal use case is exceedingly painful. One would think the code would look something like:

   zipFile.AddFile("C:/temp/someFile.txt", CompressionOption.Maximum)
zipFile.AddFile("C://temp//someFile.txt", CompressionOption.Maximum);

To add files the .NET way, one has to:

  1. Create a new URI object that will represent the name of the file inside the ZipFile.
  2. Determine the correct Mime type for the file.
  3. Create a new PackagePart using the aforementioned information.
  4. Open the source file as a Stream.
  5. Copy said stream into the PackagePart stream.

The below code shows how to do this using a very crude stream copy loop. Note that it should be much faster to use buffers than to read the stream one byte at a time.

   Dim newUri As New Uri("/someFile.txt", UriKind.Relative);
Dim part1 As ZipPackagePart = zipFile.CreatePart(newUri, _
System.Net.Mime.MediaTypeNames.Text.Plain, CompressionOption.Maximum)
Using output As Stream = part.GetStream,
input As FileStream = File.OpenRead("C:/temp/someFile.txt")
Dim value As Integer = input.ReadByte
Do Until value = -1
output.WriteByte(CByte(value))
value = input.ReadByte
Loop
End Using

While there isn't a simple way to decompress a zip file, it is far less painful than creating the file in the first place. This code lists all of the files in a zip file and dumps the text ones to the screen.

   zipFile = CType(ZipPackage.Open("C:\Temp\test.zip", IO.FileMode.Open), ZipPackage)
For Each part As ZipPackagePart In zipFile.GetParts
Console.WriteLine(part.Uri)
Console.WriteLine(vbTab & "Type:" & part.ContentType)
Console.WriteLine(vbTab & "Option:" & part.CompressionOption)
If part.ContentType.ToLower.Contains ("text/") Then
Using output As New StreamReader(part.GetStream)
Console.WriteLine(output.ReadToEnd)
End Using
End If
Next
Even when Maximum compression is chosen, the compression rate is very poor compared to that of WinZip. We tested this using a 2KB plain text file containing the readme for an application. WinZip had a 46% compression while .NET 3.0 had only a 4% compression. For WinZip the setting "Maximum (portable)" was used. For .NET 3.0, the above code was used.
 
Actually that comparison isn't fair, because buried in the docs is this note:
For the default ZipPackage subclass, the CreatePart method only supports two compressionOption values, NotCompressed or Normal compression. Other CompressionOption values of Maximum, Fast, or SuperFast use Normal compression.

One last warning, this method doesn't create standard zip files. While the files can be read by normal tools, the zip files will have an addition file called "[Content_Types].xml". Likewise, .NET 3.0 cannot read zip files unless the file contains "[Content_Types].xml". If the file is missing, it silently fails to find any files.

 In conclusion, .NET 3.0's support for the ZIP format is so highly specialized that is is useless in the general case.

Rate this Article

Adoption
Style

Hello stranger!

You need to Register an InfoQ account or or login to post comments. But there's so much more behind being registered.

Get the most out of the InfoQ experience.

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Community comments

  • .NET Zip support

    by Rob Eisenberg,

    Your message is awaiting moderation. Thank you for participating in the discussion.

    There is quite a bit of inaccuracy in this post. To begin with, Zip support has been available in .NET since version 1.1, although it was buried in some j# or vb libraries. No one really knew about it. However, .NET 2.0 introduced the System.IO.Compression namespace which contains two implementations: DeflateStream and GZipStream. I believe the 3.0 specific functionality mentioned above (System.IO.Packaging) is specifically related to XPS documents. Perhaps someone can confirm this? As to the quality of any of these algorithms, I am no expert.

  • Re: .NET Zip support

    by Jonathan Allen,

    Your message is awaiting moderation. Thank you for participating in the discussion.

    There is quite a bit of inaccuracy in this post. To begin with, Zip support has been available in .NET since version 1.1, although it was buried in some j# or vb libraries. No one really knew about it.

    Can you give me a reference for that?
    However, .NET 2.0 introduced the System.IO.Compression namespace which contains two implementations: DeflateStream and GZipStream.

    While those can be used for accessing ZIP files, the code needed to do it isn't trivial. Even just getting the file list requires manually parsing the stream to get the header information. There is a sample at MSDN
    I believe the 3.0 specific functionality mentioned above (System.IO.Packaging) is specifically related to XPS documents. Perhaps someone can confirm this?

    Most of the implementation details for that namespace are in the XPS documents.

    Much of the fault with the namespace is that is was described as a general solution in the documentation, and only when you really dig into it do you see that it is just support code for XPS.

  • Re: .NET Zip support

    by Birger Halfmeier,

    Your message is awaiting moderation. Thank you for participating in the discussion.

    There is quite a bit of inaccuracy in this post. To begin with, Zip support has been available in .NET since version 1.1, although it was buried in some j# or vb libraries. No one really knew about it.

    Can you give me a reference for that?

    Have a look at this MSDN Magazine article:
    Using the Zip Classes in the J# Class Libraries to Compress Files and Data with C#

  • Re: .NET Zip support

    by Ted Neward,

    Your message is awaiting moderation. Thank you for participating in the discussion.

    There is quite a bit of inaccuracy in this post. To begin with, Zip support has been available in .NET since version 1.1, although it was buried in some j# or vb libraries. No one really knew about it.

    Can you give me a reference for that?

    It's in J#; look at java.util.zip packages. It's a port of the ZIP support introduced in JDK 1.1, and it may not be any better than what you see in System.IO.Compression, but it is there. There was an MSDN article from some years back that demonstrated how to use those libraries from C#, by the way--a quick Google search (which I'm too lazy to run at the moment) should dig it up.

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

Allowed html: a,b,br,blockquote,i,li,pre,u,ul,p

BT