DotNetZip编码源码

Mon, 03 Apr 2006 10:26 Zip Library --------------------------------- The Microsoft .NET Framework {v2.0 v3.0 v3.5} includes new base class libraries supporting compression within streams - both the Deflate and Gzip formats are supported. But the new-for-.NET2.0 System.IO.Compression namespace provides streaming compression only - useful for communicating between cooperating parties but not directly useful for creating compressed archives, like .zip files. The built-in compression library does not know how to format zip archive headers and so on. This is a simple class library that augments the System.IO.Compression.DeflateStream class, to provide handling for Zip files. Using this library, you can write .NET applications that read and write zip-format files. The Zip Format --------------------------------- The zip format is described by PKWare, at http://www.pkware.com/business_and_developers/developer/popups/appnote.txt Every valid zipfile conforms to this specification. For example, the spec says that for each compressed file contained in the zip archive, the zipfile contains a byte array of compressed data. (The byte array is something the DeflateStream class can produce directly.) But the zipfile also contains header and "directory" information - you might call this "metadata". In other words, the zipfile must contain a list of all the compressed files in the archive. The zipfile also contains CRC checksums, and can also contain comments, and other optional attributes for each file. These are things the DeflateStream class, included in the .NET Framework Class Library, does not read or write. This Class Library --------------------------------- The library included here depends on the DeflateStream class, and extends it to support reading and writing of the metadata - the header, CRC, and other optional data - defined or required by the zip format spec. The key object in the class library is the ZipFile class. The key methods on it: - AddItem - adds a file or a directory to a zip archive - AddDirectory - adds a directory to a zip archive - AddFile - adds a file to a zip archive - Extract - extract a single element from a zip file - Read - static methods to read in an existing zipfile, for later extraction - Save - save a zipfile to disk There is also a supporting class, called ZipEntry. Applications can enumerate the entries in a ZipFile, via ZipEntry. There are other supporting classes as well. Typically apps do not directly interact with these other classes. Using the Class Library --------------------------------- Check the examples included in this package for simple apps that show how to read and write zip files. The simplest way to create a zipfile looks like this: using(ZipFile zip= new ZipFile(NameOfZipFileTocreate)) { zip.AddFile(filename); zip.Save(); } The simplest way to Extract all the entries from a zipfile looks like this: using (ZipFile zip = ZipFile.Read(NameOfExistingZipFile)) { zip.ExtractAll(args[1]); } There are a number of other options for using the class library. For example, you can read zip archives from streams, or you can create (write) zip archives to streams. Check the doc for complete information. About Directory Paths --------------------------------- One important note: the ZipFile.AddXxx methods add the file or directory you specify, including the directory. In other words, logic like this: zip.AddFile("c:\\a\\b\\c\\Hello.doc"); zip.Save(); ...will produce a zip archive that contains a single file, which is stored with the relative directory information. When you extract that file from the zip, either using this Zip library or winzip or the built-in zip support in Windows, or some other package, all those directories will be created, and the file will be written into that directory hierarchy. If you don't want that directory information in your archive, then you need to either (a) copy the file or files to be compressed into the local directory (b) change the applications current directory to where the file resides, before adding it to the zipfile. The latter involves a call to System.IO.Directory.SetCurrentDirectory(), before you call ZipFile.AddXxx(). See the doc: http://msdn2.microsoft.com/en-us/library/system.io.directory.setcurrentdirectory.aspx About the Help file -------------------------------- The .chm file contains help generated from the code. In some cases, Upon opening the .chm file for DotNetZipLib, the help items tree loads, but the contents are empty. You may see an Error: This program cannot display the webpage. If this happens, it's probable that you encounter problem with Windows protection of files downloaded from less trusted location. Within Windows Explorer, right-click on the CHM file, select properties, and Unblock it (button in lower part of properties window). License -------- This software is released under the Microsoft Permissive License of OCtober 2006. See the License.txt file for details. About Other Intellectual Property --------------------------------- I am no lawyer, but before using this library in your app, it may be worth contacting PKWare for clarification on rights and licensing. The specification for the zip format includes a paragraph that reads: PKWARE is committed to the interoperability and advancement of the .ZIP format. PKWARE offers a free license for certain technological aspects described above under certain restrictions and conditions. However, the use or implementation in a product of certain technological aspects set forth in the current APPNOTE, including those with regard to strong encryption or patching, requires a license from PKWARE. Please contact PKWARE with regard to acquiring a license. Contact pkware at: zipformat@pkware.com This example also uses a CRC utility class, in modified form, that was published on the internet without an explicit license. You can find the original CRC class at: http://www.vbaccelerator.com/home/net/code/libraries/CRC32/Crc32_zip_CRC32_CRC32_cs.asp Pre-requisites --------------------------------- to run: .NET Framework 2.0 or later to build: .NET Framework 2.0 SDK or later or Visual Studio 2005 or later Building --------------------------------- To build this example, 1. extract the contents of this zip into a new directory. 2. be sure the .NET 2.0 SDK and .NET 2.0 runtime directories are on your path. These are typically C:\Program Files\Microsoft.NET\SDK\v2.0\bin and c:\WINDOWS\Microsoft.NET\Framework\v2.0.50727 3. open a CMD prompt and CD to the zip\Library directory. 4. nmake 5. To build the examples, cd ..\Examples and type nmake again Building the Help File -------------------------------------------- If you want to build the helpfile, you need the SandCastle helpfile builder. Use the DotNetZip.shfb file with SandCastle. Limitations --------------------------------- There are numerous limitations to this library: it does not support encryption, file comments, or double-byte chars in filenames. it does not support file lengths greater than 0xffffffff. it does not support "multi-disk archives." it does not do varying compression levels. it can actually expand the size of previously compressed data, such as JPG files. there is no GUI tool and, I'm sure, many others But it is a good basic library for reading and writing zipfiles in .NET applications.. And yes, the zipfile that this example is shipped in, was produced by this example library. See Also ----------------------------