Small testing of two libraries for working with ZIP archives (C # language)

Not once had to work with zip archives using C #, in my cases - it was downloading the archive with the database, then retrieved the database to disk, if it was not a text file, then dynamically connected the database, otherwise I just read the data I needed. For these purposes, I used ZipStorer , I like it because it very simply does the work assigned to it, weighs a little and is a class that turns out to be in an executable file with a minimum of gestures. Today I decided to try something different, for which I chose two SharpZipLib and DotNetZip libraries , after which I did a little testing in order to understand which of them might be useful to me in the future.

Once I have already mentioned ZipStorer, I’ll show an example of working with it, it’s enough to add the ZipStorer.cs class and write such a method, after which everything will work:

private void Unzip( string fileInput, string saveFile)
{
// Open an existing zip file for reading
using (ZipStorer zip = ZipStorer.Open(fileInput, FileAccess.Read))
{
// Read the central directory collection
List <ZipStorer.ZipFileEntry> dir = zip.ReadCentralDir();

// Look for the desired file
string DBfile = string .Empty;
foreach (ZipStorer.ZipFileEntry entry in dir)
{
if (Path.GetFileName(entry.FilenameInZip).ToLower().IndexOf( ".txt" ) != -1)
{
DBfile = Path.GetFileName(entry.FilenameInZip);
zip.ExtractFile(entry, saveFile + DBfile);

//Now we can connect to database or read file
}
}
}
}

* This source code was highlighted with Source Code Highlighter .

Now we will find out how easy it will be to do such things for SharpZipLib and DotNetZip. Examples of working with SharpZipLib can be found at this link or an official source ; for DotNetZip, examples of work will be immediately archived with the library .

Licenses:
SharpZipLib - GPL
DotNetZip - Ms-PL
')
This time I decided not to translate information from the developer’s sites, I’ll just say that both libraries have Compact Framework support and can be used in web applications. The size of both libraries is about 200 kilobytes, although for DotNetZip I take the “Reduced” version, the full version takes about 400 kilobytes, which is twice as large. Now let's go directly to the test, unlike ZipStorer, we will archive the data, and compare the compression ratio and speed. We will archive with the default level, for DotNetZip, the CompressionLevel property is set to "Default" or "Level6", and the GetLevel () method for SharpZipLib returns "6".

Libraries have classes with the same name, so I connected the spaces in this way:

using test1 = Ionic.Zip;
using test2 = ICSharpCode.SharpZipLib.Zip;


To archive using DotNetZip, the following code is sufficient:

string ZipFileToCreate = Application.StartupPath + @"\test1.zip" ;
string FileToZip = textBoxFile.Text;
using (test1.ZipFile zip = new test1.ZipFile())
{
test1.ZipEntry ze = zip.AddFile(FileToZip);
zip.Save(ZipFileToCreate);
}

* This source code was highlighted with Source Code Highlighter .

For SharpZipLib, you need to write a little more:

string file = textBoxFile.Text;
using (test2.ZipOutputStream s = new test2.ZipOutputStream( File .Create(Application.StartupPath + @"\test2.zip" )))
{
byte [] buffer = new byte [4096];
test2.ZipEntry entry = new test2.ZipEntry(Path.GetFileName(file));
s.PutNextEntry(entry);
using ( FileStream fs = File .OpenRead(file))
{
int sourceBytes;
do
{
sourceBytes = fs.Read(buffer, 0, buffer.Length);
s.Write(buffer, 0, sourceBytes);
} while (sourceBytes > 0);
}
s.Finish();
s.Close();
}

* This source code was highlighted with Source Code Highlighter .

I took both pieces of code from official examples and deleted comments to save space. The results of the work can be found in this table:



The files for the test were generated artificially, so such a beautiful compression result, to see how the situation changes with real data, I download the ru domain list from the registrar site, the archive I downloaded will be the benchmark, its size is 28,043,005 bytes, I will extract the data from the archive and check again two libraries:

Conclusion

Based on the results, the speed characteristics are better with DotNetZip, but at the same time data compression in this library is worse, let me remind you that I did not change the compression ratio and it was 6 in both libraries. I personally visually liked DotNetZip, I’ll dwell on it , the number of those who downloaded this library is more than 100 thousand, it means that my choice is not alone.