We Keep Coding

Invalid zip file reading from a Stream in C#

I have the following code:
private static byte[] ConverterStringToByte(Stream body)
{
string fileName = "data_" + DateTime.Now.ToString("yyyyMMddhhmmss") + ".zip";
// Take out the bytes from the memory stream and safely close the stream
using (var ms = new MemoryStream())
{
body.CopyTo(ms);
using (var zipArchive = new ZipArchive(ms, ZipArchiveMode.Create, false))
{
var zipEntry = zipArchive.CreateEntry(fileName, CompressionLevel.Optimal);
using (BinaryWriter writer = new BinaryWriter(zipEntry.Open()))
{
ms.Position = 0;
writer.Write(ms.ToArray());
}
}
return ms.ToArray();
}
}
I am downloading the file successfully, however I'm getting
invalid file
when trying to open

I think it should be something like this. Not sure about fileName though, because it's the name of the file being put into archive, so I don't think it should have *.zip extension. Unless you are creating a zip of zips.
static byte[] ConverterStringToByte(Stream body)
{
string fileName = #"data_" + DateTime.Now.ToString("yyyyMMddhhmmss") + ".zip";
using (var ms = new MemoryStream())
{
using (var zipArchive = new ZipArchive(ms, ZipArchiveMode.Create, false))
{
var zipEntry = zipArchive.CreateEntry(fileName, CompressionLevel.Optimal);
using (var destStream = zipEntry.Open())
{
body.CopyTo(destStream);
}
}
return ms.ToArray();
}
}

Copy files from one Zip file to another

I am copying files from one zip file to another in certain circumstances. I am wondering if there is a better way to do it than what I came up with:
using FileStream sourceFileStream = new FileStream(source.FileName, FileMode.Open);
using FileStream targetFileStream = new FileStream(target.FileName, FileMode.Open, FileAccess.ReadWrite);
using ZipArchive sourceZip = new ZipArchive(sourceFileStream, ZipArchiveMode.Read);
using ZipArchive targetZip = new ZipArchive(targetFileStream, ZipArchiveMode.Update);
ZipArchiveEntry sourceEntry = sourceZip.GetEntry(filePathInArchive);
if (sourceEntry == null)
return;
ZipArchiveEntry targetEntry = targetZip.GetEntry(filePathInArchive);
if (targetEntry != null)
targetEntry.Delete();
targetZip.CreateEntry(filePathInArchive);
targetEntry = targetZip.GetEntry(filePathInArchive);
if (targetEntry != null)
{
Stream writer = targetEntry.Open();
Stream reader = sourceEntry.Open();
int b;
do
{
b = reader.ReadByte();
writer.WriteByte((byte)b);
} while (b != -1);
writer.Close();
reader.Close();
}
Tips and suggestions would be appreciated.

You can iterate each entry from source archive with opening its streams and using Stream.CopyTo write source entry content to target entry.
From C# 8.0 it looks compact and works fine:
static void CopyZipEntries(string sourceZipFile, string targetZipFile)
{
using FileStream sourceFS = new FileStream(sourceZipFile, FileMode.Open);
using FileStream targetFS = new FileStream(targetZipFile, FileMode.Open);
using ZipArchive sourceZIP = new ZipArchive(sourceFS, ZipArchiveMode.Read, false, Encoding.GetEncoding(1251));
using ZipArchive targetZIP = new ZipArchive(targetFS, ZipArchiveMode.Update, false, Encoding.GetEncoding(1251));
foreach (ZipArchiveEntry sourceEntry in sourceZIP.Entries)
{
// 'is' is replacement for 'null' check
if (targetZIP.GetEntry(sourceEntry.FullName) is ZipArchiveEntry existingTargetEntry)
existingTargetEntry.Delete();
using (Stream targetEntryStream = targetZIP.CreateEntry(sourceEntry.FullName).Open())
{
sourceEntry.Open().CopyTo(targetEntryStream);
}
}
}
With earlier than C# 8.0 versions it works fine too, but more braces needed:
static void CopyZipEntries(string sourceZipFile, string targetZipFile)
{
using (FileStream sourceFS = new FileStream(sourceZipFile, FileMode.Open))
{
using (FileStream targetFS = new FileStream(targetZipFile, FileMode.Open))
{
using (ZipArchive sourceZIP = new ZipArchive(sourceFS, ZipArchiveMode.Read, false, Encoding.GetEncoding(1251)))
{
using (ZipArchive targetZIP = new ZipArchive(targetFS, ZipArchiveMode.Update, false, Encoding.GetEncoding(1251)))
{
foreach (ZipArchiveEntry sourceEntry in sourceZIP.Entries)
{
if (targetZIP.GetEntry(sourceEntry.FullName) is ZipArchiveEntry existingTargetEntry)
{
existingTargetEntry.Delete();
}
using (Stream target = targetZIP.CreateEntry(sourceEntry.FullName).Open())
{
sourceEntry.Open().CopyTo(target);
}
}
}
}
}
}
}
For single specified file copy just replace bottom part from foreach loop to if condition:
static void CopyZipEntry(string fileName, string sourceZipFile, string targetZipFile)
{
// ...
// It means specified file exists in source ZIP-archive
// and we can copy it to target ZIP-archive
if (sourceZIP.GetEntry(fileName) is ZipArchiveEntry sourceEntry)
{
if (targetZIP.GetEntry(sourceEntry.FullName) is ZipArchiveEntry existingTargetEntry)
existingTargetEntry.Delete();
using (Stream targetEntryStream = targetZIP.CreateEntry(sourceEntry.FullName).Open())
{
sourceEntry.Open().CopyTo(targetEntryStream);
}
}
else
MessageBox.Show("Source ZIP-archive doesn't contains file " + fileName);
}

Thanks to the input so far, I cleaned up and improved the code. I think this looks cleaner and more reliable.
//Making sure files exist etc before this part...
string filePathInArchive = source.GetFilePath(fileId);
using FileStream sourceFileStream = new FileStream(source.FileName, FileMode.Open);
using FileStream targetFileStream = new FileStream(target.FileName, FileMode.Open, FileAccess.ReadWrite);
using ZipArchive sourceZip = new ZipArchive(sourceFileStream, ZipArchiveMode.Read, false );
using ZipArchive targetZip = new ZipArchive(targetFileStream, ZipArchiveMode.Update, false);
ZipArchiveEntry sourceEntry = sourceZip.GetEntry(filePathInArchive);
if (sourceEntry != null)
{
if (targetZip.GetEntry(filePathInArchive) is { } existingTargetEntry)
{
existingTargetEntry.Delete();
}
using var targetEntryStream = targetZip.CreateEntry(sourceEntry.FullName).Open();
sourceEntry.Open().CopyTo(targetEntryStream);
}

Download multiple files .NET CORE Web API as ArchiveZip return root directory

I make a private class to get the bytes of the file and then return it into a zip file. Basically, I just want to pick specific files inside my folder and then zip it then download it. Here's my class:
private FileResult DownloadMultipleFiles(List<byte[]> byteArrayList)
{
var zipName = $"archive-EvidenceFiles-{DateTime.Now.ToString("yyyy_MM_dd-HH_mm_ss")}.zip";
using (MemoryStream ms = new MemoryStream())
{
using (var archive = new ZipArchive(ms, ZipArchiveMode.Create, true))
{
foreach (var file in byteArrayList)
{
string fPath = Encoding.ASCII.GetString(file);
var entry = archive.CreateEntry(fPath, CompressionLevel.Fastest);
using (var zipStream = entry.Open())
{
zipStream.Write(file, 0, file.Length);
}
}
}
return File(ms.ToArray(), "application/zip", zipName);
}
}
And then, here's my controller:
[HttpGet("GetBundleFiles/{rhaId}")]
public async Task<IActionResult> GetBundleFiles(string rhaId)
{
List<byte[]> filesPath = new List<byte[]>();
var results = await _rhaFileEvidence.GetByRhaID(rhaId);
var files = results.ToList();
if (files.Count == 0)
return Ok(new { status = "null", message = "Empty data" });
files.ForEach(file =>
{
var fPath = file.FilePath;
byte[] bytes = Encoding.ASCII.GetBytes(fPath);
filesPath.Add(bytes);
});
return DownloadMultipleFiles(filesPath);
}
The controller works well, I can download the zip but when I open it, I can't get the files instead I get the root directory of the project I saved, like D:. I think I make mistake when making the memory stream or something, is there any suggestion how can I fix this? (Paste some solution code in the answer please)

you are getting the directory root because when you use archive.CreateEntry you are passing the file full path in parameter, you should be using only the file name
var entry = archive.CreateEntry(System.IO.Path.GetFileName(fPath), CompressionLevel.Fastest);
a second issue is that you actually saving the file path to your files not the content of the original file. you can update your DownloadMultipleFiles like this
private FileResult DownloadMultipleFiles(List<byte[]> byteArrayList)
{
var zipName = $"archive-EvidenceFiles-{DateTime.Now.ToString("yyyy_MM_dd-HH_mm_ss")}.zip";
using (MemoryStream ms = new MemoryStream())
{
using (var archive = new ZipArchive(ms, ZipArchiveMode.Create, true))
{
foreach (var file in byteArrayList)
{
string fPath = Encoding.ASCII.GetString(file);
var entry = archive.CreateEntry(System.IO.Path.GetFileName(fPath), CompressionLevel.Fastest);
using (var zipStream = entry.Open())
{
var bytes = System.IO.File.ReadAllBytes(fPath);
zipStream.Write(bytes, 0, bytes.Length);
}
}
}
return File(ms.ToArray(), "application/zip", zipName);
}
}

C# ZipArchive - How to nest internal .zip files without writing to disk

I need to create a zip file in memory, then send the zip file to the client. However, there are cases where the created zip file will need to contain other zip files that were also generated in memory. For instance, the file structure might look like this:
SendToClient.zip
InnerZip1.zip
File1.xml
File2.xml
InnerZip2.zip
File3.xml
File4.xml
I've been attempting to use the System.IO.Compression.ZipArchive library. I cannot use the System.IO.Compression.ZipFile library because my project's version of .NET is not compatible with it.
Here's an example of what I've tried.
public Stream GetMemoryStream() {
var memoryStream = new MemoryStream();
string fileContents = "Lorem ipsum dolor sit amet";
string entryName = "Lorem.txt";
string innerZipName = "InnerZip.zip";
using (var archive = new ZipArchive(memoryStream, ZipArchiveMode.Create, true)) {
ZipArchiveEntry entry = archive.CreateEntry(Path.Combine(innerZipName, entryName), CompressionLevel.Optimal);
using (var writer = new StreamWriter(entry.Open())) {
writer.Write(fileContents);
}
}
return memoryStream
}
However, this just puts Lorem.txt in a folder called "Inner.zip" (instead of in an actual zip file).
I can create an empty inner zip file if I create an entry called "Inner.zip" without writing to it. I can't add anything to it, though, and writing to an entry called "Inner.zip\Lorem.txt" afterward just creates a folder again (alongside the identically named empty .zip file).
I've also tried creating a separate archive, serializing it with a memory stream, then writing that to the original archive as a .zip.
public Stream CreateOuterZip() {
var memoryStream = new MemoryStream();
using (var archive = new ZipArchive(memoryStream, ZipArchiveMode.Create, true)) {
ZipArchiveEntry entry = archive.CreateEntry("Outer.zip", CompressionLevel.NoCompression);
using (var writer = new BinaryWriter(entry.Open())) {
writer.Write(GetMemoryStream().ToArray());
}
}
return memoryStream;
}
This just creates an invalid .zip file that windows doesn't know how to open, though.
Thanks in advance!

So I created a FileStream instead of a MemoryStream so the code can be tested easier
public static Stream CreateOuterZip()
{
string fileContents = "Lorem ipsum dolor sit amet";
// Final zip file
var fs = new FileStream(
Path.Combine(AppDomain.CurrentDomain.BaseDirectory, "SendToClient.zip"), FileMode.OpenOrCreate);
// Create inner zip 1
var innerZip1 = new MemoryStream();
using (var archive = new ZipArchive(innerZip1, ZipArchiveMode.Create, true))
{
var file1 = archive.CreateEntry("File1.xml");
using (var writer = new BinaryWriter(file1.Open()))
{
writer.Write(fileContents); // Change fileContents to real XML content
}
var file2 = archive.CreateEntry("File2.xml");
using (var writer = new BinaryWriter(file2.Open()))
{
writer.Write(fileContents); // Change fileContents to real XML content
}
}
// Create inner zip 2
var innerZip2 = new MemoryStream();
using (var archive = new ZipArchive(innerZip2, ZipArchiveMode.Create, true))
{
var file3 = archive.CreateEntry("File3.xml");
using (var writer = new BinaryWriter(file3.Open()))
{
writer.Write(fileContents); // Change fileContents to real XML content
}
var file4 = archive.CreateEntry("File4.xml");
using (var writer = new BinaryWriter(file4.Open()))
{
writer.Write(fileContents); // Change fileContents to real XML content
}
}
using (var archive = new ZipArchive(fs, ZipArchiveMode.Create, true))
{
// Create inner zip 1
var innerZipEntry = archive.CreateEntry("InnerZip1.zip");
innerZip1.Position = 0;
using (var s = innerZipEntry.Open())
{
innerZip1.WriteTo(s);
}
// Create inner zip 2
var innerZipEntry2 = archive.CreateEntry("InnerZip2.zip");
innerZip2.Position = 0;
using (var s = innerZipEntry2.Open())
{
innerZip2.WriteTo(s);
}
}
fs.Close();
return fs; // The file is written, can probably just close this
}
You can obviously modify this method to return a MemoryStream, or change the method to Void to just have the zip file written out to disk

You should create ZipArchive for internal zip file also. Write it to stream (memorystream). And after write this stream as general stream into main zip.
static Stream Inner() {
var memoryStream = new MemoryStream();
using (var archive = new ZipArchive(memoryStream, ZipArchiveMode.Create, true)) {
var demoFile = archive.CreateEntry("foo2.txt");
using (var entryStream = demoFile.Open())
using (var streamWriter = new StreamWriter(entryStream)) {
streamWriter.Write("Bar2!");
}
}
return memoryStream;
}
static void Main(string[] args) {
using (var memoryStream = new MemoryStream()) {
using (var archive = new ZipArchive(memoryStream, ZipArchiveMode.Create, true)) {
var demoFile = archive.CreateEntry("foo.txt");
using (var entryStream = demoFile.Open())
using (var streamWriter = new StreamWriter(entryStream)) {
streamWriter.Write("Bar!");
}
var zip = archive.CreateEntry("inner.zip");
using (var entryStream = zip.Open()) {
var inner = Inner();
inner.Seek(0, SeekOrigin.Begin);
inner.CopyTo(entryStream);
}
}
using (var fileStream = new FileStream(#"d:\test.zip", FileMode.Create)) {
memoryStream.Seek(0, SeekOrigin.Begin);
memoryStream.CopyTo(fileStream);
}
}
Thanks to this answer.

Zip package with Unicode chars in file name can't be uncompressed with System.IO.Packaging.Package?

Have some troubles with file package manipulation.
Here is short description of issue:
If a package contains a file with Unicode characters in it, compressing of it works well(can open package and be uncompressed it in windows) but if try to use this Package.Open(.., FileMode.Open), it results in error "File contains corrupted data".
Question: What am i doing wrong? Can't figure this out. Please advise.
Here is code for compressing and uncompromising binary data in-memory that i use:
public virtual byte[] UnZipFile(byte[] data, string packageFileName)
{
using (var inputStream = new MemoryStream(data))
{
using (var package = Package.Open(inputStream, FileMode.Open))
{
var fileUri = PackUriHelper.CreatePartUri(new Uri(#"/" + packageFileName, UriKind.Relative));
using (var outputStream = new MemoryStream())
{
using (var partStream = package.GetPart(fileUri).GetStream())
{
partStream.CopyTo(outputStream);
}
return outputStream.ToArray();
}
}
}
}
public virtual byte[] ZipFile(byte[] data, string packageFileName)
{
using (var outMs = new MemoryStream())
{
using (var package = Package.Open(outMs, FileMode.Create))
{
var fileUri = PackUriHelper.CreatePartUri(new Uri(#"/" + packageFileName, UriKind.Relative));
var contentType = #"data/" + packageFileName.Substring(packageFileName.LastIndexOf(".", StringComparison.Ordinal) + 1);
using (var zipStream = package.CreatePart(fileUri, contentType, CompressionOption.Normal).GetStream())
{
using (var ms = new MemoryStream(data))
{
ms.CopyTo(zipStream);
}
}
}
return outMs.ToArray();
}
}

solved by using newer lib, System.IO.Compression, thanks for pointing out this #Hans Passant