mirror of
https://github.com/adamhathcock/sharpcompress.git
synced 2026-07-08 18:16:30 +00:00
Zip: fix Zip64 streaming reader corrupting entry size/CRC and failing on non-seekable streams
The Zip64 branch of the streaming header reader assumed a data descriptor always follows a >=4GB entry. When the entry instead has back-patched sizes (no descriptor), the following header was parsed as descriptor fields, overwriting the entry's correct size/CRC with central-directory 0xFFFFFFFF sentinels and, on non-seekable async streams, leaving the reader misaligned so the next entry threw. Now it detects a header signature after the entry data and leaves the already-correct metadata untouched, parsing that header normally. The fix has been applied to both the sync and async readers.
This commit is contained in:
Pat Hartl
@@ -189,64 +189,77 @@ internal sealed partial class StreamingZipHeaderFactory
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // version
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // flags
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // compressionMethod
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // lastModifiedDate
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // lastModifiedTime
|
||||
|
||||
var crc = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
|
||||
if (crc == POST_DATA_DESCRIPTOR)
|
||||
// A Zip64 entry that does not use a post-data descriptor stores its real CRC
|
||||
// and sizes in the local header's Zip64 extra field, so a header signature
|
||||
// (the next local entry, or the central directory) follows the data directly.
|
||||
// In that case the entry's metadata is already correct and must not be
|
||||
// overwritten with bytes read from the following header. We have only consumed
|
||||
// the 4-byte signature, so fall through and parse this header normally. Because
|
||||
// no seek-back is required here, this also works for non-seekable streams.
|
||||
if (headerBytes == 0x04034b50 || headerBytes == 0x02014b50)
|
||||
{
|
||||
crc = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
}
|
||||
lastEntryHeader.Crc = crc;
|
||||
lastEntryHeader.IsCrcAvailable = true;
|
||||
|
||||
// The DataDescriptor can be either 64bit or 32bit
|
||||
var compressedSize = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
var uncompressedSize = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
|
||||
// Check if we have header or 64bit DataDescriptor
|
||||
var testHeader = !(headerBytes == 0x04034b50 || headerBytes == 0x02014b50);
|
||||
|
||||
var test64Bit = ((long)uncompressedSize << 32) | compressedSize;
|
||||
if (test64Bit == lastEntryHeader.CompressedSize && testHeader)
|
||||
{
|
||||
lastEntryHeader.UncompressedSize =
|
||||
(
|
||||
(long)
|
||||
await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false) << 32
|
||||
) | headerBytes;
|
||||
headerBytes = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
if (pos.HasValue)
|
||||
{
|
||||
lastEntryHeader.DataStartPosition =
|
||||
pos - lastEntryHeader.CompressedSize;
|
||||
}
|
||||
}
|
||||
else
|
||||
{
|
||||
lastEntryHeader.UncompressedSize = uncompressedSize;
|
||||
}
|
||||
// A data descriptor follows. Recover the CRC and sizes from it; the
|
||||
// descriptor can carry either 32-bit or 64-bit sizes.
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // version
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // flags
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // compressionMethod
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // lastModifiedDate
|
||||
_ = await _reader.ReadUInt16Async(_cancellationToken).ConfigureAwait(false); // lastModifiedTime
|
||||
|
||||
if (pos.HasValue)
|
||||
{
|
||||
lastEntryHeader.DataStartPosition = pos - lastEntryHeader.CompressedSize;
|
||||
var crc = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
|
||||
// For SeekableSharpCompressStream, seek back to just after the local header signature.
|
||||
// Plain SharpCompressStream cannot seek to arbitrary positions, so we skip this.
|
||||
// 4 = First 4 bytes of the entry header (i.e. 50 4B 03 04)
|
||||
if (_sharpCompressStream is SeekableSharpCompressStream)
|
||||
if (crc == POST_DATA_DESCRIPTOR)
|
||||
{
|
||||
crc = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
}
|
||||
lastEntryHeader.Crc = crc;
|
||||
lastEntryHeader.IsCrcAvailable = true;
|
||||
|
||||
// The DataDescriptor can be either 64bit or 32bit
|
||||
var compressedSize = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
var uncompressedSize = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
|
||||
var test64Bit = ((long)uncompressedSize << 32) | compressedSize;
|
||||
if (test64Bit == lastEntryHeader.CompressedSize)
|
||||
{
|
||||
lastEntryHeader.UncompressedSize =
|
||||
(
|
||||
(long)
|
||||
await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false) << 32
|
||||
) | headerBytes;
|
||||
headerBytes = await _reader
|
||||
.ReadUInt32Async(_cancellationToken)
|
||||
.ConfigureAwait(false);
|
||||
}
|
||||
else
|
||||
{
|
||||
lastEntryHeader.UncompressedSize = uncompressedSize;
|
||||
}
|
||||
|
||||
if (pos.HasValue)
|
||||
{
|
||||
lastEntryHeader.DataStartPosition =
|
||||
pos - lastEntryHeader.CompressedSize;
|
||||
|
||||
// 4 = First 4 bytes of the entry header (i.e. 50 4B 03 04)
|
||||
_sharpCompressStream.Position = pos.Value + 4;
|
||||
}
|
||||
}
|
||||
|
||||
@@ -109,46 +109,63 @@ internal partial class StreamingZipHeaderFactory : ZipHeaderFactory
|
||||
|
||||
headerBytes = reader.ReadUInt32();
|
||||
|
||||
var version = reader.ReadUInt16();
|
||||
var flags = (HeaderFlags)reader.ReadUInt16();
|
||||
var compressionMethod = (ZipCompressionMethod)reader.ReadUInt16();
|
||||
var lastModifiedDate = reader.ReadUInt16();
|
||||
var lastModifiedTime = reader.ReadUInt16();
|
||||
|
||||
var crc = reader.ReadUInt32();
|
||||
|
||||
if (crc == POST_DATA_DESCRIPTOR)
|
||||
// A Zip64 entry that does not use a post-data descriptor stores its real CRC
|
||||
// and sizes in the local header's Zip64 extra field, so a header signature
|
||||
// (the next local entry, or the central directory) follows the data directly.
|
||||
// In that case the entry's metadata is already correct and must not be
|
||||
// overwritten with bytes read from the following header. We have only consumed
|
||||
// the 4-byte signature, so fall through and parse this header normally.
|
||||
if (headerBytes == 0x04034b50 || headerBytes == 0x02014b50)
|
||||
{
|
||||
crc = reader.ReadUInt32();
|
||||
}
|
||||
_lastEntryHeader.Crc = crc;
|
||||
_lastEntryHeader.IsCrcAvailable = true;
|
||||
|
||||
// The DataDescriptor can be either 64bit or 32bit
|
||||
var compressed_size = reader.ReadUInt32();
|
||||
var uncompressed_size = reader.ReadUInt32();
|
||||
|
||||
// Check if we have header or 64bit DataDescriptor
|
||||
var test_header = !(headerBytes == 0x04034b50 || headerBytes == 0x02014b50);
|
||||
|
||||
var test_64bit = ((long)uncompressed_size << 32) | compressed_size;
|
||||
if (test_64bit == _lastEntryHeader.CompressedSize && test_header)
|
||||
{
|
||||
_lastEntryHeader.UncompressedSize =
|
||||
((long)reader.ReadUInt32() << 32) | headerBytes;
|
||||
headerBytes = reader.ReadUInt32();
|
||||
if (pos.HasValue)
|
||||
{
|
||||
_lastEntryHeader.DataStartPosition =
|
||||
pos - _lastEntryHeader.CompressedSize;
|
||||
}
|
||||
}
|
||||
else
|
||||
{
|
||||
_lastEntryHeader.UncompressedSize = uncompressed_size;
|
||||
}
|
||||
// A data descriptor follows. Recover the CRC and sizes from it; the
|
||||
// descriptor can carry either 32-bit or 64-bit sizes.
|
||||
_ = reader.ReadUInt16(); // version
|
||||
_ = reader.ReadUInt16(); // flags
|
||||
_ = reader.ReadUInt16(); // compressionMethod
|
||||
_ = reader.ReadUInt16(); // lastModifiedDate
|
||||
_ = reader.ReadUInt16(); // lastModifiedTime
|
||||
|
||||
if (pos.HasValue)
|
||||
{
|
||||
_lastEntryHeader.DataStartPosition = pos - _lastEntryHeader.CompressedSize;
|
||||
var crc = reader.ReadUInt32();
|
||||
|
||||
// 4 = First 4 bytes of the entry header (i.e. 50 4B 03 04)
|
||||
sharpCompressStream.Position = pos.Value + 4;
|
||||
if (crc == POST_DATA_DESCRIPTOR)
|
||||
{
|
||||
crc = reader.ReadUInt32();
|
||||
}
|
||||
_lastEntryHeader.Crc = crc;
|
||||
_lastEntryHeader.IsCrcAvailable = true;
|
||||
|
||||
// The DataDescriptor can be either 64bit or 32bit
|
||||
var compressed_size = reader.ReadUInt32();
|
||||
var uncompressed_size = reader.ReadUInt32();
|
||||
|
||||
var test_64bit = ((long)uncompressed_size << 32) | compressed_size;
|
||||
if (test_64bit == _lastEntryHeader.CompressedSize)
|
||||
{
|
||||
_lastEntryHeader.UncompressedSize =
|
||||
((long)reader.ReadUInt32() << 32) | headerBytes;
|
||||
headerBytes = reader.ReadUInt32();
|
||||
}
|
||||
else
|
||||
{
|
||||
_lastEntryHeader.UncompressedSize = uncompressed_size;
|
||||
}
|
||||
|
||||
if (pos.HasValue)
|
||||
{
|
||||
_lastEntryHeader.DataStartPosition =
|
||||
pos - _lastEntryHeader.CompressedSize;
|
||||
|
||||
// 4 = First 4 bytes of the entry header (i.e. 50 4B 03 04)
|
||||
sharpCompressStream.Position = pos.Value + 4;
|
||||
}
|
||||
}
|
||||
}
|
||||
else
|
||||
|
||||
Reference in New Issue
Block a user