private void IncrementalHash(IGrouping <string, DupItem> quickHashGroup)
{
var groups = quickHashGroup.ToArray();
var first = groups.First();
var length = first.Size / BufferSize;
if (length == 0)
{
length = 1;
}
var position = 0;
for (var i = 0; i < length; i++)
{
position += BufferSize;
foreach (var group in groups.GroupBy(g => i == 0 ? string.Empty : g.HashSections[i - 1]))
{
foreach (var groupFile in group)
{
if (groupFile.HashSections == null)
{
groupFile.HashSections = new List <string>();
}
groupFile.HashSections.Add(HashTool.HashFile(groupFile.FileName, position, BufferSize));
}
}
}
foreach (var groupFile in groups)
{
groupFile.FullHash = string.Join(string.Empty, groupFile.HashSections);
}
}
private static void ProgressiveHashSection(long position, DupItem dupItem, int bufferSize, ref long totalReadBytes)
{
if (dupItem.HashSections == null)
{
dupItem.HashSections = new List <string>();
}
dupItem.HashSections.Add(HashTool.HashFile(dupItem.FileName, position, bufferSize, bufferSize, out var readSize));
Interlocked.Add(ref totalReadBytes, readSize);
}
/// <summary>
/// 尝试秒传文件，文件大小必须大于256K
/// </summary>
/// <param name="FilePath">文件路径</param>
/// <param name="FileName">文件名称</param>
/// <param name="UploadPath">上传路径</param>
/// <returns>是否成功</returns>
public static bool RapidUpload(string FilePath, string FileName, string UploadPath)
{
FileInfo info = new FileInfo(FilePath);
if (info.Length < (256 * 1024))
{
return(false);
}
if (!SliceFile(FilePath, FileName))
{
return(false);
}
var MD5 = HashTool.HashFile(FilePath);
var SliceMD5 = HashTool.HashFile($"{Program.config.TempPath}\\{FileName},Tmp");
var CRC32 = HashTool.GetFileCRC32(FilePath);
return(RapidUpload(MD5, SliceMD5, CRC32, info.Length, UploadPath));
}
public List <Duplicate> Find(IEnumerable <string> files, int workers)
{
var result = new List <Duplicate>();
//groups with same file size
var sameSizeGroups = files.Select(f =>
{
var info = new FileInfo(f);
return(new DupItem {
FileName = f, ModifiedTime = info.LastWriteTime, Size = info.Length
});
}).GroupBy(f => f.Size).Where(g => g.Count() > 1);
foreach (var group in sameSizeGroups)
{
foreach (var file in group)
{
if (file.Size > 0)
{
//fast random byte checking
using (var stream = File.OpenRead(file.FileName))
{
var length = stream.Length;
file.Tags = new byte[3];
//first byte
stream.Seek(0, SeekOrigin.Begin);
file.Tags[0] = (byte)stream.ReadByte();
//middle byte, we need it especially for xml like files
if (length > 1)
{
stream.Seek(stream.Length / 2, SeekOrigin.Begin);
file.Tags[1] = (byte)stream.ReadByte();
}
//last byte
if (length > 2)
{
stream.Seek(0, SeekOrigin.End);
file.Tags[2] = (byte)stream.ReadByte();
}
file.QuickHash = HashTool.GetHashText(file.Tags);
}
}
}
//groups with same quick hash value
var sameQuickHashGroups = group.GroupBy(f => f.QuickHash).Where(g => g.Count() > 1);
foreach (var quickHashGroup in sameQuickHashGroups)
{
foreach (var groupFile in quickHashGroup)
{
groupFile.FullHash = HashTool.HashFile(groupFile.FileName);
}
//phew, finally.....
//group by same file hash
var sameFullHashGroups = quickHashGroup.GroupBy(g => g.FullHash).Where(g => g.Count() > 1);
result.AddRange(sameFullHashGroups.Select(fullHashGroup => new Duplicate {
Items = fullHashGroup.Select(f => new FileItem {
FileName = f.FileName, ModifiedTime = f.ModifiedTime, Size = f.Size
})
}));
}
}
return(result);
}
public DupResult Find(IEnumerable <string> files, int workers, int quickHashSize = 3, int bufferSize = 0)
{
var result = new DupResult {
Duplicates = new List <Duplicate>(), FailedToProcessFiles = new List <string>()
};
_workers = workers;
if (_workers <= 0)
{
_workers = 5;
}
if (bufferSize <= 3)
{
bufferSize = DefaultBufferSize;
}
//groups with same file size
var sameSizeGroups = files.Select(f =>
{
var info = new FileInfo(f);
return(new DupItem {
FileName = f, ModifiedTime = info.LastWriteTime, Size = info.Length
});
}).GroupBy(f => f.Size).Where(g => g.Count() > 1);
var mappedSameSizeGroupList = new ConcurrentBag <IGrouping <string, DupItem> >();
Parallel.ForEach(MapFileSizeGroups(sameSizeGroups), mappedSameSizeGroups =>
{
foreach (var group in mappedSameSizeGroups)
{
foreach (var file in group)
{
if (file.Size > 0)
{
//fast random byte checking
try
{
using (var stream = File.Open(file.FileName, FileMode.Open, FileAccess.Read, FileShare.Read))
{
var length = stream.Length;
file.Tags = new byte[3];
//first byte
stream.Seek(0, SeekOrigin.Begin);
file.Tags[0] = (byte)stream.ReadByte();
//middle byte, we need it especially for xml like files
if (length > 1)
{
stream.Seek(stream.Length / 2, SeekOrigin.Begin);
file.Tags[1] = (byte)stream.ReadByte();
}
//last byte
if (length > 2)
{
stream.Seek(0, SeekOrigin.End);
file.Tags[2] = (byte)stream.ReadByte();
}
file.QuickHash = HashTool.GetHashText(file.Tags);
}
}
catch (Exception)
{
file.Status = CompareStatus.Failed;
result.FailedToProcessFiles.Add(file.FileName);
}
}
}
//groups with same quick hash value
var sameQuickHashGroups = group.Where(f => f.Status != CompareStatus.Failed).GroupBy(f => f.QuickHash).Where(g => g.Count() > 1);
foreach (var sameQuickHashGroup in sameQuickHashGroups)
{
mappedSameSizeGroupList.Add(sameQuickHashGroup);
}
}
});
Parallel.ForEach(MapFileHashGroups(mappedSameSizeGroupList), mappedSameHashGroups =>
{
foreach (var quickHashGroup in mappedSameHashGroups)
{
foreach (var groupFile in quickHashGroup)
{
try
{
groupFile.FullHash = HashTool.HashFile(groupFile.FileName, bufferSize);
}
catch (Exception)
{
result.FailedToProcessFiles.Add(groupFile.FileName);
}
}
//phew, finally.....
//group by same file hash
var sameFullHashGroups = quickHashGroup.GroupBy(g => g.FullHash).Where(g => g.Count() > 1);
result.Duplicates.AddRange(sameFullHashGroups.Select(fullHashGroup => new Duplicate {
Items = fullHashGroup.Select(f => new FileItem {
FileName = f.FileName, ModifiedTime = f.ModifiedTime, Size = f.Size
})
}));
}
});
return(result);
}
