MSDN documentation unclear ?

[mesajflaviu]

I am trying to read directories from an EXFAT USB drive. And for this task, I have found here a good documentation: https://docs.microsoft.com/en-us/win...-specification

But seem that is unclear somehow, at least to me.

I have successfully read - Main and Backup Boot Sector Structure Table (https://docs.microsoft.com/en-us/win...ctor-structure), value by value.

And then, I have read - Cluster Heap Structure Table (https://docs.microsoft.com/en-us/win...heap-structure)

	const int nSize = dwSectorSize * 1 << nSectorsPerClusterShift;
	const int nCount = nClusterCount + 1;
	LONGLONG llPos = nClusterHeapOffset;
	LONGLONG nLimit = nClusterHeapOffset + ((nClusterCount - 1) * nSize);

	for (int i = 0; i < nCount; ++i)
	{
		file.Seek(llPos, CFile::begin);
		arrByte.RemoveAll();
		arrByte.SetSize(nSize);
		UINT nRet = file.Read(arrByte.GetData(), arrByte.GetSize());

		llPos += nSize;
		if (llPos > nLimit)
			break;
	}

and for every cluster from cluster heap, I retrieved a CByteArray of 32768 bytes. Fine by now. Now, there is another part: Directory Structure Table (https://docs.microsoft.com/en-us/win...tory-structure)

Here I am stuck. I guess I can read the directory (or files) from this cluster, one by one. Theoretically, I should read every cluster from 32 to 32 bytes, according Directory Structure Table, which gave me 1024 (32768 / 32). But in this cluster, I don't know how to consider those bytes in order to have a directory (or file). More over, I don't know how to compute N (for DirectoryEntry[N–1]).

The documentation says: N, the number of DirectoryEntry fields, is the size, in bytes, of the cluster chain which contains the given directory, divided by the size of a DirectoryEntry field, 32 bytes.

From documentation: A cluster chain is a series of clusters which provides space for recording the contents of files, directories, and other file system structures.

Now, how is: "cluster chain which contains the given directory" ? Can you help me a little bit ? I am stuck here for a while, and I don't know how to move on ...

[2kaud]

Why can't you just use the assigned drive letter when the USB drive is plugged in and the the usual FindFirst/FindNext WIN32 API's or C++17 directory_iterator()?

[mesajflaviu]

Why can't you just use the assigned drive letter when the USB drive is plugged in and the the usual FindFirst/FindNext WIN32 API's or C++17 directory_iterator()?

In the end, I intend to read even the deleted files/directory, and FindFirst/FindNext doesn't have such feature, as far as I know.

[2kaud]

In the end, I intend to read even the deleted files/directory, and FindFirst/FindNext doesn't have such feature, as far as I know.

Ah. No, FindFirst() et al doesn't provide info on deleted items. Then the example in my post #3 could be useful.

[mesajflaviu]

Ah. No, FindFirst() et al doesn't provide info on deleted items. Then the example in my post #3 could be useful.

Yes, that's right, look very promising that post ... thank you !

[2kaud]

If you really need to do this the hard way, have a look at https://rshullic.wordpress.com/2010/...ory-structure/ which contains sample c code to do this.