Compare binaryfiles with AHK

[Albireo]

It's easy to retrieve the size, datetime stamp of a file or folder, and then compare the result with another file.
But I don't found an instruction like this ..: FileCompare ...

Is it difficult to compare the binary content of two files with AHK?
Will the solution be slow?

[Guest]

Calculate a hash of a file crc32, md5 demo with Gui https://github.com/jNizM/HashCalc although I believe https://github.com/jNizM/AHK_CNG is the 'latest' code by jNizM (check the hash functions)

[jeeswg]

This is the function that I use currently:

Code: Select all

JEE_FilesMatchContents(vPath1, vPath2)
{
	FileGetSize, vSize1, % vPath1
	FileGetSize, vSize2, % vPath2
	if !(vSize1 = vSize2)
		return 0
	if (vSize1 = 0) && (vSize2 = 0)
		return 1
	;FileRead, vData1, % "*c " vPath1
	;FileRead, vData2, % "*c " vPath2
	oFile := FileOpen(vPath1, "r")
	oFile.Pos := 0
	oFile.RawRead(vData1, oFile.Length)
	oFile.Close()
	oFile := FileOpen(vPath2, "r")
	oFile.Pos := 0
	oFile.RawRead(vData2, oFile.Length)
	oFile.Close()
	return !DllCall("msvcrt\memcmp", Ptr,&vData1, Ptr,&vData2, UPtr,vSize1, "Cdecl Int")
}
Some related info:

conversion logic, v1 = -> v1 := -> v2, two-way compatibility - Page 5 - AutoHotkey Community
https://autohotkey.com/boards/viewtopic ... 80#p169780

The problem with the function above is that it loads the file into memory before comparing, I don't know if there's a better approach that operates on files directly.

[KeypressGuy]

Here's a method from Christian Ghisler that takes 3 approaches.
1. Check file sizes
If identical,
2. Compare small pieces of the files
If identical,
3. Calculate MD5

Because this method is performing small spot checks rather than reading the entire file, it often detects differences very quickly.

Wiki http://www.ghisler.ch/wiki/index.php/Au ... le_compare

[Helgef]

For comparing two files, jeeswg's function is the best suggestion here imo. There is no point in calculating hashes in this case. In case of very big files I'd avoid reading the entire file unconditionally. You could loop chunks of 50-500 mb perhaps. One could even start at a small size and then increase. If you want to compare many files against each other, hash is good.

Cheers.