Delphi - Reading a file excluding header information [closed] - file

Question

Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 9 years ago.
Improve this question
in my application i read a file, calculate its MD5 hash, cypher it with AES128 and write it inside file comments.
Thats because i need to ensure that file is not manipulated.
Sadly, when i write something inside file's "comments" properties its MD5 change and my work become useless.
So, i want to read my file excluding file's properties and only then calculate my HASH.
Now i read my file with this function:
function MD5File(const FileName: string): string;
var
IdMD5: TIdHashMessageDigest5;
FS: TFileStream;
begin
IdMD5 := TIdHashMessageDigest5.Create;
FS := TFileStream.Create(FileName, fmOpenRead or fmShareDenyWrite);
try
Result := IdMD5.HashStreamAsHex(FS)
finally
FS.Free;
IdMD5.Free;
end;
end;
How can i get the "file properties" size? i suppose they are on the header OR the footer, so if i know the size i can exclude the first or the last part of the file
edit:
more information: i store the information manually editing the file properties (right click -> properties)
i too suppose that editing those information must not alterate my MD5.. but it does for some reason!
I read the file AND calculate its MD5 with the function in my post.
My problem is that when i write any value inside the comment the file's hash change and i'm writing a string based on file hash.
i think i will just wipe the "comment" properties before calculate hash.

Answer 1

Your premise is wrong. The "Comments" data you are entering is stored within the file in the EXIF part of the .JPG file. It does not have a fixed offset, so it's not easy to exclude it from your hash. You'll need to parse the .JPG file and manually skip over the portions of the file that you do not want to include in your hash. Read up on .JPG files and EXIF information stored therein.
Or rethink your premise. If you simply want a hash over the image portion, you can load the image as a TBitMap and iterate over all the pixels in the image (a very slow process) and include their colour code in your hash, thus excluding any data that isn't an actual part of the image itself.
Both these methods will mean that you can't do what you are attempting on anything other than files for which you know the file format (unless they are all image files that you can load into a TBitmap and thus only include the pixels in your hash). You cannot, for example, allow people to change comments or other things in f.ex. Word documents or PDF documents, unless - again - you read the file manually and only calculate a hash over the parts that you don't want to allow people to alter.