I'm doing some file analysis, where I mark explored regions in the file. Now I would like to find the unexplored regions, so I know what to look at next. This is much like what defragmentation software shows for free and used regions.
Example:
In this picture, let's say that explored regions are red, unexplored regions are gray. I need to determine the gray-region boundaries from these red regions.
My current code, a custom binary reader that logs what's been read:
public class CustomBinaryReader : BinaryReader {
private readonly List<Block> _blocks;
public CustomBinaryReader([NotNull] Stream input) : this(input, Encoding.Default) { }
public CustomBinaryReader(Stream input, Encoding encoding, bool leaveOpen = true) : base(input, encoding, leaveOpen) {
_blocks = new List<Block>();
}
public override byte[] ReadBytes(int count) {
Log(count);
return base.ReadBytes(count);
}
private void Log(int count) {
_blocks.Add(new Block(BaseStream.Position, count));
}
private IEnumerable<Block> GetUnreadBlocks() {
// how to get unread blocks in the stream, from read blocks ?
throw new NotImplementedException();
}
}
And the type that defines what a region is:
public class Block {
public Block(long position, long length) {
Position = position;
Length = length;
}
public long Position { get; }
public long Length { get; }
}
Question:
Is there a class of algorithms or data structures to solve such problem (like a tree or a graph) ? If such thing does not exist, can you give me some approach or tips on how to solve such problem?
