search crashplan manifest (can be run in LINQPad in C# program mode)

parse-crashplan-csv.cs

void Main()
{
	var result = ReadLines(@"D:\temp\crashplan\full-dump3.csv")
// faster initial tests...
//		.Skip(1)
//		.Take(10)
		.Select(l => l.Split(','))
		// extract properties of each csv row
		.Select(l => new
		{
			Id = l[0],
			ParentId = l[1],
			// NOTE: paths may contain commas and they are not escaped, but this is the only column which does
			Path = String.Join(",", l.Skip(3).Take(l.Length - 10)),
			VersionTimestamp = l[l.Length - 7],
			Size = l[l.Length - 5],
			Checksum = l[l.Length - 4]
		})
		.Where(row => row.Size != "0")
		// filter non-relevant noise
		.Where(row => !row.Path.Contains("/Application Data/"))
		.Where(row => !row.Path.Contains("/AppData/"))
		.Where(row => !row.Path.Contains("/SkyDrive/"))
		.Where(row => !row.Path.Contains("dave/."))
		.Where(row => !row.Path.Contains("Dave/."))
		// group entries by filename (grouping by checksum is also useful)
		.GroupBy(row => row.Path.Split('/').Last())
		// old backup location was on K
		.Where(g => g.Any(row => row.Path.StartsWith("K")))
		// new backup location is on D (find entries which were on K, but not on D)
		.Where(g => !g.Any(row => row.Path.StartsWith("D")))
		// sort by file path
		.OrderBy(g => g.Select(item => item.Path).First())
		.Dump();
}

public IEnumerable<string> ReadLines(string filename)
{
	string line;
	
	System.IO.StreamReader file = new System.IO.StreamReader(filename);
	while((line = file.ReadLine()) != null)
	{
	   yield return line;
	}
	
	file.Close();
}