public void supports_adding_long_strings()
{
var charset = GenerateCharset();
var random = new Random(123);
var longString = GenerateRandomString(10000, charset, random);
Assert.IsFalse(_filter.MightContain(longString));
_filter.Add(longString);
Assert.IsTrue(_filter.MightContain(longString));
}
public void can_close_and_reopen()
{
_filter.Add("hello");
_filter.Dispose();
using var newFilter = new MemoryMappedFileStreamBloomFilter(_path, create: false, MemoryMappedFileBloomFilter.MinSizeKB * 1000, hasher: null);
Assert.IsTrue(newFilter.MightContain("hello"));
}
public void has_false_positives_with_probability_p(
[Values(MemoryMappedFileBloomFilter.MinSizeKB * 1000, 2 * MemoryMappedFileBloomFilter.MinSizeKB * 1000)] long size,
[Values(0.001, 0.02, 0.05, 0.1, 0.2)] double p
)
{
using var filter = new MemoryMappedFileStreamBloomFilter(GetTempFilePath(), create: true, size, hasher: null);
var n = (int)filter.CalculateOptimalNumItems(p);
var random = new Random(123);
var charset = GenerateCharset();
var list = new List <string>();
var selected = new HashSet <string>();
//generate 2n distinct items
for (int i = 0; i < 2 * n; i++)
{
while (true)
{
var length = 1 + random.Next() % 10;
var s = GenerateRandomString(length, charset, random);
if (selected.Contains(s))
{
continue;
}
list.Add(s);
selected.Add(s);
break;
}
}
//add first n distinct items to the filter
for (int i = 0; i < n; i++)
{
filter.Add(list[i]);
}
//expected number of false positives
var expectedFalsePositives = Convert.ToInt32(Math.Ceiling(n * p));
//the second n distinct items should not exist but there may be some false positives
var falsePositives = 0;
for (var i = n; i < 2 * n; i++)
{
if (filter.MightContain(list[i]))
{
falsePositives++;
}
}
//X = random variable that takes value 1 with probability p and value 0 with probability (1-p)
//var(X) = E(X^2) - E(X)^2 = p - p*p;
//var(X1 + X2 + X3 + ... + Xn) = n*var(Xi); //variance of n uncorrelated random variables
//var(X1 + X2 + X3 + ... + Xn) = n*(p-p*p);
var variance = n * (p - (p * p));
var standardDeviation = Math.Sqrt(variance);
var threeStandardDeviations = 3 * standardDeviation; //99.7%
if (falsePositives > 0)
{
Console.Out.WriteLine("n: {0:N0}, p:{1:N3}. Found {2:N0} false positives. Expected false positives: {3:N0}. Standard deviation: {4:N2}",
n, p, falsePositives, expectedFalsePositives, standardDeviation);
}
Assert.LessOrEqual(falsePositives, expectedFalsePositives + threeStandardDeviations);
Assert.GreaterOrEqual(falsePositives, Math.Max(0, expectedFalsePositives - threeStandardDeviations));
}