name	testing-standards
description	Testing conventions for Stock Indicators. Use for test naming (MethodName_StateUnderTest_ExpectedBehavior), FluentAssertions patterns, precision requirements, and test base class selection.

Testing standards

Test base class selection

Style	Base Class	Purpose
Series	`StaticSeriesTestBase`	Batch processing tests
Buffer	`BufferListTestBase`	Incremental processing tests
Stream	`StreamHubTestBase`	Real-time processing tests
Other	`TestBase`	General utility tests

Test naming convention

Pattern: MethodName_StateUnderTest_ExpectedBehavior

[TestMethod]
public void ToEma_WithSmallDataset_CalculatesCorrectly() { }

[TestMethod]
public void ToRsi_WithInsufficientData_ReturnsEmpty() { }

[TestMethod]
public void AddQuote_WithNullInput_ThrowsArgumentNull() { }

Required test coverage

MUST include these test types:

Standard tests

[TestMethod]
public void Standard()
{
    // Test against historical data with expected results
    var results = quotes.ToIndicator(14);

    // Verify specific data points
    var result = results.ElementAt(index);
    result.Value.Should().BeApproximately(expected, precision);
}

Boundary tests

[TestMethod]
public void MinimumPeriods()
{
    // Test with minimum lookback
    var results = quotes.ToIndicator(2);
    results.Should().NotBeEmpty();
}

Bad data tests

[TestMethod]
public void BadParameter()
{
    Action act = () => quotes.ToIndicator(-1);
    act.Should().Throw<ArgumentOutOfRangeException>();
}

[TestMethod]
public void NullQuotes()
{
    Action act = () => ((IReadOnlyList<Quote>)null!).ToIndicator(14);
    act.Should().Throw<ArgumentNullException>();
}

Insufficient data tests

[TestMethod]
public void InsufficientQuotes()
{
    var results = quotes.Take(5).ToList().ToIndicator(20);
    results.Should().BeEmpty();
}

FluentAssertions patterns

// Exact equality
result.Value.Should().Be(expected);

// Approximate equality with precision
result.Value.Should().BeApproximately(expected, 0.0001);

// Null checks
result.Value.Should().BeNull();
result.Value.Should().NotBeNull();

// Collection assertions
results.Should().HaveCount(502);
results.Should().BeEmpty();
results.Should().NotBeEmpty();

// Exception assertions
act.Should().Throw<ArgumentOutOfRangeException>();
act.Should().Throw<ArgumentNullException>()
   .WithParameterName("quotes");

Precision constants

Use constants from TestBase:

Constant	Value	Use Case
`Money6`	0.000001	Most calculations
`Money5`	0.00001	Lower precision
`Money4`	0.0001	Standard tolerance
`Percent`	0.01	Percentage values

result.Ema.Should().BeApproximately(123.456789, Money6);

Series parity validation

Buffer and Stream tests must validate against Series:

[TestMethod]
public void MatchesSeries()
{
    // Get Series results (canonical reference)
    var series = quotes.ToIndicator(14);

    // Get Buffer results
    var buffer = new IndicatorList(14) { quotes };

    // Validate exact parity
    buffer.IsExactly(series);
}

BufferList constraints

MUST implement these 5 base class tests:

AddQuote_IncrementsResults()
AddQuotesBatch_IncrementsResults()
QuotesCtor_OnInstantiation_IncrementsResults()
Clear_WithState_ResetsState()
PruneList_OverMaxListSize_AutoAdjustsListAndBuffers()

StreamHub constraints

All StreamHub tests must validate rollback scenarios:

Prefill warmup window before subscribing
Stream in-order including duplicates
Insert late historical quote → verify recalculation
Remove historical quote → verify recalculation
Compare to Series with strict ordering

Test data

Use standard test data from Data.GetDefault():

502 quotes of historical data
Consistent across all tests
Excel reference file: tests/_common/Data.Quotes.xlsx

Last updated: December 31, 2025