Pbt

Rapid: Property-Based Testing for Go

Rapid (pgregory.net/rapid) is a Go library for property-based testing. Tests integrate with go test via rapid.Check. Rapid generates random inputs for your code and automatically shrinks failing cases to minimal counterexamples. It has zero external dependencies.

Workflow

Follow these steps when writing property-based tests.

1. Load the Go Reference

Load references/go.md for full API details and idiomatic patterns.

2. Explore the Code Under Test

Before writing any test, understand what you're testing:

• Read the source code of the function/module under test
• Read existing tests to understand expected behavior and edge cases
• Read godoc comments and type signatures for documented contracts
• Read usage sites to see how callers use the code and what they expect

The goal is to find evidence for properties, not to invent them.

3. Identify Valuable Properties

Look for properties that are:

• Grounded in evidence from the code, docs, or usage patterns
• Non-trivial — they test real behavior, not tautologies, and do not duplicate the code being tested
• Falsifiable — a buggy implementation could actually violate them

Write one test per property. Don't cram multiple properties into one test.

If you cannot write down the expected output for a concrete input — or outputs legitimately vary between runs (concurrency, arbitrary tie-breaking, hidden seeds, floating point) — derive metamorphic relations instead of forcing an oracle: see "Metamorphic Relations" below and load references/metamorphic.md.

4. Check for Existing Tests to Evolve or Port

Before writing tests from scratch, always check existing tests:

• Existing PBTs in another framework (testing/quick, etc.) should be ported to rapid. Load references/porting.md for guidance. Don't carry over narrow generator bounds from the old framework — use broader generators unless bounds are justified by the function's contract.
• Unit tests and example-based tests can often be evolved into PBTs. Load references/evolving-tests.md for guidance. Tests with hardcoded values, table-driven tests, or multiple similar test cases are prime candidates.
• Tests that use math/rand with fixed seeds are especially good candidates — the randomness should come from rapid instead so failures produce shrinkable counterexamples.

When you evolve an existing test, modify the existing test file rather than creating a new one. Add rapid tests alongside (or replacing) the existing tests in the same file where the original tests live. Do not create a separate pbt_test.go or similar — property-based tests are tests like any other and belong with the code they're testing.

5. Write the Tests

For each property:

1. Add tests to the appropriate existing test file. If there's already a foo_test.go covering the module, add rapid tests there. Only create a new file if no relevant test file exists.
2. Choose the simplest possible generators — start with no bounds, unless bounds are logically necessary (e.g. if a number has to be non-zero it's fine to force it to be, but slices should not have maxLen set unless there is a compelling correctness reason to set them or poor performance has been observed when actually running the test)
3. Draw values using generator.Draw(t, "label")
4. Run the code under test
5. Assert the property using t.Fatalf, t.Errorf, or standard Go assertions

6. Run and Reflect

Run the tests. When a test fails, ask:

• Is this a real bug? If the code violates its own contract, flag the bug to the user and ask what to do, or fix the code if instructed to do so.
• Is the property unsound? If you asserted something the code never promised, fix the test.
• Is the generator too broad? Only if the failing input is genuinely outside the function's domain, add constraints. Investigate before constraining.

Property Categories

Use this taxonomy to identify what to test. Not every category applies to every function — pick the ones supported by evidence.

Category	Description	Example
Round-trip	encode then decode recovers the original	Unmarshal(Marshal(x)) == x
Idempotence	applying twice equals applying once	sort(sort(xs)) == sort(xs)
Commutativity	order of operations doesn't matter	a + b == b + a
Invariant preservation	an operation maintains a structural property	insert into BST preserves ordering
Oracle / reference impl	compare against a known-correct implementation	mySort(xs) == slices.Sort(xs), or comparing against an unoptimised implementation
Monotonicity	more input means more (or equal) output	len(append(xs, ys...)) >= len(xs)
Bounds / contracts	output stays within documented limits	clamp(x, lo, hi) is in [lo, hi]
No-crash / robustness	function handles all valid inputs without panicking	parse(arbitraryString) doesn't panic
Equivalence	two implementations produce the same result	iterativeFib(n) == recursiveFib(n)
Model-based	operations on real system match a simplified model	custom map ops match built-in map
Consistency	related APIs in the same library agree	StringWidth(s) == sum of RuneWidth per rune
Precision preservation	numeric values survive format conversions	strconv.Atoi(strconv.Itoa(n)) == n
Metamorphic	a relation between the outputs of two related runs replaces the missing oracle	count(q∧p) + count(q∧¬p) == count(q); add one matching record → exactly one more result

High-Value Patterns (Field-Tested)

These patterns are ranked by how often they found real bugs when tested across many popular Go libraries. See references/field-tested-patterns.md for detailed examples.

1. Model Tests (Highest Value for Data Structures)

For any data structure, the highest-value first test is a model test — run the same operations on the library under test and a known-good reference (usually a standard library type), then assert they agree after every operation.

Choose the right oracle:

• []T for sequential containers (fixed-capacity slices, ring buffers)
• map[K]V for hash maps (concurrent maps, ordered maps)
• map[K]struct{} or sorted slice for sets

2. Idempotence Tests (Highest Value for String/Text Processing)

Any normalization, case conversion, or formatting function should be idempotent: f(f(x)) == f(x). Use rapid.String() (not ASCII-only generators) because Unicode edge cases like ß → SS and combining characters are where bugs hide.

3. Parse Robustness (Universal — Test Every Parser)

Every parsing function should be tested with rapid.String(). The property is simple: it should never panic. Parsers that delegate to constructors which panic on invalid values (instead of returning errors) are a common source of bugs.

4. Roundtrip Tests (High Value for Serialization)

parse(format(x)) == x for any serialize/deserialize pair. Test with the full input domain — don't restrict to "reasonable" values. Bugs hide at boundaries like zero (e.g. scientific notation missing the coefficient), large integers (precision loss through float64 intermediaries for values > 2^53), and unusual string content (double slashes in paths, control characters).

5. Boundary Value Tests (High Value for Numeric Code)

Integer operations should be tested with math.MinInt, math.MaxInt, 0, and unconstrained ranges. Negating math.MinInt overflows, and many libraries forget to handle these. Don't add rapid.IntRange(-100, 100) — those bounds hide real bugs.

Metamorphic Relations (When There Is No Oracle)

Model and oracle tests assume you can predict the output. When you can't — verifying one output is as hard as computing it, results legitimately vary between runs (tie-breaking, floats, hidden seeds), or the code is concurrent and a sequential model can't say which racing outcome is correct — run the code twice, on an input and on a transformed version of it, and assert a relation between the two outputs instead of predicting either. Round-trip, idempotence, commutativity, and equivalence properties above are already metamorphic relations; this generalizes them.

Choose in this order: (1) expected output computable → plain assertion; (2) a short reference you trust → model test (Pattern 1); (3) an independent implementation you don't trust → equivalence test; (4) none of these → metamorphic relations, derived from two questions: what must NOT change the output (shuffle order-free input, rename keys, add records that match nothing, change a documented-invisible config knob) and what changes it predictably (strengthen a filter → subset; partition → the parts recompose; add one matching record → exactly one more result). Write 3–6 relations from different families — a few diverse relations approximate a real oracle.

Non-negotiables: every relation carries a one-line // spec: comment deriving it from the documented contract (docs and domain math, never the implementation); canonicalize before comparing (sort unordered results, strip undetermined fields); assert nothing the contract doesn't promise — result order, bit-equal floats, mid-flight counts. For concurrent code, assert exact relations only at quiescence.

Load references/metamorphic.md for the transformation catalog, float tolerance rules, the concurrent playbook (determinate workloads, conserved aggregates, invariance grids, unique-value tagging, checkpoint immutability), failure triage, unsound-relation traps, and validating the suite with mutants. Load references/metamorphic-templates.md for validated rapid templates of each pattern, ready to adapt.

Choosing Properties

Properties must be evidence-based. Find evidence in:

• Names and Type signatures: A function func Merge(a, b []T) []T implies the output length might equal the sum of input lengths.
• Godoc comments: "Returns a sorted slice" directly gives you an invariant.
• Panics and assertions in the source: These are properties the author already identified, and do not need to be duplicated in the tests, but may suggest other invariants.
• Usage patterns: If callers always assume a result is non-empty, assert that the result is always non-empty.
• Existing tests: Unit tests often encode specific instances of general properties.

Err on the side of creating more properties rather than fewer, and if they fail investigate whether the failure is legitimate behaviour or not.

Beware of properties that seem universal but aren't. Read the docs carefully before asserting a property. Examples from real testing:

• Grapheme-based string reverse is NOT an involution (reverse(reverse("\n\r")) ≠ "\n\r" because \r\n is one grapheme cluster while \n\r is two).
• A method called Difference might mean symmetric difference (A △ B), not set difference (A \ B) — check the docs.
• A function documented as "returns the largest key ≤ k" means ≤, not <.

When a property fails, investigate whether it's a real bug or a genuine edge case in the domain. A weaker property often still holds.

Generator Discipline

A common mistake agents make when writing property-based tests is over-constraining generators. This leads to tests that are weaker than they need to be.

Start With No Bounds

If the function accepts any int, use:

rapid.Int()  // no min, no max

Do NOT preemptively write:

rapid.IntRange(0, 100)  // WRONG unless justified

Edge Cases Are the Point

Don't narrow ranges to "avoid edge cases." Edge cases are exactly what PBT is for. If a function claims to work on all int values, test it on all int values — including math.MinInt, math.MaxInt, 0, -1, and 1.

Don't Set minLen to 1 by Default

Unless the function's contract explicitly requires non-empty input, test with empty collections too. If a function panics on an empty slice, that might be a bug worth knowing about.

When a Test Fails on Extreme Values

Your first reaction should be: is this a real bug?

You should assume that it is unless you have strong evidence that it is not. If in doubt, ask the user.

• If the function's documentation says it handles all integers but it overflows on math.MaxInt, that's a bug in the code, not in your test.
• Only add bounds after investigating and confirming the input is outside the function's documented domain.

When to Add Constraints

Add generator bounds only when:

1. The function's contract explicitly excludes some inputs. For example, func Sqrt(x float64) documents that x >= 0 is required.
2. You need to avoid undefined behavior. For example, division by zero.
3. A test failure has been investigated and confirmed to be outside the function's domain.
4. You are concentrating the input space to force a structural interaction — hot keys so concurrent operations actually contend, a narrow score range to force ties, a small key pool so deletes hit written keys. State the reason in a comment next to the bound; references/metamorphic-templates.md shows the pattern.

Avoid Rejection Sampling Where Possible

When a constraint involves relationships between multiple generated values, you may use t.Skip:

a := rapid.Int().Draw(t, "a")
b := rapid.Int().Draw(t, "b")
if a == b {
    t.Skip("need distinct values")
}

This example is perfectly fine, but it is better to avoid Skip altogether when you can:

e.g.

a := rapid.Int().Draw(t, "a")
b := rapid.IntMin(a).Draw(t, "b")

is better than

a := rapid.Int().Draw(t, "a")
b := rapid.Int().Draw(t, "b")
if a > b {
    t.Skip("need a <= b")
}

Even better is:

a := rapid.Int().Draw(t, "a")
b := rapid.Int().Draw(t, "b")
if a > b {
    a, b = b, a
}

It is particularly important to avoid rejection sampling in cases where the rejection rate is likely to be high.

For example rapid.Map(rapid.Int(), func(n int) int { return n * 2 }) is much better than rapid.Int().Filter(func(n int) bool { return n%2 == 0 }), as the former constructs an even number directly, while the latter throws away around 50% of test cases.

Getting Large Collections

Rapid's default collection size is small. If you need large collections (e.g., to exercise deep tree paths), draw the size separately and use SliceOfN:

// GOOD — can generate large collections, shrinks well
n := rapid.IntRange(0, 300).Draw(t, "n")
keys := rapid.SliceOfN(rapid.Int(), n, -1).Draw(t, "keys")

// BAD — rapid's default size distribution rarely produces 100+ elements
keys := rapid.SliceOf(rapid.Int()).Draw(t, "keys")

Setting minLen but not maxLen (using -1) is a shrinking optimization: rapid can shrink n to find the minimal collection size that triggers the bug, while still being able to add extra elements if needed.

Use SliceOfDistinct for Key Generation

When testing maps/sets that need unique keys:

keys := rapid.SliceOfNDistinct(rapid.Int(), -1, 30, rapid.ID[int]).Draw(t, "keys")

This avoids confusion about which value wins for duplicate keys.

Handling Randomness in Code Under Test

When the code under test requires an RNG (e.g., func Sample(weights []float64, rng *rand.Rand)), do not create a seeded RNG like rand.New(rand.NewSource(seed)) with a rapid-generated seed. This defeats shrinking — rapid can only shrink the seed integer, not the actual random decisions the RNG makes.

Instead, generate the random decisions you need directly through rapid generators. For example, if the code needs a random index, draw it with rapid.IntRange(0, n-1).

If the code's RNG is deeply embedded and cannot be easily replaced, generating a seed is acceptable as a last resort, but understand that shrinking quality will be reduced.

Common Mistakes

1. Over-constraining generators — Adding bounds "just in case." This hides bugs and makes tests less valuable. See Generator Discipline above.
2. Testing trivial properties — assert(x == x) or assert(len(s) >= 0) test nothing. Every property should be falsifiable by a buggy implementation.
3. Using the implementation as the oracle — If your test calls the same function to compute the expected result, it can never fail. Use an independent reference implementation (do not just copy the code to write this!), a simpler algorithm, or a structural property.
4. Generating too broadly then filtering almost everything — If .Filter() or t.Skip() rejects most inputs, rapid will give up. Restructure your generators instead (e.g., use rapid.Map or dependent generation).
5. Creating a separate test file for rapid tests — Property-based tests belong alongside the existing tests for the same code. Don't put them in pbt_test.go or properties_test.go — add them to the existing test files.
6. Using manually seeded RNGs — Don't generate a seed with rapid then create rand.New(rand.NewSource(seed)). Generate the random decisions you need through rapid generators so rapid can shrink them. See "Handling Randomness" above.
7. Overflowing in test code — When computing values from generated data (e.g., m[k] = k * 10), your test code itself can overflow before the library has a chance to be buggy. Use smaller intermediate types (draw int16, cast to int for multiplication) to prevent this. Distinguish "this constraint protects the library's contract" (keep it) from "this constraint prevents my test from overflowing" (use a smaller type instead).
8. Adding maxLen for performance — If a test is slow with large collections, lower the check count with -rapid.checks rather than restricting the input space. A slow test that finds bugs beats a fast test that can't. Many tree/trie bugs only manifest at 50-200+ elements.
9. Asserting more than the contract promises in cross-run relations — result order where ordering is documented arbitrary, bit-equal floats across reordered arithmetic, exact sizes while writers are active. Canonicalize and compare only the promised projection. See references/metamorphic.md.

Quick Setup

// go.mod
require pgregory.net/rapid latest

Run with go test. Rapid integrates with go test via rapid.Check(t, prop).