pangolin/server/internal/idgen/idgen_test.go

package idgen_test

import (
	"strings"
	"testing"

	"github.com/google/uuid"
	"github.com/wangjia/pangolin/server/internal/idgen"
)

// ─────────────────────────────────────────────────────────────────────────────
// UUID v7 tests
// ─────────────────────────────────────────────────────────────────────────────

// TestNewReturnsVersion7 verifies that New() returns a UUID with version 7.
func TestNewReturnsVersion7(t *testing.T) {
	for i := 0; i < 100; i++ {
		id := idgen.New()
		if id.Version() != uuid.Version(7) {
			t.Fatalf("New(): version = %d, want 7", id.Version())
		}
	}
}

// TestNewStringFormat verifies that NewString() returns a properly formatted UUID string.
func TestNewStringFormat(t *testing.T) {
	s := idgen.NewString()
	// Standard UUID format: xxxxxxxx-xxxx-xxxx-xxxx-xxxxxxxxxxxx (36 chars)
	if len(s) != 36 {
		t.Fatalf("NewString() length = %d, want 36", len(s))
	}
	parts := strings.Split(s, "-")
	if len(parts) != 5 {
		t.Fatalf("NewString() has %d hyphen-separated parts, want 5", len(parts))
	}
	expected := []int{8, 4, 4, 4, 12}
	for i, p := range parts {
		if len(p) != expected[i] {
			t.Errorf("part %d: length = %d, want %d", i, len(p), expected[i])
		}
	}
}

// TestNewUniqueness verifies that 10 000 generated UUIDs are all distinct.
func TestNewUniqueness(t *testing.T) {
	const n = 10_000
	seen := make(map[string]struct{}, n)
	for i := 0; i < n; i++ {
		s := idgen.NewString()
		if _, dup := seen[s]; dup {
			t.Fatalf("duplicate UUID at iteration %d: %s", i, s)
		}
		seen[s] = struct{}{}
	}
}

// TestNewTimeOrdered verifies that sequentially generated UUIDs are
// monotonically non-decreasing in their string representation (UUID v7 is
// time-ordered so lexicographic sort ≈ generation order).
func TestNewTimeOrdered(t *testing.T) {
	prev := idgen.NewString()
	for i := 0; i < 1000; i++ {
		next := idgen.NewString()
		if next < prev {
			t.Fatalf("UUID ordering violation: %s > %s", prev, next)
		}
		prev = next
	}
}

// ─────────────────────────────────────────────────────────────────────────────
// Crockford Base32 activation-code tests
// ─────────────────────────────────────────────────────────────────────────────

// TestGenerateCodeFormat verifies that GenerateCode returns a 16-char code
// composed entirely of the Crockford alphabet (15 data chars + 1 check char).
func TestGenerateCodeFormat(t *testing.T) {
	for i := 0; i < 1000; i++ {
		code, err := idgen.GenerateCode()
		if err != nil {
			t.Fatalf("GenerateCode error: %v", err)
		}
		if len(code) != 16 {
			t.Errorf("code %q: length = %d, want 16", code, len(code))
		}
		// A freshly generated code must canonicalize to itself.
		canonical, err := idgen.CanonicalizeCode(code)
		if err != nil {
			t.Errorf("CanonicalizeCode(%q): %v", code, err)
		}
		if canonical != code {
			t.Errorf("canonical form mismatch: got %q, want %q", canonical, code)
		}
	}
}

// TestGenerateCodeUniqueness checks that 10 000 generated codes have no
// hash collisions (birthday probability ≈ 10^−8 for 75-bit codes).
func TestGenerateCodeUniqueness(t *testing.T) {
	const n = 10_000
	seen := make(map[string]struct{}, n)
	for i := 0; i < n; i++ {
		code, err := idgen.GenerateCode()
		if err != nil {
			t.Fatalf("GenerateCode: %v", err)
		}
		h := idgen.HashCode(code)
		if _, dup := seen[h]; dup {
			t.Fatalf("hash collision at iteration %d: code=%s hash=%s", i, code, h)
		}
		seen[h] = struct{}{}
	}
}

// TestCanonicalizeCodeNormalization verifies I/L→1 and O→0 substitutions.
func TestCanonicalizeCodeNormalization(t *testing.T) {
	base, err := idgen.GenerateCode()
	if err != nil {
		t.Fatalf("GenerateCode: %v", err)
	}

	// Lower-case input must produce the upper-case canonical form.
	lower := strings.ToLower(base)
	canonical, err := idgen.CanonicalizeCode(lower)
	if err != nil {
		t.Errorf("CanonicalizeCode(lower) error: %v", err)
	}
	if canonical != base {
		t.Errorf("CanonicalizeCode(lower) = %q, want %q", canonical, base)
	}

	// I, L → 1.
	idx := strings.IndexByte(base, '1')
	if idx >= 0 && idx < 15 {
		for _, sub := range []string{"I", "L", "i", "l"} {
			variant := base[:idx] + sub + base[idx+1:]
			c, err := idgen.CanonicalizeCode(variant)
			if err != nil {
				t.Errorf("CanonicalizeCode(%q) error: %v", variant, err)
				continue
			}
			if c != base {
				t.Errorf("CanonicalizeCode(%q) = %q, want %q", variant, c, base)
			}
		}
	}

	// O → 0.
	idx = strings.IndexByte(base, '0')
	if idx >= 0 && idx < 15 {
		variant := base[:idx] + "O" + base[idx+1:]
		c, err := idgen.CanonicalizeCode(variant)
		if err != nil {
			t.Errorf("CanonicalizeCode(%q) error: %v", variant, err)
		} else if c != base {
			t.Errorf("CanonicalizeCode(%q) = %q, want %q", variant, c, base)
		}
	}
}

// TestCheckCharDetectsSingleErrors verifies that mutating any single data
// character in a valid code causes CanonicalizeCode to return an error.
func TestCheckCharDetectsSingleErrors(t *testing.T) {
	const alpha = "0123456789ABCDEFGHJKMNPQRSTVWXYZ"
	code, err := idgen.GenerateCode()
	if err != nil {
		t.Fatalf("GenerateCode: %v", err)
	}

	for pos := 0; pos < 15; pos++ {
		original := rune(code[pos])
		for _, replacement := range alpha {
			if replacement == original {
				continue
			}
			mutated := code[:pos] + string(replacement) + code[pos+1:]
			if _, err := idgen.CanonicalizeCode(mutated); err == nil {
				t.Errorf("mutating pos %d (%c→%c) not detected: code=%q mutated=%q",
					pos, original, replacement, code, mutated)
			}
		}
	}
}

// TestHashCodeConsistency verifies that HashCode is deterministic and that
// different codes produce different digests.
func TestHashCodeConsistency(t *testing.T) {
	code1, _ := idgen.GenerateCode()
	code2, _ := idgen.GenerateCode()
	for code1 == code2 {
		code2, _ = idgen.GenerateCode()
	}

	h1a := idgen.HashCode(code1)
	h1b := idgen.HashCode(code1)
	h2 := idgen.HashCode(code2)

	if h1a != h1b {
		t.Error("HashCode is not deterministic")
	}
	if h1a == h2 {
		t.Error("different codes produced the same digest")
	}
	if len(h1a) != 64 {
		t.Errorf("HashCode length = %d, want 64 (hex SHA-256)", len(h1a))
	}
}

// TestCanonicalizeCodeRejectsInvalidLength tests length validation.
func TestCanonicalizeCodeRejectsInvalidLength(t *testing.T) {
	cases := []string{"", "ABCDE", "ABCDEFGH12345678X"}
	for _, c := range cases {
		if _, err := idgen.CanonicalizeCode(c); err == nil {
			t.Errorf("CanonicalizeCode(%q) should fail for length %d", c, len(c))
		}
	}
}

// TestCanonicalizeCodeRejectsInvalidChars tests that non-Crockford characters
// at data positions (0–14) are rejected.
func TestCanonicalizeCodeRejectsInvalidChars(t *testing.T) {
	base, _ := idgen.GenerateCode()
	invalid := "!" + base[1:]
	if _, err := idgen.CanonicalizeCode(invalid); err == nil {
		t.Errorf("CanonicalizeCode(%q) should fail for invalid character", invalid)
	}
}

// TestHyphenStripping verifies that hyphens inserted for display readability
// (e.g. XXXX-XXXX-XXXX-XXXX) are stripped before validation.
func TestHyphenStripping(t *testing.T) {
	code, err := idgen.GenerateCode()
	if err != nil {
		t.Fatalf("GenerateCode: %v", err)
	}
	hyphenated := code[:4] + "-" + code[4:8] + "-" + code[8:12] + "-" + code[12:]
	canonical, err := idgen.CanonicalizeCode(hyphenated)
	if err != nil {
		t.Errorf("CanonicalizeCode(hyphenated) error: %v", err)
	}
	if canonical != code {
		t.Errorf("CanonicalizeCode(hyphenated) = %q, want %q", canonical, code)
	}
}