Artifact [fb79c8ae96]

Artifact fb79c8ae969abcee513d3d96e7a6a9ee1dba60364aa7bdb85d644166357d822c:

• File internal/content/cache_test.go — part of check-in [6273a0750d] at 2026-07-22 23:03:24 on branch trunk — fix: correct go.mod module path to go-libfossil (#76) (#117)

  * fix: correct go.mod module path to go-libfossil (#76)

  The module directive still declared github.com/danmestas/libfossil after the repository was renamed to go-libfossil. pkg.go.dev/godoc scraping matches the declared module path against its hosting path, so the stale path broke SDK doc generation.

  Rename the module path across all four in-repo modules (root, the two db/driver submodules, observer/otel, and docs/site) and every internal import, plus the release workflow, the doc-generation script, and the docs.

  No tag has been published under the go-libfossil path yet, so the cross-module requires cannot resolve from the proxy. Add replace directives pointing the in-repo requires at their local directories so GOWORK=off builds and the CI matrix resolve without a published version. These replaces are ignored by downstream consumers and can be dropped once new-path tags are cut.

  * Guard releases against replace directives; regenerate SDK docs

  The go-libfossil rename bridges the in-repo modules with `replace` directives. Go refuses a main module that carries them, so a tag cut while they are present breaks `go install .../cmd/libfossil@<tag>` and every `@version` fetch of the driver submodules. Add a Release workflow step that fails the job before anything is published if any tracked go.mod still contains a replace directive, so the bridge is a CI-enforced TODO rather than a silent landmine. The step comment records what unblocks a release: re-tag the submodules with the corrected module paths, bump the root requires, then drop the replace blocks.

  Regenerate the SDK reference docs. The previous pass only substituted URLs over the pre-existing generated files, leaving stale signatures (Repo.Timeline was still documented as taking LogOpts rather than TimelineOpts). Note that gomarkdoc must be run outside a git worktree — inside one it silently strips every source link.

  Also reword the README rename note, which read as though the import path had not changed, and re-sort the ncruces require block. (user: [email protected], size: 7527) [annotate] [blame] [check-ins using]

package content

import (
	"bytes"
	"path/filepath"
	"sync"
	"testing"

	"github.com/danmestas/go-libfossil/db"
	"github.com/danmestas/go-libfossil/internal/blob"
	_ "github.com/danmestas/go-libfossil/internal/testdriver"
)

func TestCache_HitMiss(t *testing.T) {
	d := setupTestDB(t)

	data := []byte("hello cache")
	rid, _, err := blob.Store(d, data)
	if err != nil {
		t.Fatal(err)
	}

	c := NewCache(1 << 20) // 1MB

	// First call = miss
	got, err := c.Expand(d, rid)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got, data) {
		t.Fatalf("got %q, want %q", got, data)
	}

	stats := c.Stats()
	if stats.Hits != 0 || stats.Misses != 1 {
		t.Fatalf("expected 0 hits / 1 miss, got %d / %d", stats.Hits, stats.Misses)
	}

	// Second call = hit
	got2, err := c.Expand(d, rid)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got2, data) {
		t.Fatalf("got %q, want %q", got2, data)
	}

	stats = c.Stats()
	if stats.Hits != 1 || stats.Misses != 1 {
		t.Fatalf("expected 1 hit / 1 miss, got %d / %d", stats.Hits, stats.Misses)
	}
}

func TestCache_ReturnsCopy(t *testing.T) {
	d := setupTestDB(t)

	data := []byte("immutable data")
	rid, _, _ := blob.Store(d, data)

	c := NewCache(1 << 20)

	got1, _ := c.Expand(d, rid)
	got1[0] = 'X' // mutate the returned slice

	got2, _ := c.Expand(d, rid)
	if got2[0] == 'X' {
		t.Fatal("cache returned a reference instead of a copy — caller mutation corrupted cache")
	}
}

func TestCache_Eviction(t *testing.T) {
	d := setupTestDB(t)

	// Create blobs: 100 bytes each. Cache fits ~2.
	data1 := bytes.Repeat([]byte("A"), 100)
	data2 := bytes.Repeat([]byte("B"), 100)
	data3 := bytes.Repeat([]byte("C"), 100)

	rid1, _, _ := blob.Store(d, data1)
	rid2, _, _ := blob.Store(d, data2)
	rid3, _, _ := blob.Store(d, data3)

	c := NewCache(250) // fits ~2 entries of 100 bytes each

	c.Expand(d, rid1)
	c.Expand(d, rid2)

	stats := c.Stats()
	if stats.Entries != 2 {
		t.Fatalf("expected 2 entries, got %d", stats.Entries)
	}

	// Adding rid3 should evict rid1 (LRU)
	c.Expand(d, rid3)

	stats = c.Stats()
	if stats.Entries != 2 {
		t.Fatalf("expected 2 entries after eviction, got %d", stats.Entries)
	}
	if stats.Size > 250 {
		t.Fatalf("cache size %d exceeds max 250", stats.Size)
	}
}

func TestCache_Invalidate(t *testing.T) {
	d := setupTestDB(t)

	rid, _, _ := blob.Store(d, []byte("data to invalidate"))

	c := NewCache(1 << 20)
	c.Expand(d, rid)

	if c.Stats().Entries != 1 {
		t.Fatal("expected 1 entry")
	}

	c.Invalidate(rid)

	if c.Stats().Entries != 0 {
		t.Fatal("expected 0 entries after invalidate")
	}
}

func TestCache_Clear(t *testing.T) {
	d := setupTestDB(t)

	rid, _, _ := blob.Store(d, []byte("data to clear"))

	c := NewCache(1 << 20)
	c.Expand(d, rid)
	c.Clear()

	stats := c.Stats()
	if stats.Entries != 0 || stats.Size != 0 {
		t.Fatalf("expected empty cache, got entries=%d size=%d", stats.Entries, stats.Size)
	}
}

func TestCache_NilPassthrough(t *testing.T) {
	d := setupTestDB(t)

	data := []byte("passthrough data")
	rid, _, _ := blob.Store(d, data)

	var c *Cache
	got, err := c.Expand(d, rid)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got, data) {
		t.Fatalf("got %q, want %q", got, data)
	}
}

func TestCache_ConcurrentAccess(t *testing.T) {
	d := setupTestDB(t)

	data := []byte("concurrent data")
	rid, _, _ := blob.Store(d, data)

	c := NewCache(1 << 20)

	var wg sync.WaitGroup
	for i := 0; i < 50; i++ {
		wg.Add(1)
		go func() {
			defer wg.Done()
			got, err := c.Expand(d, rid)
			if err != nil {
				t.Errorf("expand error: %v", err)
				return
			}
			if !bytes.Equal(got, data) {
				t.Errorf("got %q, want %q", got, data)
			}
		}()
	}
	wg.Wait()
}

func TestNewCache_PanicsOnZero(t *testing.T) {
	defer func() {
		if r := recover(); r == nil {
			t.Fatal("expected panic for maxBytes <= 0")
		}
	}()
	NewCache(0)
}

func TestCache_LRUOrder(t *testing.T) {
	d := setupTestDB(t)

	data := bytes.Repeat([]byte("X"), 100)
	rid1, _, _ := blob.Store(d, data)

	data2 := bytes.Repeat([]byte("Y"), 100)
	rid2, _, _ := blob.Store(d, data2)

	data3 := bytes.Repeat([]byte("Z"), 100)
	rid3, _, _ := blob.Store(d, data3)

	c := NewCache(250) // fits ~2 entries

	c.Expand(d, rid1) // order: [rid1]
	c.Expand(d, rid2) // order: [rid2, rid1]
	c.Expand(d, rid1) // touch rid1 → order: [rid1, rid2]
	c.Expand(d, rid3) // evicts rid2 (LRU), order: [rid3, rid1]

	// rid1 should still be cached (was touched)
	missesBefore := c.Stats().Misses
	c.Expand(d, rid1)
	if c.Stats().Misses != missesBefore {
		t.Fatal("rid1 should have been a cache hit (was recently touched)")
	}

	// rid2 should have been evicted
	c.Expand(d, rid2)
	if c.Stats().Misses != missesBefore+1 {
		t.Fatal("rid2 should have been a cache miss (was evicted)")
	}
}

func TestCache_BlobLargerThanMax(t *testing.T) {
	d := setupTestDB(t)

	// 500-byte blob into a 100-byte cache — blob exceeds maxSize
	big := bytes.Repeat([]byte("B"), 500)
	rid, _, _ := blob.Store(d, big)

	c := NewCache(100)

	// The data should still be returned correctly even though it can't stay cached.
	got, err := c.Expand(d, rid)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got, big) {
		t.Fatal("data mismatch for oversized blob")
	}

	// The blob is inserted then immediately evicted (size 500 > max 100).
	stats := c.Stats()
	if stats.Entries != 0 {
		t.Fatalf("expected 0 entries (oversized blob evicted), got %d", stats.Entries)
	}
}

func TestCache_ExpandError(t *testing.T) {
	d := setupTestDB(t)

	c := NewCache(1 << 20)

	// rid 99999 doesn't exist — Expand should fail
	_, err := c.Expand(d, 99999)
	if err == nil {
		t.Fatal("expected error for nonexistent rid")
	}

	// Cache should remain empty — error results must not be cached
	stats := c.Stats()
	if stats.Entries != 0 {
		t.Fatalf("expected 0 entries after failed expand, got %d", stats.Entries)
	}
}

func TestCache_DeltaChainViaCacheHit(t *testing.T) {
	d := setupTestDB(t)

	// Build a 3-deep delta chain and verify cached expansion is correct
	v1 := []byte("version one with enough padding to create proper deltas here!!")
	v2 := []byte("version TWO with enough padding to create proper deltas here!!")
	v3 := []byte("version THREE with enough padding to create proper deltas here!!")

	rid1, _, _ := blob.Store(d, v1)
	rid2, _, _ := blob.StoreDelta(d, v2, rid1)
	rid3, _, _ := blob.StoreDelta(d, v3, rid2)

	c := NewCache(1 << 20)

	// Expand the leaf of the chain (miss)
	got, err := c.Expand(d, rid3)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got, v3) {
		t.Fatalf("got %q, want %q", got, v3)
	}

	// Second expand should hit cache and return identical data
	got2, err := c.Expand(d, rid3)
	if err != nil {
		t.Fatal(err)
	}
	if !bytes.Equal(got2, v3) {
		t.Fatalf("cached result differs: got %q, want %q", got2, v3)
	}

	stats := c.Stats()
	if stats.Hits != 1 || stats.Misses != 1 {
		t.Fatalf("expected 1 hit / 1 miss, got %d / %d", stats.Hits, stats.Misses)
	}
}

func BenchmarkCache_Expand(b *testing.B) {
	path := filepath.Join(b.TempDir(), "bench.fossil")
	d, _ := db.Open(path)
	db.CreateRepoSchema(d)
	defer d.Close()

	base := bytes.Repeat([]byte("base content for cache benchmark "), 100)
	rid, _, _ := blob.Store(d, base)

	// Build a 5-deep delta chain
	for i := 0; i < 5; i++ {
		next := make([]byte, len(base))
		copy(next, base)
		copy(next[i*50:], []byte("CHANGED!"))
		newRid, _, _ := blob.StoreDelta(d, next, rid)
		rid = newRid
		base = next
	}

	c := NewCache(1 << 20)

	// Prime the cache
	c.Expand(d, rid)

	b.ResetTimer()
	for i := 0; i < b.N; i++ {
		c.Expand(d, rid)
	}
}