Make object repacking more configurable

6 files changed, 44 insertions, 15 deletions

diff --git a/plumbing/storer/object.go b/plumbing/storer/object.go
index 29e0090..e5f98d7 100644
--- a/plumbing/storer/object.go
+++ b/plumbing/storer/object.go
@@ -55,8 +55,9 @@ type EncodedObjectStorer interface {
 	// ObjectPacks returns hashes of object packs if the underlying
 	// implementation has pack files.
 	ObjectPacks() ([]plumbing.Hash, error)
-	// DeleteObjectPackAndIndex deletes an object pack and the corresponding index file if they exist.
-	DeleteObjectPackAndIndex(plumbing.Hash) error
+	// DeleteOldObjectPackAndIndex deletes an object pack and the corresponding index file if they exist.
+	// Deletion is only performed if the pack is older than the supplied time (or the time is zero).
+	DeleteOldObjectPackAndIndex(plumbing.Hash, time.Time) error
 }
 
 // DeltaObjectStorer is an EncodedObjectStorer that can return delta
diff --git a/plumbing/storer/object_test.go b/plumbing/storer/object_test.go
index 68e8e10..da0db81 100644
--- a/plumbing/storer/object_test.go
+++ b/plumbing/storer/object_test.go
@@ -165,3 +165,7 @@ func (o *MockObjectStorage) DeleteLooseObject(plumbing.Hash) error {
 func (o *MockObjectStorage) ObjectPacks() ([]plumbing.Hash, error) {
 	return nil, nil
 }
+
+func (o *MockObjectStorage) DeleteOldObjectPackAndIndex(plumbing.Hash, time.Time) error {
+	return plumbing.ErrObjectNotFound
+}
diff --git a/repository.go b/repository.go
index 51cd077..999cc6d 100644
--- a/repository.go
+++ b/repository.go
@@ -8,6 +8,7 @@ import (
 	"os"
 	"path/filepath"
 	"strings"
+	"time"
 
 	"gopkg.in/src-d/go-git.v4/config"
 	"gopkg.in/src-d/go-git.v4/internal/revision"
@@ -1013,7 +1014,18 @@ func (r *Repository) ResolveRevision(rev plumbing.Revision) (*plumbing.Hash, err
 	return &commit.Hash, nil
 }
 
-func (r *Repository) RepackObjects() (err error) {
+type RepackConfig struct {
+	// UseRefDeltas configures whether packfile encoder will use reference deltas.
+	// By default OFSDeltaObject is used.
+	UseRefDeltas bool
+	// PackWindow for packing objects.
+	PackWindow uint
+	// OnlyDeletePacksOlderThan if set to non-zero value
+	// selects only objects older than the time provided.
+	OnlyDeletePacksOlderThan time.Time
+}
+
+func (r *Repository) RepackObjects(cfg *RepackConfig) (err error) {
 	// Get the existing object packs.
 	hs, err := r.Storer.ObjectPacks()
 	if err != nil {
@@ -1021,7 +1033,7 @@ func (r *Repository) RepackObjects() (err error) {
 	}
 
 	// Create a new pack.
-	nh, err := r.createNewObjectPack()
+	nh, err := r.createNewObjectPack(cfg)
 	if err != nil {
 		return err
 	}
@@ -1032,7 +1044,7 @@ func (r *Repository) RepackObjects() (err error) {
 		if h == nh {
 			continue
 		}
-		err = r.Storer.DeleteObjectPackAndIndex(h)
+		err = r.Storer.DeleteOldObjectPackAndIndex(h, cfg.OnlyDeletePacksOlderThan)
 		if err != nil {
 			return err
 		}
@@ -1044,12 +1056,12 @@ func (r *Repository) RepackObjects() (err error) {
 // createNewObjectPack is a helper for RepackObjects taking care
 // of creating a new pack. It is used so the the PackfileWriter
 // deferred close has the right scope.
-func (r *Repository) createNewObjectPack() (h plumbing.Hash, err error) {
+func (r *Repository) createNewObjectPack(cfg *RepackConfig) (h plumbing.Hash, err error) {
 	pfw, ok := r.Storer.(storer.PackfileWriter)
 	if !ok {
 		return h, fmt.Errorf("Repository storer is not a storer.PackfileWriter")
 	}
-	wc, err := pfw.PackfileWriter(nil)
+	wc, err := pfw.PackfileWriter()
 	if err != nil {
 		return h, err
 	}
@@ -1066,8 +1078,8 @@ func (r *Repository) createNewObjectPack() (h plumbing.Hash, err error) {
 	if err != nil {
 		return h, err
 	}
-	enc := packfile.NewEncoder(wc, r.Storer, false)
-	h, err = enc.Encode(objs, 10, nil)
+	enc := packfile.NewEncoder(wc, r.Storer, cfg.UseRefDeltas)
+	h, err = enc.Encode(objs, cfg.PackWindow)
 	if err != nil {
 		return h, err
 	}
diff --git a/storage/filesystem/internal/dotgit/dotgit.go b/storage/filesystem/internal/dotgit/dotgit.go
index 09688c1..cd4f517 100644
--- a/storage/filesystem/internal/dotgit/dotgit.go
+++ b/storage/filesystem/internal/dotgit/dotgit.go
@@ -9,6 +9,7 @@ import (
 	stdioutil "io/ioutil"
 	"os"
 	"strings"
+	"time"
 
 	"gopkg.in/src-d/go-git.v4/plumbing"
 	"gopkg.in/src-d/go-git.v4/utils/ioutil"
@@ -193,8 +194,19 @@ func (d *DotGit) ObjectPackIdx(hash plumbing.Hash) (billy.File, error) {
 	return d.objectPackOpen(hash, `idx`)
 }
 
-func (d *DotGit) DeleteObjectPackAndIndex(hash plumbing.Hash) error {
-	err := d.fs.Remove(d.objectPackPath(hash, `pack`))
+func (d *DotGit) DeleteOldObjectPackAndIndex(hash plumbing.Hash, t time.Time) error {
+	path := d.objectPackPath(hash, `pack`)
+	if !t.IsZero() {
+		fi, err := d.fs.Stat(path)
+		if err != nil {
+			return err
+		}
+		// too new, skip deletion.
+		if !fi.ModTime().Before(t) {
+			return nil
+		}
+	}
+	err := d.fs.Remove(path)
 	if err != nil {
 		return err
 	}
diff --git a/storage/filesystem/object.go b/storage/filesystem/object.go
index f235204..13b8ddb 100644
--- a/storage/filesystem/object.go
+++ b/storage/filesystem/object.go
@@ -508,6 +508,6 @@ func (s *ObjectStorage) ObjectPacks() ([]plumbing.Hash, error) {
 	return s.dir.ObjectPacks()
 }
 
-func (s *ObjectStorage) DeleteObjectPackAndIndex(h plumbing.Hash) error {
-	return s.dir.DeleteObjectPackAndIndex(h)
+func (s *ObjectStorage) DeleteOldObjectPackAndIndex(h plumbing.Hash, t time.Time) error {
+	return s.dir.DeleteOldObjectPackAndIndex(h, t)
 }
diff --git a/storage/memory/storage.go b/storage/memory/storage.go
index 5049036..2d1a4b6 100644
--- a/storage/memory/storage.go
+++ b/storage/memory/storage.go
@@ -173,8 +173,8 @@ func (o *ObjectStorage) ForEachObjectHash(fun func(plumbing.Hash) error) error {
 func (o *ObjectStorage) ObjectPacks() ([]plumbing.Hash, error) {
 	return nil, nil
 }
-func (o *ObjectStorage) DeleteObjectPackAndIndex(plumbing.Hash) error {
-	return errNotSupported
+func (o *ObjectStorage) DeleteOldObjectPackAndIndex(plumbing.Hash, time.Time) error {
+	return nil
 }
 
 var errNotSupported = fmt.Errorf("Not supported")