rclone/fs/dir.go
Saksham Khanna 4d8ef7bca7
cmd/dedupe: make largest directory primary to minimize data moved (#3648)
This change makes dedupe recursively count elements in same-named directories
and make the largest one primary. This allows to minimize the amount of data
moved (or at least the amount of API calls) when dedupe merges them.
It also adds a new fs.Object interface `ParentIDer` with function `ParentID` and
implements it for the drive and opendrive backends. This function returns
parent directory ID for objects on filesystems that allow same-named dirs.
We use it to correctly count sizes of same-named directories.

Fixes #2568

Co-authored-by: Ivan Andreev <ivandeex@gmail.com>
2021-03-11 20:40:29 +03:00

113 lines
2.4 KiB
Go

package fs
import (
"context"
"time"
)
// Dir describes an unspecialized directory for directory/container/bucket lists
type Dir struct {
remote string // name of the directory
modTime time.Time // modification or creation time - IsZero for unknown
size int64 // size of directory and contents or -1 if unknown
items int64 // number of objects or -1 for unknown
id string // optional ID
parent string // optional parent directory ID
}
// NewDir creates an unspecialized Directory object
func NewDir(remote string, modTime time.Time) *Dir {
return &Dir{
remote: remote,
modTime: modTime,
size: -1,
items: -1,
}
}
// NewDirCopy creates an unspecialized copy of the Directory object passed in
func NewDirCopy(ctx context.Context, d Directory) *Dir {
return &Dir{
remote: d.Remote(),
modTime: d.ModTime(ctx),
size: d.Size(),
items: d.Items(),
id: d.ID(),
}
}
// String returns the name
func (d *Dir) String() string {
return d.remote
}
// Remote returns the remote path
func (d *Dir) Remote() string {
return d.remote
}
// SetRemote sets the remote
func (d *Dir) SetRemote(remote string) *Dir {
d.remote = remote
return d
}
// ID gets the optional ID
func (d *Dir) ID() string {
return d.id
}
// SetID sets the optional ID
func (d *Dir) SetID(id string) *Dir {
d.id = id
return d
}
// ParentID returns the IDs of the Dir parent if known
func (d *Dir) ParentID() string {
return d.parent
}
// SetParentID sets the optional parent ID of the Dir
func (d *Dir) SetParentID(parent string) *Dir {
d.parent = parent
return d
}
// ModTime returns the modification date of the file
// It should return a best guess if one isn't available
func (d *Dir) ModTime(ctx context.Context) time.Time {
if !d.modTime.IsZero() {
return d.modTime
}
return time.Now()
}
// Size returns the size of the file
func (d *Dir) Size() int64 {
return d.size
}
// SetSize sets the size of the directory
func (d *Dir) SetSize(size int64) *Dir {
d.size = size
return d
}
// Items returns the count of items in this directory or this
// directory and subdirectories if known, -1 for unknown
func (d *Dir) Items() int64 {
return d.items
}
// SetItems sets the number of items in the directory
func (d *Dir) SetItems(items int64) *Dir {
d.items = items
return d
}
// Check interfaces
var (
_ DirEntry = (*Dir)(nil)
_ Directory = (*Dir)(nil)
)