felixfoertsch/syncthing
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 10 Wiki Activity

refactor: use modern Protobuf encoder (#9817)

Browse Source 
At a high level, this is what I've done and why:

- I'm moving the protobuf generation for the `protocol`, `discovery` and
`db` packages to the modern alternatives, and using `buf` to generate
because it's nice and simple.
- After trying various approaches on how to integrate the new types with
the existing code, I opted for splitting off our own data model types
from the on-the-wire generated types. This means we can have a
`FileInfo` type with nicer ergonomics and lots of methods, while the
protobuf generated type stays clean and close to the wire protocol. It
does mean copying between the two when required, which certainly adds a
small amount of inefficiency. If we want to walk this back in the future
and use the raw generated type throughout, that's possible, this however
makes the refactor smaller (!) as it doesn't change everything about the
type for everyone at the same time.
- I have simply removed in cold blood a significant number of old
database migrations. These depended on previous generations of generated
messages of various kinds and were annoying to support in the new
fashion. The oldest supported database version now is the one from
Syncthing 1.9.0 from Sep 7, 2020.
- I changed config structs to be regular manually defined structs.

For the sake of discussion, some things I tried that turned out not to
work...

### Embedding / wrapping

Embedding the protobuf generated structs in our existing types as a data
container and keeping our methods and stuff:

```
package protocol

type FileInfo struct {
  *generated.FileInfo
}
```

This generates a lot of problems because the internal shape of the
generated struct is quite different (different names, different types,
more pointers), because initializing it doesn't work like you'd expect
(i.e., you end up with an embedded nil pointer and a panic), and because
the types of child types don't get wrapped. That is, even if we also
have a similar wrapper around a `Vector`, that's not the type you get
when accessing `someFileInfo.Version`, you get the `*generated.Vector`
that doesn't have methods, etc.

### Aliasing

```
package protocol

type FileInfo = generated.FileInfo
```

Doesn't help because you can't attach methods to it, plus all the above.

### Generating the types into the target package like we do now and
attaching methods

This fails because of the different shape of the generated type (as in
the embedding case above) plus the generated struct already has a bunch
of methods that we can't necessarily override properly (like `String()`
and a bunch of getters).

### Methods to functions

I considered just moving all the methods we attach to functions in a
specific package, so that for example

```
package protocol

func (f FileInfo) Equal(other FileInfo) bool
```

would become

```
package fileinfos

func Equal(a, b *generated.FileInfo) bool
```

and this would mostly work, but becomes quite verbose and cumbersome,
and somewhat limits discoverability (you can't see what methods are
available on the type in auto completions, etc). In the end I did this
in some cases, like in the database layer where a lot of things like
`func (fv *FileVersion) IsEmpty() bool` becomes `func fvIsEmpty(fv
*generated.FileVersion)` because they were anyway just internal methods.

Fixes #8247
This commit is contained in:
Jakob Borg
2024-12-01 16:50:17 +01:00
committed by GitHub
parent 2b8ee4c7a5
commit 77970d5113
203 changed files with 7437 additions and 28636 deletions
Download Patch File Download Diff File Expand all files Collapse all files
cmd/syncthing/cli/index_dump.go
+22 -18
View File
@@ -11,6 +11,10 @@ import (
"fmt"
"time"
"google.golang.org/protobuf/proto"
"github.com/syncthing/syncthing/internal/gen/bep"
"github.com/syncthing/syncthing/internal/gen/dbproto"
"github.com/syncthing/syncthing/lib/db"
"github.com/syncthing/syncthing/lib/protocol"
)
@@ -33,19 +37,19 @@ func indexDump() error {
name := nulString(key[1+4+4:])
fmt.Printf("[device] F:%d D:%d N:%q", folder, device, name)
var f protocol.FileInfo
err := f.Unmarshal(it.Value())
var f bep.FileInfo
err := proto.Unmarshal(it.Value(), &f)
if err != nil {
return err
}
fmt.Printf(" V:%v\n", f)
fmt.Printf(" V:%v\n", &f)
case db.KeyTypeGlobal:
folder := binary.BigEndian.Uint32(key[1:])
name := nulString(key[1+4:])
var flv db.VersionList
flv.Unmarshal(it.Value())
fmt.Printf("[global] F:%d N:%q V:%s\n", folder, name, flv)
var flv dbproto.VersionList
proto.Unmarshal(it.Value(), &flv)
fmt.Printf("[global] F:%d N:%q V:%s\n", folder, name, &flv)
case db.KeyTypeBlock:
folder := binary.BigEndian.Uint32(key[1:])
@@ -94,11 +98,11 @@ func indexDump() error {
case db.KeyTypeFolderMeta:
folder := binary.BigEndian.Uint32(key[1:])
fmt.Printf("[foldermeta] F:%d", folder)
var cs db.CountsSet
if err := cs.Unmarshal(it.Value()); err != nil {
var cs dbproto.CountsSet
if err := proto.Unmarshal(it.Value(), &cs); err != nil {
fmt.Printf(" (invalid)\n")
} else {
fmt.Printf(" V:%v\n", cs)
fmt.Printf(" V:%v\n", &cs)
}
case db.KeyTypeMiscData:
@@ -125,20 +129,20 @@ func indexDump() error {
case db.KeyTypeVersion:
fmt.Printf("[version] H:%x", key[1:])
var v protocol.Vector
err := v.Unmarshal(it.Value())
var v bep.Vector
err := proto.Unmarshal(it.Value(), &v)
if err != nil {
fmt.Printf(" (invalid)\n")
} else {
fmt.Printf(" V:%v\n", v)
fmt.Printf(" V:%v\n", &v)
}
case db.KeyTypePendingFolder:
device := binary.BigEndian.Uint32(key[1:])
folder := string(key[5:])
var of db.ObservedFolder
of.Unmarshal(it.Value())
fmt.Printf("[pendingFolder] D:%d F:%s V:%v\n", device, folder, of)
var of dbproto.ObservedFolder
proto.Unmarshal(it.Value(), &of)
fmt.Printf("[pendingFolder] D:%d F:%s V:%v\n", device, folder, &of)
case db.KeyTypePendingDevice:
device := "<invalid>"
@@ -146,9 +150,9 @@ func indexDump() error {
if err == nil {
device = dev.String()
}
var od db.ObservedDevice
od.Unmarshal(it.Value())
fmt.Printf("[pendingDevice] D:%v V:%v\n", device, od)
var od dbproto.ObservedDevice
proto.Unmarshal(it.Value(), &od)
fmt.Printf("[pendingDevice] D:%v V:%v\n", device, &od)
default:
fmt.Printf("[??? %d]\n %x\n %x\n", key[0], key, it.Value())
cmd/syncthing/cli/index_idxck.go
+106 -27
View File
@@ -13,6 +13,10 @@ import (
"fmt"
"sort"
"google.golang.org/protobuf/proto"
"github.com/syncthing/syncthing/internal/gen/bep"
"github.com/syncthing/syncthing/internal/gen/dbproto"
"github.com/syncthing/syncthing/lib/db"
"github.com/syncthing/syncthing/lib/protocol"
)
@@ -42,12 +46,12 @@ func indexCheck() (err error) {
folders := make(map[uint32]string)
devices := make(map[uint32]string)
deviceToIDs := make(map[string]uint32)
fileInfos := make(map[fileInfoKey]protocol.FileInfo)
globals := make(map[globalKey]db.VersionList)
fileInfos := make(map[fileInfoKey]*bep.FileInfo)
globals := make(map[globalKey]*dbproto.VersionList)
sequences := make(map[sequenceKey]string)
needs := make(map[globalKey]struct{})
blocklists := make(map[string]struct{})
versions := make(map[string]protocol.Vector)
versions := make(map[string]*bep.Vector)
usedBlocklists := make(map[string]struct{})
usedVersions := make(map[string]struct{})
var localDeviceKey uint32
@@ -74,26 +78,26 @@ func indexCheck() (err error) {
device := binary.BigEndian.Uint32(key[1+4:])
name := nulString(key[1+4+4:])
var f protocol.FileInfo
err := f.Unmarshal(it.Value())
var f bep.FileInfo
err := proto.Unmarshal(it.Value(), &f)
if err != nil {
fmt.Println("Unable to unmarshal FileInfo:", err)
success = false
continue
}
fileInfos[fileInfoKey{folder, device, name}] = f
fileInfos[fileInfoKey{folder, device, name}] = &f
case db.KeyTypeGlobal:
folder := binary.BigEndian.Uint32(key[1:])
name := nulString(key[1+4:])
var flv db.VersionList
if err := flv.Unmarshal(it.Value()); err != nil {
var flv dbproto.VersionList
if err := proto.Unmarshal(it.Value(), &flv); err != nil {
fmt.Println("Unable to unmarshal VersionList:", err)
success = false
continue
}
globals[globalKey{folder, name}] = flv
globals[globalKey{folder, name}] = &flv
case db.KeyTypeFolderIdx:
key := binary.BigEndian.Uint32(it.Key()[1:])
@@ -124,13 +128,13 @@ func indexCheck() (err error) {
case db.KeyTypeVersion:
hash := string(key[1:])
var v protocol.Vector
if err := v.Unmarshal(it.Value()); err != nil {
var v bep.Vector
if err := proto.Unmarshal(it.Value(), &v); err != nil {
fmt.Println("Unable to unmarshal Vector:", err)
success = false
continue
}
versions[hash] = v
versions[hash] = &v
}
}
@@ -248,25 +252,27 @@ func indexCheck() (err error) {
if fi.VersionHash != nil {
fiv = versions[string(fi.VersionHash)]
}
if !fiv.Equal(version) {
if !protocol.VectorFromWire(fiv).Equal(version) {
fmt.Printf("VersionList %q, folder %q, entry %d, FileInfo version mismatch, %v (VersionList) != %v (FileInfo)\n", gk.name, folder, i, version, fi.Version)
success = false
}
if fi.IsInvalid() != invalid {
fmt.Printf("VersionList %q, folder %q, entry %d, FileInfo invalid mismatch, %v (VersionList) != %v (FileInfo)\n", gk.name, folder, i, invalid, fi.IsInvalid())
ffi := protocol.FileInfoFromDB(fi)
if ffi.IsInvalid() != invalid {
fmt.Printf("VersionList %q, folder %q, entry %d, FileInfo invalid mismatch, %v (VersionList) != %v (FileInfo)\n", gk.name, folder, i, invalid, ffi.IsInvalid())
success = false
}
if fi.IsDeleted() != deleted {
fmt.Printf("VersionList %q, folder %q, entry %d, FileInfo deleted mismatch, %v (VersionList) != %v (FileInfo)\n", gk.name, folder, i, deleted, fi.IsDeleted())
if ffi.IsDeleted() != deleted {
fmt.Printf("VersionList %q, folder %q, entry %d, FileInfo deleted mismatch, %v (VersionList) != %v (FileInfo)\n", gk.name, folder, i, deleted, ffi.IsDeleted())
success = false
}
}
for i, fv := range vl.RawVersions {
for i, fv := range vl.Versions {
ver := protocol.VectorFromWire(fv.Version)
for _, device := range fv.Devices {
checkGlobal(i, device, fv.Version, false, fv.Deleted)
checkGlobal(i, device, ver, false, fv.Deleted)
}
for _, device := range fv.InvalidDevices {
checkGlobal(i, device, fv.Version, true, fv.Deleted)
checkGlobal(i, device, ver, true, fv.Deleted)
}
}
@@ -276,10 +282,10 @@ func indexCheck() (err error) {
if needsLocally(vl) {
_, ok := needs[gk]
if !ok {
fv, _ := vl.GetGlobal()
devB, _ := fv.FirstDevice()
fv, _ := vlGetGlobal(vl)
devB, _ := fvFirstDevice(fv)
dev := deviceToIDs[string(devB)]
fi := fileInfos[fileInfoKey{gk.folder, dev, gk.name}]
fi := protocol.FileInfoFromDB(fileInfos[fileInfoKey{gk.folder, dev, gk.name}])
if !fi.IsDeleted() && !fi.IsIgnored() {
fmt.Printf("Missing need entry for needed file %q, folder %q\n", gk.name, folder)
}
@@ -345,11 +351,84 @@ func indexCheck() (err error) {
return nil
}
func needsLocally(vl db.VersionList) bool {
gfv, gok := vl.GetGlobal()
func needsLocally(vl *dbproto.VersionList) bool {
gfv, gok := vlGetGlobal(vl)
if !gok { // That's weird, but we hardly need something non-existent
return false
}
fv, ok := vl.Get(protocol.LocalDeviceID[:])
return db.Need(gfv, ok, fv.Version)
fv, ok := vlGet(vl, protocol.LocalDeviceID[:])
return db.Need(gfv, ok, protocol.VectorFromWire(fv.Version))
}
// Get returns a FileVersion that contains the given device and whether it has
// been found at all.
func vlGet(vl *dbproto.VersionList, device []byte) (*dbproto.FileVersion, bool) {
_, i, _, ok := vlFindDevice(vl, device)
if !ok {
return &dbproto.FileVersion{}, false
}
return vl.Versions[i], true
}
// GetGlobal returns the current global FileVersion. The returned FileVersion
// may be invalid, if all FileVersions are invalid. Returns false only if
// VersionList is empty.
func vlGetGlobal(vl *dbproto.VersionList) (*dbproto.FileVersion, bool) {
i := vlFindGlobal(vl)
if i == -1 {
return nil, false
}
return vl.Versions[i], true
}
// findGlobal returns the first version that isn't invalid, or if all versions are
// invalid just the first version (i.e. 0) or -1, if there's no versions at all.
func vlFindGlobal(vl *dbproto.VersionList) int {
for i := range vl.Versions {
if !fvIsInvalid(vl.Versions[i]) {
return i
}
}
if len(vl.Versions) == 0 {
return -1
}
return 0
}
// findDevice returns whether the device is in InvalidVersions or Versions and
// in InvalidDevices or Devices (true for invalid), the positions in the version
// and device slices and whether it has been found at all.
func vlFindDevice(vl *dbproto.VersionList, device []byte) (bool, int, int, bool) {
for i, v := range vl.Versions {
if j := deviceIndex(v.Devices, device); j != -1 {
return false, i, j, true
}
if j := deviceIndex(v.InvalidDevices, device); j != -1 {
return true, i, j, true
}
}
return false, -1, -1, false
}
func deviceIndex(devices [][]byte, device []byte) int {
for i, dev := range devices {
if bytes.Equal(device, dev) {
return i
}
}
return -1
}
func fvFirstDevice(fv *dbproto.FileVersion) ([]byte, bool) {
if len(fv.Devices) != 0 {
return fv.Devices[0], true
}
if len(fv.InvalidDevices) != 0 {
return fv.InvalidDevices[0], true
}
return nil, false
}
func fvIsInvalid(fv *dbproto.FileVersion) bool {
return fv == nil || len(fv.Devices) == 0
}