2021-05-17 00:39:33 +08:00
// Package bisync implements bisync
// Copyright (c) 2017-2020 Chris Nelson
package bisync
import (
"context"
"crypto/md5"
"encoding/hex"
2021-11-04 18:12:57 +08:00
"errors"
"fmt"
2021-05-17 00:39:33 +08:00
"io"
"os"
"path/filepath"
"strings"
"time"
"github.com/rclone/rclone/cmd"
"github.com/rclone/rclone/cmd/bisync/bilib"
"github.com/rclone/rclone/fs"
"github.com/rclone/rclone/fs/config"
"github.com/rclone/rclone/fs/config/flags"
"github.com/rclone/rclone/fs/filter"
"github.com/rclone/rclone/fs/hash"
"github.com/spf13/cobra"
)
2023-10-07 04:38:47 +08:00
// TestFunc allows mocking errors during tests
type TestFunc func ( )
2021-05-17 00:39:33 +08:00
// Options keep bisync options
type Options struct {
2023-07-11 18:40:01 +08:00
Resync bool
CheckAccess bool
CheckFilename string
CheckSync CheckSyncMode
2023-07-11 19:09:06 +08:00
CreateEmptySrcDirs bool
2023-07-11 18:40:01 +08:00
RemoveEmptyDirs bool
MaxDelete int // percentage from 0 to 100
Force bool
FiltersFile string
Workdir string
2023-11-12 23:34:38 +08:00
OrigBackupDir string
BackupDir1 string
BackupDir2 string
2023-07-11 18:40:01 +08:00
DryRun bool
NoCleanup bool
SaveQueues bool // save extra debugging files (test only flag)
IgnoreListingChecksum bool
2023-07-11 18:57:49 +08:00
Resilient bool
2023-10-07 04:38:47 +08:00
TestFn TestFunc // test-only option, for mocking errors
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
2023-11-11 11:56:28 +08:00
Retries int
2021-05-17 00:39:33 +08:00
}
// Default values
const (
DefaultMaxDelete int = 50
DefaultCheckFilename string = "RCLONE_TEST"
)
// DefaultWorkdir is default working directory
var DefaultWorkdir = filepath . Join ( config . GetCacheDir ( ) , "bisync" )
// CheckSyncMode controls when to compare final listings
type CheckSyncMode int
// CheckSync modes
const (
CheckSyncTrue CheckSyncMode = iota // Compare final listings (default)
CheckSyncFalse // Disable comparison of final listings
CheckSyncOnly // Only compare listings from the last run, do not sync
)
func ( x CheckSyncMode ) String ( ) string {
switch x {
case CheckSyncTrue :
return "true"
case CheckSyncFalse :
return "false"
case CheckSyncOnly :
return "only"
}
return "unknown"
}
// Set a CheckSync mode from a string
func ( x * CheckSyncMode ) Set ( s string ) error {
switch strings . ToLower ( s ) {
case "true" :
* x = CheckSyncTrue
case "false" :
* x = CheckSyncFalse
case "only" :
* x = CheckSyncOnly
default :
2021-11-04 18:12:57 +08:00
return fmt . Errorf ( "unknown check-sync mode for bisync: %q" , s )
2021-05-17 00:39:33 +08:00
}
return nil
}
// Type of the CheckSync value
func ( x * CheckSyncMode ) Type ( ) string {
return "string"
}
// Opt keeps command line options
var Opt Options
func init ( ) {
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
2023-11-11 11:56:28 +08:00
Opt . Retries = 3
2021-05-17 00:39:33 +08:00
cmd . Root . AddCommand ( commandDefinition )
cmdFlags := commandDefinition . Flags ( )
2023-11-12 23:34:38 +08:00
// when adding new flags, remember to also update the rc params:
// cmd/bisync/rc.go cmd/bisync/help.go (not docs/content/rc.md)
2023-07-11 01:34:10 +08:00
flags . BoolVarP ( cmdFlags , & Opt . Resync , "resync" , "1" , Opt . Resync , "Performs the resync run. Path1 files may overwrite Path2 versions. Consider using --verbose or --dry-run first." , "" )
flags . BoolVarP ( cmdFlags , & Opt . CheckAccess , "check-access" , "" , Opt . CheckAccess , makeHelp ( "Ensure expected {CHECKFILE} files are found on both Path1 and Path2 filesystems, else abort." ) , "" )
flags . StringVarP ( cmdFlags , & Opt . CheckFilename , "check-filename" , "" , Opt . CheckFilename , makeHelp ( "Filename for --check-access (default: {CHECKFILE})" ) , "" )
flags . BoolVarP ( cmdFlags , & Opt . Force , "force" , "" , Opt . Force , "Bypass --max-delete safety check and run the sync. Consider using with --verbose" , "" )
flags . FVarP ( cmdFlags , & Opt . CheckSync , "check-sync" , "" , "Controls comparison of final listings: true|false|only (default: true)" , "" )
2023-07-11 19:09:06 +08:00
flags . BoolVarP ( cmdFlags , & Opt . CreateEmptySrcDirs , "create-empty-src-dirs" , "" , Opt . CreateEmptySrcDirs , "Sync creation and deletion of empty directories. (Not compatible with --remove-empty-dirs)" , "" )
flags . BoolVarP ( cmdFlags , & Opt . RemoveEmptyDirs , "remove-empty-dirs" , "" , Opt . RemoveEmptyDirs , "Remove ALL empty directories at the final cleanup step." , "" )
2023-07-11 01:34:10 +08:00
flags . StringVarP ( cmdFlags , & Opt . FiltersFile , "filters-file" , "" , Opt . FiltersFile , "Read filtering patterns from a file" , "" )
flags . StringVarP ( cmdFlags , & Opt . Workdir , "workdir" , "" , Opt . Workdir , makeHelp ( "Use custom working dir - useful for testing. (default: {WORKDIR})" ) , "" )
2023-11-12 23:34:38 +08:00
flags . StringVarP ( cmdFlags , & Opt . BackupDir1 , "backup-dir1" , "" , Opt . BackupDir1 , "--backup-dir for Path1. Must be a non-overlapping path on the same remote." , "" )
flags . StringVarP ( cmdFlags , & Opt . BackupDir2 , "backup-dir2" , "" , Opt . BackupDir2 , "--backup-dir for Path2. Must be a non-overlapping path on the same remote." , "" )
2023-07-11 01:34:10 +08:00
flags . BoolVarP ( cmdFlags , & tzLocal , "localtime" , "" , tzLocal , "Use local time in listings (default: UTC)" , "" )
flags . BoolVarP ( cmdFlags , & Opt . NoCleanup , "no-cleanup" , "" , Opt . NoCleanup , "Retain working files (useful for troubleshooting and testing)." , "" )
2023-07-11 18:40:01 +08:00
flags . BoolVarP ( cmdFlags , & Opt . IgnoreListingChecksum , "ignore-listing-checksum" , "" , Opt . IgnoreListingChecksum , "Do not use checksums for listings (add --ignore-checksum to additionally skip post-copy checksum checks)" , "" )
2023-07-11 18:57:49 +08:00
flags . BoolVarP ( cmdFlags , & Opt . Resilient , "resilient" , "" , Opt . Resilient , "Allow future runs to retry after certain less-serious errors, instead of requiring --resync. Use at your own risk!" , "" )
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
bisync: high-level retries if --resilient
Before this change, bisync had no ability to retry in the event of sync errors.
After this change, bisync will retry if --resilient is passed, but only in one
direction at a time. We can safely retry in one direction because the source is
still intact, even if the dest was left in a messy state. If the first
direction still fails after our final retry, we abort and do NOT continue in
the other direction, to prevent the messy dest from polluting the source. If
the first direction succeeds, we do then allow retries in the other direction.
The number of retries is controllable by --retries (default 3)
2023-11-11 11:56:28 +08:00
flags . IntVarP ( cmdFlags , & Opt . Retries , "retries" , "" , Opt . Retries , "Retry operations this many times if they fail" , "" )
2021-05-17 00:39:33 +08:00
}
// bisync command definition
var commandDefinition = & cobra . Command {
Use : "bisync remote1:path1 remote2:path2" ,
Short : shortHelp ,
Long : longHelp ,
2022-11-27 06:40:49 +08:00
Annotations : map [ string ] string {
"versionIntroduced" : "v1.58" ,
2023-07-11 01:34:10 +08:00
"groups" : "Filter,Copy,Important" ,
2022-11-27 06:40:49 +08:00
} ,
2021-05-17 00:39:33 +08:00
RunE : func ( command * cobra . Command , args [ ] string ) error {
cmd . CheckArgs ( 2 , 2 , command , args )
fs1 , file1 , fs2 , file2 := cmd . NewFsSrcDstFiles ( args )
if file1 != "" || file2 != "" {
return errors . New ( "paths must be existing directories" )
}
ctx := context . Background ( )
opt := Opt
opt . applyContext ( ctx )
if tzLocal {
TZ = time . Local
}
commonHashes := fs1 . Hashes ( ) . Overlap ( fs2 . Hashes ( ) )
isDropbox1 := strings . HasPrefix ( fs1 . String ( ) , "Dropbox" )
isDropbox2 := strings . HasPrefix ( fs2 . String ( ) , "Dropbox" )
if commonHashes == hash . Set ( 0 ) && ( isDropbox1 || isDropbox2 ) {
ci := fs . GetConfig ( ctx )
if ! ci . DryRun && ! ci . RefreshTimes {
fs . Debugf ( nil , "Using flag --refresh-times is recommended" )
}
}
fs . Logf ( nil , "bisync is EXPERIMENTAL. Don't use in production!" )
cmd . Run ( false , true , command , func ( ) error {
err := Bisync ( ctx , fs1 , fs2 , & opt )
if err == ErrBisyncAborted {
os . Exit ( 2 )
}
return err
} )
return nil
} ,
}
func ( opt * Options ) applyContext ( ctx context . Context ) {
maxDelete := DefaultMaxDelete
ci := fs . GetConfig ( ctx )
if ci . MaxDelete >= 0 {
maxDelete = int ( ci . MaxDelete )
}
if maxDelete < 0 {
maxDelete = 0
}
if maxDelete > 100 {
maxDelete = 100
}
opt . MaxDelete = maxDelete
// reset MaxDelete for fs/operations, bisync handles this parameter specially
ci . MaxDelete = - 1
opt . DryRun = ci . DryRun
}
func ( opt * Options ) setDryRun ( ctx context . Context ) context . Context {
ctxNew , ci := fs . AddConfig ( ctx )
ci . DryRun = opt . DryRun
return ctxNew
}
func ( opt * Options ) applyFilters ( ctx context . Context ) ( context . Context , error ) {
filtersFile := opt . FiltersFile
if filtersFile == "" {
return ctx , nil
}
f , err := os . Open ( filtersFile )
if err != nil {
2021-11-04 18:12:57 +08:00
return ctx , fmt . Errorf ( "specified filters file does not exist: %s" , filtersFile )
2021-05-17 00:39:33 +08:00
}
fs . Infof ( nil , "Using filters file %s" , filtersFile )
hasher := md5 . New ( )
if _ , err := io . Copy ( hasher , f ) ; err != nil {
_ = f . Close ( )
return ctx , err
}
gotHash := hex . EncodeToString ( hasher . Sum ( nil ) )
_ = f . Close ( )
hashFile := filtersFile + ".md5"
2022-08-20 22:38:02 +08:00
wantHash , err := os . ReadFile ( hashFile )
2021-05-17 00:39:33 +08:00
if err != nil && ! opt . Resync {
2021-11-04 18:12:57 +08:00
return ctx , fmt . Errorf ( "filters file md5 hash not found (must run --resync): %s" , filtersFile )
2021-05-17 00:39:33 +08:00
}
if gotHash != string ( wantHash ) && ! opt . Resync {
2021-11-04 18:12:57 +08:00
return ctx , fmt . Errorf ( "filters file has changed (must run --resync): %s" , filtersFile )
2021-05-17 00:39:33 +08:00
}
if opt . Resync {
2023-07-11 16:35:01 +08:00
if opt . DryRun {
fs . Infof ( nil , "Skipped storing filters file hash to %s as --dry-run is set" , hashFile )
} else {
2023-07-11 18:40:01 +08:00
fs . Infof ( nil , "Storing filters file hash to %s" , hashFile )
if err := os . WriteFile ( hashFile , [ ] byte ( gotHash ) , bilib . PermSecure ) ; err != nil {
return ctx , err
2023-07-11 16:35:01 +08:00
}
2021-05-17 00:39:33 +08:00
}
}
// Prepend our filter file first in the list
filterOpt := filter . GetConfig ( ctx ) . Opt
filterOpt . FilterFrom = append ( [ ] string { filtersFile } , filterOpt . FilterFrom ... )
newFilter , err := filter . NewFilter ( & filterOpt )
if err != nil {
2021-11-04 18:12:57 +08:00
return ctx , fmt . Errorf ( "invalid filters file: %s: %w" , filtersFile , err )
2021-05-17 00:39:33 +08:00
}
return filter . ReplaceConfig ( ctx , newFilter ) , nil
}