status-go/protocol/communities/manager.go

5109 lines
152 KiB
Go
Raw Normal View History

package communities
import (
"context"
"crypto/ecdsa"
"database/sql"
"fmt"
"io/ioutil"
"net"
"os"
"sort"
"strings"
"sync"
2021-01-11 10:32:51 +00:00
"time"
"github.com/anacrolix/torrent"
"github.com/anacrolix/torrent/bencode"
"github.com/anacrolix/torrent/metainfo"
"github.com/golang/protobuf/proto"
"github.com/waku-org/go-waku/waku/v2/protocol/relay"
"github.com/google/uuid"
"github.com/pkg/errors"
"go.uber.org/zap"
gethcommon "github.com/ethereum/go-ethereum/common"
"github.com/ethereum/go-ethereum/common/hexutil"
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
"github.com/status-im/status-go/account"
"github.com/status-im/status-go/eth-node/crypto"
"github.com/status-im/status-go/eth-node/types"
"github.com/status-im/status-go/images"
"github.com/status-im/status-go/params"
2021-01-11 10:32:51 +00:00
"github.com/status-im/status-go/protocol/common"
community_token "github.com/status-im/status-go/protocol/communities/token"
"github.com/status-im/status-go/protocol/encryption"
2021-01-11 10:32:51 +00:00
"github.com/status-im/status-go/protocol/ens"
"github.com/status-im/status-go/protocol/protobuf"
2021-01-11 10:32:51 +00:00
"github.com/status-im/status-go/protocol/requests"
"github.com/status-im/status-go/protocol/transport"
"github.com/status-im/status-go/services/communitytokens"
"github.com/status-im/status-go/services/wallet/bigint"
walletcommon "github.com/status-im/status-go/services/wallet/common"
"github.com/status-im/status-go/services/wallet/thirdparty"
"github.com/status-im/status-go/services/wallet/token"
"github.com/status-im/status-go/signal"
)
var defaultAnnounceList = [][]string{
{"udp://tracker.opentrackr.org:1337/announce"},
{"udp://tracker.openbittorrent.com:6969/announce"},
}
var pieceLength = 100 * 1024
const maxArchiveSizeInBytes = 30000000
var memberPermissionsCheckInterval = 1 * time.Hour
var validateInterval = 2 * time.Minute
// Used for testing only
func SetValidateInterval(duration time.Duration) {
validateInterval = duration
}
// errors
var (
ErrTorrentTimedout = errors.New("torrent has timed out")
ErrCommunityRequestAlreadyRejected = errors.New("that user was already rejected from the community")
)
type Manager struct {
persistence *Persistence
encryptor *encryption.Protocol
ensSubscription chan []*ens.VerificationRecord
subscriptions []chan *Subscription
ensVerifier *ens.Verifier
ownerVerifier OwnerVerifier
identity *ecdsa.PrivateKey
installationID string
accountsManager account.Manager
tokenManager TokenManager
collectiblesManager CollectiblesManager
logger *zap.Logger
stdoutLogger *zap.Logger
transport *transport.Transport
timesource common.TimeSource
quit chan struct{}
torrentConfig *params.TorrentConfig
torrentClient *torrent.Client
walletConfig *params.WalletConfig
communityTokensService communitytokens.ServiceInterface
historyArchiveTasksWaitGroup sync.WaitGroup
historyArchiveTasks sync.Map // stores `chan struct{}`
periodicMembersReevaluationTasks sync.Map // stores `chan struct{}`
torrentTasks map[string]metainfo.Hash
historyArchiveDownloadTasks map[string]*HistoryArchiveDownloadTask
stopped bool
RekeyInterval time.Duration
PermissionChecker PermissionChecker
}
type HistoryArchiveDownloadTask struct {
CancelChan chan struct{}
Waiter sync.WaitGroup
m sync.RWMutex
Cancelled bool
}
func (t *HistoryArchiveDownloadTask) IsCancelled() bool {
t.m.RLock()
defer t.m.RUnlock()
return t.Cancelled
}
func (t *HistoryArchiveDownloadTask) Cancel() {
t.m.Lock()
defer t.m.Unlock()
t.Cancelled = true
close(t.CancelChan)
}
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
type managerOptions struct {
accountsManager account.Manager
tokenManager TokenManager
collectiblesManager CollectiblesManager
walletConfig *params.WalletConfig
communityTokensService communitytokens.ServiceInterface
permissionChecker PermissionChecker
}
type TokenManager interface {
GetBalancesByChain(ctx context.Context, accounts, tokens []gethcommon.Address, chainIDs []uint64) (map[uint64]map[gethcommon.Address]map[gethcommon.Address]*hexutil.Big, error)
2023-09-11 11:45:30 +00:00
FindOrCreateTokenByAddress(ctx context.Context, chainID uint64, address gethcommon.Address) *token.Token
GetAllChainIDs() ([]uint64, error)
}
type DefaultTokenManager struct {
tokenManager *token.Manager
}
func NewDefaultTokenManager(tm *token.Manager) *DefaultTokenManager {
return &DefaultTokenManager{tokenManager: tm}
}
type BalancesByChain = map[uint64]map[gethcommon.Address]map[gethcommon.Address]*hexutil.Big
func (m *DefaultTokenManager) GetAllChainIDs() ([]uint64, error) {
networks, err := m.tokenManager.RPCClient.NetworkManager.Get(false)
if err != nil {
return nil, err
}
chainIDs := make([]uint64, 0)
for _, network := range networks {
chainIDs = append(chainIDs, network.ChainID)
}
return chainIDs, nil
}
type CollectiblesManager interface {
FetchBalancesByOwnerAndContractAddress(chainID walletcommon.ChainID, ownerAddress gethcommon.Address, contractAddresses []gethcommon.Address) (thirdparty.TokenBalancesPerContractAddress, error)
}
func (m *DefaultTokenManager) GetBalancesByChain(ctx context.Context, accounts, tokenAddresses []gethcommon.Address, chainIDs []uint64) (BalancesByChain, error) {
clients, err := m.tokenManager.RPCClient.EthClients(chainIDs)
if err != nil {
return nil, err
}
resp, err := m.tokenManager.GetBalancesByChain(context.Background(), clients, accounts, tokenAddresses)
return resp, err
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
}
2023-09-11 11:45:30 +00:00
func (m *DefaultTokenManager) FindOrCreateTokenByAddress(ctx context.Context, chainID uint64, address gethcommon.Address) *token.Token {
return m.tokenManager.FindOrCreateTokenByAddress(ctx, chainID, address)
}
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
type ManagerOption func(*managerOptions)
func WithAccountManager(accountsManager account.Manager) ManagerOption {
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
return func(opts *managerOptions) {
opts.accountsManager = accountsManager
}
}
func WithPermissionChecker(permissionChecker PermissionChecker) ManagerOption {
return func(opts *managerOptions) {
opts.permissionChecker = permissionChecker
}
}
func WithCollectiblesManager(collectiblesManager CollectiblesManager) ManagerOption {
return func(opts *managerOptions) {
opts.collectiblesManager = collectiblesManager
}
}
func WithTokenManager(tokenManager TokenManager) ManagerOption {
return func(opts *managerOptions) {
opts.tokenManager = tokenManager
}
}
func WithWalletConfig(walletConfig *params.WalletConfig) ManagerOption {
return func(opts *managerOptions) {
opts.walletConfig = walletConfig
}
}
func WithCommunityTokensService(communityTokensService communitytokens.ServiceInterface) ManagerOption {
return func(opts *managerOptions) {
opts.communityTokensService = communityTokensService
}
}
type OwnerVerifier interface {
SafeGetSignerPubKey(ctx context.Context, chainID uint64, communityID string) (string, error)
}
func NewManager(identity *ecdsa.PrivateKey, installationID string, db *sql.DB, encryptor *encryption.Protocol, logger *zap.Logger, ensverifier *ens.Verifier, ownerVerifier OwnerVerifier, transport *transport.Transport, timesource common.TimeSource, torrentConfig *params.TorrentConfig, opts ...ManagerOption) (*Manager, error) {
2021-01-11 10:32:51 +00:00
if identity == nil {
return nil, errors.New("empty identity")
}
if timesource == nil {
return nil, errors.New("no timesource")
}
var err error
if logger == nil {
if logger, err = zap.NewDevelopment(); err != nil {
return nil, errors.Wrap(err, "failed to create a logger")
}
}
stdoutLogger, err := zap.NewDevelopment()
if err != nil {
return nil, errors.Wrap(err, "failed to create archive logger")
}
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
managerConfig := managerOptions{}
for _, opt := range opts {
opt(&managerConfig)
}
2021-01-11 10:32:51 +00:00
manager := &Manager{
logger: logger,
stdoutLogger: stdoutLogger,
encryptor: encryptor,
identity: identity,
installationID: installationID,
ownerVerifier: ownerVerifier,
quit: make(chan struct{}),
transport: transport,
timesource: timesource,
torrentConfig: torrentConfig,
torrentTasks: make(map[string]metainfo.Hash),
historyArchiveDownloadTasks: make(map[string]*HistoryArchiveDownloadTask),
persistence: &Persistence{
logger: logger,
db: db,
timesource: timesource,
},
2021-01-11 10:32:51 +00:00
}
feat: add verified wallet accounts to community requests This commit extends the `CommunityRequestToJoin` with `RevealedAddresses` which represent wallet addresses and signatures provided by the sender, to proof a community owner ownership of those wallet addresses. **Note: This only works with keystore files maanged by status-go** At high level, the follwing happens: 1. User instructs Status to send a request to join to a community. By adding a password hash to the instruction, Status will try to unlock the users keystore and verify each wallet account. 2. For every verified wallet account, a signature is created for the following payload, using each wallet's private key ``` keccak256(chatkey + communityID + requestToJoinID) ``` A map of walletAddress->signature is then attached to the community request to join, which will be sent to the community owner 3. The owner node receives the request, and if the community requires users to hold tokens to become a member, it will check and verify whether the given wallet addresses are indeed owned by the sender. If any signature provided by the request cannot be recovered, the request is immediately declined by the owner. 4. The verified addresses are then added to the owner node's database such that, once the request should be accepted, the addresses can be used to check on chain whether they own the necessary funds to fulfill the community's permissions The checking of required funds is **not** part of this commit. It will be added in a follow-up commit.
2023-03-17 09:19:40 +00:00
if managerConfig.accountsManager != nil {
manager.accountsManager = managerConfig.accountsManager
}
if managerConfig.collectiblesManager != nil {
manager.collectiblesManager = managerConfig.collectiblesManager
}
if managerConfig.tokenManager != nil {
manager.tokenManager = managerConfig.tokenManager
}
if managerConfig.walletConfig != nil {
manager.walletConfig = managerConfig.walletConfig
}
if managerConfig.communityTokensService != nil {
manager.communityTokensService = managerConfig.communityTokensService
}
if ensverifier != nil {
2021-01-11 10:32:51 +00:00
sub := ensverifier.Subscribe()
2021-01-11 10:32:51 +00:00
manager.ensSubscription = sub
manager.ensVerifier = ensverifier
2021-01-11 10:32:51 +00:00
}
if managerConfig.permissionChecker != nil {
manager.PermissionChecker = managerConfig.permissionChecker
} else {
manager.PermissionChecker = &DefaultPermissionChecker{
tokenManager: manager.tokenManager,
collectiblesManager: manager.collectiblesManager,
logger: logger,
ensVerifier: ensverifier,
}
}
2021-01-11 10:32:51 +00:00
return manager, nil
}
func (m *Manager) LogStdout(msg string, fields ...zap.Field) {
m.stdoutLogger.Info(msg, fields...)
m.logger.Debug(msg, fields...)
}
type archiveMDSlice []*archiveMetadata
type archiveMetadata struct {
hash string
from uint64
}
func (md archiveMDSlice) Len() int {
return len(md)
}
func (md archiveMDSlice) Swap(i, j int) {
md[i], md[j] = md[j], md[i]
}
func (md archiveMDSlice) Less(i, j int) bool {
return md[i].from > md[j].from
}
type Subscription struct {
Community *Community
CreatingHistoryArchivesSignal *signal.CreatingHistoryArchivesSignal
HistoryArchivesCreatedSignal *signal.HistoryArchivesCreatedSignal
NoHistoryArchivesCreatedSignal *signal.NoHistoryArchivesCreatedSignal
HistoryArchivesSeedingSignal *signal.HistoryArchivesSeedingSignal
HistoryArchivesUnseededSignal *signal.HistoryArchivesUnseededSignal
HistoryArchiveDownloadedSignal *signal.HistoryArchiveDownloadedSignal
DownloadingHistoryArchivesStartedSignal *signal.DownloadingHistoryArchivesStartedSignal
DownloadingHistoryArchivesFinishedSignal *signal.DownloadingHistoryArchivesFinishedSignal
ImportingHistoryArchiveMessagesSignal *signal.ImportingHistoryArchiveMessagesSignal
CommunityEventsMessage *CommunityEventsMessage
CommunityEventsMessageInvalidClock *CommunityEventsMessageInvalidClockSignal
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
AcceptedRequestsToJoin []types.HexBytes
RejectedRequestsToJoin []types.HexBytes
CommunityPrivilegedMemberSyncMessage *CommunityPrivilegedMemberSyncMessage
TokenCommunityValidated *CommunityResponse
2021-01-11 10:32:51 +00:00
}
type CommunityResponse struct {
Community *Community `json:"community"`
Changes *CommunityChanges `json:"changes"`
RequestsToJoin []*RequestToJoin `json:"requestsToJoin"`
}
type CommunityEventsMessageInvalidClockSignal struct {
Community *Community
CommunityEventsMessage *CommunityEventsMessage
}
func (m *Manager) Subscribe() chan *Subscription {
subscription := make(chan *Subscription, 100)
m.subscriptions = append(m.subscriptions, subscription)
return subscription
}
2021-01-11 10:32:51 +00:00
func (m *Manager) Start() error {
m.stopped = false
2021-01-11 10:32:51 +00:00
if m.ensVerifier != nil {
m.runENSVerificationLoop()
}
if m.ownerVerifier != nil {
m.runOwnerVerificationLoop()
}
if m.torrentConfig != nil && m.torrentConfig.Enabled {
err := m.StartTorrentClient()
if err != nil {
m.LogStdout("couldn't start torrent client", zap.Error(err))
}
}
2021-01-11 10:32:51 +00:00
return nil
}
func (m *Manager) runENSVerificationLoop() {
go func() {
for {
select {
case <-m.quit:
m.logger.Debug("quitting ens verification loop")
return
case records, more := <-m.ensSubscription:
if !more {
m.logger.Debug("no more ens records, quitting")
return
}
m.logger.Info("received records", zap.Any("records", records))
}
}
}()
}
// Only for testing
func (m *Manager) CommunitiesToValidate() (map[string][]communityToValidate, error) { // nolint: golint
return m.persistence.getCommunitiesToValidate()
}
func (m *Manager) runOwnerVerificationLoop() {
m.logger.Info("starting owner verification loop")
go func() {
for {
select {
case <-m.quit:
m.logger.Debug("quitting owner verification loop")
return
case <-time.After(validateInterval):
// If ownerverifier is nil, we skip, this is useful for testing
if m.ownerVerifier == nil {
continue
}
communitiesToValidate, err := m.persistence.getCommunitiesToValidate()
2021-01-11 10:32:51 +00:00
if err != nil {
m.logger.Error("failed to fetch communities to validate", zap.Error(err))
continue
}
for id, communities := range communitiesToValidate {
m.logger.Info("validating communities", zap.String("id", id), zap.Int("count", len(communities)))
for _, communityToValidate := range communities {
signer, description, err := UnwrapCommunityDescriptionMessage(communityToValidate.payload)
if err != nil {
m.logger.Error("failed to unwrap community", zap.Error(err))
continue
}
chainID := CommunityDescriptionTokenOwnerChainID(description)
if chainID == 0 {
// This should not happen
m.logger.Error("chain id is 0, ignoring")
continue
}
m.logger.Info("validating community", zap.String("id", types.EncodeHex(communityToValidate.id)), zap.String("signer", common.PubkeyToHex(signer)))
ctx, cancel := context.WithTimeout(context.Background(), time.Second*3)
defer cancel()
owner, err := m.ownerVerifier.SafeGetSignerPubKey(ctx, chainID, id)
if err != nil {
m.logger.Error("failed to get owner", zap.Error(err))
continue
}
ownerPK, err := common.HexToPubkey(owner)
if err != nil {
m.logger.Error("failed to convert pk string to ecdsa", zap.Error(err))
continue
}
// TODO: handle shards
response, err := m.HandleCommunityDescriptionMessage(signer, description, communityToValidate.payload, ownerPK)
if err != nil {
m.logger.Error("failed to handle community", zap.Error(err))
err = m.persistence.DeleteCommunityToValidate(communityToValidate.id, communityToValidate.clock)
if err != nil {
m.logger.Error("failed to delete community to validate", zap.Error(err))
}
continue
}
if response != nil {
m.logger.Info("community validated", zap.String("id", types.EncodeHex(communityToValidate.id)), zap.String("signer", common.PubkeyToHex(signer)))
m.publish(&Subscription{TokenCommunityValidated: response})
err := m.persistence.DeleteCommunitiesToValidateByCommunityID(communityToValidate.id)
if err != nil {
m.logger.Error("failed to delete communities to validate", zap.Error(err))
}
break
}
}
}
2021-01-11 10:32:51 +00:00
}
}
}()
}
func (m *Manager) Stop() error {
m.stopped = true
2021-01-11 10:32:51 +00:00
close(m.quit)
for _, c := range m.subscriptions {
close(c)
}
m.StopTorrentClient()
return nil
}
func (m *Manager) SetTorrentConfig(config *params.TorrentConfig) {
m.torrentConfig = config
}
// getTCPandUDPport will return the same port number given if != 0,
// otherwise, it will attempt to find a free random tcp and udp port using
// the same number for both protocols
func (m *Manager) getTCPandUDPport(portNumber int) (int, error) {
if portNumber != 0 {
return portNumber, nil
}
// Find free port
for i := 0; i < 10; i++ {
tcpAddr, err := net.ResolveTCPAddr("tcp", net.JoinHostPort("localhost", "0"))
if err != nil {
m.logger.Warn("unable to resolve tcp addr: %v", zap.Error(err))
continue
}
tcpListener, err := net.ListenTCP("tcp", tcpAddr)
if err != nil {
tcpListener.Close()
m.logger.Warn("unable to listen on addr", zap.Stringer("addr", tcpAddr), zap.Error(err))
continue
}
port := tcpListener.Addr().(*net.TCPAddr).Port
tcpListener.Close()
udpAddr, err := net.ResolveUDPAddr("udp", net.JoinHostPort("localhost", fmt.Sprintf("%d", port)))
if err != nil {
m.logger.Warn("unable to resolve udp addr: %v", zap.Error(err))
continue
}
udpListener, err := net.ListenUDP("udp", udpAddr)
if err != nil {
udpListener.Close()
m.logger.Warn("unable to listen on addr", zap.Stringer("addr", udpAddr), zap.Error(err))
continue
}
udpListener.Close()
return port, nil
}
return 0, fmt.Errorf("no free port found")
}
func (m *Manager) StartTorrentClient() error {
if m.torrentConfig == nil {
return fmt.Errorf("can't start torrent client: missing torrentConfig")
}
if m.TorrentClientStarted() {
return nil
}
port, err := m.getTCPandUDPport(m.torrentConfig.Port)
if err != nil {
return err
}
config := torrent.NewDefaultClientConfig()
config.SetListenAddr(":" + fmt.Sprint(port))
config.Seed = true
config.DataDir = m.torrentConfig.DataDir
if _, err := os.Stat(m.torrentConfig.DataDir); os.IsNotExist(err) {
err := os.MkdirAll(m.torrentConfig.DataDir, 0700)
if err != nil {
return err
}
}
m.logger.Info("Starting torrent client", zap.Any("port", port))
// Instantiating the client will make it bootstrap and listen eagerly,
// so no go routine is needed here
client, err := torrent.NewClient(config)
if err != nil {
return err
}
m.torrentClient = client
return nil
}
func (m *Manager) StopTorrentClient() []error {
if m.TorrentClientStarted() {
m.StopHistoryArchiveTasksIntervals()
m.logger.Info("Stopping torrent client")
errs := m.torrentClient.Close()
if len(errs) > 0 {
return errs
}
m.torrentClient = nil
}
return make([]error, 0)
}
func (m *Manager) TorrentClientStarted() bool {
return m.torrentClient != nil
}
func (m *Manager) publish(subscription *Subscription) {
if m.stopped {
return
}
for _, s := range m.subscriptions {
select {
case s <- subscription:
default:
m.logger.Warn("subscription channel full, dropping message")
}
}
}
func (m *Manager) All() ([]*Community, error) {
communities, err := m.persistence.AllCommunities(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
for _, c := range communities {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
}
return communities, nil
}
type CommunityShard struct {
CommunityID string `json:"communityID"`
Shard *common.Shard `json:"shard"`
}
type KnownCommunitiesResponse struct {
ContractCommunities []string `json:"contractCommunities"` // TODO: use CommunityShard
ContractFeaturedCommunities []string `json:"contractFeaturedCommunities"` // TODO: use CommunityShard
2023-05-05 15:55:32 +00:00
Descriptions map[string]*Community `json:"communities"`
UnknownCommunities []string `json:"unknownCommunities"` // TODO: use CommunityShard
}
func (m *Manager) GetStoredDescriptionForCommunities(communityIDs []types.HexBytes) (response *KnownCommunitiesResponse, err error) {
response = &KnownCommunitiesResponse{
Descriptions: make(map[string]*Community),
}
for i := range communityIDs {
communityID := communityIDs[i].String()
var community *Community
community, err = m.GetByID(communityIDs[i])
if err != nil {
return
}
// TODO: use CommunityShard instead of appending the communityID
response.ContractCommunities = append(response.ContractCommunities, communityID) // , CommunityShard{
// CommunityID: communityID,
// Shard: community.Shard(),
// }
if community != nil {
response.Descriptions[community.IDString()] = community
} else {
response.UnknownCommunities = append(response.UnknownCommunities, communityID) // CommunityShard{
// CommunityID: communityID,
// Shard: community.Shard(),
// })
}
}
return
}
func (m *Manager) Joined() ([]*Community, error) {
communities, err := m.persistence.JoinedCommunities(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
for _, c := range communities {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
}
return communities, nil
}
func (m *Manager) Spectated() ([]*Community, error) {
communities, err := m.persistence.SpectatedCommunities(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
for _, c := range communities {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
}
return communities, nil
}
func (m *Manager) JoinedAndPendingCommunitiesWithRequests() ([]*Community, error) {
communities, err := m.persistence.JoinedAndPendingCommunitiesWithRequests(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
for _, c := range communities {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
}
return communities, nil
}
2022-04-11 16:14:08 +00:00
func (m *Manager) DeletedCommunities() ([]*Community, error) {
communities, err := m.persistence.DeletedCommunities(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
for _, c := range communities {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
}
return communities, nil
2022-04-11 16:14:08 +00:00
}
func (m *Manager) Controlled() ([]*Community, error) {
communities, err := m.persistence.CommunitiesWithPrivateKey(&m.identity.PublicKey, m.installationID)
if err != nil {
return nil, err
}
controlled := make([]*Community, 0, len(communities))
for _, c := range communities {
if c.IsControlNode() {
err = m.initializeCommunity(c)
if err != nil {
return nil, err
}
controlled = append(controlled, c)
}
}
return controlled, nil
}
// CreateCommunity takes a description, generates an ID for it, saves it and return it
func (m *Manager) CreateCommunity(request *requests.CreateCommunity, publish bool) (*Community, error) {
description, err := request.ToCommunityDescription()
if err != nil {
return nil, err
}
description.Members = make(map[string]*protobuf.CommunityMember)
description.Members[common.PubkeyToHex(&m.identity.PublicKey)] = &protobuf.CommunityMember{Roles: []protobuf.CommunityMember_Roles{protobuf.CommunityMember_ROLE_OWNER}}
err = ValidateCommunityDescription(description)
if err != nil {
return nil, err
}
description.Clock = 1
key, err := crypto.GenerateKey()
if err != nil {
return nil, err
}
description.ID = types.EncodeHex(crypto.CompressPubkey(&key.PublicKey))
config := Config{
ID: &key.PublicKey,
PrivateKey: key,
ControlNode: &key.PublicKey,
ControlDevice: true,
Logger: m.logger,
Joined: true,
MemberIdentity: &m.identity.PublicKey,
CommunityDescription: description,
Shard: nil,
}
community, err := New(config, m.timesource)
if err != nil {
return nil, err
}
2021-01-11 10:32:51 +00:00
// We join any community we create
community.Join()
2020-12-17 14:36:09 +00:00
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
// Mark this device as the control node
syncControlNode := &protobuf.SyncCommunityControlNode{
Clock: 1,
InstallationId: m.installationID,
}
err = m.SaveSyncControlNode(community.ID(), syncControlNode)
if err != nil {
return nil, err
}
if publish {
m.publish(&Subscription{Community: community})
}
2020-12-17 14:36:09 +00:00
return community, nil
}
func (m *Manager) CreateCommunityTokenPermission(request *requests.CreateCommunityTokenPermission) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
community, changes, err := m.createCommunityTokenPermission(request, community)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) EditCommunityTokenPermission(request *requests.EditCommunityTokenPermission) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
tokenPermission := request.ToCommunityTokenPermission()
changes, err := community.UpsertTokenPermission(&tokenPermission)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) ReevaluateMembers(community *Community) (map[protobuf.CommunityMember_Roles][]*ecdsa.PublicKey, error) {
becomeMemberPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_MEMBER)
becomeAdminPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_ADMIN)
becomeTokenMasterPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_TOKEN_MASTER)
hasMemberPermissions := len(becomeMemberPermissions) > 0
newPrivilegedRoles := make(map[protobuf.CommunityMember_Roles][]*ecdsa.PublicKey)
newPrivilegedRoles[protobuf.CommunityMember_ROLE_TOKEN_MASTER] = []*ecdsa.PublicKey{}
newPrivilegedRoles[protobuf.CommunityMember_ROLE_ADMIN] = []*ecdsa.PublicKey{}
for memberKey := range community.Members() {
memberPubKey, err := common.HexToPubkey(memberKey)
if err != nil {
return nil, err
}
if memberKey == common.PubkeyToHex(&m.identity.PublicKey) || community.IsMemberOwner(memberPubKey) {
continue
}
isCurrentRoleTokenMaster := community.IsMemberTokenMaster(memberPubKey)
isCurrentRoleAdmin := community.IsMemberAdmin(memberPubKey)
requestID := CalculateRequestID(memberKey, community.ID())
revealedAccounts, err := m.persistence.GetRequestToJoinRevealedAddresses(requestID)
if err != nil {
return nil, err
}
memberHasWallet := len(revealedAccounts) > 0
// Check if user has privilege role without sharing the account to controlNode
// or user treated as a member without wallet in closed community
if !memberHasWallet && (hasMemberPermissions || isCurrentRoleTokenMaster || isCurrentRoleAdmin) {
_, err = community.RemoveUserFromOrg(memberPubKey)
if err != nil {
return nil, err
}
continue
}
accountsAndChainIDs := revealedAccountsToAccountsAndChainIDsCombination(revealedAccounts)
isNewRoleTokenMaster, err := m.ReevaluatePrivilegedMember(community, becomeTokenMasterPermissions, accountsAndChainIDs, memberPubKey,
protobuf.CommunityMember_ROLE_TOKEN_MASTER, isCurrentRoleTokenMaster)
if err != nil {
return nil, err
}
if isNewRoleTokenMaster {
if !isCurrentRoleTokenMaster {
newPrivilegedRoles[protobuf.CommunityMember_ROLE_TOKEN_MASTER] =
append(newPrivilegedRoles[protobuf.CommunityMember_ROLE_TOKEN_MASTER], memberPubKey)
}
// Skip further validation if user has TokenMaster permissions
continue
}
isNewRoleAdmin, err := m.ReevaluatePrivilegedMember(community, becomeAdminPermissions, accountsAndChainIDs, memberPubKey,
protobuf.CommunityMember_ROLE_ADMIN, isCurrentRoleAdmin)
if err != nil {
return nil, err
}
if isNewRoleAdmin {
if !isCurrentRoleAdmin {
newPrivilegedRoles[protobuf.CommunityMember_ROLE_ADMIN] =
append(newPrivilegedRoles[protobuf.CommunityMember_ROLE_TOKEN_MASTER], memberPubKey)
}
// Skip further validation if user has Admin permissions
continue
}
if hasMemberPermissions {
permissionResponse, err := m.PermissionChecker.CheckPermissions(becomeMemberPermissions, accountsAndChainIDs, true)
if err != nil {
return nil, err
}
if !permissionResponse.Satisfied {
_, err = community.RemoveUserFromOrg(memberPubKey)
if err != nil {
return nil, err
}
// Skip channels validation if user has been removed
continue
}
}
// Validate channel permissions
for channelID := range community.Chats() {
chatID := community.IDString() + channelID
viewOnlyPermissions := community.ChannelTokenPermissionsByType(chatID, protobuf.CommunityTokenPermission_CAN_VIEW_CHANNEL)
viewAndPostPermissions := community.ChannelTokenPermissionsByType(chatID, protobuf.CommunityTokenPermission_CAN_VIEW_AND_POST_CHANNEL)
if len(viewOnlyPermissions) == 0 && len(viewAndPostPermissions) == 0 {
// ensure all members are added back if channel permissions were removed
_, err = community.PopulateChatWithAllMembers(channelID)
if err != nil {
return nil, err
}
continue
}
response, err := m.checkChannelPermissions(viewOnlyPermissions, viewAndPostPermissions, accountsAndChainIDs, true)
if err != nil {
return nil, err
}
isMemberAlreadyInChannel := community.IsMemberInChat(memberPubKey, channelID)
if response.ViewOnlyPermissions.Satisfied || response.ViewAndPostPermissions.Satisfied {
if !isMemberAlreadyInChannel {
_, err := community.AddMemberToChat(channelID, memberPubKey, []protobuf.CommunityMember_Roles{})
if err != nil {
return nil, err
}
}
} else if isMemberAlreadyInChannel {
_, err := community.RemoveUserFromChat(memberPubKey, channelID)
if err != nil {
return nil, err
}
}
}
}
return newPrivilegedRoles, m.saveAndPublish(community)
}
func (m *Manager) ReevaluateMembersPeriodically(communityID types.HexBytes) {
if _, exists := m.periodicMembersReevaluationTasks.Load(communityID.String()); exists {
return
}
cancel := make(chan struct{})
m.periodicMembersReevaluationTasks.Store(communityID.String(), cancel)
ticker := time.NewTicker(memberPermissionsCheckInterval)
defer ticker.Stop()
for {
select {
case <-ticker.C:
community, err := m.GetByID(communityID)
if err != nil {
m.logger.Debug("can't validate member permissions, community was not found", zap.Error(err))
m.periodicMembersReevaluationTasks.Delete(communityID.String())
}
if err = m.ReevaluateCommunityMembersPermissions(community); err != nil {
m.logger.Debug("failed to check member permissions", zap.Error(err))
continue
}
case <-cancel:
m.periodicMembersReevaluationTasks.Delete(communityID.String())
return
}
}
}
func (m *Manager) DeleteCommunityTokenPermission(request *requests.DeleteCommunityTokenPermission) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
changes, err := community.DeleteTokenPermission(request.PermissionID)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) ReevaluateCommunityMembersPermissions(community *Community) error {
if community == nil {
return ErrOrgNotFound
}
// TODO: Control node needs to be notified to do a permission check if TokenMasters did airdrop
// of the token which is using in a community permissions
if !community.IsControlNode() {
return ErrNotEnoughPermissions
}
newPrivilegedMembers, err := m.ReevaluateMembers(community)
if err != nil {
return err
}
return m.shareRequestsToJoinWithNewPrivilegedMembers(community, newPrivilegedMembers)
}
2022-09-29 11:50:23 +00:00
func (m *Manager) DeleteCommunity(id types.HexBytes) error {
err := m.persistence.DeleteCommunity(id)
if err != nil {
return err
}
return m.persistence.DeleteCommunitySettings(id)
}
func (m *Manager) UpdateShard(community *Community, shard *common.Shard) error {
community.config.Shard = shard
return m.persistence.SaveCommunity(community)
}
// SetShard assigns a shard to a community
func (m *Manager) SetShard(communityID types.HexBytes, shard *common.Shard) (*Community, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.IsControlNode() {
return nil, errors.New("not admin or owner")
}
err = m.UpdateShard(community, shard)
if err != nil {
return nil, err
}
m.publish(&Subscription{Community: community})
return community, nil
}
func (m *Manager) UpdatePubsubTopicPrivateKey(community *Community, privKey *ecdsa.PrivateKey) error {
community.SetPubsubTopicPrivateKey(privKey)
if privKey != nil {
if err := m.transport.StorePubsubTopicKey(community.PubsubTopic(), privKey); err != nil {
return err
}
}
return nil
}
// EditCommunity takes a description, updates the community with the description,
// saves it and returns it
func (m *Manager) EditCommunity(request *requests.EditCommunity) (*Community, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
newDescription, err := request.ToCommunityDescription()
if err != nil {
return nil, fmt.Errorf("Can't create community description: %v", err)
}
// If permissions weren't explicitly set on original request, use existing ones
if newDescription.Permissions.Access == protobuf.CommunityPermissions_UNKNOWN_ACCESS {
newDescription.Permissions.Access = community.config.CommunityDescription.Permissions.Access
}
// Use existing images for the entries that were not updated
// NOTE: This will NOT allow deletion of the community image; it will need to
// be handled separately.
for imageName := range community.config.CommunityDescription.Identity.Images {
_, exists := newDescription.Identity.Images[imageName]
if !exists {
// If no image was set in ToCommunityDescription then Images is nil.
if newDescription.Identity.Images == nil {
newDescription.Identity.Images = make(map[string]*protobuf.IdentityImage)
}
newDescription.Identity.Images[imageName] = community.config.CommunityDescription.Identity.Images[imageName]
}
}
// TODO: handle delete image (if needed)
err = ValidateCommunityDescription(newDescription)
if err != nil {
return nil, err
}
if !(community.IsControlNode() || community.hasPermissionToSendCommunityEvent(protobuf.CommunityEvent_COMMUNITY_EDIT)) {
return nil, ErrNotAuthorized
}
// Edit the community values
community.Edit(newDescription)
if err != nil {
return nil, err
}
if community.IsControlNode() {
community.increaseClock()
} else {
err := community.addNewCommunityEvent(community.ToCommunityEditCommunityEvent(newDescription))
if err != nil {
return nil, err
}
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) RemovePrivateKey(id types.HexBytes) (*Community, error) {
community, err := m.GetByID(id)
if err != nil {
return community, err
}
if !community.IsControlNode() {
return community, ErrNotControlNode
}
community.config.PrivateKey = nil
err = m.persistence.SaveCommunity(community)
if err != nil {
return community, err
}
return community, nil
}
2021-01-11 10:32:51 +00:00
func (m *Manager) ExportCommunity(id types.HexBytes) (*ecdsa.PrivateKey, error) {
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if !community.IsControlNode() {
return nil, ErrNotControlNode
}
2020-12-17 14:36:09 +00:00
return community.config.PrivateKey, nil
}
func (m *Manager) ImportCommunity(key *ecdsa.PrivateKey, clock uint64) (*Community, error) {
2020-12-17 14:36:09 +00:00
communityID := crypto.CompressPubkey(&key.PublicKey)
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
if community == nil {
createCommunityRequest := requests.CreateCommunity{
Membership: protobuf.CommunityPermissions_MANUAL_ACCEPT,
Name: "unknown imported",
}
description, err := createCommunityRequest.ToCommunityDescription()
if err != nil {
return nil, err
}
err = ValidateCommunityDescription(description)
if err != nil {
return nil, err
2020-12-17 14:36:09 +00:00
}
description.Clock = 1
description.ID = types.EncodeHex(communityID)
2020-12-17 14:36:09 +00:00
config := Config{
ID: &key.PublicKey,
PrivateKey: key,
ControlNode: &key.PublicKey,
ControlDevice: true,
2020-12-17 14:36:09 +00:00
Logger: m.logger,
Joined: true,
MemberIdentity: &m.identity.PublicKey,
2020-12-17 14:36:09 +00:00
CommunityDescription: description,
}
community, err = New(config, m.timesource)
2020-12-17 14:36:09 +00:00
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
} else {
community.config.PrivateKey = key
community.config.ControlDevice = true
2020-12-17 14:36:09 +00:00
}
community.Join()
2020-12-17 14:36:09 +00:00
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
// Mark this device as the control node
syncControlNode := &protobuf.SyncCommunityControlNode{
Clock: clock,
InstallationId: m.installationID,
}
err = m.SaveSyncControlNode(community.ID(), syncControlNode)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
return community, nil
}
func (m *Manager) CreateChat(communityID types.HexBytes, chat *protobuf.CommunityChat, publish bool, thirdPartyID string) (*CommunityChanges, error) {
2021-01-11 10:32:51 +00:00
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
if community == nil {
return nil, ErrOrgNotFound
}
chatID := uuid.New().String()
if thirdPartyID != "" {
chatID = chatID + thirdPartyID
}
2020-12-17 14:36:09 +00:00
changes, err := community.CreateChat(chatID, chat)
if err != nil {
return nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return changes, nil
}
func (m *Manager) EditChat(communityID types.HexBytes, chatID string, chat *protobuf.CommunityChat) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
// Remove communityID prefix from chatID if exists
if strings.HasPrefix(chatID, communityID.String()) {
chatID = strings.TrimPrefix(chatID, communityID.String())
}
changes, err := community.EditChat(chatID, chat)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) DeleteChat(communityID types.HexBytes, chatID string) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
// Remove communityID prefix from chatID if exists
if strings.HasPrefix(chatID, communityID.String()) {
chatID = strings.TrimPrefix(chatID, communityID.String())
}
changes, err := community.DeleteChat(chatID)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) CreateCategory(request *requests.CreateCommunityCategory, publish bool) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
categoryID := uuid.New().String()
if request.ThirdPartyID != "" {
categoryID = categoryID + request.ThirdPartyID
}
// Remove communityID prefix from chatID if exists
for i, cid := range request.ChatIDs {
if strings.HasPrefix(cid, request.CommunityID.String()) {
request.ChatIDs[i] = strings.TrimPrefix(cid, request.CommunityID.String())
}
}
changes, err := community.CreateCategory(categoryID, request.CategoryName, request.ChatIDs)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) EditCategory(request *requests.EditCommunityCategory) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
// Remove communityID prefix from chatID if exists
for i, cid := range request.ChatIDs {
if strings.HasPrefix(cid, request.CommunityID.String()) {
request.ChatIDs[i] = strings.TrimPrefix(cid, request.CommunityID.String())
}
}
changes, err := community.EditCategory(request.CategoryID, request.CategoryName, request.ChatIDs)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) EditChatFirstMessageTimestamp(communityID types.HexBytes, chatID string, timestamp uint32) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
// Remove communityID prefix from chatID if exists
if strings.HasPrefix(chatID, communityID.String()) {
chatID = strings.TrimPrefix(chatID, communityID.String())
}
changes, err := community.UpdateChatFirstMessageTimestamp(chatID, timestamp)
if err != nil {
return nil, nil, err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, nil, err
}
// Advertise changes
m.publish(&Subscription{Community: community})
return community, changes, nil
}
func (m *Manager) ReorderCategories(request *requests.ReorderCommunityCategories) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
changes, err := community.ReorderCategories(request.CategoryID, request.Position)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) ReorderChat(request *requests.ReorderCommunityChat) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
// Remove communityID prefix from chatID if exists
if strings.HasPrefix(request.ChatID, request.CommunityID.String()) {
request.ChatID = strings.TrimPrefix(request.ChatID, request.CommunityID.String())
}
changes, err := community.ReorderChat(request.CategoryID, request.ChatID, request.Position)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) DeleteCategory(request *requests.DeleteCommunityCategory) (*Community, *CommunityChanges, error) {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return nil, nil, err
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
changes, err := community.DeleteCategory(request.CategoryID)
if err != nil {
return nil, nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, nil, err
}
return changes.Community, changes, nil
}
func (m *Manager) GenerateRequestsToJoinForAutoApprovalOnNewOwnership(communityID types.HexBytes, kickedMembers map[string]*protobuf.CommunityMember) error {
var requestsToJoin []*RequestToJoin
clock := uint64(time.Now().Unix())
for pubKeyStr := range kickedMembers {
requestToJoin := &RequestToJoin{
PublicKey: pubKeyStr,
Clock: clock,
CommunityID: communityID,
State: RequestToJoinStateAwaitingAddresses,
Our: true,
RevealedAccounts: make([]*protobuf.RevealedAccount, 0),
}
requestToJoin.CalculateID()
requestsToJoin = append(requestsToJoin, requestToJoin)
}
return m.persistence.SaveRequestsToJoin(requestsToJoin)
}
func (m *Manager) Queue(signer *ecdsa.PublicKey, community *Community, clock uint64, payload []byte) error {
m.logger.Info("queuing community", zap.String("id", community.IDString()), zap.String("signer", common.PubkeyToHex(signer)))
communityToValidate := communityToValidate{
id: community.ID(),
clock: clock,
payload: payload,
validateAt: uint64(time.Now().UnixNano()),
signer: crypto.CompressPubkey(signer),
}
err := m.persistence.SaveCommunityToValidate(communityToValidate)
if err != nil {
m.logger.Error("failed to save community", zap.Error(err))
return err
}
return nil
}
func (m *Manager) HandleCommunityDescriptionMessage(signer *ecdsa.PublicKey, description *protobuf.CommunityDescription, payload []byte, verifiedOwner *ecdsa.PublicKey) (*CommunityResponse, error) {
2023-03-29 11:36:13 +00:00
if signer == nil {
return nil, errors.New("signer can't be nil")
}
var id []byte
var err error
if len(description.ID) != 0 {
id, err = types.DecodeHex(description.ID)
if err != nil {
return nil, err
}
} else {
// Backward compatibility
id = crypto.CompressPubkey(signer)
}
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
// Workaround for https://github.com/status-im/status-desktop/issues/12188
HydrateChannelsMembers(types.EncodeHex(id), description)
// We should queue only if the community has a token owner, and the owner has been verified
hasTokenOwnership := HasTokenOwnership(description)
shouldQueue := hasTokenOwnership && verifiedOwner == nil
2020-12-17 14:36:09 +00:00
if community == nil {
pubKey, err := crypto.DecompressPubkey(id)
if err != nil {
return nil, err
}
config := Config{
CommunityDescription: description,
Logger: m.logger,
CommunityDescriptionProtocolMessage: payload,
MemberIdentity: &m.identity.PublicKey,
ID: pubKey,
ControlNode: signer,
}
community, err = New(config, m.timesource)
if err != nil {
return nil, err
}
// A new community, we need to check if we need to validate async.
// That would be the case if it has a contract. We queue everything and process separately.
if shouldQueue {
return nil, m.Queue(signer, community, description.Clock, payload)
}
} else {
// only queue if already known control node is different than the signer
// and if the clock is greater
shouldQueue = shouldQueue && !common.IsPubKeyEqual(community.ControlNode(), signer) &&
community.config.CommunityDescription.Clock < description.Clock
if shouldQueue {
return nil, m.Queue(signer, community, description.Clock, payload)
}
}
if hasTokenOwnership {
// Override verified owner
if verifiedOwner != nil {
m.logger.Info("updating verified owner", zap.String("communityID", community.IDString()), zap.String("owner", common.PubkeyToHex(verifiedOwner)))
// If we are not the verified owner anymore, drop the private key
if !common.IsPubKeyEqual(verifiedOwner, &m.identity.PublicKey) {
community.config.PrivateKey = nil
}
// new control node will be set in the 'UpdateCommunityDescription'
if !common.IsPubKeyEqual(verifiedOwner, signer) {
return nil, ErrNotAuthorized
}
} else if !common.IsPubKeyEqual(community.ControlNode(), signer) {
return nil, ErrNotAuthorized
}
} else if !common.IsPubKeyEqual(community.PublicKey(), signer) {
return nil, ErrNotAuthorized
}
return m.handleCommunityDescriptionMessageCommon(community, description, payload, verifiedOwner)
}
func (m *Manager) handleCommunityDescriptionMessageCommon(community *Community, description *protobuf.CommunityDescription, payload []byte, newControlNode *ecdsa.PublicKey) (*CommunityResponse, error) {
changes, err := community.UpdateCommunityDescription(description, payload, newControlNode)
if err != nil {
return nil, err
}
if err = m.HandleCommunityTokensMetadataByPrivilegedMembers(community); err != nil {
return nil, err
}
hasCommunityArchiveInfo, err := m.persistence.HasCommunityArchiveInfo(community.ID())
if err != nil {
return nil, err
}
cdMagnetlinkClock := community.config.CommunityDescription.ArchiveMagnetlinkClock
if !hasCommunityArchiveInfo {
err = m.persistence.SaveCommunityArchiveInfo(community.ID(), cdMagnetlinkClock, 0)
if err != nil {
return nil, err
}
} else {
magnetlinkClock, err := m.persistence.GetMagnetlinkMessageClock(community.ID())
if err != nil {
return nil, err
}
if cdMagnetlinkClock > magnetlinkClock {
err = m.persistence.UpdateMagnetlinkMessageClock(community.ID(), cdMagnetlinkClock)
if err != nil {
return nil, err
}
}
}
pkString := common.PubkeyToHex(&m.identity.PublicKey)
if m.tokenManager != nil && description.CommunityTokensMetadata != nil && len(description.CommunityTokensMetadata) > 0 {
for _, tokenMetadata := range description.CommunityTokensMetadata {
if tokenMetadata.TokenType != protobuf.CommunityTokenType_ERC20 {
continue
}
for chainID, address := range tokenMetadata.ContractAddresses {
2023-09-11 11:45:30 +00:00
_ = m.tokenManager.FindOrCreateTokenByAddress(context.Background(), chainID, gethcommon.HexToAddress(address))
}
}
}
2021-01-11 10:32:51 +00:00
// If the community require membership, we set whether we should leave/join the community after a state change
if community.ManualAccept() || community.AutoAccept() {
2021-01-11 10:32:51 +00:00
if changes.HasNewMember(pkString) {
hasPendingRequest, err := m.persistence.HasPendingRequestsToJoinForUserAndCommunity(pkString, changes.Community.ID())
if err != nil {
return nil, err
}
// If there's any pending request, we should join the community
// automatically
changes.ShouldMemberJoin = hasPendingRequest
}
if changes.HasMemberLeft(pkString) {
// If we joined previously the community, that means we have been kicked
changes.MemberKicked = community.Joined()
2021-01-11 10:32:51 +00:00
}
}
err = m.persistence.DeleteCommunityEvents(community.ID())
if err != nil {
return nil, err
}
community.config.EventsData = nil
// Set Joined if we are part of the member list
if !community.Joined() && community.hasMember(&m.identity.PublicKey) {
changes.ShouldMemberJoin = true
}
2020-12-17 14:36:09 +00:00
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
2021-01-11 10:32:51 +00:00
// We mark our requests as completed, though maybe we should mark
// any request for any user that has been added as completed
if err := m.markRequestToJoinAsAccepted(&m.identity.PublicKey, community); err != nil {
2021-01-11 10:32:51 +00:00
return nil, err
}
// Check if there's a change and we should be joining
return &CommunityResponse{
Community: community,
Changes: changes,
}, nil
}
func (m *Manager) signEvents(community *Community) error {
for i := range community.config.EventsData.Events {
communityEvent := &community.config.EventsData.Events[i]
if communityEvent.Signature == nil || len(communityEvent.Signature) == 0 {
err := communityEvent.Sign(m.identity)
if err != nil {
return err
}
}
}
return nil
}
func (m *Manager) validateAndFilterEvents(community *Community, events []CommunityEvent) []CommunityEvent {
validatedEvents := make([]CommunityEvent, 0, len(events))
validateEvent := func(event *CommunityEvent) error {
signer, err := event.RecoverSigner()
if err != nil {
return err
}
err = community.ValidateEvent(event, signer)
if err != nil {
return err
}
return nil
}
for i := range events {
if err := validateEvent(&events[i]); err == nil {
validatedEvents = append(validatedEvents, events[i])
} else {
m.logger.Warn("invalid community event", zap.Error(err))
}
}
return validatedEvents
}
func (m *Manager) HandleCommunityEventsMessage(signer *ecdsa.PublicKey, message *protobuf.CommunityEventsMessage) (*CommunityResponse, error) {
if signer == nil {
return nil, errors.New("signer can't be nil")
}
eventsMessage, err := CommunityEventsMessageFromProtobuf(message)
if err != nil {
return nil, err
}
community, err := m.GetByID(eventsMessage.CommunityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.IsPrivilegedMember(signer) {
return nil, errors.New("user has not permissions to send events")
}
originCommunity := community.CreateDeepCopy()
eventsMessage.Events = m.validateAndFilterEvents(community, eventsMessage.Events)
err = community.UpdateCommunityByEvents(eventsMessage)
if err != nil {
if err == ErrInvalidCommunityEventClock && community.IsControlNode() {
// send updated CommunityDescription to the event sender on top of which he must apply his changes
eventsMessage.EventsBaseCommunityDescription = community.config.CommunityDescriptionProtocolMessage
m.publish(&Subscription{
CommunityEventsMessageInvalidClock: &CommunityEventsMessageInvalidClockSignal{
Community: community,
CommunityEventsMessage: eventsMessage,
}})
}
return nil, err
}
additionalCommunityResponse, err := m.handleAdditionalAdminChanges(community)
if err != nil {
return nil, err
}
if err = m.HandleCommunityTokensMetadataByPrivilegedMembers(community); err != nil {
return nil, err
}
// Control node applies events and publish updated CommunityDescription
if community.IsControlNode() {
community.config.EventsData = nil // clear events, they are already applied
community.increaseClock()
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
m.publish(&Subscription{Community: community})
} else {
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
err := m.persistence.SaveCommunityEvents(community)
if err != nil {
return nil, err
}
}
return &CommunityResponse{
Community: community,
Changes: EvaluateCommunityChanges(originCommunity, community),
RequestsToJoin: additionalCommunityResponse.RequestsToJoin,
}, nil
}
// Creates new CommunityEventsMessage by re-applying our rejected events on top of latest known CommunityDescription.
// Returns nil if none of our events were rejected.
func (m *Manager) HandleCommunityEventsMessageRejected(signer *ecdsa.PublicKey, message *protobuf.CommunityEventsMessageRejected) (*CommunityEventsMessage, error) {
if signer == nil {
return nil, errors.New("signer can't be nil")
}
id := crypto.CompressPubkey(signer)
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
eventsMessage, err := CommunityEventsMessageFromProtobuf(message.Msg)
if err != nil {
return nil, err
}
communityDescription, err := validateAndGetEventsMessageCommunityDescription(eventsMessage.EventsBaseCommunityDescription, signer)
if err != nil {
return nil, err
}
// the privileged member did not receive updated CommunityDescription so his events
// will be send on top of outdated CommunityDescription
if communityDescription.Clock != community.Clock() {
return nil, errors.New("resend rejected community events aborted, client node has outdated community description")
}
eventsMessage.Events = m.validateAndFilterEvents(community, eventsMessage.Events)
myRejectedEvents := make([]CommunityEvent, 0)
for _, rejectedEvent := range eventsMessage.Events {
rejectedEventSigner, err := rejectedEvent.RecoverSigner()
if err != nil {
continue
}
if rejectedEventSigner.Equal(m.identity.Public()) {
myRejectedEvents = append(myRejectedEvents, rejectedEvent)
}
}
if len(myRejectedEvents) == 0 {
return nil, nil
}
// Re-apply rejected events on top of latest known `CommunityDescription`
community.config.EventsData = &EventsData{
EventsBaseCommunityDescription: community.config.CommunityDescriptionProtocolMessage,
Events: myRejectedEvents,
}
reapplyEventsMessage := community.ToCommunityEventsMessage()
return reapplyEventsMessage, nil
}
func (m *Manager) handleAdditionalAdminChanges(community *Community) (*CommunityResponse, error) {
communityResponse := CommunityResponse{
RequestsToJoin: make([]*RequestToJoin, 0),
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if !(community.IsControlNode() || community.HasPermissionToSendCommunityEvents()) {
// we're a normal user/member node, so there's nothing for us to do here
return &communityResponse, nil
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
for i := range community.config.EventsData.Events {
communityEvent := &community.config.EventsData.Events[i]
switch communityEvent.Type {
case protobuf.CommunityEvent_COMMUNITY_REQUEST_TO_JOIN_ACCEPT:
requestsToJoin, err := m.handleCommunityEventRequestAccepted(community, communityEvent)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if err != nil {
return nil, err
}
if requestsToJoin != nil {
communityResponse.RequestsToJoin = append(communityResponse.RequestsToJoin, requestsToJoin...)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
case protobuf.CommunityEvent_COMMUNITY_REQUEST_TO_JOIN_REJECT:
requestsToJoin, err := m.handleCommunityEventRequestRejected(community, communityEvent)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if err != nil {
return nil, err
}
if requestsToJoin != nil {
communityResponse.RequestsToJoin = append(communityResponse.RequestsToJoin, requestsToJoin...)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
default:
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
return &communityResponse, nil
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
func (m *Manager) saveOrUpdateRequestToJoin(communityID types.HexBytes, requestToJoin *RequestToJoin) (bool, error) {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
updated := false
existingRequestToJoin, err := m.persistence.GetRequestToJoin(requestToJoin.ID)
if err != nil && err != sql.ErrNoRows {
return updated, err
}
if existingRequestToJoin != nil {
// node already knows about this request to join, so let's compare clocks
// and update it if necessary
if existingRequestToJoin.Clock <= requestToJoin.Clock {
pk, err := common.HexToPubkey(existingRequestToJoin.PublicKey)
if err != nil {
return updated, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
err = m.persistence.SetRequestToJoinState(common.PubkeyToHex(pk), communityID, requestToJoin.State)
if err != nil {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
return updated, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
updated = true
}
} else {
err := m.persistence.SaveRequestToJoin(requestToJoin)
if err != nil {
return updated, err
}
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
return updated, nil
}
func (m *Manager) handleCommunityEventRequestAccepted(community *Community, communityEvent *CommunityEvent) ([]*RequestToJoin, error) {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
acceptedRequestsToJoin := make([]types.HexBytes, 0)
requestsToJoin := make([]*RequestToJoin, 0)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
for signer, request := range communityEvent.AcceptedRequestsToJoin {
requestToJoin := &RequestToJoin{
PublicKey: signer,
Clock: request.Clock,
ENSName: request.EnsName,
CommunityID: request.CommunityId,
State: RequestToJoinStateAcceptedPending,
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
requestToJoin.CalculateID()
existingRequestToJoin, err := m.persistence.GetRequestToJoin(requestToJoin.ID)
if err != nil && err != sql.ErrNoRows {
return nil, err
}
if existingRequestToJoin != nil {
alreadyProcessedByControlNode := existingRequestToJoin.State == RequestToJoinStateAccepted || existingRequestToJoin.State == RequestToJoinStateDeclined
if alreadyProcessedByControlNode || existingRequestToJoin.State == RequestToJoinStateCanceled {
continue
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
requestUpdated, err := m.saveOrUpdateRequestToJoin(community.ID(), requestToJoin)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if err != nil {
return nil, err
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
// If request to join exists in control node, add request to acceptedRequestsToJoin.
// Otherwise keep the request as RequestToJoinStateAcceptedPending,
// as privileged users don't have revealed addresses. This can happen if control node received
// community event message before user request to join.
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if community.IsControlNode() && requestUpdated {
acceptedRequestsToJoin = append(acceptedRequestsToJoin, requestToJoin.ID)
}
requestsToJoin = append(requestsToJoin, requestToJoin)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
if community.IsControlNode() {
m.publish(&Subscription{AcceptedRequestsToJoin: acceptedRequestsToJoin})
}
return requestsToJoin, nil
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
func (m *Manager) handleCommunityEventRequestRejected(community *Community, communityEvent *CommunityEvent) ([]*RequestToJoin, error) {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
rejectedRequestsToJoin := make([]types.HexBytes, 0)
requestsToJoin := make([]*RequestToJoin, 0)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
for signer, request := range communityEvent.RejectedRequestsToJoin {
requestToJoin := &RequestToJoin{
PublicKey: signer,
Clock: request.Clock,
ENSName: request.EnsName,
CommunityID: request.CommunityId,
State: RequestToJoinStateDeclinedPending,
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
requestToJoin.CalculateID()
existingRequestToJoin, err := m.persistence.GetRequestToJoin(requestToJoin.ID)
if err != nil && err != sql.ErrNoRows {
return nil, err
}
if existingRequestToJoin != nil {
alreadyProcessedByControlNode := existingRequestToJoin.State == RequestToJoinStateAccepted || existingRequestToJoin.State == RequestToJoinStateDeclined
if alreadyProcessedByControlNode || existingRequestToJoin.State == RequestToJoinStateCanceled {
continue
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
requestUpdated, err := m.saveOrUpdateRequestToJoin(community.ID(), requestToJoin)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if err != nil {
return nil, err
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
// If request to join exists in control node, add request to rejectedRequestsToJoin.
// Otherwise keep the request as RequestToJoinStateDeclinedPending,
// as privileged users don't have revealed addresses. This can happen if control node received
// community event message before user request to join.
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if community.IsControlNode() && requestUpdated {
rejectedRequestsToJoin = append(rejectedRequestsToJoin, requestToJoin.ID)
}
requestsToJoin = append(requestsToJoin, requestToJoin)
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if community.IsControlNode() {
m.publish(&Subscription{RejectedRequestsToJoin: rejectedRequestsToJoin})
}
return requestsToJoin, nil
}
// markRequestToJoinAsAccepted marks all the pending requests to join as completed
2021-01-11 10:32:51 +00:00
// if we are members
func (m *Manager) markRequestToJoinAsAccepted(pk *ecdsa.PublicKey, community *Community) error {
2021-01-11 10:32:51 +00:00
if community.HasMember(pk) {
return m.persistence.SetRequestToJoinState(common.PubkeyToHex(pk), community.ID(), RequestToJoinStateAccepted)
}
return nil
}
func (m *Manager) markRequestToJoinAsCanceled(pk *ecdsa.PublicKey, community *Community) error {
return m.persistence.SetRequestToJoinState(common.PubkeyToHex(pk), community.ID(), RequestToJoinStateCanceled)
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
func (m *Manager) markRequestToJoinAsAcceptedPending(pk *ecdsa.PublicKey, community *Community) error {
return m.persistence.SetRequestToJoinState(common.PubkeyToHex(pk), community.ID(), RequestToJoinStateAcceptedPending)
}
func (m *Manager) DeletePendingRequestToJoin(request *RequestToJoin) error {
community, err := m.GetByID(request.CommunityID)
if err != nil {
return err
}
err = m.persistence.DeletePendingRequestToJoin(request.ID)
if err != nil {
return err
}
err = m.saveAndPublish(community)
if err != nil {
return err
}
return nil
}
// UpdateClockInRequestToJoin method is used for testing
func (m *Manager) UpdateClockInRequestToJoin(id types.HexBytes, clock uint64) error {
return m.persistence.UpdateClockInRequestToJoin(id, clock)
}
func (m *Manager) SetMuted(id types.HexBytes, muted bool) error {
return m.persistence.SetMuted(id, muted)
}
func (m *Manager) MuteCommunityTill(communityID []byte, muteTill time.Time) error {
return m.persistence.MuteCommunityTill(communityID, muteTill)
}
func (m *Manager) CancelRequestToJoin(request *requests.CancelRequestToJoinCommunity) (*RequestToJoin, *Community, error) {
dbRequest, err := m.persistence.GetRequestToJoin(request.ID)
if err != nil {
return nil, nil, err
}
community, err := m.GetByID(dbRequest.CommunityID)
if err != nil {
return nil, nil, err
}
pk, err := common.HexToPubkey(dbRequest.PublicKey)
if err != nil {
return nil, nil, err
}
dbRequest.State = RequestToJoinStateCanceled
if err := m.markRequestToJoinAsCanceled(pk, community); err != nil {
return nil, nil, err
}
return dbRequest, community, nil
}
func (m *Manager) CheckPermissionToJoin(id []byte, addresses []gethcommon.Address) (*CheckPermissionToJoinResponse, error) {
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
return m.PermissionChecker.CheckPermissionToJoin(community, addresses)
}
func (m *Manager) accountsSatisfyPermissionsToJoin(community *Community, accounts []*protobuf.RevealedAccount) (bool, protobuf.CommunityMember_Roles, error) {
accountsAndChainIDs := revealedAccountsToAccountsAndChainIDsCombination(accounts)
becomeAdminPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_ADMIN)
becomeMemberPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_MEMBER)
becomeTokenMasterPermissions := community.TokenPermissionsByType(protobuf.CommunityTokenPermission_BECOME_TOKEN_MASTER)
if m.accountsHasPrivilegedPermission(becomeTokenMasterPermissions, accountsAndChainIDs) {
return true, protobuf.CommunityMember_ROLE_TOKEN_MASTER, nil
}
if m.accountsHasPrivilegedPermission(becomeAdminPermissions, accountsAndChainIDs) {
return true, protobuf.CommunityMember_ROLE_ADMIN, nil
}
if len(becomeMemberPermissions) > 0 {
permissionResponse, err := m.PermissionChecker.CheckPermissions(becomeMemberPermissions, accountsAndChainIDs, true)
if err != nil {
return false, protobuf.CommunityMember_ROLE_NONE, err
}
return permissionResponse.Satisfied, protobuf.CommunityMember_ROLE_NONE, nil
}
return true, protobuf.CommunityMember_ROLE_NONE, nil
}
func (m *Manager) accountsSatisfyPermissionsToJoinChannels(community *Community, accounts []*protobuf.RevealedAccount) (map[string]*protobuf.CommunityChat, error) {
result := make(map[string]*protobuf.CommunityChat)
accountsAndChainIDs := revealedAccountsToAccountsAndChainIDsCombination(accounts)
for channelID, channel := range community.config.CommunityDescription.Chats {
channelViewOnlyPermissions := community.ChannelTokenPermissionsByType(community.IDString()+channelID, protobuf.CommunityTokenPermission_CAN_VIEW_CHANNEL)
channelViewAndPostPermissions := community.ChannelTokenPermissionsByType(community.IDString()+channelID, protobuf.CommunityTokenPermission_CAN_VIEW_AND_POST_CHANNEL)
channelPermissions := append(channelViewOnlyPermissions, channelViewAndPostPermissions...)
if len(channelPermissions) > 0 {
permissionResponse, err := m.PermissionChecker.CheckPermissions(channelPermissions, accountsAndChainIDs, true)
if err != nil {
return nil, err
}
if permissionResponse.Satisfied {
result[channelID] = channel
}
} else {
result[channelID] = channel
}
}
return result, nil
}
func (m *Manager) AcceptRequestToJoin(dbRequest *RequestToJoin) (*Community, error) {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
pk, err := common.HexToPubkey(dbRequest.PublicKey)
2021-01-11 10:32:51 +00:00
if err != nil {
return nil, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
community, err := m.GetByID(dbRequest.CommunityID)
if err != nil {
return nil, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if community.IsControlNode() {
revealedAccounts, err := m.persistence.GetRequestToJoinRevealedAddresses(dbRequest.ID)
if err != nil {
return nil, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
permissionsSatisfied, role, err := m.accountsSatisfyPermissionsToJoin(community, revealedAccounts)
if err != nil {
return nil, err
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if !permissionsSatisfied {
return community, ErrNoPermissionToJoin
}
memberRoles := []protobuf.CommunityMember_Roles{}
if role != protobuf.CommunityMember_ROLE_NONE {
memberRoles = []protobuf.CommunityMember_Roles{role}
}
_, err = community.AddMember(pk, memberRoles)
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
if err != nil {
return nil, err
}
channels, err := m.accountsSatisfyPermissionsToJoinChannels(community, revealedAccounts)
if err != nil {
return nil, err
}
for channelID := range channels {
_, err = community.AddMemberToChat(channelID, pk, memberRoles)
if err != nil {
return nil, err
}
}
dbRequest.State = RequestToJoinStateAccepted
if err := m.markRequestToJoinAsAccepted(pk, community); err != nil {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
return nil, err
}
if err = m.shareAcceptedRequestToJoinWithPrivilegedMembers(community, dbRequest); err != nil {
return nil, err
}
// if accepted member has a privilege role, share with him requests to join
memberRole := community.MemberRole(pk)
if memberRole == protobuf.CommunityMember_ROLE_OWNER || memberRole == protobuf.CommunityMember_ROLE_ADMIN ||
memberRole == protobuf.CommunityMember_ROLE_TOKEN_MASTER {
newPrivilegedMember := make(map[protobuf.CommunityMember_Roles][]*ecdsa.PublicKey)
newPrivilegedMember[memberRole] = []*ecdsa.PublicKey{pk}
if err = m.shareRequestsToJoinWithNewPrivilegedMembers(community, newPrivilegedMember); err != nil {
return nil, err
}
}
} else if community.hasPermissionToSendCommunityEvent(protobuf.CommunityEvent_COMMUNITY_REQUEST_TO_JOIN_ACCEPT) {
// admins do not perform permission checks, they merely mark the
// request as accepted (pending) and forward their decision to the control node
acceptedRequestsToJoin := make(map[string]*protobuf.CommunityRequestToJoin)
acceptedRequestsToJoin[dbRequest.PublicKey] = dbRequest.ToCommunityRequestToJoinProtobuf()
adminChanges := &CommunityEventChanges{
AcceptedRequestsToJoin: acceptedRequestsToJoin,
}
err := community.addNewCommunityEvent(community.ToCommunityRequestToJoinAcceptCommunityEvent(adminChanges))
if err != nil {
return nil, err
}
dbRequest.State = RequestToJoinStateAcceptedPending
if err := m.markRequestToJoinAsAcceptedPending(pk, community); err != nil {
return nil, err
}
} else {
return nil, ErrNotAuthorized
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) GetRequestToJoin(ID types.HexBytes) (*RequestToJoin, error) {
return m.persistence.GetRequestToJoin(ID)
2021-01-11 10:32:51 +00:00
}
func (m *Manager) DeclineRequestToJoin(dbRequest *RequestToJoin) (*Community, error) {
community, err := m.GetByID(dbRequest.CommunityID)
if err != nil {
return nil, err
}
adminEventCreated, err := community.DeclineRequestToJoin(dbRequest)
if err != nil {
return nil, err
}
requestToJoinState := RequestToJoinStateDeclined
if adminEventCreated {
requestToJoinState = RequestToJoinStateDeclinedPending // can only be declined by control node
}
dbRequest.State = requestToJoinState
err = m.persistence.SetRequestToJoinState(dbRequest.PublicKey, dbRequest.CommunityID, requestToJoinState)
if err != nil {
return nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return community, nil
2021-01-11 10:32:51 +00:00
}
func (m *Manager) shouldUserRetainDeclined(signer *ecdsa.PublicKey, community *Community, requestClock uint64) (bool, error) {
requestID := CalculateRequestID(common.PubkeyToHex(signer), types.HexBytes(community.IDString()))
request, err := m.persistence.GetRequestToJoin(requestID)
if err != nil {
if err == sql.ErrNoRows {
return false, nil
}
return false, err
}
return request.ShouldRetainDeclined(requestClock)
}
func (m *Manager) HandleCommunityCancelRequestToJoin(signer *ecdsa.PublicKey, request *protobuf.CommunityCancelRequestToJoin) (*RequestToJoin, error) {
community, err := m.GetByID(request.CommunityId)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
retainDeclined, err := m.shouldUserRetainDeclined(signer, community, request.Clock)
if err != nil {
return nil, err
}
if retainDeclined {
return nil, ErrCommunityRequestAlreadyRejected
}
err = m.markRequestToJoinAsCanceled(signer, community)
if err != nil {
return nil, err
}
requestToJoin, err := m.persistence.GetRequestToJoinByPk(common.PubkeyToHex(signer), community.ID(), RequestToJoinStateCanceled)
if err != nil {
return nil, err
}
return requestToJoin, nil
}
func (m *Manager) HandleCommunityRequestToJoin(signer *ecdsa.PublicKey, receiver *ecdsa.PublicKey, request *protobuf.CommunityRequestToJoin) (*Community, *RequestToJoin, error) {
community, err := m.GetByID(request.CommunityId)
2021-01-11 10:32:51 +00:00
if err != nil {
return nil, nil, err
2021-01-11 10:32:51 +00:00
}
if community == nil {
return nil, nil, ErrOrgNotFound
}
err = community.ValidateRequestToJoin(signer, request)
if err != nil {
return nil, nil, err
2021-01-11 10:32:51 +00:00
}
requestToJoin := &RequestToJoin{
PublicKey: common.PubkeyToHex(signer),
Clock: request.Clock,
ENSName: request.EnsName,
CommunityID: request.CommunityId,
State: RequestToJoinStatePending,
RevealedAccounts: request.RevealedAccounts,
2021-01-11 10:32:51 +00:00
}
requestToJoin.CalculateID()
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
existingRequestToJoin, err := m.persistence.GetRequestToJoin(requestToJoin.ID)
if err != nil && err != sql.ErrNoRows {
return nil, nil, err
2021-01-11 10:32:51 +00:00
}
if existingRequestToJoin == nil {
err = m.SaveRequestToJoin(requestToJoin)
if err != nil {
return nil, nil, err
}
} else {
retainDeclined, err := existingRequestToJoin.ShouldRetainDeclined(request.Clock)
if err != nil {
return nil, nil, err
}
if retainDeclined {
return nil, nil, ErrCommunityRequestAlreadyRejected
}
switch existingRequestToJoin.State {
case RequestToJoinStatePending, RequestToJoinStateDeclined, RequestToJoinStateCanceled:
// Another request have been received, save request back to pending state
err = m.SaveRequestToJoin(requestToJoin)
if err != nil {
return nil, nil, err
}
}
}
if community.IsControlNode() {
// verify if revealed addresses indeed belong to requester
for _, revealedAccount := range request.RevealedAccounts {
recoverParams := account.RecoverParams{
Message: types.EncodeHex(crypto.Keccak256(crypto.CompressPubkey(signer), community.ID(), requestToJoin.ID)),
Signature: types.EncodeHex(revealedAccount.Signature),
}
matching, err := m.accountsManager.CanRecover(recoverParams, types.HexToAddress(revealedAccount.Address))
if err != nil {
return nil, nil, err
}
if !matching {
// if ownership of only one wallet address cannot be verified,
// we mark the request as cancelled and stop
requestToJoin.State = RequestToJoinStateDeclined
return community, requestToJoin, nil
}
}
// Save revealed addresses + signatures so they can later be added
// to the control node's local table of known revealed addresses
err = m.persistence.SaveRequestToJoinRevealedAddresses(requestToJoin.ID, requestToJoin.RevealedAccounts)
if err != nil {
return nil, nil, err
}
if existingRequestToJoin != nil {
// request to join was already processed by privileged user
// and waits to get confirmation for its decision
if existingRequestToJoin.State == RequestToJoinStateDeclinedPending {
requestToJoin.State = RequestToJoinStateDeclined
return community, requestToJoin, nil
} else if existingRequestToJoin.State == RequestToJoinStateAcceptedPending {
requestToJoin.State = RequestToJoinStateAccepted
return community, requestToJoin, nil
} else if existingRequestToJoin.State == RequestToJoinStateAwaitingAddresses {
// community ownership changed, accept request automatically
requestToJoin.State = RequestToJoinStateAccepted
return community, requestToJoin, nil
}
}
// If user is already a member, then accept request automatically
// It may happen when member removes itself from community and then tries to rejoin
// More specifically, CommunityRequestToLeave may be delivered later than CommunityRequestToJoin, or not delivered at all
acceptAutomatically := community.AutoAccept() || community.HasMember(signer)
if acceptAutomatically {
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
// Don't check permissions here,
// it will be done further in the processing pipeline.
requestToJoin.State = RequestToJoinStateAccepted
return community, requestToJoin, nil
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
}
return community, requestToJoin, nil
2021-01-11 10:32:51 +00:00
}
func (m *Manager) HandleCommunityEditSharedAddresses(signer *ecdsa.PublicKey, request *protobuf.CommunityEditSharedAddresses) error {
community, err := m.GetByID(request.CommunityId)
if err != nil {
return err
}
if community == nil {
return ErrOrgNotFound
}
if err := community.ValidateEditSharedAddresses(signer, request); err != nil {
return err
}
// verify if revealed addresses indeed belong to requester
for _, revealedAccount := range request.RevealedAccounts {
recoverParams := account.RecoverParams{
Message: types.EncodeHex(crypto.Keccak256(crypto.CompressPubkey(signer), community.ID())),
Signature: types.EncodeHex(revealedAccount.Signature),
}
matching, err := m.accountsManager.CanRecover(recoverParams, types.HexToAddress(revealedAccount.Address))
if err != nil {
return err
}
if !matching {
// if ownership of only one wallet address cannot be verified we stop
return errors.New("wrong wallet address used")
}
}
requestToJoin := &RequestToJoin{
PublicKey: common.PubkeyToHex(signer),
CommunityID: community.ID(),
RevealedAccounts: request.RevealedAccounts,
}
requestToJoin.CalculateID()
err = m.persistence.RemoveRequestToJoinRevealedAddresses(requestToJoin.ID)
if err != nil {
return err
}
err = m.persistence.SaveRequestToJoinRevealedAddresses(requestToJoin.ID, requestToJoin.RevealedAccounts)
if err != nil {
return err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return err
}
if community.IsControlNode() {
m.publish(&Subscription{Community: community})
}
return nil
}
type CheckPermissionsResponse struct {
Satisfied bool `json:"satisfied"`
Permissions map[string]*PermissionTokenCriteriaResult `json:"permissions"`
ValidCombinations []*AccountChainIDsCombination `json:"validCombinations"`
}
type CheckPermissionToJoinResponse = CheckPermissionsResponse
type PermissionTokenCriteriaResult struct {
Criteria []bool `json:"criteria"`
}
type AccountChainIDsCombination struct {
Address gethcommon.Address `json:"address"`
ChainIDs []uint64 `json:"chainIds"`
}
func (c *CheckPermissionsResponse) calculateSatisfied() {
if len(c.Permissions) == 0 {
c.Satisfied = true
return
}
c.Satisfied = false
for _, p := range c.Permissions {
satisfied := true
for _, criteria := range p.Criteria {
if !criteria {
satisfied = false
break
}
}
if satisfied {
c.Satisfied = true
}
}
}
func calculateChainIDsSet(accountsAndChainIDs []*AccountChainIDsCombination, requirementsChainIDs map[uint64]bool) []uint64 {
revealedAccountsChainIDs := make([]uint64, 0)
revealedAccountsChainIDsMap := make(map[uint64]bool)
// we want all chainIDs provided by revealed addresses that also exist
// in the token requirements
for _, accountAndChainIDs := range accountsAndChainIDs {
for _, chainID := range accountAndChainIDs.ChainIDs {
if requirementsChainIDs[chainID] && !revealedAccountsChainIDsMap[chainID] {
revealedAccountsChainIDsMap[chainID] = true
revealedAccountsChainIDs = append(revealedAccountsChainIDs, chainID)
}
}
}
return revealedAccountsChainIDs
}
type CollectiblesByChain = map[uint64]map[gethcommon.Address]thirdparty.TokenBalancesPerContractAddress
func (m *Manager) GetOwnedERC721Tokens(walletAddresses []gethcommon.Address, tokenRequirements map[uint64]map[string]*protobuf.TokenCriteria, chainIDs []uint64) (CollectiblesByChain, error) {
if m.collectiblesManager == nil {
return nil, errors.New("no collectibles manager")
}
ownedERC721Tokens := make(CollectiblesByChain)
for chainID, erc721Tokens := range tokenRequirements {
skipChain := true
for _, cID := range chainIDs {
if chainID == cID {
skipChain = false
}
}
if skipChain {
continue
}
contractAddresses := make([]gethcommon.Address, 0)
for contractAddress := range erc721Tokens {
contractAddresses = append(contractAddresses, gethcommon.HexToAddress(contractAddress))
}
if _, exists := ownedERC721Tokens[chainID]; !exists {
ownedERC721Tokens[chainID] = make(map[gethcommon.Address]thirdparty.TokenBalancesPerContractAddress)
}
for _, owner := range walletAddresses {
balances, err := m.collectiblesManager.FetchBalancesByOwnerAndContractAddress(walletcommon.ChainID(chainID), owner, contractAddresses)
if err != nil {
m.logger.Info("couldn't fetch owner assets", zap.Error(err))
return nil, err
}
ownedERC721Tokens[chainID][owner] = balances
}
}
return ownedERC721Tokens, nil
}
func (m *Manager) CheckChannelPermissions(communityID types.HexBytes, chatID string, addresses []gethcommon.Address) (*CheckChannelPermissionsResponse, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
if chatID == "" {
return nil, errors.New(fmt.Sprintf("couldn't check channel permissions, invalid chat id: %s", chatID))
}
viewOnlyPermissions := community.ChannelTokenPermissionsByType(chatID, protobuf.CommunityTokenPermission_CAN_VIEW_CHANNEL)
viewAndPostPermissions := community.ChannelTokenPermissionsByType(chatID, protobuf.CommunityTokenPermission_CAN_VIEW_AND_POST_CHANNEL)
allChainIDs, err := m.tokenManager.GetAllChainIDs()
if err != nil {
return nil, err
}
accountsAndChainIDs := combineAddressesAndChainIDs(addresses, allChainIDs)
response, err := m.checkChannelPermissions(viewOnlyPermissions, viewAndPostPermissions, accountsAndChainIDs, false)
if err != nil {
return nil, err
}
err = m.persistence.SaveCheckChannelPermissionResponse(communityID.String(), chatID, response)
if err != nil {
return nil, err
}
return response, nil
}
type CheckChannelPermissionsResponse struct {
ViewOnlyPermissions *CheckChannelViewOnlyPermissionsResult `json:"viewOnlyPermissions"`
ViewAndPostPermissions *CheckChannelViewAndPostPermissionsResult `json:"viewAndPostPermissions"`
}
type CheckChannelViewOnlyPermissionsResult struct {
Satisfied bool `json:"satisfied"`
Permissions map[string]*PermissionTokenCriteriaResult `json:"permissions"`
}
type CheckChannelViewAndPostPermissionsResult struct {
Satisfied bool `json:"satisfied"`
Permissions map[string]*PermissionTokenCriteriaResult `json:"permissions"`
}
func (m *Manager) checkChannelPermissions(viewOnlyPermissions []*CommunityTokenPermission, viewAndPostPermissions []*CommunityTokenPermission, accountsAndChainIDs []*AccountChainIDsCombination, shortcircuit bool) (*CheckChannelPermissionsResponse, error) {
response := &CheckChannelPermissionsResponse{
ViewOnlyPermissions: &CheckChannelViewOnlyPermissionsResult{
Satisfied: false,
Permissions: make(map[string]*PermissionTokenCriteriaResult),
},
ViewAndPostPermissions: &CheckChannelViewAndPostPermissionsResult{
Satisfied: false,
Permissions: make(map[string]*PermissionTokenCriteriaResult),
},
}
viewOnlyPermissionsResponse, err := m.PermissionChecker.CheckPermissions(viewOnlyPermissions, accountsAndChainIDs, shortcircuit)
if err != nil {
return nil, err
}
viewAndPostPermissionsResponse, err := m.PermissionChecker.CheckPermissions(viewAndPostPermissions, accountsAndChainIDs, shortcircuit)
if err != nil {
return nil, err
}
hasViewOnlyPermissions := len(viewOnlyPermissions) > 0
hasViewAndPostPermissions := len(viewAndPostPermissions) > 0
if (hasViewAndPostPermissions && !hasViewOnlyPermissions) || (hasViewOnlyPermissions && hasViewAndPostPermissions && viewAndPostPermissionsResponse.Satisfied) {
response.ViewOnlyPermissions.Satisfied = viewAndPostPermissionsResponse.Satisfied
} else {
response.ViewOnlyPermissions.Satisfied = viewOnlyPermissionsResponse.Satisfied
}
response.ViewOnlyPermissions.Permissions = viewOnlyPermissionsResponse.Permissions
if (hasViewOnlyPermissions && !viewOnlyPermissionsResponse.Satisfied) ||
(hasViewOnlyPermissions && !hasViewAndPostPermissions) {
response.ViewAndPostPermissions.Satisfied = false
} else {
response.ViewAndPostPermissions.Satisfied = viewAndPostPermissionsResponse.Satisfied
}
response.ViewAndPostPermissions.Permissions = viewAndPostPermissionsResponse.Permissions
return response, nil
}
func (m *Manager) CheckAllChannelsPermissions(communityID types.HexBytes, addresses []gethcommon.Address) (*CheckAllChannelsPermissionsResponse, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
channels := community.Chats()
allChainIDs, err := m.tokenManager.GetAllChainIDs()
if err != nil {
return nil, err
}
accountsAndChainIDs := combineAddressesAndChainIDs(addresses, allChainIDs)
response := &CheckAllChannelsPermissionsResponse{
Channels: make(map[string]*CheckChannelPermissionsResponse),
}
for channelID := range channels {
viewOnlyPermissions := community.ChannelTokenPermissionsByType(community.IDString()+channelID, protobuf.CommunityTokenPermission_CAN_VIEW_CHANNEL)
viewAndPostPermissions := community.ChannelTokenPermissionsByType(community.IDString()+channelID, protobuf.CommunityTokenPermission_CAN_VIEW_AND_POST_CHANNEL)
checkChannelPermissionsResponse, err := m.checkChannelPermissions(viewOnlyPermissions, viewAndPostPermissions, accountsAndChainIDs, false)
if err != nil {
return nil, err
}
err = m.persistence.SaveCheckChannelPermissionResponse(community.IDString(), community.IDString()+channelID, checkChannelPermissionsResponse)
if err != nil {
return nil, err
}
response.Channels[community.IDString()+channelID] = checkChannelPermissionsResponse
}
return response, nil
}
func (m *Manager) GetCheckChannelPermissionResponses(communityID types.HexBytes) (*CheckAllChannelsPermissionsResponse, error) {
response, err := m.persistence.GetCheckChannelPermissionResponses(communityID.String())
if err != nil {
return nil, err
}
return &CheckAllChannelsPermissionsResponse{Channels: response}, nil
}
type CheckAllChannelsPermissionsResponse struct {
Channels map[string]*CheckChannelPermissionsResponse `json:"channels"`
}
func (m *Manager) HandleCommunityRequestToJoinResponse(signer *ecdsa.PublicKey, request *protobuf.CommunityRequestToJoinResponse) (*RequestToJoin, error) {
pkString := common.PubkeyToHex(&m.identity.PublicKey)
community, err := m.GetByID(request.CommunityId)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
communityDescriptionBytes, err := proto.Marshal(request.Community)
if err != nil {
return nil, err
}
// We need to wrap `request.Community` in an `ApplicationMetadataMessage`
// of type `CommunityDescription` because `UpdateCommunityDescription` expects this.
//
// This is merely for marsheling/unmarsheling, hence we attaching a `Signature`
// is not needed.
metadataMessage := &protobuf.ApplicationMetadataMessage{
Payload: communityDescriptionBytes,
Type: protobuf.ApplicationMetadataMessage_COMMUNITY_DESCRIPTION,
}
appMetadataMsg, err := proto.Marshal(metadataMessage)
if err != nil {
return nil, err
}
isControlNodeSigner := common.IsPubKeyEqual(community.PublicKey(), signer)
if !isControlNodeSigner {
return nil, ErrNotAuthorized
}
_, err = community.UpdateCommunityDescription(request.Community, appMetadataMsg, nil)
if err != nil {
return nil, err
}
if err = m.HandleCommunityTokensMetadataByPrivilegedMembers(community); err != nil {
return nil, err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
if request.Accepted {
err = m.markRequestToJoinAsAccepted(&m.identity.PublicKey, community)
if err != nil {
return nil, err
}
} else {
err = m.persistence.SetRequestToJoinState(pkString, community.ID(), RequestToJoinStateDeclined)
if err != nil {
return nil, err
}
}
return m.persistence.GetRequestToJoinByPkAndCommunityID(pkString, community.ID())
}
func (m *Manager) HandleCommunityRequestToLeave(signer *ecdsa.PublicKey, proto *protobuf.CommunityRequestToLeave) error {
requestToLeave := NewRequestToLeave(common.PubkeyToHex(signer), proto)
if err := m.persistence.SaveRequestToLeave(requestToLeave); err != nil {
return err
}
// Ensure corresponding requestToJoin clock is older than requestToLeave
requestToJoin, err := m.persistence.GetRequestToJoin(requestToLeave.ID)
if err != nil {
return err
}
if requestToJoin.Clock > requestToLeave.Clock {
return ErrOldRequestToLeave
}
return nil
}
func UnwrapCommunityDescriptionMessage(payload []byte) (*ecdsa.PublicKey, *protobuf.CommunityDescription, error) {
applicationMetadataMessage := &protobuf.ApplicationMetadataMessage{}
err := proto.Unmarshal(payload, applicationMetadataMessage)
if err != nil {
return nil, nil, err
}
if applicationMetadataMessage.Type != protobuf.ApplicationMetadataMessage_COMMUNITY_DESCRIPTION {
return nil, nil, ErrInvalidMessage
}
signer, err := applicationMetadataMessage.RecoverKey()
if err != nil {
return nil, nil, err
}
description := &protobuf.CommunityDescription{}
err = proto.Unmarshal(applicationMetadataMessage.Payload, description)
if err != nil {
return nil, nil, err
}
return signer, description, nil
}
func (m *Manager) JoinCommunity(id types.HexBytes, forceJoin bool) (*Community, error) {
2021-01-11 10:32:51 +00:00
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
if community == nil {
return nil, ErrOrgNotFound
}
if !forceJoin && community.Joined() {
// Nothing to do, we are already joined
return community, ErrOrgAlreadyJoined
}
2020-12-17 14:36:09 +00:00
community.Join()
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
return community, nil
}
func (m *Manager) SpectateCommunity(id types.HexBytes) (*Community, error) {
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
community.Spectate()
if err = m.persistence.SaveCommunity(community); err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) GetMagnetlinkMessageClock(communityID types.HexBytes) (uint64, error) {
return m.persistence.GetMagnetlinkMessageClock(communityID)
}
func (m *Manager) GetRequestToJoinIDByPkAndCommunityID(pk *ecdsa.PublicKey, communityID []byte) ([]byte, error) {
return m.persistence.GetRequestToJoinIDByPkAndCommunityID(common.PubkeyToHex(pk), communityID)
}
func (m *Manager) UpdateCommunityDescriptionMagnetlinkMessageClock(communityID types.HexBytes, clock uint64) error {
community, err := m.GetByIDString(communityID.String())
if err != nil {
return err
}
community.config.CommunityDescription.ArchiveMagnetlinkClock = clock
return m.persistence.SaveCommunity(community)
}
func (m *Manager) UpdateMagnetlinkMessageClock(communityID types.HexBytes, clock uint64) error {
return m.persistence.UpdateMagnetlinkMessageClock(communityID, clock)
}
func (m *Manager) UpdateLastSeenMagnetlink(communityID types.HexBytes, magnetlinkURI string) error {
return m.persistence.UpdateLastSeenMagnetlink(communityID, magnetlinkURI)
}
func (m *Manager) GetLastSeenMagnetlink(communityID types.HexBytes) (string, error) {
return m.persistence.GetLastSeenMagnetlink(communityID)
}
2021-01-11 10:32:51 +00:00
func (m *Manager) LeaveCommunity(id types.HexBytes) (*Community, error) {
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
if community == nil {
return nil, ErrOrgNotFound
}
community.RemoveOurselvesFromOrg(&m.identity.PublicKey)
2020-12-17 14:36:09 +00:00
community.Leave()
if err = m.persistence.SaveCommunity(community); err != nil {
return nil, err
}
2020-12-17 14:36:09 +00:00
return community, nil
}
// Same as LeaveCommunity, but we want to stay spectating
func (m *Manager) KickedOutOfCommunity(id types.HexBytes) (*Community, error) {
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
community.RemoveOurselvesFromOrg(&m.identity.PublicKey)
community.Leave()
community.Spectate()
if err = m.persistence.SaveCommunity(community); err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) AddMemberOwnerToCommunity(communityID types.HexBytes, pk *ecdsa.PublicKey) (*Community, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
_, err = community.AddMember(pk, []protobuf.CommunityMember_Roles{protobuf.CommunityMember_ROLE_OWNER})
if err != nil {
return nil, err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
m.publish(&Subscription{Community: community})
return community, nil
}
2021-01-11 10:32:51 +00:00
func (m *Manager) RemoveUserFromCommunity(id types.HexBytes, pk *ecdsa.PublicKey) (*Community, error) {
community, err := m.GetByID(id)
2020-12-21 15:10:52 +00:00
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
_, err = community.RemoveUserFromOrg(pk)
if err != nil {
return nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) UnbanUserFromCommunity(request *requests.UnbanUserFromCommunity) (*Community, error) {
id := request.CommunityID
publicKey, err := common.HexToPubkey(request.User.String())
if err != nil {
return nil, err
}
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
_, err = community.UnbanUserFromCommunity(publicKey)
if err != nil {
return nil, err
}
err = m.saveAndPublish(community)
if err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) AddRoleToMember(request *requests.AddRoleToMember) (*Community, error) {
id := request.CommunityID
publicKey, err := common.HexToPubkey(request.User.String())
if err != nil {
return nil, err
}
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.hasMember(publicKey) {
return nil, ErrMemberNotFound
}
_, err = community.AddRoleToMember(publicKey, request.Role)
if err != nil {
return nil, err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
m.publish(&Subscription{Community: community})
return community, nil
}
func (m *Manager) RemoveRoleFromMember(request *requests.RemoveRoleFromMember) (*Community, error) {
id := request.CommunityID
publicKey, err := common.HexToPubkey(request.User.String())
if err != nil {
return nil, err
}
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.hasMember(publicKey) {
return nil, ErrMemberNotFound
}
_, err = community.RemoveRoleFromMember(publicKey, request.Role)
if err != nil {
return nil, err
}
err = m.persistence.SaveCommunity(community)
if err != nil {
return nil, err
}
m.publish(&Subscription{Community: community})
return community, nil
}
func (m *Manager) BanUserFromCommunity(request *requests.BanUserFromCommunity) (*Community, error) {
id := request.CommunityID
publicKey, err := common.HexToPubkey(request.User.String())
if err != nil {
return nil, err
}
community, err := m.GetByID(id)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
_, err = community.BanUserFromCommunity(publicKey)
if err != nil {
return nil, err
}
err = m.saveAndPublish(community)
2020-12-21 15:10:52 +00:00
if err != nil {
return nil, err
}
return community, nil
}
// Apply events to raw community
func (m *Manager) initializeCommunity(community *Community) error {
err := community.updateCommunityDescriptionByEvents()
if err != nil {
return err
}
if m.transport != nil && m.transport.WakuVersion() == 2 {
privKey, err := m.transport.RetrievePubsubTopicKey(community.PubsubTopic())
if err != nil {
return err
}
community.config.PubsubTopicPrivateKey = privKey
}
// Workaround for https://github.com/status-im/status-desktop/issues/12188
HydrateChannelsMembers(community.IDString(), community.config.CommunityDescription)
return nil
}
2021-01-11 10:32:51 +00:00
func (m *Manager) GetByID(id []byte) (*Community, error) {
community, err := m.persistence.GetByID(&m.identity.PublicKey, m.installationID, id)
if err != nil {
return nil, err
}
if community == nil {
return nil, nil
}
err = m.initializeCommunity(community)
if err != nil {
return nil, err
}
return community, nil
2021-01-11 10:32:51 +00:00
}
func (m *Manager) GetByIDString(idString string) (*Community, error) {
id, err := types.DecodeHex(idString)
if err != nil {
return nil, err
}
2021-01-11 10:32:51 +00:00
return m.GetByID(id)
}
func (m *Manager) SaveRequestToJoinRevealedAddresses(requestID types.HexBytes, revealedAccounts []*protobuf.RevealedAccount) error {
return m.persistence.SaveRequestToJoinRevealedAddresses(requestID, revealedAccounts)
}
func (m *Manager) RemoveRequestToJoinRevealedAddresses(requestID types.HexBytes) error {
return m.persistence.RemoveRequestToJoinRevealedAddresses(requestID)
}
func (m *Manager) SaveRequestToJoinAndCommunity(requestToJoin *RequestToJoin, community *Community) (*Community, *RequestToJoin, error) {
if err := m.persistence.SaveRequestToJoin(requestToJoin); err != nil {
return nil, nil, err
}
community.config.RequestedToJoinAt = uint64(time.Now().Unix())
community.AddRequestToJoin(requestToJoin)
// Save revealed addresses to our own table so that we can retrieve them later when editing
if err := m.SaveRequestToJoinRevealedAddresses(requestToJoin.ID, requestToJoin.RevealedAccounts); err != nil {
return nil, nil, err
}
return community, requestToJoin, nil
}
func (m *Manager) CheckCommunityForJoining(communityID types.HexBytes) (*Community, error) {
community, err := m.GetByID(communityID)
2021-01-11 10:32:51 +00:00
if err != nil {
return nil, err
2021-01-11 10:32:51 +00:00
}
if community == nil {
return nil, ErrOrgNotFound
}
// We don't allow requesting access if already joined
if community.Joined() {
return nil, ErrAlreadyJoined
2021-01-11 10:32:51 +00:00
}
err = community.updateCommunityDescriptionByEvents()
if err != nil {
return nil, err
}
return community, nil
}
func (m *Manager) CreateRequestToJoin(request *requests.RequestToJoinCommunity) *RequestToJoin {
2021-01-11 10:32:51 +00:00
clock := uint64(time.Now().Unix())
requestToJoin := &RequestToJoin{
PublicKey: common.PubkeyToHex(&m.identity.PublicKey),
Clock: clock,
ENSName: request.ENSName,
CommunityID: request.CommunityID,
State: RequestToJoinStatePending,
Our: true,
RevealedAccounts: make([]*protobuf.RevealedAccount, 0),
2021-01-11 10:32:51 +00:00
}
requestToJoin.CalculateID()
addSignature := len(request.Signatures) == len(request.AddressesToReveal)
for i := range request.AddressesToReveal {
revealedAcc := &protobuf.RevealedAccount{
Address: request.AddressesToReveal[i],
IsAirdropAddress: types.HexToAddress(request.AddressesToReveal[i]) == types.HexToAddress(request.AirdropAddress),
}
if addSignature {
revealedAcc.Signature = request.Signatures[i]
}
requestToJoin.RevealedAccounts = append(requestToJoin.RevealedAccounts, revealedAcc)
}
return requestToJoin
2021-01-11 10:32:51 +00:00
}
func (m *Manager) SaveRequestToJoin(request *RequestToJoin) error {
return m.persistence.SaveRequestToJoin(request)
}
func (m *Manager) CanceledRequestsToJoinForUser(pk *ecdsa.PublicKey) ([]*RequestToJoin, error) {
return m.persistence.CanceledRequestsToJoinForUser(common.PubkeyToHex(pk))
}
func (m *Manager) PendingRequestsToJoin() ([]*RequestToJoin, error) {
return m.persistence.PendingRequestsToJoin()
}
2021-01-11 10:32:51 +00:00
func (m *Manager) PendingRequestsToJoinForUser(pk *ecdsa.PublicKey) ([]*RequestToJoin, error) {
return m.persistence.RequestsToJoinForUserByState(common.PubkeyToHex(pk), RequestToJoinStatePending)
}
func (m *Manager) AwaitingAddressesRequestsToJoinForUser(pk *ecdsa.PublicKey) ([]*RequestToJoin, error) {
return m.persistence.RequestsToJoinForUserByState(common.PubkeyToHex(pk), RequestToJoinStateAwaitingAddresses)
2021-01-11 10:32:51 +00:00
}
func (m *Manager) PendingRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
m.logger.Info("fetching pending invitations", zap.String("community-id", id.String()))
return m.persistence.PendingRequestsToJoinForCommunity(id)
}
func (m *Manager) DeclinedRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
m.logger.Info("fetching declined invitations", zap.String("community-id", id.String()))
return m.persistence.DeclinedRequestsToJoinForCommunity(id)
}
func (m *Manager) CanceledRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
m.logger.Info("fetching canceled invitations", zap.String("community-id", id.String()))
return m.persistence.CanceledRequestsToJoinForCommunity(id)
}
func (m *Manager) AcceptedRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
m.logger.Info("fetching canceled invitations", zap.String("community-id", id.String()))
return m.persistence.AcceptedRequestsToJoinForCommunity(id)
}
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
func (m *Manager) AcceptedPendingRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
return m.persistence.AcceptedPendingRequestsToJoinForCommunity(id)
}
func (m *Manager) DeclinedPendingRequestsToJoinForCommunity(id types.HexBytes) ([]*RequestToJoin, error) {
return m.persistence.DeclinedPendingRequestsToJoinForCommunity(id)
}
func (m *Manager) GetPubsubTopic(communityID string) (string, error) {
community, err := m.GetByIDString(communityID)
if err != nil {
return "", err
}
if community == nil {
return relay.DefaultWakuTopic, nil
}
return transport.GetPubsubTopic(community.Shard().TransportShard()), nil
refactor: EventSenders forward RequestToJoin decision to control node This is a bigger change in how community membership requests are handled among admins, token masters, owners, and control nodes. Prior to this commit, all privileged users, also known as `EventSenders`, were able to accept and reject community membership requests and those changes would be applied by all users. This commit changes this behaviour such that: 1. EventSenders can make a decision (accept, reject), but merely forward their decision to the control node, which ultimately has to confirm it 2. EventSenders are no longer removing or adding members to and from communities 3. When an eventsender signaled a decision, the membership request will enter a pending state (acceptedPending or rejectedPending) 4. Once a decision was made by one eventsender, no other eventsender can override that decision This implementation is covered with a bunch of tests: - Ensure that decision made by event sender is shared with other event senders - `testAcceptMemberRequestToJoinResponseSharedWithOtherEventSenders()` - `testRejectMemberRequestToJoinResponseSharedWithOtherEventSenders()` - Ensure memebrship request stays pending, until control node has confirmed decision by event senders - `testAcceptMemberRequestToJoinNotConfirmedByControlNode()` - `testRejectMemberRequestToJoinNotConfirmedByControlNode()` - Ensure that decision made by event sender cannot be overriden by other event senders - `testEventSenderCannotOverrideRequestToJoinState()` These test cases live in three test suites for different event sender types respectively - `OwnerWithoutCommunityKeyCommunityEventsSuite` - `TokenMasterCommunityEventsSuite` - `AdminCommunityEventsSuite` In addition to the changes mentioned above, there's also a smaller changes that ensures membership requests to *not* attached revealed wallet addresses when the requests are sent to event senders (in addition to control nodes). Requests send to a control node will still include revealed addresses as the control node needs them to verify token permissions. This commit does not yet handle the case of event senders attempting to kick and ban members. Similar to accepting and rejecting membership requests, kicking and banning need a new pending state. However, we don't track such state in local databases yet so those two cases will be handled in future commit to not have this commit grow larger.
2023-08-02 12:04:47 +00:00
}
func (m *Manager) RequestsToJoinForCommunityAwaitingAddresses(id types.HexBytes) ([]*RequestToJoin, error) {
m.logger.Info("fetching ownership changed invitations", zap.String("community-id", id.String()))
return m.persistence.RequestsToJoinForCommunityAwaitingAddresses(id)
}
2021-01-11 10:32:51 +00:00
func (m *Manager) CanPost(pk *ecdsa.PublicKey, communityID string, chatID string, grant []byte) (bool, error) {
community, err := m.GetByIDString(communityID)
if err != nil {
return false, err
}
2020-12-17 14:36:09 +00:00
if community == nil {
return false, nil
}
2020-12-17 14:36:09 +00:00
return community.CanPost(pk, chatID, grant)
}
2022-05-27 09:14:40 +00:00
func (m *Manager) IsEncrypted(communityID string) (bool, error) {
community, err := m.GetByIDString(communityID)
if err != nil {
return false, err
}
return community.Encrypted(), nil
}
func (m *Manager) IsChannelEncrypted(communityID string, chatID string) (bool, error) {
community, err := m.GetByIDString(communityID)
if err != nil {
return false, err
}
2022-05-27 09:14:40 +00:00
2023-10-26 15:09:43 +00:00
channelID := strings.TrimPrefix(chatID, communityID)
return community.ChannelEncrypted(channelID), nil
2022-05-27 09:14:40 +00:00
}
func (m *Manager) ShouldHandleSyncCommunity(community *protobuf.SyncInstallationCommunity) (bool, error) {
return m.persistence.ShouldHandleSyncCommunity(community)
}
func (m *Manager) ShouldHandleSyncCommunitySettings(communitySettings *protobuf.SyncCommunitySettings) (bool, error) {
return m.persistence.ShouldHandleSyncCommunitySettings(communitySettings)
}
func (m *Manager) HandleSyncCommunitySettings(syncCommunitySettings *protobuf.SyncCommunitySettings) (*CommunitySettings, error) {
id, err := types.DecodeHex(syncCommunitySettings.CommunityId)
if err != nil {
return nil, err
}
settings, err := m.persistence.GetCommunitySettingsByID(id)
if err != nil {
return nil, err
}
if settings == nil {
settings = &CommunitySettings{
CommunityID: syncCommunitySettings.CommunityId,
HistoryArchiveSupportEnabled: syncCommunitySettings.HistoryArchiveSupportEnabled,
Clock: syncCommunitySettings.Clock,
}
}
if syncCommunitySettings.Clock > settings.Clock {
settings.CommunityID = syncCommunitySettings.CommunityId
settings.HistoryArchiveSupportEnabled = syncCommunitySettings.HistoryArchiveSupportEnabled
settings.Clock = syncCommunitySettings.Clock
}
err = m.persistence.SaveCommunitySettings(*settings)
if err != nil {
return nil, err
}
return settings, nil
}
func (m *Manager) SetSyncClock(id []byte, clock uint64) error {
return m.persistence.SetSyncClock(id, clock)
}
func (m *Manager) SetPrivateKey(id []byte, privKey *ecdsa.PrivateKey) error {
return m.persistence.SetPrivateKey(id, privKey)
}
2022-03-28 10:10:40 +00:00
func (m *Manager) GetSyncedRawCommunity(id []byte) (*RawCommunityRow, error) {
return m.persistence.getSyncedRawCommunity(id)
}
func (m *Manager) GetCommunitySettingsByID(id types.HexBytes) (*CommunitySettings, error) {
return m.persistence.GetCommunitySettingsByID(id)
}
func (m *Manager) GetCommunitiesSettings() ([]CommunitySettings, error) {
return m.persistence.GetCommunitiesSettings()
}
func (m *Manager) SaveCommunitySettings(settings CommunitySettings) error {
return m.persistence.SaveCommunitySettings(settings)
}
func (m *Manager) CommunitySettingsExist(id types.HexBytes) (bool, error) {
return m.persistence.CommunitySettingsExist(id)
}
func (m *Manager) DeleteCommunitySettings(id types.HexBytes) error {
return m.persistence.DeleteCommunitySettings(id)
}
func (m *Manager) UpdateCommunitySettings(settings CommunitySettings) error {
return m.persistence.UpdateCommunitySettings(settings)
}
func (m *Manager) GetOwnedCommunitiesChatIDs() (map[string]bool, error) {
ownedCommunities, err := m.Controlled()
if err != nil {
return nil, err
}
chatIDs := make(map[string]bool)
for _, c := range ownedCommunities {
if c.Joined() {
for _, id := range c.ChatIDs() {
chatIDs[id] = true
}
}
}
return chatIDs, nil
}
func (m *Manager) GetCommunityChatsFilters(communityID types.HexBytes) ([]*transport.Filter, error) {
chatIDs, err := m.persistence.GetCommunityChatIDs(communityID)
if err != nil {
return nil, err
}
filters := []*transport.Filter{}
for _, cid := range chatIDs {
filters = append(filters, m.transport.FilterByChatID(cid))
}
return filters, nil
}
func (m *Manager) GetCommunityChatsTopics(communityID types.HexBytes) ([]types.TopicType, error) {
filters, err := m.GetCommunityChatsFilters(communityID)
if err != nil {
return nil, err
}
topics := []types.TopicType{}
for _, filter := range filters {
topics = append(topics, filter.ContentTopic)
}
return topics, nil
}
func (m *Manager) StoreWakuMessage(message *types.Message) error {
return m.persistence.SaveWakuMessage(message)
}
2022-09-29 11:50:23 +00:00
func (m *Manager) StoreWakuMessages(messages []*types.Message) error {
return m.persistence.SaveWakuMessages(messages)
}
func (m *Manager) GetLatestWakuMessageTimestamp(topics []types.TopicType) (uint64, error) {
return m.persistence.GetLatestWakuMessageTimestamp(topics)
}
func (m *Manager) GetOldestWakuMessageTimestamp(topics []types.TopicType) (uint64, error) {
return m.persistence.GetOldestWakuMessageTimestamp(topics)
}
func (m *Manager) GetLastMessageArchiveEndDate(communityID types.HexBytes) (uint64, error) {
return m.persistence.GetLastMessageArchiveEndDate(communityID)
}
func (m *Manager) GetHistoryArchivePartitionStartTimestamp(communityID types.HexBytes) (uint64, error) {
filters, err := m.GetCommunityChatsFilters(communityID)
if err != nil {
m.LogStdout("failed to get community chats filters", zap.Error(err))
return 0, err
}
if len(filters) == 0 {
// If we don't have chat filters, we likely don't have any chats
// associated to this community, which means there's nothing more
// to do here
return 0, nil
}
topics := []types.TopicType{}
for _, filter := range filters {
topics = append(topics, filter.ContentTopic)
}
lastArchiveEndDateTimestamp, err := m.GetLastMessageArchiveEndDate(communityID)
if err != nil {
m.LogStdout("failed to get last archive end date", zap.Error(err))
return 0, err
}
if lastArchiveEndDateTimestamp == 0 {
// If we don't have a tracked last message archive end date, it
// means we haven't created an archive before, which means
// the next thing to look at is the oldest waku message timestamp for
// this community
lastArchiveEndDateTimestamp, err = m.GetOldestWakuMessageTimestamp(topics)
if err != nil {
m.LogStdout("failed to get oldest waku message timestamp", zap.Error(err))
return 0, err
}
if lastArchiveEndDateTimestamp == 0 {
// This means there's no waku message stored for this community so far
// (even after requesting possibly missed messages), so no messages exist yet that can be archived
m.LogStdout("can't find valid `lastArchiveEndTimestamp`")
return 0, nil
}
}
return lastArchiveEndDateTimestamp, nil
}
func (m *Manager) CreateAndSeedHistoryArchive(communityID types.HexBytes, topics []types.TopicType, startDate time.Time, endDate time.Time, partition time.Duration, encrypt bool) error {
m.UnseedHistoryArchiveTorrent(communityID)
feat(CommunitiesManager): introduce `CreateHistoryArchiveTorrentFromMessages` API Prior to this commit we had a `CreateHistoryArchiveTorrent()` API which takes a `startDate`, an `endDate` and a `partition` to create a bunch of message archives, given a certain time range. The function expects the messages to live in the database, which means, all messages that need to be archived have to be saved there at some point. This turns out to be an issue when importing communities from third party services, where, sometimes, there are several thousands of messages including attachment payloads, that have to be save to the database first. There are only two options to get the messages into the database: 1. Make one write operation with all messages - this slow, takes a long time and blocks the database until done 2. Create message chunks and perform multiple write operations - this is also slow, takes long but makes the database a bit more responsive as it's many smaller operations instead of one big one Option 2) turned out to not be super feasible either as sometimes, inserting even a single such message can take up to 10 seconds (depending on payload) Which brings me to the third option. **A third option** is to not store those imported messages as waku message into the database, just to later query them again to create the archives, but instead create the archives right away from all the messages that have been loaded into memory. This is significantly faster and doesn't block the database. To make this possible, this commit introduces a `CreateHistoryArchiveTorrentFromMessages()` API, and a `CreateHistoryArchiveTorrentFromDB()` API which can be used for different use cases.
2022-10-20 14:37:04 +00:00
_, err := m.CreateHistoryArchiveTorrentFromDB(communityID, topics, startDate, endDate, partition, encrypt)
if err != nil {
return err
}
return m.SeedHistoryArchiveTorrent(communityID)
}
func (m *Manager) StartHistoryArchiveTasksInterval(community *Community, interval time.Duration) {
id := community.IDString()
if _, exists := m.historyArchiveTasks.Load(id); exists {
m.LogStdout("history archive tasks interval already in progres", zap.String("id", id))
return
}
cancel := make(chan struct{})
m.historyArchiveTasks.Store(id, cancel)
m.historyArchiveTasksWaitGroup.Add(1)
ticker := time.NewTicker(interval)
defer ticker.Stop()
m.LogStdout("starting history archive tasks interval", zap.String("id", id))
for {
select {
case <-ticker.C:
m.LogStdout("starting archive task...", zap.String("id", id))
lastArchiveEndDateTimestamp, err := m.GetHistoryArchivePartitionStartTimestamp(community.ID())
if err != nil {
m.LogStdout("failed to get last archive end date", zap.Error(err))
continue
}
if lastArchiveEndDateTimestamp == 0 {
// This means there are no waku messages for this community,
// so nothing to do here
m.LogStdout("couldn't determine archive start date - skipping")
continue
}
topics, err := m.GetCommunityChatsTopics(community.ID())
if err != nil {
m.LogStdout("failed to get community chat topics ", zap.Error(err))
continue
}
ts := time.Now().Unix()
to := time.Unix(ts, 0)
lastArchiveEndDate := time.Unix(int64(lastArchiveEndDateTimestamp), 0)
err = m.CreateAndSeedHistoryArchive(community.ID(), topics, lastArchiveEndDate, to, interval, community.Encrypted())
if err != nil {
m.LogStdout("failed to create and seed history archive", zap.Error(err))
continue
}
case <-cancel:
m.UnseedHistoryArchiveTorrent(community.ID())
m.historyArchiveTasks.Delete(id)
m.historyArchiveTasksWaitGroup.Done()
return
}
}
}
func (m *Manager) StopHistoryArchiveTasksIntervals() {
m.historyArchiveTasks.Range(func(_, task interface{}) bool {
close(task.(chan struct{})) // Need to cast to the chan
return true
})
// Stoping archive interval tasks is async, so we need
// to wait for all of them to be closed before we shutdown
// the torrent client
m.historyArchiveTasksWaitGroup.Wait()
}
func (m *Manager) StopHistoryArchiveTasksInterval(communityID types.HexBytes) {
task, exists := m.historyArchiveTasks.Load(communityID.String())
if exists {
m.logger.Info("Stopping history archive tasks interval", zap.Any("id", communityID.String()))
close(task.(chan struct{})) // Need to cast to the chan
}
}
type EncodedArchiveData struct {
padding int
bytes []byte
}
feat(CommunitiesManager): introduce `CreateHistoryArchiveTorrentFromMessages` API Prior to this commit we had a `CreateHistoryArchiveTorrent()` API which takes a `startDate`, an `endDate` and a `partition` to create a bunch of message archives, given a certain time range. The function expects the messages to live in the database, which means, all messages that need to be archived have to be saved there at some point. This turns out to be an issue when importing communities from third party services, where, sometimes, there are several thousands of messages including attachment payloads, that have to be save to the database first. There are only two options to get the messages into the database: 1. Make one write operation with all messages - this slow, takes a long time and blocks the database until done 2. Create message chunks and perform multiple write operations - this is also slow, takes long but makes the database a bit more responsive as it's many smaller operations instead of one big one Option 2) turned out to not be super feasible either as sometimes, inserting even a single such message can take up to 10 seconds (depending on payload) Which brings me to the third option. **A third option** is to not store those imported messages as waku message into the database, just to later query them again to create the archives, but instead create the archives right away from all the messages that have been loaded into memory. This is significantly faster and doesn't block the database. To make this possible, this commit introduces a `CreateHistoryArchiveTorrentFromMessages()` API, and a `CreateHistoryArchiveTorrentFromDB()` API which can be used for different use cases.
2022-10-20 14:37:04 +00:00
func (m *Manager) CreateHistoryArchiveTorrentFromMessages(communityID types.HexBytes, messages []*types.Message, topics []types.TopicType, startDate time.Time, endDate time.Time, partition time.Duration, encrypt bool) ([]string, error) {
return m.CreateHistoryArchiveTorrent(communityID, messages, topics, startDate, endDate, partition, encrypt)
}
func (m *Manager) CreateHistoryArchiveTorrentFromDB(communityID types.HexBytes, topics []types.TopicType, startDate time.Time, endDate time.Time, partition time.Duration, encrypt bool) ([]string, error) {
return m.CreateHistoryArchiveTorrent(communityID, make([]*types.Message, 0), topics, startDate, endDate, partition, encrypt)
}
func (m *Manager) CreateHistoryArchiveTorrent(communityID types.HexBytes, msgs []*types.Message, topics []types.TopicType, startDate time.Time, endDate time.Time, partition time.Duration, encrypt bool) ([]string, error) {
loadFromDB := len(msgs) == 0
from := startDate
to := from.Add(partition)
if to.After(endDate) {
to = endDate
}
archiveDir := m.torrentConfig.DataDir + "/" + communityID.String()
torrentDir := m.torrentConfig.TorrentDir
indexPath := archiveDir + "/index"
dataPath := archiveDir + "/data"
wakuMessageArchiveIndexProto := &protobuf.WakuMessageArchiveIndex{}
wakuMessageArchiveIndex := make(map[string]*protobuf.WakuMessageArchiveIndexMetadata)
archiveIDs := make([]string, 0)
if _, err := os.Stat(archiveDir); os.IsNotExist(err) {
err := os.MkdirAll(archiveDir, 0700)
if err != nil {
return archiveIDs, err
}
}
if _, err := os.Stat(torrentDir); os.IsNotExist(err) {
err := os.MkdirAll(torrentDir, 0700)
if err != nil {
return archiveIDs, err
}
}
_, err := os.Stat(indexPath)
if err == nil {
wakuMessageArchiveIndexProto, err = m.LoadHistoryArchiveIndexFromFile(m.identity, communityID)
if err != nil {
return archiveIDs, err
}
}
var offset uint64 = 0
for hash, metadata := range wakuMessageArchiveIndexProto.Archives {
offset = offset + metadata.Size
wakuMessageArchiveIndex[hash] = metadata
}
var encodedArchives []*EncodedArchiveData
topicsAsByteArrays := topicsAsByteArrays(topics)
m.publish(&Subscription{CreatingHistoryArchivesSignal: &signal.CreatingHistoryArchivesSignal{
CommunityID: communityID.String(),
}})
m.LogStdout("creating archives",
zap.Any("startDate", startDate),
zap.Any("endDate", endDate),
zap.Duration("partition", partition),
)
for {
if from.Equal(endDate) || from.After(endDate) {
break
}
m.LogStdout("creating message archive",
zap.Any("from", from),
zap.Any("to", to),
)
feat(CommunitiesManager): introduce `CreateHistoryArchiveTorrentFromMessages` API Prior to this commit we had a `CreateHistoryArchiveTorrent()` API which takes a `startDate`, an `endDate` and a `partition` to create a bunch of message archives, given a certain time range. The function expects the messages to live in the database, which means, all messages that need to be archived have to be saved there at some point. This turns out to be an issue when importing communities from third party services, where, sometimes, there are several thousands of messages including attachment payloads, that have to be save to the database first. There are only two options to get the messages into the database: 1. Make one write operation with all messages - this slow, takes a long time and blocks the database until done 2. Create message chunks and perform multiple write operations - this is also slow, takes long but makes the database a bit more responsive as it's many smaller operations instead of one big one Option 2) turned out to not be super feasible either as sometimes, inserting even a single such message can take up to 10 seconds (depending on payload) Which brings me to the third option. **A third option** is to not store those imported messages as waku message into the database, just to later query them again to create the archives, but instead create the archives right away from all the messages that have been loaded into memory. This is significantly faster and doesn't block the database. To make this possible, this commit introduces a `CreateHistoryArchiveTorrentFromMessages()` API, and a `CreateHistoryArchiveTorrentFromDB()` API which can be used for different use cases.
2022-10-20 14:37:04 +00:00
var messages []types.Message
if loadFromDB {
messages, err = m.persistence.GetWakuMessagesByFilterTopic(topics, uint64(from.Unix()), uint64(to.Unix()))
if err != nil {
return archiveIDs, err
}
} else {
for _, msg := range msgs {
if int64(msg.Timestamp) >= from.Unix() && int64(msg.Timestamp) < to.Unix() {
messages = append(messages, *msg)
}
}
}
if len(messages) == 0 {
// No need to create an archive with zero messages
m.LogStdout("no messages in this partition")
from = to
to = to.Add(partition)
if to.After(endDate) {
to = endDate
}
continue
}
// Not only do we partition messages, we also chunk them
// roughly by size, such that each chunk will not exceed a given
// size and archive data doesn't get too big
messageChunks := make([][]types.Message, 0)
currentChunkSize := 0
currentChunk := make([]types.Message, 0)
for _, msg := range messages {
msgSize := len(msg.Payload) + len(msg.Sig)
if msgSize > maxArchiveSizeInBytes {
// we drop messages this big
continue
}
if currentChunkSize+msgSize > maxArchiveSizeInBytes {
messageChunks = append(messageChunks, currentChunk)
currentChunk = make([]types.Message, 0)
currentChunkSize = 0
}
currentChunk = append(currentChunk, msg)
currentChunkSize = currentChunkSize + msgSize
}
messageChunks = append(messageChunks, currentChunk)
for _, messages := range messageChunks {
wakuMessageArchive := m.createWakuMessageArchive(from, to, messages, topicsAsByteArrays)
encodedArchive, err := proto.Marshal(wakuMessageArchive)
if err != nil {
return archiveIDs, err
}
if encrypt {
messageSpec, err := m.encryptor.BuildHashRatchetMessage(communityID, encodedArchive)
if err != nil {
return archiveIDs, err
}
encodedArchive, err = proto.Marshal(messageSpec.Message)
if err != nil {
return archiveIDs, err
}
}
rawSize := len(encodedArchive)
padding := 0
size := 0
if rawSize > pieceLength {
size = rawSize + pieceLength - (rawSize % pieceLength)
padding = size - rawSize
} else {
padding = pieceLength - rawSize
size = rawSize + padding
}
wakuMessageArchiveIndexMetadata := &protobuf.WakuMessageArchiveIndexMetadata{
Metadata: wakuMessageArchive.Metadata,
Offset: offset,
Size: uint64(size),
Padding: uint64(padding),
}
wakuMessageArchiveIndexMetadataBytes, err := proto.Marshal(wakuMessageArchiveIndexMetadata)
if err != nil {
return archiveIDs, err
}
archiveID := crypto.Keccak256Hash(wakuMessageArchiveIndexMetadataBytes).String()
archiveIDs = append(archiveIDs, archiveID)
wakuMessageArchiveIndex[archiveID] = wakuMessageArchiveIndexMetadata
encodedArchives = append(encodedArchives, &EncodedArchiveData{bytes: encodedArchive, padding: padding})
offset = offset + uint64(rawSize) + uint64(padding)
}
from = to
to = to.Add(partition)
if to.After(endDate) {
to = endDate
}
}
if len(encodedArchives) > 0 {
dataBytes := make([]byte, 0)
for _, encodedArchiveData := range encodedArchives {
dataBytes = append(dataBytes, encodedArchiveData.bytes...)
dataBytes = append(dataBytes, make([]byte, encodedArchiveData.padding)...)
}
wakuMessageArchiveIndexProto.Archives = wakuMessageArchiveIndex
indexBytes, err := proto.Marshal(wakuMessageArchiveIndexProto)
if err != nil {
return archiveIDs, err
}
if encrypt {
messageSpec, err := m.encryptor.BuildHashRatchetMessage(communityID, indexBytes)
if err != nil {
return archiveIDs, err
}
indexBytes, err = proto.Marshal(messageSpec.Message)
if err != nil {
return archiveIDs, err
}
}
err = os.WriteFile(indexPath, indexBytes, 0644) // nolint: gosec
if err != nil {
return archiveIDs, err
}
2022-09-29 11:50:23 +00:00
file, err := os.OpenFile(dataPath, os.O_WRONLY|os.O_CREATE|os.O_APPEND, 0644)
if err != nil {
return archiveIDs, err
}
defer file.Close()
_, err = file.Write(dataBytes)
if err != nil {
return archiveIDs, err
}
metaInfo := metainfo.MetaInfo{
AnnounceList: defaultAnnounceList,
}
metaInfo.SetDefaults()
metaInfo.CreatedBy = common.PubkeyToHex(&m.identity.PublicKey)
info := metainfo.Info{
PieceLength: int64(pieceLength),
}
err = info.BuildFromFilePath(archiveDir)
if err != nil {
return archiveIDs, err
}
metaInfo.InfoBytes, err = bencode.Marshal(info)
if err != nil {
return archiveIDs, err
}
metaInfoBytes, err := bencode.Marshal(metaInfo)
if err != nil {
return archiveIDs, err
}
err = os.WriteFile(m.torrentFile(communityID.String()), metaInfoBytes, 0644) // nolint: gosec
if err != nil {
return archiveIDs, err
}
m.LogStdout("torrent created", zap.Any("from", startDate.Unix()), zap.Any("to", endDate.Unix()))
m.publish(&Subscription{
HistoryArchivesCreatedSignal: &signal.HistoryArchivesCreatedSignal{
CommunityID: communityID.String(),
From: int(startDate.Unix()),
To: int(endDate.Unix()),
},
})
} else {
m.LogStdout("no archives created")
m.publish(&Subscription{
NoHistoryArchivesCreatedSignal: &signal.NoHistoryArchivesCreatedSignal{
CommunityID: communityID.String(),
From: int(startDate.Unix()),
To: int(endDate.Unix()),
},
})
}
lastMessageArchiveEndDate, err := m.persistence.GetLastMessageArchiveEndDate(communityID)
if err != nil {
return archiveIDs, err
}
if lastMessageArchiveEndDate > 0 {
err = m.persistence.UpdateLastMessageArchiveEndDate(communityID, uint64(from.Unix()))
} else {
err = m.persistence.SaveLastMessageArchiveEndDate(communityID, uint64(from.Unix()))
}
if err != nil {
return archiveIDs, err
}
return archiveIDs, nil
}
func (m *Manager) SeedHistoryArchiveTorrent(communityID types.HexBytes) error {
m.UnseedHistoryArchiveTorrent(communityID)
id := communityID.String()
torrentFile := m.torrentFile(id)
metaInfo, err := metainfo.LoadFromFile(torrentFile)
if err != nil {
return err
}
info, err := metaInfo.UnmarshalInfo()
if err != nil {
return err
}
hash := metaInfo.HashInfoBytes()
m.torrentTasks[id] = hash
if err != nil {
return err
}
torrent, err := m.torrentClient.AddTorrent(metaInfo)
if err != nil {
return err
}
torrent.DownloadAll()
m.publish(&Subscription{
HistoryArchivesSeedingSignal: &signal.HistoryArchivesSeedingSignal{
CommunityID: communityID.String(),
},
})
magnetLink := metaInfo.Magnet(nil, &info).String()
m.LogStdout("seeding torrent", zap.String("id", id), zap.String("magnetLink", magnetLink))
return nil
}
func (m *Manager) UnseedHistoryArchiveTorrent(communityID types.HexBytes) {
id := communityID.String()
2022-09-29 11:50:23 +00:00
hash, exists := m.torrentTasks[id]
if exists {
torrent, ok := m.torrentClient.Torrent(hash)
if ok {
m.logger.Debug("Unseeding and dropping torrent for community: ", zap.Any("id", id))
torrent.Drop()
delete(m.torrentTasks, id)
m.publish(&Subscription{
HistoryArchivesUnseededSignal: &signal.HistoryArchivesUnseededSignal{
CommunityID: id,
},
})
}
}
}
func (m *Manager) IsSeedingHistoryArchiveTorrent(communityID types.HexBytes) bool {
id := communityID.String()
hash := m.torrentTasks[id]
torrent, ok := m.torrentClient.Torrent(hash)
return ok && torrent.Seeding()
}
func (m *Manager) GetHistoryArchiveDownloadTask(communityID string) *HistoryArchiveDownloadTask {
return m.historyArchiveDownloadTasks[communityID]
}
func (m *Manager) DeleteHistoryArchiveDownloadTask(communityID string) {
delete(m.historyArchiveDownloadTasks, communityID)
}
func (m *Manager) AddHistoryArchiveDownloadTask(communityID string, task *HistoryArchiveDownloadTask) {
m.historyArchiveDownloadTasks[communityID] = task
}
type HistoryArchiveDownloadTaskInfo struct {
TotalDownloadedArchivesCount int
TotalArchivesCount int
Cancelled bool
}
func (m *Manager) DownloadHistoryArchivesByMagnetlink(communityID types.HexBytes, magnetlink string, cancelTask chan struct{}) (*HistoryArchiveDownloadTaskInfo, error) {
id := communityID.String()
ml, err := metainfo.ParseMagnetUri(magnetlink)
if err != nil {
return nil, err
}
m.logger.Debug("adding torrent via magnetlink for community", zap.String("id", id), zap.String("magnetlink", magnetlink))
torrent, err := m.torrentClient.AddMagnet(magnetlink)
if err != nil {
return nil, err
}
downloadTaskInfo := &HistoryArchiveDownloadTaskInfo{
TotalDownloadedArchivesCount: 0,
TotalArchivesCount: 0,
Cancelled: false,
}
m.torrentTasks[id] = ml.InfoHash
timeout := time.After(20 * time.Second)
m.LogStdout("fetching torrent info", zap.String("magnetlink", magnetlink))
select {
case <-timeout:
return nil, ErrTorrentTimedout
case <-cancelTask:
m.LogStdout("cancelled fetching torrent info")
downloadTaskInfo.Cancelled = true
return downloadTaskInfo, nil
case <-torrent.GotInfo():
files := torrent.Files()
i, ok := findIndexFile(files)
if !ok {
// We're dealing with a malformed torrent, so don't do anything
return nil, errors.New("malformed torrent data")
}
indexFile := files[i]
indexFile.Download()
m.LogStdout("downloading history archive index")
ticker := time.NewTicker(100 * time.Millisecond)
defer ticker.Stop()
for {
2022-09-29 11:50:23 +00:00
select {
case <-cancelTask:
m.LogStdout("cancelled downloading archive index")
downloadTaskInfo.Cancelled = true
return downloadTaskInfo, nil
2022-09-29 11:50:23 +00:00
case <-ticker.C:
if indexFile.BytesCompleted() == indexFile.Length() {
2022-09-29 11:50:23 +00:00
index, err := m.LoadHistoryArchiveIndexFromFile(m.identity, communityID)
if err != nil {
return nil, err
}
existingArchiveIDs, err := m.persistence.GetDownloadedMessageArchiveIDs(communityID)
if err != nil {
return nil, err
}
if len(existingArchiveIDs) == len(index.Archives) {
m.LogStdout("download cancelled, no new archives")
return downloadTaskInfo, nil
}
downloadTaskInfo.TotalDownloadedArchivesCount = len(existingArchiveIDs)
downloadTaskInfo.TotalArchivesCount = len(index.Archives)
archiveHashes := make(archiveMDSlice, 0, downloadTaskInfo.TotalArchivesCount)
2022-09-29 11:50:23 +00:00
for hash, metadata := range index.Archives {
archiveHashes = append(archiveHashes, &archiveMetadata{hash: hash, from: metadata.Metadata.From})
}
2022-09-29 11:50:23 +00:00
sort.Sort(sort.Reverse(archiveHashes))
m.publish(&Subscription{
DownloadingHistoryArchivesStartedSignal: &signal.DownloadingHistoryArchivesStartedSignal{
CommunityID: communityID.String(),
},
})
2022-09-29 11:50:23 +00:00
for _, hd := range archiveHashes {
2022-09-29 11:50:23 +00:00
hash := hd.hash
hasArchive := false
for _, existingHash := range existingArchiveIDs {
if existingHash == hash {
hasArchive = true
break
}
2022-09-29 11:50:23 +00:00
}
if hasArchive {
continue
}
metadata := index.Archives[hash]
2022-09-29 11:50:23 +00:00
startIndex := int(metadata.Offset) / pieceLength
endIndex := startIndex + int(metadata.Size)/pieceLength
downloadMsg := fmt.Sprintf("downloading data for message archive (%d/%d)", downloadTaskInfo.TotalDownloadedArchivesCount+1, downloadTaskInfo.TotalArchivesCount)
m.LogStdout(downloadMsg, zap.String("hash", hash))
2022-09-29 11:50:23 +00:00
m.LogStdout("pieces (start, end)", zap.Any("startIndex", startIndex), zap.Any("endIndex", endIndex-1))
torrent.DownloadPieces(startIndex, endIndex)
piecesCompleted := make(map[int]bool)
for i = startIndex; i < endIndex; i++ {
piecesCompleted[i] = false
}
2022-09-29 11:50:23 +00:00
psc := torrent.SubscribePieceStateChanges()
downloadTicker := time.NewTicker(1 * time.Second)
defer downloadTicker.Stop()
downloadLoop:
for {
2022-09-29 11:50:23 +00:00
select {
case <-downloadTicker.C:
done := true
for i = startIndex; i < endIndex; i++ {
piecesCompleted[i] = torrent.PieceState(i).Complete
if !piecesCompleted[i] {
done = false
}
}
if done {
psc.Close()
break downloadLoop
}
case <-cancelTask:
m.LogStdout("downloading archive data interrupted")
downloadTaskInfo.Cancelled = true
return downloadTaskInfo, nil
}
}
downloadTaskInfo.TotalDownloadedArchivesCount++
2022-09-29 11:50:23 +00:00
err = m.persistence.SaveMessageArchiveID(communityID, hash)
if err != nil {
m.LogStdout("couldn't save message archive ID", zap.Error(err))
2022-09-29 11:50:23 +00:00
continue
}
2022-09-29 11:50:23 +00:00
m.publish(&Subscription{
HistoryArchiveDownloadedSignal: &signal.HistoryArchiveDownloadedSignal{
CommunityID: communityID.String(),
From: int(metadata.Metadata.From),
To: int(metadata.Metadata.To),
},
})
}
m.publish(&Subscription{
2022-09-29 11:50:23 +00:00
HistoryArchivesSeedingSignal: &signal.HistoryArchivesSeedingSignal{
CommunityID: communityID.String(),
},
})
m.LogStdout("finished downloading archives")
return downloadTaskInfo, nil
}
}
}
}
}
func (m *Manager) GetMessageArchiveIDsToImport(communityID types.HexBytes) ([]string, error) {
return m.persistence.GetMessageArchiveIDsToImport(communityID)
}
func (m *Manager) ExtractMessagesFromHistoryArchive(communityID types.HexBytes, archiveID string) ([]*protobuf.WakuMessage, error) {
id := communityID.String()
index, err := m.LoadHistoryArchiveIndexFromFile(m.identity, communityID)
if err != nil {
return nil, err
}
dataFile, err := os.Open(m.archiveDataFile(id))
if err != nil {
return nil, err
}
defer dataFile.Close()
m.LogStdout("extracting messages from history archive", zap.String("archive id", archiveID))
metadata := index.Archives[archiveID]
_, err = dataFile.Seek(int64(metadata.Offset), 0)
if err != nil {
m.LogStdout("failed to seek archive data file", zap.Error(err))
return nil, err
}
data := make([]byte, metadata.Size-metadata.Padding)
m.LogStdout("loading history archive data into memory", zap.Float64("data_size_MB", float64(metadata.Size-metadata.Padding)/1024.0/1024.0))
_, err = dataFile.Read(data)
if err != nil {
m.LogStdout("failed failed to read archive data", zap.Error(err))
return nil, err
}
archive := &protobuf.WakuMessageArchive{}
err = proto.Unmarshal(data, archive)
if err != nil {
// The archive data might eb encrypted so we try to decrypt instead first
var protocolMessage encryption.ProtocolMessage
err := proto.Unmarshal(data, &protocolMessage)
if err != nil {
m.LogStdout("failed to unmarshal protocol message", zap.Error(err))
return nil, err
}
pk, err := crypto.DecompressPubkey(communityID)
if err != nil {
m.logger.Debug("failed to decompress community pubkey", zap.Error(err))
return nil, err
}
decryptedBytes, err := m.encryptor.HandleMessage(m.identity, pk, &protocolMessage, make([]byte, 0))
if err != nil {
m.LogStdout("failed to decrypt message archive", zap.Error(err))
return nil, err
}
err = proto.Unmarshal(decryptedBytes.DecryptedMessage, archive)
if err != nil {
m.LogStdout("failed to unmarshal message archive data", zap.Error(err))
return nil, err
}
}
return archive.Messages, nil
}
func (m *Manager) SetMessageArchiveIDImported(communityID types.HexBytes, hash string, imported bool) error {
return m.persistence.SetMessageArchiveIDImported(communityID, hash, imported)
}
func (m *Manager) GetHistoryArchiveMagnetlink(communityID types.HexBytes) (string, error) {
id := communityID.String()
torrentFile := m.torrentFile(id)
metaInfo, err := metainfo.LoadFromFile(torrentFile)
if err != nil {
return "", err
}
info, err := metaInfo.UnmarshalInfo()
if err != nil {
return "", err
}
return metaInfo.Magnet(nil, &info).String(), nil
}
func (m *Manager) createWakuMessageArchive(from time.Time, to time.Time, messages []types.Message, topics [][]byte) *protobuf.WakuMessageArchive {
var wakuMessages []*protobuf.WakuMessage
for _, msg := range messages {
topic := types.TopicTypeToByteArray(msg.Topic)
wakuMessage := &protobuf.WakuMessage{
Sig: msg.Sig,
Timestamp: uint64(msg.Timestamp),
Topic: topic,
Payload: msg.Payload,
Padding: msg.Padding,
Hash: msg.Hash,
ThirdPartyId: msg.ThirdPartyID,
}
wakuMessages = append(wakuMessages, wakuMessage)
}
metadata := protobuf.WakuMessageArchiveMetadata{
From: uint64(from.Unix()),
To: uint64(to.Unix()),
ContentTopic: topics,
}
wakuMessageArchive := &protobuf.WakuMessageArchive{
Metadata: &metadata,
Messages: wakuMessages,
}
return wakuMessageArchive
}
func (m *Manager) LoadHistoryArchiveIndexFromFile(myKey *ecdsa.PrivateKey, communityID types.HexBytes) (*protobuf.WakuMessageArchiveIndex, error) {
wakuMessageArchiveIndexProto := &protobuf.WakuMessageArchiveIndex{}
indexPath := m.archiveIndexFile(communityID.String())
indexData, err := os.ReadFile(indexPath)
if err != nil {
return nil, err
}
err = proto.Unmarshal(indexData, wakuMessageArchiveIndexProto)
if err != nil {
return nil, err
}
if len(wakuMessageArchiveIndexProto.Archives) == 0 && len(indexData) > 0 {
// This means we're dealing with an encrypted index file, so we have to decrypt it first
var protocolMessage encryption.ProtocolMessage
err := proto.Unmarshal(indexData, &protocolMessage)
if err != nil {
return nil, err
}
pk, err := crypto.DecompressPubkey(communityID)
if err != nil {
return nil, err
}
decryptedBytes, err := m.encryptor.HandleMessage(myKey, pk, &protocolMessage, make([]byte, 0))
if err != nil {
return nil, err
}
err = proto.Unmarshal(decryptedBytes.DecryptedMessage, wakuMessageArchiveIndexProto)
if err != nil {
return nil, err
}
}
return wakuMessageArchiveIndexProto, nil
}
func (m *Manager) TorrentFileExists(communityID string) bool {
_, err := os.Stat(m.torrentFile(communityID))
return err == nil
}
func (m *Manager) torrentFile(communityID string) string {
return m.torrentConfig.TorrentDir + "/" + communityID + ".torrent"
}
func (m *Manager) archiveIndexFile(communityID string) string {
return m.torrentConfig.DataDir + "/" + communityID + "/index"
}
func (m *Manager) archiveDataFile(communityID string) string {
return m.torrentConfig.DataDir + "/" + communityID + "/data"
}
func topicsAsByteArrays(topics []types.TopicType) [][]byte {
var topicsAsByteArrays [][]byte
for _, t := range topics {
topic := types.TopicTypeToByteArray(t)
topicsAsByteArrays = append(topicsAsByteArrays, topic)
}
return topicsAsByteArrays
}
func findIndexFile(files []*torrent.File) (index int, ok bool) {
for i, f := range files {
if f.DisplayPath() == "index" {
return i, true
}
}
return 0, false
}
func (m *Manager) GetCommunityToken(communityID string, chainID int, address string) (*community_token.CommunityToken, error) {
return m.persistence.GetCommunityToken(communityID, chainID, address)
}
func (m *Manager) GetCommunityTokens(communityID string) ([]*community_token.CommunityToken, error) {
return m.persistence.GetCommunityTokens(communityID)
}
func (m *Manager) GetAllCommunityTokens() ([]*community_token.CommunityToken, error) {
return m.persistence.GetAllCommunityTokens()
}
func (m *Manager) ImageToBase64(uri string) string {
if uri == "" {
return ""
}
file, err := os.Open(uri)
if err != nil {
m.logger.Error(err.Error())
return ""
}
defer file.Close()
payload, err := ioutil.ReadAll(file)
if err != nil {
m.logger.Error(err.Error())
return ""
}
base64img, err := images.GetPayloadDataURI(payload)
if err != nil {
m.logger.Error(err.Error())
return ""
}
return base64img
}
func (m *Manager) SaveCommunityToken(token *community_token.CommunityToken, croppedImage *images.CroppedImage) (*community_token.CommunityToken, error) {
community, err := m.GetByIDString(token.CommunityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if croppedImage != nil && croppedImage.ImagePath != "" {
bytes, err := images.OpenAndAdjustImage(*croppedImage, true)
if err != nil {
return nil, err
}
base64img, err := images.GetPayloadDataURI(bytes)
if err != nil {
return nil, err
}
token.Base64Image = base64img
} else if !images.IsPayloadDataURI(token.Base64Image) {
// if image is already base64 do not convert (owner and master tokens have already base64 image)
token.Base64Image = m.ImageToBase64(token.Base64Image)
}
return token, m.persistence.AddCommunityToken(token)
}
func (m *Manager) AddCommunityToken(token *community_token.CommunityToken, clock uint64) (*Community, error) {
if token == nil {
return nil, errors.New("Token is absent in database")
}
community, err := m.GetByIDString(token.CommunityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.MemberCanManageToken(&m.identity.PublicKey, token) {
return nil, ErrInvalidManageTokensPermission
}
tokenMetadata := &protobuf.CommunityTokenMetadata{
ContractAddresses: map[uint64]string{uint64(token.ChainID): token.Address},
Description: token.Description,
Image: token.Base64Image,
Symbol: token.Symbol,
TokenType: token.TokenType,
Name: token.Name,
Decimals: uint32(token.Decimals),
}
_, err = community.AddCommunityTokensMetadata(tokenMetadata)
if err != nil {
return nil, err
}
if community.IsControlNode() && (token.PrivilegesLevel == community_token.MasterLevel || token.PrivilegesLevel == community_token.OwnerLevel) {
permissionType := protobuf.CommunityTokenPermission_BECOME_TOKEN_OWNER
if token.PrivilegesLevel == community_token.MasterLevel {
permissionType = protobuf.CommunityTokenPermission_BECOME_TOKEN_MASTER
}
contractAddresses := make(map[uint64]string)
contractAddresses[uint64(token.ChainID)] = token.Address
tokenCriteria := &protobuf.TokenCriteria{
ContractAddresses: contractAddresses,
Type: protobuf.CommunityTokenType_ERC721,
Symbol: token.Symbol,
Name: token.Name,
Amount: "1",
Decimals: uint64(token.Decimals),
}
request := &requests.CreateCommunityTokenPermission{
CommunityID: community.ID(),
Type: permissionType,
TokenCriteria: []*protobuf.TokenCriteria{tokenCriteria},
IsPrivate: true,
ChatIds: []string{},
}
community, _, err = m.createCommunityTokenPermission(request, community)
if err != nil {
return nil, err
}
if token.PrivilegesLevel == community_token.OwnerLevel {
_, err = m.promoteSelfToControlNode(community, clock)
if err != nil {
return nil, err
}
}
}
return community, m.saveAndPublish(community)
}
func (m *Manager) UpdateCommunityTokenState(chainID int, contractAddress string, deployState community_token.DeployState) error {
return m.persistence.UpdateCommunityTokenState(chainID, contractAddress, deployState)
}
func (m *Manager) UpdateCommunityTokenAddress(chainID int, oldContractAddress string, newContractAddress string) error {
return m.persistence.UpdateCommunityTokenAddress(chainID, oldContractAddress, newContractAddress)
}
func (m *Manager) UpdateCommunityTokenSupply(chainID int, contractAddress string, supply *bigint.BigInt) error {
return m.persistence.UpdateCommunityTokenSupply(chainID, contractAddress, supply)
}
func (m *Manager) RemoveCommunityToken(chainID int, contractAddress string) error {
return m.persistence.RemoveCommunityToken(chainID, contractAddress)
}
func (m *Manager) SetCommunityActiveMembersCount(communityID string, activeMembersCount uint64) error {
community, err := m.GetByIDString(communityID)
if err != nil {
return err
}
if community == nil {
return ErrOrgNotFound
}
updated, err := community.SetActiveMembersCount(activeMembersCount)
if err != nil {
return err
}
if updated {
if err = m.persistence.SaveCommunity(community); err != nil {
return err
}
m.publish(&Subscription{Community: community})
}
return nil
}
// UpdateCommunity takes a Community persists it and republishes it.
// The clock is incremented meaning even a no change update will be republished by the admin, and parsed by the member.
func (m *Manager) UpdateCommunity(c *Community) error {
c.increaseClock()
err := m.persistence.SaveCommunity(c)
if err != nil {
return err
}
m.publish(&Subscription{Community: c})
return nil
}
func combineAddressesAndChainIDs(addresses []gethcommon.Address, chainIDs []uint64) []*AccountChainIDsCombination {
combinations := make([]*AccountChainIDsCombination, 0)
for _, address := range addresses {
combinations = append(combinations, &AccountChainIDsCombination{
Address: address,
ChainIDs: chainIDs,
})
}
return combinations
}
func revealedAccountsToAccountsAndChainIDsCombination(revealedAccounts []*protobuf.RevealedAccount) []*AccountChainIDsCombination {
accountsAndChainIDs := make([]*AccountChainIDsCombination, 0)
for _, revealedAccount := range revealedAccounts {
accountsAndChainIDs = append(accountsAndChainIDs, &AccountChainIDsCombination{
Address: gethcommon.HexToAddress(revealedAccount.Address),
ChainIDs: revealedAccount.ChainIds,
})
}
return accountsAndChainIDs
}
func (m *Manager) accountsHasPrivilegedPermission(privilegedPermissions []*CommunityTokenPermission, accounts []*AccountChainIDsCombination) bool {
if len(privilegedPermissions) > 0 {
permissionResponse, err := m.PermissionChecker.CheckPermissions(privilegedPermissions, accounts, true)
if err != nil {
m.logger.Warn("check privileged permission failed: %v", zap.Error(err))
return false
}
return permissionResponse.Satisfied
}
return false
}
func (m *Manager) saveAndPublish(community *Community) error {
err := m.persistence.SaveCommunity(community)
if err != nil {
return err
}
if community.IsControlNode() {
m.publish(&Subscription{Community: community})
return nil
} else if community.HasPermissionToSendCommunityEvents() {
err := m.signEvents(community)
if err != nil {
return err
}
err = m.persistence.SaveCommunityEvents(community)
if err != nil {
return err
}
m.publish(&Subscription{CommunityEventsMessage: community.ToCommunityEventsMessage()})
return nil
}
return nil
}
func (m *Manager) GetRevealedAddresses(communityID types.HexBytes, memberPk string) ([]*protobuf.RevealedAccount, error) {
requestID := CalculateRequestID(memberPk, communityID)
return m.persistence.GetRequestToJoinRevealedAddresses(requestID)
}
func (m *Manager) ReevaluatePrivilegedMember(community *Community, tokenPermissions []*CommunityTokenPermission,
accountsAndChainIDs []*AccountChainIDsCombination, memberPubKey *ecdsa.PublicKey,
privilegedRole protobuf.CommunityMember_Roles, alreadyHasPrivilegedRole bool) (bool, error) {
hasPrivilegedRolePermissions := len(tokenPermissions) > 0
removeCurrentRole := false
if hasPrivilegedRolePermissions {
permissionResponse, err := m.PermissionChecker.CheckPermissions(tokenPermissions, accountsAndChainIDs, true)
if err != nil {
return alreadyHasPrivilegedRole, err
} else if permissionResponse.Satisfied && !alreadyHasPrivilegedRole {
_, err = community.AddRoleToMember(memberPubKey, privilegedRole)
if err != nil {
return alreadyHasPrivilegedRole, err
}
alreadyHasPrivilegedRole = true
} else if !permissionResponse.Satisfied && alreadyHasPrivilegedRole {
removeCurrentRole = true
alreadyHasPrivilegedRole = false
}
}
// Remove privileged role if user does not pass role permissions check or
// Community does not have permissions but user has a role
if removeCurrentRole || (!hasPrivilegedRolePermissions && alreadyHasPrivilegedRole) {
_, err := community.RemoveRoleFromMember(memberPubKey, privilegedRole)
if err != nil {
return alreadyHasPrivilegedRole, err
}
alreadyHasPrivilegedRole = false
}
if alreadyHasPrivilegedRole {
// Make sure privileged user is added to every channel
for channelID := range community.Chats() {
if !community.IsMemberInChat(memberPubKey, channelID) {
_, err := community.AddMemberToChat(channelID, memberPubKey, []protobuf.CommunityMember_Roles{privilegedRole})
if err != nil {
return alreadyHasPrivilegedRole, err
}
}
}
}
return alreadyHasPrivilegedRole, nil
}
func (m *Manager) HandleCommunityTokensMetadataByPrivilegedMembers(community *Community) error {
if community.HasPermissionToSendCommunityEvents() || community.IsControlNode() {
if err := m.HandleCommunityTokensMetadata(community); err != nil {
return err
}
}
return nil
}
func (m *Manager) HandleCommunityTokensMetadata(community *Community) error {
communityID := community.IDString()
communityTokens := community.CommunityTokensMetadata()
if len(communityTokens) == 0 {
return nil
}
for _, tokenMetadata := range communityTokens {
for chainID, address := range tokenMetadata.ContractAddresses {
exists, err := m.persistence.HasCommunityToken(communityID, address, int(chainID))
if err != nil {
return err
}
if !exists {
// Fetch community token to make sure it's stored in the DB, discard result
communityToken, err := m.FetchCommunityToken(community, tokenMetadata, chainID, address)
if err != nil {
return err
}
err = m.persistence.AddCommunityToken(communityToken)
if err != nil {
return err
}
}
}
}
return nil
}
func (m *Manager) FetchCommunityToken(community *Community, tokenMetadata *protobuf.CommunityTokenMetadata, chainID uint64, contractAddress string) (*community_token.CommunityToken, error) {
communityID := community.IDString()
communityToken := &community_token.CommunityToken{
CommunityID: communityID,
Address: contractAddress,
TokenType: tokenMetadata.TokenType,
Name: tokenMetadata.Name,
Symbol: tokenMetadata.Symbol,
Description: tokenMetadata.Description,
Transferable: true,
RemoteSelfDestruct: false,
ChainID: int(chainID),
DeployState: community_token.Deployed,
Base64Image: tokenMetadata.Image,
Decimals: int(tokenMetadata.Decimals),
}
switch tokenMetadata.TokenType {
case protobuf.CommunityTokenType_ERC721:
contractData, err := m.communityTokensService.GetCollectibleContractData(chainID, contractAddress)
if err != nil {
return nil, err
}
communityToken.Supply = contractData.TotalSupply
communityToken.Transferable = contractData.Transferable
communityToken.RemoteSelfDestruct = contractData.RemoteBurnable
communityToken.InfiniteSupply = contractData.InfiniteSupply
case protobuf.CommunityTokenType_ERC20:
contractData, err := m.communityTokensService.GetAssetContractData(chainID, contractAddress)
if err != nil {
return nil, err
}
communityToken.Supply = contractData.TotalSupply
communityToken.InfiniteSupply = contractData.InfiniteSupply
}
communityToken.PrivilegesLevel = getPrivilegesLevel(chainID, contractAddress, community.TokenPermissions())
return communityToken, nil
}
func getPrivilegesLevel(chainID uint64, tokenAddress string, tokenPermissions map[string]*CommunityTokenPermission) community_token.PrivilegesLevel {
for _, permission := range tokenPermissions {
if permission.Type == protobuf.CommunityTokenPermission_BECOME_TOKEN_MASTER || permission.Type == protobuf.CommunityTokenPermission_BECOME_TOKEN_OWNER {
for _, tokenCriteria := range permission.TokenCriteria {
value, exist := tokenCriteria.ContractAddresses[chainID]
if exist && value == tokenAddress {
if permission.Type == protobuf.CommunityTokenPermission_BECOME_TOKEN_OWNER {
return community_token.OwnerLevel
}
return community_token.MasterLevel
}
}
}
}
return community_token.CommunityLevel
}
func (m *Manager) ValidateCommunityPrivilegedUserSyncMessage(message *protobuf.CommunityPrivilegedUserSyncMessage) error {
if message == nil {
return errors.New("invalid CommunityPrivilegedUserSyncMessage message")
}
if message.CommunityId == nil || len(message.CommunityId) == 0 {
return errors.New("invalid CommunityId in CommunityPrivilegedUserSyncMessage message")
}
switch message.Type {
case protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ACCEPT_REQUEST_TO_JOIN:
fallthrough
case protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_REJECT_REQUEST_TO_JOIN:
if message.RequestToJoin == nil || len(message.RequestToJoin) == 0 {
return errors.New("invalid request to join in CommunityPrivilegedUserSyncMessage message")
}
for _, requestToJoinProto := range message.RequestToJoin {
if len(requestToJoinProto.CommunityId) == 0 {
return errors.New("no communityId in request to join in CommunityPrivilegedUserSyncMessage message")
}
}
case protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ALL_SYNC_REQUESTS_TO_JOIN:
if message.SyncRequestsToJoin == nil || len(message.SyncRequestsToJoin) == 0 {
return errors.New("invalid sync requests to join in CommunityPrivilegedUserSyncMessage message")
}
}
return nil
}
func (m *Manager) createCommunityTokenPermission(request *requests.CreateCommunityTokenPermission, community *Community) (*Community, *CommunityChanges, error) {
if community == nil {
return nil, nil, ErrOrgNotFound
}
tokenPermission := request.ToCommunityTokenPermission()
tokenPermission.Id = uuid.New().String()
changes, err := community.UpsertTokenPermission(&tokenPermission)
if err != nil {
return nil, nil, err
}
return community, changes, nil
}
func (m *Manager) PromoteSelfToControlNode(communityID types.HexBytes, clock uint64) (*CommunityChanges, error) {
community, err := m.GetByID(communityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
ownerChanged, err := m.promoteSelfToControlNode(community, clock)
if err != nil {
return nil, err
}
if ownerChanged {
return community.RemoveAllUsersFromOrg(), m.saveAndPublish(community)
}
return community.emptyCommunityChanges(), m.saveAndPublish(community)
}
func (m *Manager) promoteSelfToControlNode(community *Community, clock uint64) (bool, error) {
ownerChanged := false
community.setPrivateKey(m.identity)
if !community.ControlNode().Equal(&m.identity.PublicKey) {
ownerChanged = true
community.setControlNode(&m.identity.PublicKey)
}
// Mark this device as the control node
syncControlNode := &protobuf.SyncCommunityControlNode{
Clock: clock,
InstallationId: m.installationID,
}
err := m.SaveSyncControlNode(community.ID(), syncControlNode)
if err != nil {
return false, err
}
community.config.ControlDevice = true
_, err = community.AddRoleToMember(&m.identity.PublicKey, protobuf.CommunityMember_ROLE_OWNER)
if err != nil {
return false, err
}
community.increaseClock()
return ownerChanged, nil
}
func (m *Manager) shareRequestsToJoinWithNewPrivilegedMembers(community *Community, newPrivilegedMembers map[protobuf.CommunityMember_Roles][]*ecdsa.PublicKey) error {
requestsToJoin, err := m.GetCommunityRequestsToJoinWithRevealedAddresses(community.ID())
if err != nil {
return err
}
var syncRequestsWithoutRevealedAccounts []*protobuf.SyncCommunityRequestsToJoin
var syncRequestsWithRevealedAccounts []*protobuf.SyncCommunityRequestsToJoin
for _, request := range requestsToJoin {
syncRequestsWithRevealedAccounts = append(syncRequestsWithRevealedAccounts, request.ToSyncProtobuf())
requestProtoWithoutAccounts := request.ToSyncProtobuf()
requestProtoWithoutAccounts.RevealedAccounts = []*protobuf.RevealedAccount{}
syncRequestsWithoutRevealedAccounts = append(syncRequestsWithoutRevealedAccounts, requestProtoWithoutAccounts)
}
syncMsgWithoutRevealedAccounts := &protobuf.CommunityPrivilegedUserSyncMessage{
Type: protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ALL_SYNC_REQUESTS_TO_JOIN,
CommunityId: community.ID(),
SyncRequestsToJoin: syncRequestsWithoutRevealedAccounts,
}
syncMsgWitRevealedAccounts := &protobuf.CommunityPrivilegedUserSyncMessage{
Type: protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ALL_SYNC_REQUESTS_TO_JOIN,
CommunityId: community.ID(),
SyncRequestsToJoin: syncRequestsWithRevealedAccounts,
}
subscriptionMsg := &CommunityPrivilegedMemberSyncMessage{
CommunityPrivateKey: community.PrivateKey(),
}
for role, members := range newPrivilegedMembers {
if len(members) == 0 {
continue
}
subscriptionMsg.Receivers = members
switch role {
case protobuf.CommunityMember_ROLE_ADMIN:
subscriptionMsg.CommunityPrivilegedUserSyncMessage = syncMsgWithoutRevealedAccounts
case protobuf.CommunityMember_ROLE_OWNER:
fallthrough
case protobuf.CommunityMember_ROLE_TOKEN_MASTER:
subscriptionMsg.CommunityPrivilegedUserSyncMessage = syncMsgWitRevealedAccounts
}
m.publish(&Subscription{CommunityPrivilegedMemberSyncMessage: subscriptionMsg})
}
return nil
}
func (m *Manager) shareAcceptedRequestToJoinWithPrivilegedMembers(community *Community, requestsToJoin *RequestToJoin) error {
pk, err := common.HexToPubkey(requestsToJoin.PublicKey)
if err != nil {
return err
}
acceptedRequestsToJoinWithoutRevealedAccounts := make(map[string]*protobuf.CommunityRequestToJoin)
acceptedRequestsToJoinWithRevealedAccounts := make(map[string]*protobuf.CommunityRequestToJoin)
acceptedRequestsToJoinWithRevealedAccounts[requestsToJoin.PublicKey] = requestsToJoin.ToCommunityRequestToJoinProtobuf()
requestsToJoin.RevealedAccounts = make([]*protobuf.RevealedAccount, 0)
acceptedRequestsToJoinWithoutRevealedAccounts[requestsToJoin.PublicKey] = requestsToJoin.ToCommunityRequestToJoinProtobuf()
msgWithRevealedAccounts := &protobuf.CommunityPrivilegedUserSyncMessage{
Type: protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ACCEPT_REQUEST_TO_JOIN,
CommunityId: community.ID(),
RequestToJoin: acceptedRequestsToJoinWithRevealedAccounts,
}
msgWithoutRevealedAccounts := &protobuf.CommunityPrivilegedUserSyncMessage{
Type: protobuf.CommunityPrivilegedUserSyncMessage_CONTROL_NODE_ACCEPT_REQUEST_TO_JOIN,
CommunityId: community.ID(),
RequestToJoin: acceptedRequestsToJoinWithoutRevealedAccounts,
}
// do not sent to ourself and to the accepted user
skipMembers := make(map[string]struct{})
skipMembers[common.PubkeyToHex(&m.identity.PublicKey)] = struct{}{}
skipMembers[common.PubkeyToHex(pk)] = struct{}{}
subscriptionMsg := &CommunityPrivilegedMemberSyncMessage{
CommunityPrivateKey: community.PrivateKey(),
}
fileredPrivilegedMembers := community.GetFilteredPrivilegedMembers(skipMembers)
for role, members := range fileredPrivilegedMembers {
if len(members) == 0 {
continue
}
subscriptionMsg.Receivers = members
switch role {
case protobuf.CommunityMember_ROLE_ADMIN:
subscriptionMsg.CommunityPrivilegedUserSyncMessage = msgWithoutRevealedAccounts
case protobuf.CommunityMember_ROLE_OWNER:
fallthrough
case protobuf.CommunityMember_ROLE_TOKEN_MASTER:
subscriptionMsg.CommunityPrivilegedUserSyncMessage = msgWithRevealedAccounts
}
m.publish(&Subscription{CommunityPrivilegedMemberSyncMessage: subscriptionMsg})
}
return nil
}
func (m *Manager) GetCommunityRequestsToJoinWithRevealedAddresses(communityID types.HexBytes) ([]*RequestToJoin, error) {
return m.persistence.GetCommunityRequestsToJoinWithRevealedAddresses(communityID)
}
func (m *Manager) SaveCommunity(community *Community) error {
return m.persistence.SaveCommunity(community)
}
func (m *Manager) CreateCommunityTokenDeploymentSignature(ctx context.Context, chainID uint64, addressFrom string, communityID string) ([]byte, error) {
community, err := m.GetByIDString(communityID)
if err != nil {
return nil, err
}
if community == nil {
return nil, ErrOrgNotFound
}
if !community.IsControlNode() {
return nil, ErrNotControlNode
}
digest, err := m.communityTokensService.DeploymentSignatureDigest(chainID, addressFrom, communityID)
if err != nil {
return nil, err
}
return crypto.Sign(digest, community.PrivateKey())
}
func (m *Manager) GetSyncControlNode(id types.HexBytes) (*protobuf.SyncCommunityControlNode, error) {
return m.persistence.GetSyncControlNode(id)
}
func (m *Manager) SaveSyncControlNode(id types.HexBytes, syncControlNode *protobuf.SyncCommunityControlNode) error {
return m.persistence.SaveSyncControlNode(id, syncControlNode.Clock, syncControlNode.InstallationId)
}
func (m *Manager) SetSyncControlNode(id types.HexBytes, syncControlNode *protobuf.SyncCommunityControlNode) error {
existingSyncControlNode, err := m.GetSyncControlNode(id)
if err != nil {
return err
}
if existingSyncControlNode == nil || existingSyncControlNode.Clock < syncControlNode.Clock {
return m.SaveSyncControlNode(id, syncControlNode)
}
return nil
}
func (m *Manager) GetCommunityRequestToJoinWithRevealedAddresses(pubKey string, communityID types.HexBytes) (*RequestToJoin, error) {
return m.persistence.GetCommunityRequestToJoinWithRevealedAddresses(pubKey, communityID)
}
func (m *Manager) SafeGetSignerPubKey(chainID uint64, communityID string) (string, error) {
ctx, cancel := context.WithTimeout(context.Background(), time.Second*3)
defer cancel()
return m.ownerVerifier.SafeGetSignerPubKey(ctx, chainID, communityID)
}