blockstore.go

// package blockstore implements a thin wrapper over a datastore, giving a
// clean interface for Getting and Putting block objects.
package blockstore

import (
	"context"
	"errors"
	"sync"
	"sync/atomic"

	blocks "github.com/ipfs/go-ipfs/blocks"
	dshelp "github.com/ipfs/go-ipfs/thirdparty/ds-help"

	ds "gx/ipfs/QmRWDav6mzWseLWeYfVd5fvUKiVe9xNH29YfMF438fG364/go-datastore"
	dsns "gx/ipfs/QmRWDav6mzWseLWeYfVd5fvUKiVe9xNH29YfMF438fG364/go-datastore/namespace"
	dsq "gx/ipfs/QmRWDav6mzWseLWeYfVd5fvUKiVe9xNH29YfMF438fG364/go-datastore/query"
	logging "gx/ipfs/QmSpJByNKFX1sCsHBEp3R73FL4NF6FnQTEGyNAXHm2GS52/go-log"
	cid "gx/ipfs/QmcTcsTvfaeEBRFo1TkFgT8sRmgi1n1LTZpecfVP8fzpGD/go-cid"
)

var log = logging.Logger("blockstore")

// BlockPrefix namespaces blockstore datastores
var BlockPrefix = ds.NewKey("blocks")

var ValueTypeMismatch = errors.New("the retrieved value is not a Block")
var ErrHashMismatch = errors.New("block in storage has different hash than requested")

var ErrNotFound = errors.New("blockstore: block not found")

// Blockstore wraps a Datastore
type Blockstore interface {
	DeleteBlock(*cid.Cid) error
	Has(*cid.Cid) (bool, error)
	Get(*cid.Cid) (blocks.Block, error)
	Put(blocks.Block) error
	PutMany([]blocks.Block) error

	AllKeysChan(ctx context.Context) (<-chan *cid.Cid, error)
}

type GCLocker interface {
	// GCLock locks the blockstore for garbage collection. No operations
	// that expect to finish with a pin should ocurr simultaneously.
	// Reading during GC is safe, and requires no lock.
	GCLock() Unlocker

	// PinLock locks the blockstore for sequences of puts expected to finish
	// with a pin (before GC). Multiple put->pin sequences can write through
	// at the same time, but no GC should not happen simulatenously.
	// Reading during Pinning is safe, and requires no lock.
	PinLock() Unlocker

	// GcRequested returns true if GCLock has been called and is waiting to
	// take the lock
	GCRequested() bool
}

type GCBlockstore interface {
	Blockstore
	GCLocker
}

func NewGCBlockstore(bs Blockstore, gcl GCLocker) GCBlockstore {
	return gcBlockstore{bs, gcl}
}

type gcBlockstore struct {
	Blockstore
	GCLocker
}

func NewBlockstore(d ds.Batching) *blockstore {
	var dsb ds.Batching
	dd := dsns.Wrap(d, BlockPrefix)
	dsb = dd
	return &blockstore{
		datastore: dsb,
	}
}

type blockstore struct {
	datastore ds.Batching

	lk      sync.RWMutex
	gcreq   int32
	gcreqlk sync.Mutex

	rehash bool
}

func (bs *blockstore) HashOnRead(enabled bool) {
	bs.rehash = enabled
}

func (bs *blockstore) Get(k *cid.Cid) (blocks.Block, error) {
	if k == nil {
		log.Error("nil cid in blockstore")
		return nil, ErrNotFound
	}

	maybeData, err := bs.datastore.Get(dshelp.CidToDsKey(k))
	if err == ds.ErrNotFound {
		return nil, ErrNotFound
	}
	if err != nil {
		return nil, err
	}
	bdata, ok := maybeData.([]byte)
	if !ok {
		return nil, ValueTypeMismatch
	}

	if bs.rehash {
		rbcid, err := k.Prefix().Sum(bdata)
		if err != nil {
			return nil, err
		}

		if !rbcid.Equals(k) {
			return nil, ErrHashMismatch
		}

		return blocks.NewBlockWithCid(bdata, rbcid)
	} else {
		return blocks.NewBlockWithCid(bdata, k)
	}
}

func (bs *blockstore) Put(block blocks.Block) error {
	k := dshelp.CidToDsKey(block.Cid())

	// Has is cheaper than Put, so see if we already have it
	exists, err := bs.datastore.Has(k)
	if err == nil && exists {
		return nil // already stored.
	}
	return bs.datastore.Put(k, block.RawData())
}

func (bs *blockstore) PutMany(blocks []blocks.Block) error {
	t, err := bs.datastore.Batch()
	if err != nil {
		return err
	}
	for _, b := range blocks {
		k := dshelp.CidToDsKey(b.Cid())
		exists, err := bs.datastore.Has(k)
		if err == nil && exists {
			continue
		}

		err = t.Put(k, b.RawData())
		if err != nil {
			return err
		}
	}
	return t.Commit()
}

func (bs *blockstore) Has(k *cid.Cid) (bool, error) {
	return bs.datastore.Has(dshelp.CidToDsKey(k))
}

func (s *blockstore) DeleteBlock(k *cid.Cid) error {
	return s.datastore.Delete(dshelp.CidToDsKey(k))
}

// AllKeysChan runs a query for keys from the blockstore.
// this is very simplistic, in the future, take dsq.Query as a param?
//
// AllKeysChan respects context
func (bs *blockstore) AllKeysChan(ctx context.Context) (<-chan *cid.Cid, error) {

	// KeysOnly, because that would be _a lot_ of data.
	q := dsq.Query{KeysOnly: true}
	// datastore/namespace does *NOT* fix up Query.Prefix
	q.Prefix = BlockPrefix.String()
	res, err := bs.datastore.Query(q)
	if err != nil {
		return nil, err
	}

	// this function is here to compartmentalize
	get := func() (*cid.Cid, bool) {
		select {
		case <-ctx.Done():
			return nil, false
		case e, more := <-res.Next():
			if !more {
				return nil, false
			}
			if e.Error != nil {
				log.Debug("blockstore.AllKeysChan got err:", e.Error)
				return nil, false
			}

			// need to convert to key.Key using key.KeyFromDsKey.
			c, err := dshelp.DsKeyToCid(ds.RawKey(e.Key))
			if err != nil {
				log.Warningf("error parsing key from DsKey: ", err)
				return nil, true
			}

			return c, true
		}
	}

	output := make(chan *cid.Cid, dsq.KeysOnlyBufSize)
	go func() {
		defer func() {
			res.Process().Close() // ensure exit (signals early exit, too)
			close(output)
		}()

		for {
			k, ok := get()
			if !ok {
				return
			}
			if k == nil {
				continue
			}

			select {
			case <-ctx.Done():
				return
			case output <- k:
			}
		}
	}()

	return output, nil
}

func NewGCLocker() *gclocker {
	return &gclocker{}
}

type gclocker struct {
	lk      sync.RWMutex
	gcreq   int32
	gcreqlk sync.Mutex
}

type Unlocker interface {
	Unlock()
}

type unlocker struct {
	unlock func()
}

func (u *unlocker) Unlock() {
	u.unlock()
	u.unlock = nil // ensure its not called twice
}

func (bs *gclocker) GCLock() Unlocker {
	atomic.AddInt32(&bs.gcreq, 1)
	bs.lk.Lock()
	atomic.AddInt32(&bs.gcreq, -1)
	return &unlocker{bs.lk.Unlock}
}

func (bs *gclocker) PinLock() Unlocker {
	bs.lk.RLock()
	return &unlocker{bs.lk.RUnlock}
}

func (bs *gclocker) GCRequested() bool {
	return atomic.LoadInt32(&bs.gcreq) > 0
}