pfs.go

package client

import (
	"bytes"
	"io"

	"github.com/pachyderm/pachyderm/src/client/pfs"
	"github.com/pachyderm/pachyderm/src/client/pkg/grpcutil"

	"github.com/gogo/protobuf/types"
)

// NewRepo creates a pfs.Repo.
func NewRepo(repoName string) *pfs.Repo {
	return &pfs.Repo{Name: repoName}
}

// NewCommit creates a pfs.Commit.
func NewCommit(repoName string, commitID string) *pfs.Commit {
	return &pfs.Commit{
		Repo: NewRepo(repoName),
		ID:   commitID,
	}
}

// NewFile creates a pfs.File.
func NewFile(repoName string, commitID string, path string) *pfs.File {
	return &pfs.File{
		Commit: NewCommit(repoName, commitID),
		Path:   path,
	}
}

// NewBlock creates a pfs.Block.
func NewBlock(hash string) *pfs.Block {
	return &pfs.Block{
		Hash: hash,
	}
}

// CommitTypes alias pfs.CommitType_*
const (
	CommitTypeNone  = pfs.CommitType_COMMIT_TYPE_NONE
	CommitTypeRead  = pfs.CommitType_COMMIT_TYPE_READ
	CommitTypeWrite = pfs.CommitType_COMMIT_TYPE_WRITE
)

// CommitStatus alias pfs.CommitStatus_*
const (
	CommitStatusNormal    = pfs.CommitStatus_NORMAL
	CommitStatusArchived  = pfs.CommitStatus_ARCHIVED
	CommitStatusCancelled = pfs.CommitStatus_CANCELLED
	CommitStatusAll       = pfs.CommitStatus_ALL
)

// CreateRepo creates a new Repo object in pfs with the given name. Repos are
// the top level data object in pfs and should be used to store data of a
// similar type. For example rather than having a single Repo for an entire
// project you might have seperate Repos for logs, metrics, database dumps etc.
func (c APIClient) CreateRepo(repoName string) error {
	_, err := c.PfsAPIClient.CreateRepo(
		c.ctx(),
		&pfs.CreateRepoRequest{
			Repo: NewRepo(repoName),
		},
	)
	return sanitizeErr(err)
}

// InspectRepo returns info about a specific Repo.
func (c APIClient) InspectRepo(repoName string) (*pfs.RepoInfo, error) {
	repoInfo, err := c.PfsAPIClient.InspectRepo(
		c.ctx(),
		&pfs.InspectRepoRequest{
			Repo: NewRepo(repoName),
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return repoInfo, nil
}

// ListRepo returns info about all Repos.
// provenance specifies a set of provenance repos, only repos which have ALL of
// the specified repos as provenance will be returned unless provenance is nil
// in which case it is ignored.
func (c APIClient) ListRepo(provenance []string) ([]*pfs.RepoInfo, error) {
	request := &pfs.ListRepoRequest{}
	for _, repoName := range provenance {
		request.Provenance = append(request.Provenance, NewRepo(repoName))
	}
	repoInfos, err := c.PfsAPIClient.ListRepo(
		c.ctx(),
		request,
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return repoInfos.RepoInfo, nil
}

// DeleteRepo deletes a repo and reclaims the storage space it was using. Note
// that as of 1.0 we do not reclaim the blocks that the Repo was referencing,
// this is because they may also be referenced by other Repos and deleting them
// would make those Repos inaccessible. This will be resolved in later
// versions.
// If "force" is set to true, the repo will be removed regardless of errors.
// This argument should be used with care.
func (c APIClient) DeleteRepo(repoName string, force bool) error {
	_, err := c.PfsAPIClient.DeleteRepo(
		c.ctx(),
		&pfs.DeleteRepoRequest{
			Repo:  NewRepo(repoName),
			Force: force,
		},
	)
	return err
}

// StartCommit begins the process of committing data to a Repo. Once started
// you can write to the Commit with PutFile and when all the data has been
// written you must finish the Commit with FinishCommit. NOTE, data is not
// persisted until FinishCommit is called.
// parentCommit specifies the parent Commit, upon creation the new Commit will
// appear identical to the parent Commit, data can safely be added to the new
// commit without affecting the contents of the parent Commit. You may pass ""
// as parentCommit in which case the new Commit will have no parent and will
// initially appear empty.
// branch is a more convenient way to build linear chains of commits. When a
// commit is started with a non empty branch the value of branch becomes an
// alias for the created Commit. This enables a more intuitive access pattern.
// When the commit is started on a branch the previous head of the branch is
// used as the parent of the commit.
func (c APIClient) StartCommit(repoName string, parentCommit string) (*pfs.Commit, error) {
	commit, err := c.PfsAPIClient.StartCommit(
		c.ctx(),
		&pfs.StartCommitRequest{
			Parent: &pfs.Commit{
				Repo: &pfs.Repo{
					Name: repoName,
				},
				ID: parentCommit,
			},
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commit, nil
}

// ForkCommit is the same as StartCommit except that the commit is created
// on a new branch.
func (c APIClient) ForkCommit(repoName string, parentCommit string, branch string) (*pfs.Commit, error) {
	commit, err := c.PfsAPIClient.ForkCommit(
		c.ctx(),
		&pfs.ForkCommitRequest{
			Parent: &pfs.Commit{
				Repo: &pfs.Repo{
					Name: repoName,
				},
				ID: parentCommit,
			},
			Branch: branch,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commit, nil
}

// FinishCommit ends the process of committing data to a Repo and persists the
// Commit. Once a Commit is finished the data becomes immutable and future
// attempts to write to it with PutFile will error.
func (c APIClient) FinishCommit(repoName string, commitID string) error {
	_, err := c.PfsAPIClient.FinishCommit(
		c.ctx(),
		&pfs.FinishCommitRequest{
			Commit: NewCommit(repoName, commitID),
		},
	)
	return sanitizeErr(err)
}

// ArchiveCommit marks a commit as archived. Archived commits are not listed in
// ListCommit unless commit status is set to Archived or All. Archived commits
// are not considered by FlushCommit either.
func (c APIClient) ArchiveCommit(repoName string, commitID string) error {
	_, err := c.PfsAPIClient.ArchiveCommit(
		c.ctx(),
		&pfs.ArchiveCommitRequest{
			Commits: []*pfs.Commit{NewCommit(repoName, commitID)},
		},
	)
	return sanitizeErr(err)
}

// CancelCommit ends the process of committing data to a repo. It differs from
// FinishCommit in that the Commit will not be used as a source for downstream
// pipelines. CancelCommit is used primarily by PPS for the output commits of
// errant jobs.
func (c APIClient) CancelCommit(repoName string, commitID string) error {
	_, err := c.PfsAPIClient.FinishCommit(
		c.ctx(),
		&pfs.FinishCommitRequest{
			Commit: NewCommit(repoName, commitID),
			Cancel: true,
		},
	)
	return sanitizeErr(err)
}

// InspectCommit returns info about a specific Commit.
func (c APIClient) InspectCommit(repoName string, commitID string) (*pfs.CommitInfo, error) {
	commitInfo, err := c.PfsAPIClient.InspectCommit(
		c.ctx(),
		&pfs.InspectCommitRequest{
			Commit: NewCommit(repoName, commitID),
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commitInfo, nil
}

// ListCommitByRepo lists commits in the given repos.
//
// repoNames defines a set of Repos to consider commits from, if repoNames is left
// nil or empty then the result will be empty.
//
// provenance specifies a set of provenance commits, only commits which have
// ALL of the specified commits as provenance will be returned unless
// provenance is nil in which case it is ignored.
//
// commitType specifies the type of commit you want returned, normally CommitTypeRead is the most useful option
//
// status specifies the status of commit you want returned.  By default, cancelled
// or archived commits are not returned.
//
// block, when set to true, will cause ListCommit to block until at least 1 new CommitInfo is available.
// Using repoNames and block you can get subscription semantics from ListCommit.
// commitStatus, controls the statuses of the returned commits. The default
// value `Normal` will filter out archived and cancelled commits.
func (c APIClient) ListCommitByRepo(repoNames []string, provenance []*pfs.Commit,
	commitType pfs.CommitType, status pfs.CommitStatus, block bool) ([]*pfs.CommitInfo, error) {
	var include []*pfs.Commit
	for _, repoName := range repoNames {
		include = append(include, &pfs.Commit{
			Repo: NewRepo(repoName),
		})
	}
	return c.ListCommit(include, nil, provenance, commitType, status, block)
}

// ListCommit lists commits.
//
// exclude and include are filters that either include or exclude the ancestors of the
// given commits.  A commit is considered the ancestor of itself.
// For instance, ListCommit(include("foo/2")) returns commits foo/0, foo/1, and foo/2,
// if they exist.  In contrast, ListCommit(exclude("foo/2")) returns commits that are
// *not* foo/0, foo/1, or foo/2.
//
// To get all commits on a given branch, simply include a commit whose ID is the branch
// name: ListCommit(include("foo"))
//
// To get all commits in a repo, use ListCommitByRepo.
//
// To get all commits, simply don't provide include or exclude.
//
// provenance specifies a set of provenance commits, only commits which have
// ALL of the specified commits as provenance will be returned unless
// provenance is nil in which case it is ignored.
//
// commitType specifies the type of commit you want returned, normally CommitTypeRead is the most useful option
//
// status specifies the status of commit you want returned.  By default, cancelled
// or archived commits are not returned.
//
// block, when set to true, will cause ListCommit to block until at least 1 new CommitInfo is available.
// Using fromCommits and block you can get subscription semantics from ListCommit.
// commitStatus, controls the statuses of the returned commits. The default
// value `Normal` will filter out archived and cancelled commits.
func (c APIClient) ListCommit(exclude []*pfs.Commit, include []*pfs.Commit,
	provenance []*pfs.Commit, commitType pfs.CommitType, status pfs.CommitStatus,
	block bool) ([]*pfs.CommitInfo, error) {
	commitInfos, err := c.PfsAPIClient.ListCommit(
		c.ctx(),
		&pfs.ListCommitRequest{
			Exclude:    exclude,
			Include:    include,
			Provenance: provenance,
			CommitType: commitType,
			Status:     status,
			Block:      block,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commitInfos.CommitInfo, nil
}

// ListBranch lists the active branches on a Repo.
func (c APIClient) ListBranch(repoName string, status pfs.CommitStatus) ([]string, error) {
	branches, err := c.PfsAPIClient.ListBranch(
		c.ctx(),
		&pfs.ListBranchRequest{
			Repo:   NewRepo(repoName),
			Status: status,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return branches.Branches, nil
}

// DeleteCommit deletes a commit.
// Note it is currently not implemented.
func (c APIClient) DeleteCommit(repoName string, commitID string) error {
	_, err := c.PfsAPIClient.DeleteCommit(
		c.ctx(),
		&pfs.DeleteCommitRequest{
			Commit: NewCommit(repoName, commitID),
		},
	)
	return sanitizeErr(err)
}

// FlushCommit blocks until all of the commits which have a set of commits as
// provenance have finished. For commits to be considered they must have all of
// the specified commits as provenance. This in effect waits for all of the
// jobs that are triggered by a set of commits to complete.
// It returns an error if any of the commits it's waiting on are cancelled due
// to one of the jobs encountering an error during runtime.
// If toRepos is not nil then only the commits up to and including those repos
// will be considered, otherwise all repos are considered.
// Note that it's never necessary to call FlushCommit to run jobs, they'll run
// no matter what, FlushCommit just allows you to wait for them to complete and
// see their output once they do.
func (c APIClient) FlushCommit(commits []*pfs.Commit, toRepos []*pfs.Repo) ([]*pfs.CommitInfo, error) {
	commitInfos, err := c.PfsAPIClient.FlushCommit(
		c.ctx(),
		&pfs.FlushCommitRequest{
			Commit: commits,
			ToRepo: toRepos,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commitInfos.CommitInfo, nil
}

// PutBlock takes a reader and splits the data in it into blocks.
// Blocks are guaranteed to be new line delimited.
// Blocks are content addressed and are thus identified by hashes of the content.
// NOTE: this is lower level function that's used internally and might not be
// useful to users.
func (c APIClient) PutBlock(delimiter pfs.Delimiter, reader io.Reader) (blockRefs *pfs.BlockRefs, retErr error) {
	writer, err := c.newPutBlockWriteCloser(delimiter)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	defer func() {
		err := writer.Close()
		if err != nil && retErr == nil {
			retErr = err
		}
		if retErr == nil {
			blockRefs = writer.blockRefs
		}
	}()
	if _, err := io.Copy(writer, reader); err != nil {
		return nil, sanitizeErr(err)
	}
	// Return value gets set in deferred function
	return nil, nil
}

// GetBlock returns the content of a block using it's hash.
// offset specifies a number of bytes that should be skipped in the beginning of the block.
// size limits the total amount of data returned, note you will get fewer bytes
// than size if you pass a value larger than the size of the block.
// If size is set to 0 then all of the data will be returned.
// NOTE: this is lower level function that's used internally and might not be
// useful to users.
func (c APIClient) GetBlock(hash string, offset uint64, size uint64) (io.Reader, error) {
	apiGetBlockClient, err := c.BlockAPIClient.GetBlock(
		c.ctx(),
		&pfs.GetBlockRequest{
			Block:       NewBlock(hash),
			OffsetBytes: offset,
			SizeBytes:   size,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return grpcutil.NewStreamingBytesReader(apiGetBlockClient), nil
}

// DeleteBlock deletes a block from the block store.
// NOTE: this is lower level function that's used internally and might not be
// useful to users.
func (c APIClient) DeleteBlock(block *pfs.Block) error {
	_, err := c.BlockAPIClient.DeleteBlock(
		c.ctx(),
		&pfs.DeleteBlockRequest{
			Block: block,
		},
	)
	return sanitizeErr(err)
}

// InspectBlock returns info about a specific Block.
func (c APIClient) InspectBlock(hash string) (*pfs.BlockInfo, error) {
	blockInfo, err := c.BlockAPIClient.InspectBlock(
		c.ctx(),
		&pfs.InspectBlockRequest{
			Block: NewBlock(hash),
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return blockInfo, nil
}

// ListBlock returns info about all Blocks.
func (c APIClient) ListBlock() ([]*pfs.BlockInfo, error) {
	blockInfos, err := c.BlockAPIClient.ListBlock(
		c.ctx(),
		&pfs.ListBlockRequest{},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return blockInfos.BlockInfo, nil
}

// PutObject puts a value into the object store and tags it with 0 or more tags.
func (c APIClient) PutObject(r io.Reader, tags ...string) (object *pfs.Object, retErr error) {
	w, err := c.newPutObjectWriteCloser(tags...)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	defer func() {
		if err := w.Close(); err != nil && retErr == nil {
			retErr = sanitizeErr(err)
		}
		if retErr == nil {
			object = w.object
		}
	}()
	if _, err := io.CopyBuffer(w, r, make([]byte, MaxMsgSize/2)); err != nil {
		return nil, sanitizeErr(err)
	}
	// return value set by deferred function
	return nil, nil
}

// GetObject gets an object out of the object store by hash.
func (c APIClient) GetObject(hash string, writer io.Writer) error {
	getObjectClient, err := c.ObjectAPIClient.GetObject(
		c.ctx(),
		&pfs.Object{Hash: hash},
	)
	if err != nil {
		return sanitizeErr(err)
	}
	if err := grpcutil.WriteFromStreamingBytesClient(getObjectClient, writer); err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// ReadObject gets an object by hash and returns it directly as []byte.
func (c APIClient) ReadObject(hash string) ([]byte, error) {
	var buffer bytes.Buffer
	if err := c.GetObject(hash, &buffer); err != nil {
		return nil, err
	}
	return buffer.Bytes(), nil
}

// TagObject applies a tag to an existing object.
func (c APIClient) TagObject(hash string, tags ...string) error {
	var _tags []*pfs.Tag
	for _, tag := range tags {
		_tags = append(_tags, &pfs.Tag{Name: tag})
	}
	if _, err := c.ObjectAPIClient.TagObject(
		c.ctx(),
		&pfs.TagObjectRequest{
			Object: &pfs.Object{Hash: hash},
			Tags:   _tags,
		},
	); err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// InspectObject returns info about an Object.
func (c APIClient) InspectObject(hash string) (*pfs.ObjectInfo, error) {
	value, err := c.ObjectAPIClient.InspectObject(
		c.ctx(),
		&pfs.Object{Hash: hash},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return value, nil
}

// GetTag gets an object out of the object store by tag.
func (c APIClient) GetTag(tag string, writer io.Writer) error {
	getTagClient, err := c.ObjectAPIClient.GetTag(
		c.ctx(),
		&pfs.Tag{Name: tag},
	)
	if err != nil {
		return sanitizeErr(err)
	}
	if err := grpcutil.WriteFromStreamingBytesClient(getTagClient, writer); err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// ReadTag gets an object by tag and returns it directly as []byte.
func (c APIClient) ReadTag(tag string) ([]byte, error) {
	var buffer bytes.Buffer
	if err := c.GetTag(tag, &buffer); err != nil {
		return nil, err
	}
	return buffer.Bytes(), nil
}

// Compact forces compaction of objects.
func (c APIClient) Compact() error {
	_, err := c.ObjectAPIClient.Compact(
		c.ctx(),
		&types.Empty{},
	)
	return err
}

// PutFileWriter writes a file to PFS.
// NOTE: PutFileWriter returns an io.WriteCloser you must call Close on it when
// you are done writing.
func (c APIClient) PutFileWriter(repoName string, commitID string, path string, delimiter pfs.Delimiter) (io.WriteCloser, error) {
	return c.newPutFileWriteCloser(repoName, commitID, path, delimiter)
}

// PutFile writes a file to PFS from a reader.
func (c APIClient) PutFile(repoName string, commitID string, path string, reader io.Reader) (_ int, retErr error) {
	c.streamSemaphore <- struct{}{}
	defer func() { <-c.streamSemaphore }()
	return c.PutFileWithDelimiter(repoName, commitID, path, pfs.Delimiter_LINE, reader)
}

//PutFileWithDelimiter writes a file to PFS from a reader
// delimiter is used to tell PFS how to break the input into blocks
func (c APIClient) PutFileWithDelimiter(repoName string, commitID string, path string, delimiter pfs.Delimiter, reader io.Reader) (_ int, retErr error) {
	writer, err := c.PutFileWriter(repoName, commitID, path, delimiter)
	if err != nil {
		return 0, sanitizeErr(err)
	}
	defer func() {
		if err := writer.Close(); err != nil && retErr == nil {
			retErr = err
		}
	}()
	written, err := io.Copy(writer, reader)
	return int(written), err
}

// PutFileURL puts a file using the content found at a URL.
// The URL is sent to the server which performs the request.
// recursive allow for recursive scraping of some types URLs for example on s3:// urls.
func (c APIClient) PutFileURL(repoName string, commitID string, path string, url string, recursive bool) (retErr error) {
	putFileClient, err := c.PfsAPIClient.PutFile(c.ctx())
	if err != nil {
		return sanitizeErr(err)
	}
	defer func() {
		if _, err := putFileClient.CloseAndRecv(); err != nil && retErr == nil {
			retErr = sanitizeErr(err)
		}
	}()
	if err := putFileClient.Send(&pfs.PutFileRequest{
		File:      NewFile(repoName, commitID, path),
		FileType:  pfs.FileType_FILE_TYPE_REGULAR,
		Url:       url,
		Recursive: recursive,
	}); err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// GetFile returns the contents of a file at a specific Commit.
// offset specifies a number of bytes that should be skipped in the beginning of the file.
// size limits the total amount of data returned, note you will get fewer bytes
// than size if you pass a value larger than the size of the file.
// If size is set to 0 then all of the data will be returned.
// fromCommitID lets you get only the data which was added after this Commit.
// shard allows you to downsample the data, returning only a subset of the
// blocks in the file. shard may be left nil in which case the entire file will be returned
func (c APIClient) GetFile(repoName string, commitID string, path string, offset int64,
	size int64, fromCommitID string, fullFile bool, shard *pfs.Shard, writer io.Writer) error {
	c.streamSemaphore <- struct{}{}
	defer func() { <-c.streamSemaphore }()
	return c.getFile(repoName, commitID, path, offset, size, fromCommitID, fullFile, shard, writer)
}

func (c APIClient) getFile(repoName string, commitID string, path string, offset int64,
	size int64, fromCommitID string, fullFile bool, shard *pfs.Shard, writer io.Writer) error {
	apiGetFileClient, err := c.PfsAPIClient.GetFile(
		c.ctx(),
		&pfs.GetFileRequest{
			File:        NewFile(repoName, commitID, path),
			Shard:       shard,
			OffsetBytes: offset,
			SizeBytes:   size,
			DiffMethod:  newDiffMethod(repoName, fromCommitID, fullFile),
		},
	)
	if err != nil {
		return sanitizeErr(err)
	}
	if err := grpcutil.WriteFromStreamingBytesClient(apiGetFileClient, writer); err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// InspectFile returns info about a specific file.  fromCommitID lets you get
// only info which was added after this Commit.  shard allows you to downsample
// the data, returning info about only a subset of the blocks in the file.
// shard may be left nil in which case info about the entire file will be
// returned
func (c APIClient) InspectFile(repoName string, commitID string, path string,
	fromCommitID string, fullFile bool, shard *pfs.Shard) (*pfs.FileInfo, error) {
	return c.inspectFile(repoName, commitID, path, fromCommitID, fullFile, shard)
}

func (c APIClient) inspectFile(repoName string, commitID string, path string,
	fromCommitID string, fullFile bool, shard *pfs.Shard) (*pfs.FileInfo, error) {
	fileInfo, err := c.PfsAPIClient.InspectFile(
		c.ctx(),
		&pfs.InspectFileRequest{
			File:       NewFile(repoName, commitID, path),
			Shard:      shard,
			DiffMethod: newDiffMethod(repoName, fromCommitID, fullFile),
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return fileInfo, nil
}

// ListFile returns info about all files in a Commit.
// fromCommitID lets you get only info which was added after this Commit.
// shard allows you to downsample the data, returning info about only a subset
// of the blocks in the files or only a subset of files. shard may be left nil
// in which case info about all the files and all the blocks in those files
// will be returned.
// recurse causes ListFile to accurately report the size of data stored in directories, it makes the call more expensive
func (c APIClient) ListFile(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard, recurse bool) ([]*pfs.FileInfo, error) {
	req := &pfs.ListFileRequest{
		File:       NewFile(repoName, commitID, path),
		Shard:      shard,
		DiffMethod: newDiffMethod(repoName, fromCommitID, fullFile),
	}
	if recurse {
		req.Mode = pfs.ListFileMode_ListFile_RECURSE
	} else {
		req.Mode = pfs.ListFileMode_ListFile_NORMAL
	}
	fileInfos, err := c.PfsAPIClient.ListFile(
		c.ctx(),
		req,
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return fileInfos.FileInfo, nil
}

// ListFileFast is the same as ListFile except that it doesn't compute the sizes
// of the files.  As a result it's faster than ListFile.
func (c APIClient) ListFileFast(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard) ([]*pfs.FileInfo, error) {
	req := &pfs.ListFileRequest{
		File:       NewFile(repoName, commitID, path),
		Shard:      shard,
		DiffMethod: newDiffMethod(repoName, fromCommitID, fullFile),
		Mode:       pfs.ListFileMode_ListFile_FAST,
	}
	fileInfos, err := c.PfsAPIClient.ListFile(
		c.ctx(),
		req,
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return fileInfos.FileInfo, nil
}

// WalkFn is the type of the function called for each file in Walk.
// Returning a non-nil error from WalkFn will result in Walk aborting and
// returning said error.
type WalkFn func(*pfs.FileInfo) error

// Walk walks the pfs filesystem rooted at path. walkFn will be called for each
// file found under path, this includes both regular files and directories.
func (c APIClient) Walk(repoName string, commitID string, path string, fromCommitID string, fullFile bool, shard *pfs.Shard, walkFn WalkFn) error {
	fileInfos, err := c.ListFileFast(repoName, commitID, path, fromCommitID, fullFile, shard)
	if err != nil {
		return err
	}
	for _, fileInfo := range fileInfos {
		if err := walkFn(fileInfo); err != nil {
			return err
		}
		if fileInfo.FileType == pfs.FileType_FILE_TYPE_DIR {
			if err := c.Walk(repoName, commitID, fileInfo.File.Path, fromCommitID, fullFile, shard, walkFn); err != nil {
				return err
			}
		}
	}
	return nil
}

// DeleteFile deletes a file from a Commit.
// DeleteFile leaves a tombstone in the Commit, assuming the file isn't written
// to later attempting to get the file from the finished commit will result in
// not found error.
// The file will of course remain intact in the Commit's parent.
func (c APIClient) DeleteFile(repoName string, commitID string, path string) error {
	_, err := c.PfsAPIClient.DeleteFile(
		c.ctx(),
		&pfs.DeleteFileRequest{
			File: NewFile(repoName, commitID, path),
		},
	)
	return err
}

// MakeDirectory creates a directory in PFS.
// Note directories are created implicitly by PutFile, so you technically never
// need this function unless you want to create an empty directory.
func (c APIClient) MakeDirectory(repoName string, commitID string, path string) (retErr error) {
	putFileClient, err := c.PfsAPIClient.PutFile(c.ctx())
	if err != nil {
		return sanitizeErr(err)
	}
	defer func() {
		if _, err := putFileClient.CloseAndRecv(); err != nil && retErr == nil {
			retErr = sanitizeErr(err)
		}
	}()
	return sanitizeErr(putFileClient.Send(
		&pfs.PutFileRequest{
			File:     NewFile(repoName, commitID, path),
			FileType: pfs.FileType_FILE_TYPE_DIR,
		},
	))
}

// SquashCommit copies the content of `fromCommits` to `to`, which needs to be an
// open commit.
func (c APIClient) SquashCommit(repo string, fromCommits []string, to string) error {

	var realFromCommits []*pfs.Commit
	for _, commitID := range fromCommits {
		realFromCommits = append(realFromCommits, NewCommit(repo, commitID))
	}

	_, err := c.PfsAPIClient.SquashCommit(
		c.ctx(),
		&pfs.SquashCommitRequest{
			FromCommits: realFromCommits,
			ToCommit:    NewCommit(repo, to),
		},
	)
	if err != nil {
		return sanitizeErr(err)
	}
	return nil
}

// ReplayCommit replays a series of commits on top of commit "to".  The replayed commits
// are the ancestors of the commits in "fromCommits", with no duplicates in case of
// common ancestors.
func (c APIClient) ReplayCommit(repo string, fromCommits []string, to string) ([]*pfs.Commit, error) {
	var realFromCommits []*pfs.Commit
	for _, commitID := range fromCommits {
		realFromCommits = append(realFromCommits, NewCommit(repo, commitID))
	}
	commits, err := c.PfsAPIClient.ReplayCommit(
		c.ctx(),
		&pfs.ReplayCommitRequest{
			FromCommits: realFromCommits,
			ToBranch:    to,
		},
	)
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return commits.Commit, nil
}

// ArchiveAll archives all commits in all repos.
func (c APIClient) ArchiveAll() error {
	_, err := c.PfsAPIClient.ArchiveAll(
		c.ctx(),
		&types.Empty{},
	)
	return sanitizeErr(err)
}

type putFileWriteCloser struct {
	request       *pfs.PutFileRequest
	putFileClient pfs.API_PutFileClient
	sent          bool
}

func (c APIClient) newPutFileWriteCloser(repoName string, commitID string, path string, delimiter pfs.Delimiter) (*putFileWriteCloser, error) {
	putFileClient, err := c.PfsAPIClient.PutFile(c.ctx())
	if err != nil {
		return nil, err
	}
	return &putFileWriteCloser{
		request: &pfs.PutFileRequest{
			File:      NewFile(repoName, commitID, path),
			FileType:  pfs.FileType_FILE_TYPE_REGULAR,
			Delimiter: delimiter,
		},
		putFileClient: putFileClient,
	}, nil
}

func (w *putFileWriteCloser) Write(p []byte) (int, error) {
	bytesWritten := 0
	for {
		// Buffer the write so that we don't exceed the grpc
		// MaxMsgSize. This value includes the whole payload
		// including headers, so we're conservative and halve it
		ceil := bytesWritten + MaxMsgSize/2
		if ceil > len(p) {
			ceil = len(p)
		}
		actualP := p[bytesWritten:ceil]
		if len(actualP) == 0 {
			break
		}
		w.request.Value = actualP
		if err := w.putFileClient.Send(w.request); err != nil {
			return 0, sanitizeErr(err)
		}
		w.sent = true
		w.request.Value = nil
		// File is only needed on the first request
		w.request.File = nil
		bytesWritten += len(actualP)
	}
	return bytesWritten, nil
}

func (w *putFileWriteCloser) Close() error {
	// we always send at least one request, otherwise it's impossible to create
	// an empty file
	if !w.sent {
		if err := w.putFileClient.Send(w.request); err != nil {
			return err
		}
	}
	_, err := w.putFileClient.CloseAndRecv()
	return sanitizeErr(err)
}

type putBlockWriteCloser struct {
	request        *pfs.PutBlockRequest
	putBlockClient pfs.BlockAPI_PutBlockClient
	blockRefs      *pfs.BlockRefs
}

func (c APIClient) newPutBlockWriteCloser(delimiter pfs.Delimiter) (*putBlockWriteCloser, error) {
	putBlockClient, err := c.BlockAPIClient.PutBlock(c.ctx())
	if err != nil {
		return nil, sanitizeErr(err)
	}
	return &putBlockWriteCloser{
		request: &pfs.PutBlockRequest{
			Delimiter: delimiter,
		},
		putBlockClient: putBlockClient,
		blockRefs:      &pfs.BlockRefs{},
	}, nil
}

func (w *putBlockWriteCloser) Write(p []byte) (int, error) {
	w.request.Value = p
	if err := w.putBlockClient.Send(w.request); err != nil {
		return 0, sanitizeErr(err)
	}
	return len(p), nil
}

func (w *putBlockWriteCloser) Close() error {
	var err error
	w.blockRefs, err = w.putBlockClient.CloseAndRecv()
	return sanitizeErr(err)
}

type putObjectWriteCloser struct {
	request         *pfs.PutObjectRequest
	putObjectClient pfs.ObjectAPI_PutObjectClient
	object          *pfs.Object
}

func (c APIClient) newPutObjectWriteCloser(tags ...string) (*putObjectWriteCloser, error) {
	putObjectClient, err := c.ObjectAPIClient.PutObject(c.ctx())
	if err != nil {
		return nil, sanitizeErr(err)
	}
	var _tags []*pfs.Tag
	for _, tag := range tags {
		_tags = append(_tags, &pfs.Tag{Name: tag})
	}
	return &putObjectWriteCloser{
		request: &pfs.PutObjectRequest{
			Tags: _tags,
		},
		putObjectClient: putObjectClient,
	}, nil
}

func (w *putObjectWriteCloser) Write(p []byte) (int, error) {
	w.request.Value = p
	if err := w.putObjectClient.Send(w.request); err != nil {
		return 0, sanitizeErr(err)
	}
	w.request.Tags = nil
	return len(p), nil
}

func (w *putObjectWriteCloser) Close() error {
	var err error
	w.object, err = w.putObjectClient.CloseAndRecv()
	return sanitizeErr(err)
}

func newDiffMethod(repoName string, fromCommitID string, fullFile bool) *pfs.DiffMethod {
	if fromCommitID != "" {
		return &pfs.DiffMethod{
			FromCommit: NewCommit(repoName, fromCommitID),
			FullFile:   fullFile,
		}
	}
	return nil
}