wings/server/filesystem.go

package server

import (
	"bytes"
	"context"
	"encoding/json"
	"fmt"
	"github.com/gabriel-vasile/mimetype"
	"github.com/karrick/godirwalk"
	"github.com/pkg/errors"
	"github.com/pterodactyl/wings/config"
	"github.com/pterodactyl/wings/server/backup"
	ignore "github.com/sabhiram/go-gitignore"
	"golang.org/x/sync/errgroup"
	"io"
	"io/ioutil"
	"os"
	"path"
	"path/filepath"
	"sort"
	"strconv"
	"strings"
	"sync"
	"sync/atomic"
	"syscall"
	"time"
)

// Error returned when there is a bad path provided to one of the FS calls.
type PathResolutionError struct{}

var ErrNotEnoughDiskSpace = errors.New("not enough disk space is available to perform this operation")

// Returns the error response in a string form that can be more easily consumed.
func (pre PathResolutionError) Error() string {
	return "invalid path resolution"
}

func IsPathResolutionError(err error) bool {
	_, ok := err.(PathResolutionError)

	return ok
}

type Filesystem struct {
	mu           sync.Mutex
	lookupTimeMu sync.RWMutex

	lastLookupTime   time.Time
	lookupInProgress int32
	disk             int64

	Server *Server
}

// Returns the root path that contains all of a server's data.
func (fs *Filesystem) Path() string {
	return filepath.Join(config.Get().System.Data, fs.Server.Id())
}

// Normalizes a directory being passed in to ensure the user is not able to escape
// from their data directory. After normalization if the directory is still within their home
// path it is returned. If they managed to "escape" an error will be returned.
//
// This logic is actually copied over from the SFTP server code. Ideally that eventually
// either gets ported into this application, or is able to make use of this package.
func (fs *Filesystem) SafePath(p string) (string, error) {
	var nonExistentPathResolution string

	// Start with a cleaned up path before checking the more complex bits.
	r := fs.unsafeFilePath(p)

	// At the same time, evaluate the symlink status and determine where this file or folder
	// is truly pointing to.
	p, err := filepath.EvalSymlinks(r)
	if err != nil && !os.IsNotExist(err) {
		return "", err
	} else if os.IsNotExist(err) {
		// The requested directory doesn't exist, so at this point we need to iterate up the
		// path chain until we hit a directory that _does_ exist and can be validated.
		parts := strings.Split(filepath.Dir(r), "/")

		var try string
		// Range over all of the path parts and form directory pathings from the end
		// moving up until we have a valid resolution or we run out of paths to try.
		for k := range parts {
			try = strings.Join(parts[:(len(parts)-k)], "/")

			if !fs.unsafeIsInDataDirectory(try) {
				break
			}

			t, err := filepath.EvalSymlinks(try)
			if err == nil {
				nonExistentPathResolution = t
				break
			}
		}
	}

	// If the new path doesn't start with their root directory there is clearly an escape
	// attempt going on, and we should NOT resolve this path for them.
	if nonExistentPathResolution != "" {
		if !fs.unsafeIsInDataDirectory(nonExistentPathResolution) {
			return "", PathResolutionError{}
		}

		// If the nonExistentPathResolution variable is not empty then the initial path requested
		// did not exist and we looped through the pathway until we found a match. At this point
		// we've confirmed the first matched pathway exists in the root server directory, so we
		// can go ahead and just return the path that was requested initially.
		return r, nil
	}

	// If the requested directory from EvalSymlinks begins with the server root directory go
	// ahead and return it. If not we'll return an error which will block any further action
	// on the file.
	if fs.unsafeIsInDataDirectory(p) {
		return p, nil
	}

	return "", PathResolutionError{}
}

// Generate a path to the file by cleaning it up and appending the root server path to it. This
// DOES NOT guarantee that the file resolves within the server data directory. You'll want to use
// the fs.unsafeIsInDataDirectory(p) function to confirm.
func (fs *Filesystem) unsafeFilePath(p string) string {
	// Calling filepath.Clean on the joined directory will resolve it to the absolute path,
	// removing any ../ type of resolution arguments, and leaving us with a direct path link.
	//
	// This will also trim the existing root path off the beginning of the path passed to
	// the function since that can get a bit messy.
	return filepath.Clean(filepath.Join(fs.Path(), strings.TrimPrefix(p, fs.Path())))
}

// Check that that path string starts with the server data directory path. This function DOES NOT
// validate that the rest of the path does not end up resolving out of this directory, or that the
// targeted file or folder is not a symlink doing the same thing.
func (fs *Filesystem) unsafeIsInDataDirectory(p string) bool {
	return strings.HasPrefix(strings.TrimSuffix(p, "/")+"/", strings.TrimSuffix(fs.Path(), "/")+"/")
}

// Helper function to keep some of the codebase a little cleaner. Returns a "safe" version of the path
// joined with a file. This is important because you cannot just assume that appending a file to a cleaned
// path will result in a cleaned path to that file. For example, imagine you have the following scenario:
//
// my_bad_file -> symlink:/etc/passwd
//
// cleaned := SafePath("../../etc") -> "/"
// filepath.Join(cleaned, my_bad_file) -> "/my_bad_file"
//
// You might think that "/my_bad_file" is fine since it isn't pointing to the original "../../etc/my_bad_file".
// However, this doesn't account for symlinks where the file might be pointing outside of the directory, so
// calling a function such as Chown against it would chown the symlinked location, and not the file within the
// Wings daemon.
func (fs *Filesystem) SafeJoin(dir string, f os.FileInfo) (string, error) {
	if f.Mode()&os.ModeSymlink != 0 {
		return fs.SafePath(filepath.Join(dir, f.Name()))
	}

	return filepath.Join(dir, f.Name()), nil
}

// Executes the fs.SafePath function in parallel against an array of paths. If any of the calls
// fails an error will be returned.
func (fs *Filesystem) ParallelSafePath(paths []string) ([]string, error) {
	var cleaned []string

	// Simple locker function to avoid racy appends to the array of cleaned paths.
	var m = new(sync.Mutex)
	var push = func(c string) {
		m.Lock()
		cleaned = append(cleaned, c)
		m.Unlock()
	}

	// Create an error group that we can use to run processes in parallel while retaining
	// the ability to cancel the entire process immediately should any of it fail.
	g, ctx := errgroup.WithContext(context.Background())

	// Iterate over all of the paths and generate a cleaned path, if there is an error for any
	// of the files, abort the process.
	for _, p := range paths {
		// Create copy so we can use it within the goroutine correctly.
		pi := p

		// Recursively call this function to continue digging through the directory tree within
		// a separate goroutine. If the context is canceled abort this process.
		g.Go(func() error {
			select {
			case <-ctx.Done():
				return ctx.Err()
			default:
				// If the callback returns true, go ahead and keep walking deeper. This allows
				// us to programmatically continue deeper into directories, or stop digging
				// if that pathway knows it needs nothing else.
				if c, err := fs.SafePath(pi); err != nil {
					return err
				} else {
					push(c)
				}

				return nil
			}
		})
	}

	// Block until all of the routines finish and have returned a value.
	return cleaned, g.Wait()
}

type SpaceCheckingOpts struct {
	AllowStaleResponse bool
}

// Determines if the directory a file is trying to be added to has enough space available
// for the file to be written to.
//
// Because determining the amount of space being used by a server is a taxing operation we
// will load it all up into a cache and pull from that as long as the key is not expired.
//
// This operation will potentially block unless allowStaleValue is set to true. See the
// documentation on DiskUsage for how this affects the call.
func (fs *Filesystem) HasSpaceAvailable(allowStaleValue bool) bool {
	size, err := fs.DiskUsage(allowStaleValue)
	if err != nil {
		fs.Server.Log().WithField("error", err).Warn("failed to determine root server directory size")
	}

	// Determine if their folder size, in bytes, is smaller than the amount of space they've
	// been allocated.
	fs.Server.Proc().SetDisk(size)

	space := fs.Server.DiskSpace()
	// If space is -1 or 0 just return true, means they're allowed unlimited.
	//
	// Technically we could skip disk space calculation because we don't need to check if the server exceeds it's limit
	// but because this method caches the disk usage it would be best to calculate the disk usage and always
	// return true.
	if space <= 0 {
		return true
	}

	return (size / 1000.0 / 1000.0) <= space
}

// Internal helper function to allow other parts of the codebase to check the total used disk space
// as needed without overly taxing the system. This will prioritize the value from the cache to avoid
// excessive IO usage. We will only walk the filesystem and determine the size of the directory if there
// is no longer a cached value.
//
// If "allowStaleValue" is set to true, a stale value MAY be returned to the caller if there is an
// expired cache value AND there is currently another lookup in progress. If there is no cached value but
// no other lookup is in progress, a fresh disk space response will be returned to the caller.
//
// This is primarily to avoid a bunch of I/O operations from piling up on the server, especially on servers
// with a large amount of files.
func (fs *Filesystem) DiskUsage(allowStaleValue bool) (int64, error) {
	// Check if cache is expired.
	fs.lookupTimeMu.RLock()
	isValidInCache := fs.lastLookupTime.After(time.Now().Add(time.Second * time.Duration(-1*config.Get().System.DiskCheckInterval)))
	fs.lookupTimeMu.RUnlock()

	if !isValidInCache {
		// If we are now allowing a stale response go ahead  and perform the lookup and return the fresh
		// value. This is a blocking operation to the calling process.
		if !allowStaleValue {
			return fs.updateCachedDiskUsage()
		} else if atomic.LoadInt32(&fs.lookupInProgress) == 0 {
			// Otherwise, if we allow a stale value and there isn't a valid item in the cache and we aren't
			// currently performing a lookup, just do the disk usage calculation in the background.
			go func(fs *Filesystem) {
				if _, err := fs.updateCachedDiskUsage(); err != nil {
					fs.Server.Log().WithField("error", errors.WithStack(err)).Warn("failed to determine disk usage in go-routine")
				}
			}(fs)
		}
	}

	// Return the currently cached value back to the calling function.
	return atomic.LoadInt64(&fs.disk), nil
}

// Updates the currently used disk space for a server.
func (fs *Filesystem) updateCachedDiskUsage() (int64, error) {
	// Obtain an exclusive lock on this process so that we don't unintentionally run it at the same
	// time as another running process. Once the lock is available it'll read from the cache for the
	// second call rather than hitting the disk in parallel.
	fs.mu.Lock()
	defer fs.mu.Unlock()

	// Signal that we're currently updating the disk size so that other calls to the disk checking
	// functions can determine if they should queue up additional calls to this function. Ensure that
	// we always set this back to 0 when this process is done executing.
	atomic.StoreInt32(&fs.lookupInProgress, 1)
	defer atomic.StoreInt32(&fs.lookupInProgress, 0)

	// If there is no size its either because there is no data (in which case running this function
	// will have effectively no impact), or there is nothing in the cache, in which case we need to
	// grab the size of their data directory. This is a taxing operation, so we want to store it in
	// the cache once we've gotten it.
	size, err := fs.DirectorySize("/")

	// Always cache the size, even if there is an error. We want to always return that value
	// so that we don't cause an endless loop of determining the disk size if there is a temporary
	// error encountered.
	fs.lookupTimeMu.Lock()
	fs.lastLookupTime = time.Now()
	fs.lookupTimeMu.Unlock()

	atomic.StoreInt64(&fs.disk, size)

	return size, err
}

// Determines the directory size of a given location by running parallel tasks to iterate
// through all of the folders. Returns the size in bytes. This can be a fairly taxing operation
// on locations with tons of files, so it is recommended that you cache the output.
func (fs *Filesystem) DirectorySize(dir string) (int64, error) {
	d, err := fs.SafePath(dir)
	if err != nil {
		return 0, errors.WithStack(err)
	}

	var size int64
	var st syscall.Stat_t

	err = godirwalk.Walk(d, &godirwalk.Options{
		Unsorted: true,
		Callback: func(p string, e *godirwalk.Dirent) error {
			// If this is a symlink then resolve the final destination of it before trying to continue walking
			// over its contents. If it resolves outside the server data directory just skip everything else for
			// it. Otherwise, allow it to continue.
			if e.IsSymlink() {
				if _, err := fs.SafePath(p); err != nil {
					if IsPathResolutionError(err) {
						return godirwalk.SkipThis
					}

					return err
				}
			}

			if !e.IsDir() {
				syscall.Lstat(p, &st)
				atomic.AddInt64(&size, st.Size)
			}

			return nil
		},
	})

	return size, errors.WithStack(err)
}

// Reads a file on the system and returns it as a byte representation in a file
// reader. This is not the most memory efficient usage since it will be reading the
// entirety of the file into memory.
func (fs *Filesystem) Readfile(p string) (io.Reader, error) {
	cleaned, err := fs.SafePath(p)
	if err != nil {
		return nil, err
	}

	b, err := ioutil.ReadFile(cleaned)
	if err != nil {
		return nil, err
	}

	return bytes.NewReader(b), nil
}

// Writes a file to the system. If the file does not already exist one will be created.
func (fs *Filesystem) Writefile(p string, r io.Reader) error {
	cleaned, err := fs.SafePath(p)
	if err != nil {
		return errors.WithStack(err)
	}

	var currentSize int64

	// If the file does not exist on the system already go ahead and create the pathway
	// to it and an empty file. We'll then write to it later on after this completes.
	if stat, err := os.Stat(cleaned); err != nil {
		if !os.IsNotExist(err) {
			return errors.WithStack(err)
		}

		if err := os.MkdirAll(filepath.Dir(cleaned), 0755); err != nil {
			return errors.WithStack(err)
		}

		if err := fs.Chown(filepath.Dir(cleaned)); err != nil {
			return errors.WithStack(err)
		}
	} else {
		if stat.IsDir() {
			return errors.New("cannot write file contents to a directory")
		}

		currentSize = stat.Size()
	}

	o := &fileOpener{}
	// This will either create the file if it does not already exist, or open and
	// truncate the existing file.
	file, err := o.open(cleaned, os.O_RDWR|os.O_CREATE|os.O_TRUNC, 0644)
	if err != nil {
		return errors.WithStack(err)
	}
	defer file.Close()

	buf := make([]byte, 1024*4)
	sz, err := io.CopyBuffer(file, r, buf)

	// Adjust the disk usage to account for the old size and the new size of the file.
	atomic.AddInt64(&fs.disk, sz-currentSize)

	// Finally, chown the file to ensure the permissions don't end up out-of-whack
	// if we had just created it.
	return fs.Chown(cleaned)
}

// Defines the stat struct object.
type Stat struct {
	Info     os.FileInfo
	Mimetype string
}

func (s *Stat) MarshalJSON() ([]byte, error) {
	return json.Marshal(struct {
		Name      string `json:"name"`
		Created   string `json:"created"`
		Modified  string `json:"modified"`
		Mode      string `json:"mode"`
		Size      int64  `json:"size"`
		Directory bool   `json:"directory"`
		File      bool   `json:"file"`
		Symlink   bool   `json:"symlink"`
		Mime      string `json:"mime"`
	}{
		Name:      s.Info.Name(),
		Created:   s.CTime().Format(time.RFC3339),
		Modified:  s.Info.ModTime().Format(time.RFC3339),
		Mode:      s.Info.Mode().String(),
		Size:      s.Info.Size(),
		Directory: s.Info.IsDir(),
		File:      !s.Info.IsDir(),
		Symlink:   s.Info.Mode().Perm()&os.ModeSymlink != 0,
		Mime:      s.Mimetype,
	})
}

// Stats a file or folder and returns the base stat object from go along with the
// MIME data that can be used for editing files.
func (fs *Filesystem) Stat(p string) (*Stat, error) {
	cleaned, err := fs.SafePath(p)
	if err != nil {
		return nil, err
	}

	return fs.unsafeStat(cleaned)
}

func (fs *Filesystem) unsafeStat(p string) (*Stat, error) {
	s, err := os.Stat(p)
	if err != nil {
		return nil, err
	}

	var m *mimetype.MIME
	if !s.IsDir() {
		m, err = mimetype.DetectFile(p)
		if err != nil {
			return nil, err
		}
	}

	st := &Stat{
		Info:     s,
		Mimetype: "inode/directory",
	}

	if m != nil {
		st.Mimetype = m.String()
	}

	return st, nil
}

// Creates a new directory (name) at a specified path (p) for the server.
func (fs *Filesystem) CreateDirectory(name string, p string) error {
	cleaned, err := fs.SafePath(path.Join(p, name))
	if err != nil {
		return errors.WithStack(err)
	}

	return os.MkdirAll(cleaned, 0755)
}

// Moves (or renames) a file or directory.
func (fs *Filesystem) Rename(from string, to string) error {
	cleanedFrom, err := fs.SafePath(from)
	if err != nil {
		return errors.WithStack(err)
	}

	cleanedTo, err := fs.SafePath(to)
	if err != nil {
		return errors.WithStack(err)
	}

	// If the target file or directory already exists the rename function will fail, so just
	// bail out now.
	if _, err := os.Stat(cleanedTo); err == nil {
		return os.ErrExist
	}

	if cleanedTo == fs.Path() {
		return errors.New("attempting to rename into an invalid directory space")
	}

	d := strings.TrimSuffix(cleanedTo, path.Base(cleanedTo))
	// Ensure that the directory we're moving into exists correctly on the system. Only do this if
	// we're not at the root directory level.
	if d != fs.Path() {
		if mkerr := os.MkdirAll(d, 0644); mkerr != nil {
			return errors.Wrap(mkerr, "failed to create directory structure for file rename")
		}
	}

	return os.Rename(cleanedFrom, cleanedTo)
}

// Recursively iterates over a file or directory and sets the permissions on all of the
// underlying files. Iterate over all of the files and directories. If it is a file just
// go ahead and perform the chown operation. Otherwise dig deeper into the directory until
// we've run out of directories to dig into.
func (fs *Filesystem) Chown(path string) error {
	cleaned, err := fs.SafePath(path)
	if err != nil {
		return errors.WithStack(err)
	}

	uid := config.Get().System.User.Uid
	gid := config.Get().System.User.Gid

	// Start by just chowning the initial path that we received.
	if err := os.Chown(cleaned, uid, gid); err != nil {
		return errors.WithStack(err)
	}

	// If this is not a directory we can now return from the function, there is nothing
	// left that we need to do.
	if st, _ := os.Stat(cleaned); !st.IsDir() {
		return nil
	}

	// If this was a directory, begin walking over its contents recursively and ensure that all
	// of the subfiles and directories get their permissions updated as well.
	return godirwalk.Walk(cleaned, &godirwalk.Options{
		Unsorted: true,
		Callback: func(p string, e *godirwalk.Dirent) error {
			// Do not attempt to chmod a symlink. Go's os.Chown function will affect the symlink
			// so if it points to a location outside the data directory the user would be able to
			// (un)intentionally modify that files permissions.
			if e.IsSymlink() {
				if e.IsDir() {
					return godirwalk.SkipThis
				}

				return nil
			}

			return os.Chown(p, uid, gid)
		},
	})
}

// Copies a given file to the same location and appends a suffix to the file to indicate that
// it has been copied.
//
// @todo need to get an exclusive lock on the file.
func (fs *Filesystem) Copy(p string) error {
	cleaned, err := fs.SafePath(p)
	if err != nil {
		return errors.WithStack(err)
	}

	if s, err := os.Stat(cleaned); err != nil {
		return errors.WithStack(err)
	} else if s.IsDir() || !s.Mode().IsRegular() {
		// If this is a directory or not a regular file, just throw a not-exist error
		// since anything calling this function should understand what that means.
		return os.ErrNotExist
	}

	base := filepath.Base(cleaned)
	relative := strings.TrimSuffix(strings.TrimPrefix(cleaned, fs.Path()), base)
	extension := filepath.Ext(base)
	name := strings.TrimSuffix(base, filepath.Ext(base))

	// Begin looping up to 50 times to try and create a unique copy file name. This will take
	// an input of "file.txt" and generate "file copy.txt". If that name is already taken, it will
	// then try to write "file copy 2.txt" and so on, until reaching 50 loops. At that point we
	// won't waste anymore time, just use the current timestamp and make that copy.
	//
	// Could probably make this more efficient by checking if there are any files matching the copy
	// pattern, and trying to find the highest number and then incrementing it by one rather than
	// looping endlessly.
	var i int
	copySuffix := " copy"
	for i = 0; i < 51; i++ {
		if i > 0 {
			copySuffix = " copy " + strconv.Itoa(i)
		}

		tryName := fmt.Sprintf("%s%s%s", name, copySuffix, extension)
		tryLocation, err := fs.SafePath(path.Join(relative, tryName))
		if err != nil {
			return errors.WithStack(err)
		}

		// If the file exists, continue to the next loop, otherwise we're good to start a copy.
		if _, err := os.Stat(tryLocation); err != nil && !os.IsNotExist(err) {
			return errors.WithStack(err)
		} else if os.IsNotExist(err) {
			break
		}

		if i == 50 {
			copySuffix = "." + time.Now().Format(time.RFC3339)
		}
	}

	finalPath, err := fs.SafePath(path.Join(relative, fmt.Sprintf("%s%s%s", name, copySuffix, extension)))
	if err != nil {
		return errors.WithStack(err)
	}

	source, err := os.Open(cleaned)
	if err != nil {
		return errors.WithStack(err)
	}
	defer source.Close()

	dest, err := os.Create(finalPath)
	if err != nil {
		return errors.WithStack(err)
	}
	defer dest.Close()

	buf := make([]byte, 1024*4)
	if _, err := io.CopyBuffer(dest, source, buf); err != nil {
		return errors.WithStack(err)
	}

	return nil
}

// Deletes a file or folder from the system. Prevents the user from accidentally
// (or maliciously) removing their root server data directory.
func (fs *Filesystem) Delete(p string) error {
	// This is one of the few (only?) places in the codebase where we're explicitly not using
	// the SafePath functionality when working with user provided input. If we did, you would
	// not be able to delete a file that is a symlink pointing to a location outside of the data
	// directory.
	//
	// We also want to avoid resolving a symlink that points _within_ the data directory and thus
	// deleting the actual source file for the symlink rather than the symlink itself. For these
	// purposes just resolve the actual file path using filepath.Join() and confirm that the path
	// exists within the data directory.
	resolved := fs.unsafeFilePath(p)
	if !fs.unsafeIsInDataDirectory(resolved) {
		return PathResolutionError{}
	}

	// Block any whoopsies.
	if resolved == fs.Path() {
		return errors.New("cannot delete root server directory")
	}

	if st, err := os.Stat(resolved); err != nil {
		if !os.IsNotExist(err) {
			fs.Server.Log().WithField("error", err).WithField("path", resolved).Warn("error while attempting to stat file before deletion")
		}
	} else {
		if !st.IsDir() {
			atomic.SwapInt64(&fs.disk, -st.Size())
		} else {
			go func(st os.FileInfo, resolved string) {
				if s, err := fs.DirectorySize(resolved); err == nil {
					atomic.AddInt64(&fs.disk, -s)
				}
			}(st, resolved)
		}
	}

	return os.RemoveAll(resolved)
}

// Lists the contents of a given directory and returns stat information about each
// file and folder within it.
func (fs *Filesystem) ListDirectory(p string) ([]*Stat, error) {
	cleaned, err := fs.SafePath(p)
	if err != nil {
		return nil, err
	}

	files, err := ioutil.ReadDir(cleaned)
	if err != nil {
		return nil, err
	}

	var wg sync.WaitGroup

	// You must initialize the output of this directory as a non-nil value otherwise
	// when it is marshaled into a JSON object you'll just get 'null' back, which will
	// break the panel badly.
	out := make([]*Stat, len(files))

	// Iterate over all of the files and directories returned and perform an async process
	// to get the mime-type for them all.
	for i, file := range files {
		wg.Add(1)

		go func(idx int, f os.FileInfo) {
			defer wg.Done()

			var m *mimetype.MIME
			var d = "inode/directory"
			if !f.IsDir() {
				cleanedp, _ := fs.SafeJoin(cleaned, f)
				if cleanedp != "" {
					m, _ = mimetype.DetectFile(filepath.Join(cleaned, f.Name()))
				} else {
					// Just pass this for an unknown type because the file could not safely be resolved within
					// the server data path.
					d = "application/octet-stream"
				}
			}

			st := &Stat{
				Info:     f,
				Mimetype: d,
			}

			if m != nil {
				st.Mimetype = m.String()
			}

			out[idx] = st
		}(i, file)
	}

	wg.Wait()

	// Sort the output alphabetically to begin with since we've run the output
	// through an asynchronous process and the order is gonna be very random.
	sort.SliceStable(out, func(i, j int) bool {
		if out[i].Info.Name() == out[j].Info.Name() || out[i].Info.Name() > out[j].Info.Name() {
			return true
		}

		return false
	})

	// Then, sort it so that directories are listed first in the output. Everything
	// will continue to be alphabetized at this point.
	sort.SliceStable(out, func(i, j int) bool {
		return out[i].Info.IsDir()
	})

	return out, nil
}

// Ensures that the data directory for the server instance exists.
func (fs *Filesystem) EnsureDataDirectory() error {
	if _, err := os.Stat(fs.Path()); err != nil && !os.IsNotExist(err) {
		return errors.WithStack(err)
	} else if err != nil {
		// Create the server data directory because it does not currently exist
		// on the system.
		if err := os.MkdirAll(fs.Path(), 0700); err != nil {
			return errors.WithStack(err)
		}

		if err := fs.Chown("/"); err != nil {
			fs.Server.Log().WithField("error", err).Warn("failed to chown server data directory")
		}
	}

	return nil
}

// Given a directory, iterate through all of the files and folders within it and determine
// if they should be included in the output based on an array of ignored matches. This uses
// standard .gitignore formatting to make that determination.
//
// If no ignored files are passed through you'll get the entire directory listing.
func (fs *Filesystem) GetIncludedFiles(dir string, ignored []string) (*backup.IncludedFiles, error) {
	cleaned, err := fs.SafePath(dir)
	if err != nil {
		return nil, err
	}

	i, err := ignore.CompileIgnoreLines(ignored...)
	if err != nil {
		return nil, err
	}

	// Walk through all of the files and directories on a server. This callback only returns
	// files found, and will keep walking deeper and deeper into directories.
	inc := new(backup.IncludedFiles)

	err = godirwalk.Walk(cleaned, &godirwalk.Options{
		Unsorted: true,
		Callback: func(p string, e *godirwalk.Dirent) error {
			sp := p
			if e.IsSymlink() {
				sp, err = fs.SafePath(p)
				if err != nil {
					if IsPathResolutionError(err) {
						return godirwalk.SkipThis
					}

					return err
				}
			}

			// Only push files into the result array since archives can't create an empty directory within them.
			if !e.IsDir() {
				// Avoid unnecessary parsing if there are no ignored files, nothing will match anyways
				// so no reason to call the function.
				if len(ignored) == 0 || !i.MatchesPath(strings.TrimPrefix(sp, fs.Path()+"/")) {
					inc.Push(sp)
				}
			}

			// We can't just abort if the path is technically ignored. It is possible there is a nested
			// file or folder that should not be excluded, so in this case we need to just keep going
			// until we get to a final state.
			return nil
		},
	})

	return inc, errors.WithStack(err)
}

// Compresses all of the files matching the given paths in the specified directory. This function
// also supports passing nested paths to only compress certain files and folders when working in
// a larger directory. This effectively creates a local backup, but rather than ignoring specific
// files and folders, it takes an allow-list of files and folders.
//
// All paths are relative to the dir that is passed in as the first argument, and the compressed
// file will be placed at that location named `archive-{date}.tar.gz`.
func (fs *Filesystem) CompressFiles(dir string, paths []string) (os.FileInfo, error) {
	cleanedRootDir, err := fs.SafePath(dir)
	if err != nil {
		return nil, err
	}

	// Take all of the paths passed in and merge them together with the root directory we've gotten.
	for i, p := range paths {
		paths[i] = filepath.Join(cleanedRootDir, p)
	}

	cleaned, err := fs.ParallelSafePath(paths)
	if err != nil {
		return nil, err
	}

	inc := new(backup.IncludedFiles)
	// Iterate over all of the cleaned paths and merge them into a large object of final file
	// paths to pass into the archiver. As directories are encountered this will drop into them
	// and look for all of the files.
	for _, p := range cleaned {
		f, err := os.Stat(p)
		if err != nil {
			fs.Server.Log().WithField("error", err).WithField("path", p).Debug("failed to stat file or directory for compression")
			continue
		}

		if !f.IsDir() {
			inc.Push(p)
		} else {
			err := godirwalk.Walk(p, &godirwalk.Options{
				Unsorted: true,
				Callback: func(p string, e *godirwalk.Dirent) error {
					sp := p
					if e.IsSymlink() {
						// Ensure that any symlinks are properly resolved to their final destination. If
						// that destination is outside the server directory skip over this entire item, otherwise
						// use the resolved location for the rest of this function.
						sp, err = fs.SafePath(p)
						if err != nil {
							if IsPathResolutionError(err) {
								return godirwalk.SkipThis
							}

							return err
						}
					}

					if !e.IsDir() {
						inc.Push(sp)
					}

					return nil
				},
			})

			if err != nil {
				return nil, err
			}
		}
	}

	a := &backup.Archive{TrimPrefix: fs.Path(), Files: inc}

	d := path.Join(cleanedRootDir, fmt.Sprintf("archive-%s.tar.gz", strings.ReplaceAll(time.Now().Format(time.RFC3339), ":", "")))

	return a.Create(d, context.Background())
}

// Handle errors encountered when walking through directories.
//
// If there is a path resolution error just skip the item entirely. Only return this for a
// directory, otherwise return nil. Returning this error for a file will stop the walking
// for the remainder of the directory. This is assuming an os.FileInfo struct was even returned.
func (fs *Filesystem) handleWalkerError(err error, f os.FileInfo) error {
	if !IsPathResolutionError(err) {
		return err
	}

	if f != nil && f.IsDir() {
		return filepath.SkipDir
	}

	return nil
}

type fileOpener struct {
	busy        uint
}

// Attempts to open a given file up to "attempts" number of times, using a backoff. If the file
// cannot be opened because of a "text file busy" error, we will attempt until the number of attempts
// has been exhaused, at which point we will abort with an error.
func (fo *fileOpener) open(path string, flags int, perm os.FileMode) (*os.File, error) {
	for {
		f, err := os.OpenFile(path, flags, perm)

		// If there is an error because the text file is busy, go ahead and sleep for a few
		// hundred milliseconds and then try again up to three times before just returning the
		// error back to the caller.
		//
		// Based on code from: https://github.com/golang/go/issues/22220#issuecomment-336458122
		if err != nil && fo.busy < 3 && strings.Contains(err.Error(), "text file busy") {
			time.Sleep(100 * time.Millisecond << fo.busy)
			fo.busy++
			continue
		}

		return f, err
	}
}