docspell-cli/mount.go

package main

import (
	"context"
	"encoding/json"
	"fmt"
	"os"
	"os/exec"
	"path/filepath"
	"sort"
	"strings"
	"time"

	"bazil.org/fuse"
	"bazil.org/fuse/fs"
	"bazil.org/fuse/fuseutil"
	"github.com/integrii/flaggy"
)

type docspellFS struct {
	root *dirNode
}

type dirNode struct {
	name     string
	children map[string]fs.Node
}

type fileNode struct {
	name         string
	attachmentID string
	size         uint64
	knownSize    bool
	modTime      time.Time
	client       *DocspellClient
	data         []byte
}

type staticFileNode struct {
	name    string
	data    []byte
	modTime time.Time
}

type mountMetadata struct {
	Items map[string]mountedItemMetadata `json:"items"`
}

type mountedItemMetadata struct {
	FriendlyPath string `json:"friendlyPath"`
	ByIDPath     string `json:"byIDPath"`
}

const mountMetadataFileName = ".docspell-cli-metadata.json"

func runMountCommand(args []string, cfg config, password string) error {
	options, err := parseMountOptions(args, cfg)
	if err != nil {
		return err
	}
	if options.mountPoint == "" {
		return fmt.Errorf("mount point missing: pass one or set MountPath in config")
	}
	if os.Getenv("DOCSPELL_CLI_MOUNT_FOREGROUND") != "1" {
		return startMountInBackground(args, options.mountPoint, false)
	}
	if err := os.MkdirAll(options.mountPoint, 0755); err != nil {
		return err
	}

	client := NewDocspellClient(cfg.DocspellURL, options.authHeader)
	if err := client.Login(context.Background(), cfg.User, password); err != nil {
		return err
	}
	items, err := client.SearchItems(context.Background(), options.query, options.limit)
	if err != nil {
		return fmt.Errorf("search items: %w", err)
	}
	root := buildMountTree(context.Background(), client, items)

	conn, err := fuse.Mount(
		options.mountPoint,
		fuse.FSName("docspell"),
		fuse.Subtype("docspell-cli"),
		fuse.ReadOnly(),
	)
	if err != nil {
		return err
	}
	defer conn.Close()

	fmt.Printf("Mounted %d items read-only at %s. Unmount with: fusermount -u %s\n", len(items), options.mountPoint, options.mountPoint)
	return fs.Serve(conn, &docspellFS{root: root})
}

type mountOptions struct {
	query      string
	limit      int
	authHeader string
	mountPoint string
}

func parseMountOptions(args []string, cfg config) (mountOptions, error) {
	options := mountOptions{
		query:      cfg.MountQuery,
		limit:      1000,
		authHeader: defaultDocspellAuthHeader,
		mountPoint: cfg.MountPath,
	}
	parser := flaggy.NewParser("mount")
	parser.Description = "Expose Docspell search results as a read-only FUSE filesystem"
	parser.DisableShowVersionWithVersion()
	parser.ShowHelpOnUnexpected = true
	parser.String(&options.query, "", "query", "Docspell item query to expose")
	parser.Int(&options.limit, "", "limit", "number of items to fetch per API request")
	parser.String(&options.authHeader, "", "auth-header", "HTTP header used for the Docspell auth token")
	parser.AddPositionalValue(&options.mountPoint, "mountpoint", 1, false, "mount point, overriding mountPath from config")
	return options, parser.ParseArgs(args)
}

func startMountInBackground(args []string, mountPoint string, quiet bool) error {
	mountPoint = cleanMountPath(mountPoint)
	if isMounted(mountPoint) {
		if !quiet {
			fmt.Printf("Already mounted at %s\n", mountPoint)
		}
		return nil
	}
	cmd := exec.Command(os.Args[0], append([]string{"mount"}, args...)...)
	cmd.Env = append(os.Environ(), "DOCSPELL_CLI_MOUNT_FOREGROUND=1")
	logFile, err := mountLogFile()
	if err != nil {
		return err
	}
	defer logFile.Close()
	cmd.Stdout = logFile
	cmd.Stderr = logFile
	if err := cmd.Start(); err != nil {
		return err
	}
	if err := cmd.Process.Release(); err != nil {
		return err
	}
	if !quiet {
		fmt.Printf("Mounting %s in background. Unmount with: fusermount -u %s\n", mountPoint, mountPoint)
	}
	return nil
}

func runGetPathCommand(args []string, cfg config, password string) error {
	itemIDOrURL, err := parseSingleArgumentCommand("get-path", "item-id-or-url", "Docspell item ID or web URL", args)
	if err != nil {
		return err
	}
	mountPoint, err := ensureMounted(cfg, password)
	if err != nil {
		return err
	}
	id := itemIDFromURLOrID(itemIDOrURL)
	if path, err := friendlyLocalPathForItemID(mountPoint, id); err == nil && path != "" {
		fmt.Println(path)
		return nil
	}
	matches, err := filepath.Glob(filepath.Join(mountPoint, "by-id", idShard(id), id+".*"))
	if err != nil {
		return err
	}
	if len(matches) == 0 {
		return fmt.Errorf("entry %s not found under mounted by-id tree", id)
	}
	fmt.Println(matches[0])
	return nil
}

func friendlyLocalPathForItemID(mountPoint, id string) (string, error) {
	metadata, err := readMountMetadata(mountPoint)
	if err != nil {
		return "", err
	}
	item, ok := metadata.Items[id]
	if !ok || item.FriendlyPath == "" {
		return "", nil
	}
	return filepath.Join(mountPoint, filepath.FromSlash(item.FriendlyPath)), nil
}

func readMountMetadata(mountPoint string) (mountMetadata, error) {
	var metadata mountMetadata
	data, err := os.ReadFile(filepath.Join(mountPoint, mountMetadataFileName))
	if err != nil {
		return metadata, err
	}
	err = json.Unmarshal(data, &metadata)
	return metadata, err
}

func runGetURLCommand(args []string, cfg config, password string) error {
	id, err := parseSingleArgumentCommand("get-url", "local-path-or-item-id", "local mounted path, item ID, or item URL", args)
	if err != nil {
		return err
	}
	if strings.Contains(id, string(os.PathSeparator)) || strings.HasPrefix(id, ".") {
		id, err = itemIDFromLocalPath(id, cfg)
		if err != nil {
			return err
		}
	}
	fmt.Printf("%s/app/item/%s\n", strings.TrimRight(cfg.DocspellURL, "/"), itemIDFromURLOrID(id))
	return nil
}

func parseSingleArgumentCommand(commandName, argName, argDescription string, args []string) (string, error) {
	var value string
	parser := flaggy.NewParser(commandName)
	parser.DisableShowVersionWithVersion()
	parser.ShowHelpOnUnexpected = true
	parser.AddPositionalValue(&value, argName, 1, true, argDescription)
	return value, parser.ParseArgs(args)
}

func ensureMounted(cfg config, password string) (string, error) {
	mountPoint := cleanMountPath(cfg.MountPath)
	if mountPoint == "" {
		return "", fmt.Errorf("MountPath missing in config")
	}
	if isMounted(mountPoint) {
		if isStaleMount(mountPoint) {
			_ = unmount(mountPoint)
		} else {
			return mountPoint, nil
		}
	}
	if isMounted(mountPoint) {
		return mountPoint, nil
	}
	if err := startMountInBackground([]string{mountPoint}, mountPoint, true); err != nil {
		return "", err
	}
	for i := 0; i < 300; i++ {
		if isMounted(mountPoint) {
			return mountPoint, nil
		}
		time.Sleep(100 * time.Millisecond)
	}
	return "", fmt.Errorf("mount did not become ready at %s; see %s", mountPoint, mountLogPath())
}

func isStaleMount(mountPoint string) bool {
	_, err := os.ReadDir(mountPoint)
	return err != nil && strings.Contains(err.Error(), "transport endpoint is not connected")
}

func unmount(mountPoint string) error {
	if err := exec.Command("fusermount", "-u", mountPoint).Run(); err == nil {
		return nil
	}
	return exec.Command("fusermount3", "-u", mountPoint).Run()
}

func isMounted(mountPoint string) bool {
	data, err := os.ReadFile("/proc/mounts")
	if err != nil {
		return false
	}
	mountPoint = cleanMountPath(mountPoint)
	for _, line := range strings.Split(string(data), "\n") {
		fields := strings.Fields(line)
		if len(fields) >= 2 && cleanMountPath(fields[1]) == mountPoint {
			return true
		}
	}
	return false
}

func cleanMountPath(path string) string {
	if path == "" {
		return ""
	}
	abs, err := filepath.Abs(path)
	if err != nil {
		return filepath.Clean(path)
	}
	return filepath.Clean(abs)
}

func mountLogPath() string {
	cacheDir, err := os.UserCacheDir()
	if err != nil || cacheDir == "" {
		return filepath.Join(os.TempDir(), "docspell-cli-mount.log")
	}
	return filepath.Join(cacheDir, "docspell-cli", "mount.log")
}

func mountLogFile() (*os.File, error) {
	path := mountLogPath()
	if err := os.MkdirAll(filepath.Dir(path), 0755); err != nil {
		return nil, err
	}
	return os.OpenFile(path, os.O_CREATE|os.O_WRONLY|os.O_APPEND, 0644)
}

func itemIDFromURLOrID(s string) string {
	s = strings.TrimSpace(s)
	if idx := strings.LastIndex(s, "/app/item/"); idx >= 0 {
		return strings.Trim(strings.TrimPrefix(s[idx:], "/app/item/"), "/")
	}
	return strings.Trim(s, "/")
}

func itemIDFromLocalPath(path string, cfg config) (string, error) {
	abs, err := filepath.Abs(path)
	if err != nil {
		return "", err
	}
	mountPoint, err := filepath.Abs(cfg.MountPath)
	if err != nil {
		return "", err
	}
	rel, err := filepath.Rel(mountPoint, abs)
	if err != nil || strings.HasPrefix(rel, "..") {
		return "", fmt.Errorf("path is not under MountPath %s", cfg.MountPath)
	}
	parts := strings.Split(rel, string(os.PathSeparator))
	if len(parts) >= 3 && parts[0] == "by-id" {
		return strings.TrimSuffix(parts[2], filepath.Ext(parts[2])), nil
	}
	metadata, err := readMountMetadata(mountPoint)
	if err != nil {
		return "", err
	}
	rel = filepath.ToSlash(filepath.Clean(rel))
	for id, item := range metadata.Items {
		if filepath.ToSlash(filepath.Clean(filepath.FromSlash(item.FriendlyPath))) == rel {
			return id, nil
		}
	}
	return "", fmt.Errorf("could not map mounted path to item id")
}

func buildMountTree(ctx context.Context, client *DocspellClient, items []SearchItem) *dirNode {
	root := &dirNode{name: "", children: map[string]fs.Node{}}
	byID := &dirNode{name: "by-id", children: map[string]fs.Node{}}
	root.children["by-id"] = byID
	metadata := mountMetadata{Items: map[string]mountedItemMetadata{}}
	metadataModTime := time.Now()
	for _, item := range items {
		folderName := "No Folder"
		if item.Folder != nil && item.Folder.Name != "" {
			folderName = item.Folder.Name
		}
		date := item.Date
		if date == 0 {
			date = item.ItemDate
		}
		yearMonth := "No Date"
		modTime := time.Now()
		if date > 0 {
			modTime = time.UnixMilli(date)
			yearMonth = modTime.Format("2006/01")
		}
		dirParts := append([]string{folderName}, strings.Split(yearMonth, "/")...)
		dir := ensureDir(root, dirParts[0], dirParts[1:]...)
		base := sanitizeName(item.Name)
		if base == "" {
			base = item.ID
		}
		for i, att := range item.Attachments {
			ext := filepath.Ext(att.Name)
			if ext == "" {
				ext = ".pdf"
			}
			name := base + ext
			if len(item.Attachments) > 1 {
				name = fmt.Sprintf("%s - %02d%s", base, i+1, ext)
			}
			size := att.Size
			if size <= 0 {
				if headSize, err := client.AttachmentSize(ctx, att.ID); err == nil && headSize > 0 {
					size = headSize
				}
			}
			name = uniqueName(dir.children, sanitizeName(name))
			node := &fileNode{name: name, attachmentID: att.ID, size: uint64(size), knownSize: size > 0, modTime: modTime, client: client}
			dir.children[name] = node
			friendlyRelPath := filepath.ToSlash(filepath.Join(append(sanitizePathParts(dirParts), name)...))

			byIDName := item.ID + ext
			if len(item.Attachments) > 1 {
				if i > 0 {
					byIDName = fmt.Sprintf("%s-%d%s", item.ID, i+1, ext)
				}
			}
			shard := idShard(item.ID)
			shardDir := ensureDir(byID, shard)
			byIDName = uniqueName(shardDir.children, sanitizeName(byIDName))
			shardDir.children[byIDName] = node
			if i == 0 {
				metadata.Items[item.ID] = mountedItemMetadata{
					FriendlyPath: friendlyRelPath,
					ByIDPath:     filepath.ToSlash(filepath.Join("by-id", shard, byIDName)),
				}
			}
		}
	}
	metadataData, err := json.MarshalIndent(metadata, "", "  ")
	if err == nil {
		root.children[mountMetadataFileName] = &staticFileNode{name: mountMetadataFileName, data: append(metadataData, '\n'), modTime: metadataModTime}
	}
	return root
}

func sanitizePathParts(parts []string) []string {
	sanitized := make([]string, len(parts))
	for i, part := range parts {
		part = sanitizeName(part)
		if part == "" {
			part = "_"
		}
		sanitized[i] = part
	}
	return sanitized
}

func idShard(id string) string {
	id = sanitizeName(id)
	if len(id) >= 2 {
		return id[:2]
	}
	if id != "" {
		return id
	}
	return "__"
}

func ensureDir(root *dirNode, first string, rest ...string) *dirNode {
	cur := root
	parts := append([]string{first}, rest...)
	for _, part := range parts {
		part = sanitizeName(part)
		if part == "" {
			part = "_"
		}
		next, ok := cur.children[part].(*dirNode)
		if !ok {
			next = &dirNode{name: part, children: map[string]fs.Node{}}
			cur.children[part] = next
		}
		cur = next
	}
	return cur
}

func sanitizeName(s string) string {
	s = strings.TrimSpace(s)
	s = strings.ReplaceAll(s, "/", "_")
	s = strings.ReplaceAll(s, "\x00", "_")
	return s
}

func uniqueName(children map[string]fs.Node, name string) string {
	if _, ok := children[name]; !ok {
		return name
	}
	ext := filepath.Ext(name)
	base := strings.TrimSuffix(name, ext)
	for i := 2; ; i++ {
		candidate := fmt.Sprintf("%s (%d)%s", base, i, ext)
		if _, ok := children[candidate]; !ok {
			return candidate
		}
	}
}

func (f *docspellFS) Root() (fs.Node, error) { return f.root, nil }

func (d *dirNode) Attr(ctx context.Context, a *fuse.Attr) error {
	a.Mode = os.ModeDir | 0555
	return nil
}

func (d *dirNode) Lookup(ctx context.Context, name string) (fs.Node, error) {
	child, ok := d.children[name]
	if !ok {
		return nil, fuse.ENOENT
	}
	return child, nil
}

func (d *dirNode) ReadDirAll(ctx context.Context) ([]fuse.Dirent, error) {
	names := make([]string, 0, len(d.children))
	for name := range d.children {
		names = append(names, name)
	}
	sort.Strings(names)
	entries := make([]fuse.Dirent, 0, len(names))
	for _, name := range names {
		typ := fuse.DT_File
		if _, ok := d.children[name].(*dirNode); ok {
			typ = fuse.DT_Dir
		}
		entries = append(entries, fuse.Dirent{Name: name, Type: typ})
	}
	return entries, nil
}

func (f *fileNode) Attr(ctx context.Context, a *fuse.Attr) error {
	a.Mode = 0444
	if f.knownSize {
		a.Size = f.size
	}
	a.Mtime = f.modTime
	a.Ctime = f.modTime
	return nil
}

func (f *fileNode) Open(ctx context.Context, req *fuse.OpenRequest, resp *fuse.OpenResponse) error {
	resp.Flags |= fuse.OpenDirectIO
	return nil
}

func (f *fileNode) Read(ctx context.Context, req *fuse.ReadRequest, resp *fuse.ReadResponse) error {
	data, err := f.load(ctx)
	if err != nil {
		return err
	}
	fuseutil.HandleRead(req, resp, data)
	return nil
}

func (f *fileNode) ReadAll(ctx context.Context) ([]byte, error) {
	return f.load(ctx)
}

func (f *staticFileNode) Attr(ctx context.Context, a *fuse.Attr) error {
	a.Mode = 0444
	a.Size = uint64(len(f.data))
	a.Mtime = f.modTime
	a.Ctime = f.modTime
	return nil
}

func (f *staticFileNode) Open(ctx context.Context, req *fuse.OpenRequest, resp *fuse.OpenResponse) error {
	resp.Flags |= fuse.OpenDirectIO
	return nil
}

func (f *staticFileNode) Read(ctx context.Context, req *fuse.ReadRequest, resp *fuse.ReadResponse) error {
	fuseutil.HandleRead(req, resp, f.data)
	return nil
}

func (f *staticFileNode) ReadAll(ctx context.Context) ([]byte, error) {
	return f.data, nil
}

func (f *fileNode) load(ctx context.Context) ([]byte, error) {
	if f.data != nil {
		return f.data, nil
	}
	data, err := f.client.DownloadAttachment(ctx, f.attachmentID)
	if err != nil {
		return nil, err
	}
	f.data = data
	f.size = uint64(len(data))
	f.knownSize = true
	return data, nil
}