johno/beads
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

doctor: harden corruption repair and JSONL config

Browse Source 
Co-authored-by: factory-droid[bot] <138933559+factory-droid[bot]@users.noreply.github.com>
This commit is contained in:
Jordan Hubbard
2025-12-26 04:29:29 -04:00
parent b089aaa0d6
commit 1a4f06ef8c
17 changed files with 264 additions and 80 deletions
Download Patch File Download Diff File Expand all files Collapse all files

7
cmd/bd/doctor/fix/common.go
View File

@@ -12,6 +12,13 @@ import (
// This prevents fork bombs when tests call functions that execute bd subcommands.
var ErrTestBinary = fmt.Errorf("running as test binary - cannot execute bd subcommands")
func newBdCmd(bdBinary string, args ...string) *exec.Cmd {
fullArgs := append([]string{"--no-daemon"}, args...)
cmd := exec.Command(bdBinary, fullArgs...) // #nosec G204 -- bdBinary from validated executable path
cmd.Env = append(os.Environ(), "BEADS_NO_DAEMON=1")
return cmd
}
// getBdBinary returns the path to the bd binary to use for fix operations.
// It prefers the current executable to avoid command injection attacks.
// Returns ErrTestBinary if running as a test binary to prevent fork bombs.

3
cmd/bd/doctor/fix/daemon.go
View File

@@ -3,7 +3,6 @@ package fix
import (
"fmt"
"os"
"os/exec"
"path/filepath"
)
@@ -36,7 +35,7 @@ func Daemon(path string) error {
}
// Run bd daemons killall to clean up stale daemons
cmd := exec.Command(bdBinary, "daemons", "killall") // #nosec G204 -- bdBinary from validated executable path
cmd := newBdCmd(bdBinary, "daemons", "killall")
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr

26
cmd/bd/doctor/fix/database_config.go
View File

@@ -32,6 +32,13 @@ func DatabaseConfig(path string) error {
fixed := false
// Never treat system JSONL files as a JSONL export configuration.
if isSystemJSONLFilename(cfg.JSONLExport) {
fmt.Printf(" Updating jsonl_export: %s → issues.jsonl\n", cfg.JSONLExport)
cfg.JSONLExport = "issues.jsonl"
fixed = true
}
// Check if configured JSONL exists
if cfg.JSONLExport != "" {
jsonlPath := cfg.JSONLPath(beadsDir)
@@ -99,7 +106,15 @@ func findActualJSONLFile(beadsDir string) string {
strings.Contains(lowerName, ".orig") ||
strings.Contains(lowerName, ".bak") ||
strings.Contains(lowerName, "~") ||
strings.HasPrefix(lowerName, "backup_") {
strings.HasPrefix(lowerName, "backup_") ||
// System files are not JSONL exports.
name == "deletions.jsonl" ||
name == "interactions.jsonl" ||
name == "molecules.jsonl" ||
// Git merge conflict artifacts (e.g., issues.base.jsonl, issues.left.jsonl)
strings.Contains(lowerName, ".base.jsonl") ||
strings.Contains(lowerName, ".left.jsonl") ||
strings.Contains(lowerName, ".right.jsonl") {
continue
}
@@ -121,6 +136,15 @@ func findActualJSONLFile(beadsDir string) string {
return candidates[0]
}
func isSystemJSONLFilename(name string) bool {
switch name {
case "deletions.jsonl", "interactions.jsonl", "molecules.jsonl":
return true
default:
return false
}
}
// LegacyJSONLConfig migrates from legacy beads.jsonl to canonical issues.jsonl.
// This renames the file, updates metadata.json, and updates .gitattributes if present.
// bd-6xd: issues.jsonl is the canonical filename

50
cmd/bd/doctor/fix/database_config_test.go
View File

@@ -220,3 +220,53 @@ func TestLegacyJSONLConfig_UpdatesGitattributes(t *testing.T) {
t.Errorf("Expected .gitattributes to reference issues.jsonl, got: %q", string(content))
}
}
// TestFindActualJSONLFile_SkipsSystemFiles ensures system JSONL files are never treated as JSONL exports.
func TestFindActualJSONLFile_SkipsSystemFiles(t *testing.T) {
tmpDir := t.TempDir()
// Only system files → no candidates.
if err := os.WriteFile(filepath.Join(tmpDir, "interactions.jsonl"), []byte(`{"id":"x"}`), 0644); err != nil {
t.Fatal(err)
}
if got := findActualJSONLFile(tmpDir); got != "" {
t.Fatalf("expected empty result, got %q", got)
}
// System + legacy export → legacy wins.
if err := os.WriteFile(filepath.Join(tmpDir, "beads.jsonl"), []byte(`{"id":"x"}`), 0644); err != nil {
t.Fatal(err)
}
if got := findActualJSONLFile(tmpDir); got != "beads.jsonl" {
t.Fatalf("expected beads.jsonl, got %q", got)
}
}
func TestDatabaseConfigFix_RejectsSystemJSONLExport(t *testing.T) {
tmpDir := t.TempDir()
beadsDir := filepath.Join(tmpDir, ".beads")
if err := os.Mkdir(beadsDir, 0755); err != nil {
t.Fatalf("Failed to create .beads dir: %v", err)
}
if err := os.WriteFile(filepath.Join(beadsDir, "interactions.jsonl"), []byte(`{"id":"x"}`), 0644); err != nil {
t.Fatalf("Failed to create interactions.jsonl: %v", err)
}
cfg := &configfile.Config{Database: "beads.db", JSONLExport: "interactions.jsonl"}
if err := cfg.Save(beadsDir); err != nil {
t.Fatalf("Failed to save config: %v", err)
}
if err := DatabaseConfig(tmpDir); err != nil {
t.Fatalf("DatabaseConfig failed: %v", err)
}
updated, err := configfile.Load(beadsDir)
if err != nil {
t.Fatalf("Failed to load updated config: %v", err)
}
if updated.JSONLExport != "issues.jsonl" {
t.Fatalf("expected issues.jsonl, got %q", updated.JSONLExport)
}
}

60
cmd/bd/doctor/fix/database_integrity.go
View File

@@ -1,22 +1,18 @@
package fix
import (
"bufio"
"encoding/json"
"fmt"
"os"
"os/exec"
"path/filepath"
"time"
"github.com/steveyegge/beads/internal/beads"
"github.com/steveyegge/beads/internal/configfile"
"github.com/steveyegge/beads/internal/utils"
)
// DatabaseIntegrity attempts to recover from database corruption by:
// 1. Backing up the corrupt database (and WAL/SHM if present)
// 2. Re-initializing the database from the git-tracked JSONL export
// 2. Re-initializing the database from the working tree JSONL export
//
// This is intentionally conservative: it will not delete JSONL, and it preserves the
// original DB as a backup for forensic recovery.
@@ -74,61 +70,35 @@ func DatabaseIntegrity(path string) error {
}
}
// Rebuild via bd init, pointing at the same db path.
// Rebuild by importing from the working tree JSONL into a fresh database.
bdBinary, err := getBdBinary()
if err != nil {
return err
}
args := []string{"--db", dbPath, "init", "--quiet", "--force", "--skip-hooks", "--skip-merge-driver"}
if prefix := detectPrefixFromJSONL(jsonlPath); prefix != "" {
args = append(args, "--prefix", prefix)
}
cmd := exec.Command(bdBinary, args...) // #nosec G204 -- bdBinary is a validated executable path
// Use import (not init) so we always hydrate from the working tree JSONL, not git-tracked blobs.
args := []string{"--db", dbPath, "import", "-i", jsonlPath, "--force", "--no-git-history"}
cmd := newBdCmd(bdBinary, args...)
cmd.Dir = absPath
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr
cmd.Env = os.Environ()
if err := cmd.Run(); err != nil {
// Best-effort rollback: if init didn't recreate the db, restore the backup.
if _, statErr := os.Stat(dbPath); os.IsNotExist(statErr) {
_ = os.Rename(backupDB, dbPath)
// Best-effort rollback: attempt to restore the backup, preserving any partial init output.
failedTS := time.Now().UTC().Format("20060102T150405Z")
if _, statErr := os.Stat(dbPath); statErr == nil {
failedDB := dbPath + "." + failedTS + ".failed.init.db"
_ = os.Rename(dbPath, failedDB)
for _, suffix := range []string{"-wal", "-shm", "-journal"} {
_ = os.Rename(backupDB+suffix, dbPath+suffix)
_ = os.Rename(dbPath+suffix, failedDB+suffix)
}
}
_ = os.Rename(backupDB, dbPath)
for _, suffix := range []string{"-wal", "-shm", "-journal"} {
_ = os.Rename(backupDB+suffix, dbPath+suffix)
}
return fmt.Errorf("failed to rebuild database from JSONL: %w (backup: %s)", err, backupDB)
}
return nil
}
func detectPrefixFromJSONL(jsonlPath string) string {
f, err := os.Open(jsonlPath) // #nosec G304 -- jsonlPath is within the workspace
if err != nil {
return ""
}
defer f.Close()
scanner := bufio.NewScanner(f)
for scanner.Scan() {
line := scanner.Bytes()
if len(line) == 0 {
continue
}
var issue struct {
ID string `json:"id"`
}
if err := json.Unmarshal(line, &issue); err != nil {
continue
}
if issue.ID == "" {
continue
}
return utils.ExtractIssuePrefix(issue.ID)
}
return ""
}

2
cmd/bd/doctor/fix/hooks.go
View File

@@ -28,7 +28,7 @@ func GitHooks(path string) error {
}
// Run bd hooks install
cmd := exec.Command(bdBinary, "hooks", "install") // #nosec G204 -- bdBinary from validated executable path
cmd := newBdCmd(bdBinary, "hooks", "install")
cmd.Dir = path // Set working directory without changing process dir
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr

7
cmd/bd/doctor/fix/migrate.go
View File

@@ -3,7 +3,6 @@ package fix
import (
"fmt"
"os"
"os/exec"
"path/filepath"
)
@@ -28,7 +27,7 @@ func DatabaseVersion(path string) error {
if _, err := os.Stat(dbPath); os.IsNotExist(err) {
// No database - this is a fresh clone, run bd init
fmt.Println("→ No database found, running 'bd init' to hydrate from JSONL...")
cmd := exec.Command(bdBinary, "init") // #nosec G204 -- bdBinary from validated executable path
cmd := newBdCmd(bdBinary, "init")
cmd.Dir = path
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr
@@ -41,8 +40,8 @@ func DatabaseVersion(path string) error {
}
// Database exists - run bd migrate
cmd := exec.Command(bdBinary, "migrate") // #nosec G204 -- bdBinary from validated executable path
cmd.Dir = path // Set working directory without changing process dir
cmd := newBdCmd(bdBinary, "migrate")
cmd.Dir = path // Set working directory without changing process dir
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr

11
cmd/bd/doctor/fix/repo_fingerprint.go
View File

@@ -3,7 +3,6 @@ package fix
import (
"fmt"
"os"
"os/exec"
"path/filepath"
"strings"
)
@@ -31,9 +30,9 @@ func readLineUnbuffered() (string, error) {
// RepoFingerprint fixes repo fingerprint mismatches by prompting the user
// for which action to take. This is interactive because the consequences
// differ significantly between options:
// 1. Update repo ID (if URL changed or bd upgraded)
// 2. Reinitialize database (if wrong database was copied)
// 3. Skip (do nothing)
// 1. Update repo ID (if URL changed or bd upgraded)
// 2. Reinitialize database (if wrong database was copied)
// 3. Skip (do nothing)
func RepoFingerprint(path string) error {
// Validate workspace
if err := validateBeadsWorkspace(path); err != nil {
@@ -67,7 +66,7 @@ func RepoFingerprint(path string) error {
case "1":
// Run bd migrate --update-repo-id
fmt.Println(" → Running 'bd migrate --update-repo-id'...")
cmd := exec.Command(bdBinary, "migrate", "--update-repo-id") // #nosec G204 -- bdBinary from validated executable path
cmd := newBdCmd(bdBinary, "migrate", "--update-repo-id")
cmd.Dir = path
cmd.Stdin = os.Stdin // Allow user to respond to migrate's confirmation prompt
cmd.Stdout = os.Stdout
@@ -105,7 +104,7 @@ func RepoFingerprint(path string) error {
_ = os.Remove(dbPath + "-shm")
fmt.Println(" → Running 'bd init'...")
cmd := exec.Command(bdBinary, "init", "--quiet") // #nosec G204 -- bdBinary from validated executable path
cmd := newBdCmd(bdBinary, "init", "--quiet")
cmd.Dir = path
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr

34
cmd/bd/doctor/fix/sync.go
View File

@@ -6,7 +6,6 @@ import (
"encoding/json"
"fmt"
"os"
"os/exec"
"path/filepath"
_ "github.com/ncruces/go-sqlite3/driver"
@@ -102,21 +101,36 @@ func DBJSONLSync(path string) error {
return err
}
// Run the appropriate sync command
var cmd *exec.Cmd
if syncDirection == "export" {
// Export DB to JSONL file (must specify -o to write to file, not stdout)
jsonlOutputPath := filepath.Join(beadsDir, "issues.jsonl")
cmd = exec.Command(bdBinary, "export", "-o", jsonlOutputPath, "--force") // #nosec G204 -- bdBinary from validated executable path
} else {
cmd = exec.Command(bdBinary, "sync", "--import-only") // #nosec G204 -- bdBinary from validated executable path
exportCmd := newBdCmd(bdBinary, "export", "-o", jsonlOutputPath, "--force")
exportCmd.Dir = path // Set working directory without changing process dir
exportCmd.Stdout = os.Stdout
exportCmd.Stderr = os.Stderr
if err := exportCmd.Run(); err != nil {
return fmt.Errorf("failed to export database to JSONL: %w", err)
}
// Staleness check uses last_import_time. After exporting, JSONL mtime is newer,
// so mark the DB as fresh by running a no-op import (skip existing issues).
markFreshCmd := newBdCmd(bdBinary, "import", "-i", jsonlOutputPath, "--force", "--skip-existing", "--no-git-history")
markFreshCmd.Dir = path
markFreshCmd.Stdout = os.Stdout
markFreshCmd.Stderr = os.Stderr
if err := markFreshCmd.Run(); err != nil {
return fmt.Errorf("failed to mark database as fresh after export: %w", err)
}
return nil
}
cmd.Dir = path // Set working directory without changing process dir
cmd.Stdout = os.Stdout
cmd.Stderr = os.Stderr
importCmd := newBdCmd(bdBinary, "sync", "--import-only")
importCmd.Dir = path // Set working directory without changing process dir
importCmd.Stdout = os.Stdout
importCmd.Stderr = os.Stderr
if err := cmd.Run(); err != nil {
if err := importCmd.Run(); err != nil {
return fmt.Errorf("failed to sync database with JSONL: %w", err)
}

3
cmd/bd/doctor/fix/sync_branch.go
View File

@@ -32,8 +32,7 @@ func SyncBranchConfig(path string) error {
}
// Set sync.branch using bd config set
// #nosec G204 - bdBinary is controlled by getBdBinary() which returns os.Executable()
setCmd := exec.Command(bdBinary, "config", "set", "sync.branch", currentBranch)
setCmd := newBdCmd(bdBinary, "config", "set", "sync.branch", currentBranch)
setCmd.Dir = path
if output, err := setCmd.CombinedOutput(); err != nil {
return fmt.Errorf("failed to set sync.branch: %w\nOutput: %s", err, string(output))