Files
s3/weed/mount/wfs_save_test.go
zooqueen dd74d3e6d6 Replace AGPL MinIO fork with Apache-2.0 SeaweedFS fork; rebrand CLI weed->s3
hanzoai/s3 was an AGPL-3.0 MinIO fork whose network-copyleft blocks embedding
and reselling as a managed service. Replace it wholesale with a fork of
SeaweedFS (github.com/seaweedfs/seaweedfs, Apache-2.0, Go) which ships a
built-in S3 gateway + Filer.

- License: AGPL-3.0 -> Apache-2.0 (upstream LICENSE kept verbatim, incl.
  "Copyright 2025 Chris Lu"); NOTICE added crediting SeaweedFS. No AGPL/BUSL/
  SSPL/GPL or enterprise ("ee") files remain.
- Module: go.mod path set to github.com/hanzoai/s3 with a self-replace
  (replace github.com/seaweedfs/seaweedfs => ./) so the ~1800 internal Go files
  keep their upstream import paths with zero churn. Internal identifiers and
  import paths are intentionally NOT renamed.
- CLI rebrand: the binary is built as `s3` (was `weed`); root usage, per-command
  help/usage/flag strings, shell autocomplete program name, and the product name
  in CLI output now read "Hanzo S3" / `s3`. Users never see `weed`.
  (e.g. `s3 server`, `s3 filer`, `s3 master`, `s3 volume`, `s3 s3`, `s3 mount`.)
- Build: `make` (or `go build -o s3 ./weed`) builds the `s3` binary green;
  `cmd/weed-db` and `cmd/weed-sql` also build; `go vet ./weed/command/` clean.
- README/LLM.md rewritten for Hanzo S3; AGENTS.md/CLAUDE.md symlinks preserved.
- Dropped a 33MB prebuilt Mach-O `stress` binary upstream had committed; `.gitignore`
  updated to ignore the `s3` build output.
2026-06-18 23:41:25 -07:00

74 lines
2.5 KiB
Go

package mount
import (
"strings"
"syscall"
"testing"
"unicode/utf8"
"github.com/seaweedfs/go-fuse/v2/fuse"
)
// TestSanitizeFuseName_InvalidBytesReplaced reproduces the filename from
// seaweedfs#9139: GNOME Trash "partial" files carry raw binary bytes
// (\x10\x98=\\\x8a\x7f) that are not valid UTF-8. The sanitizer must return a
// UTF-8-valid string so the subsequent proto marshal cannot fail.
func TestSanitizeFuseName_InvalidBytesReplaced(t *testing.T) {
raw := "\x10\x98=\\\x8a\x7f.trashinfo.9a51454f.partial"
out := sanitizeFuseName(raw)
if !utf8.ValidString(out) {
t.Fatalf("sanitizeFuseName returned non-UTF-8: %q", out)
}
// The replacement must be URL-safe: these sanitized names flow into HTTP
// URLs (volume-server uploads, filer HTTP API, S3/WebDAV gateways), so a
// '?' would be interpreted as the query-string delimiter and split the
// path. Explicitly assert the replacement char is neither '?' nor U+FFFD.
if strings.ContainsRune(out, '?') {
t.Fatalf("sanitizer produced non-URL-safe '?' in %q", out)
}
if strings.ContainsRune(out, 0xFFFD) {
t.Fatalf("expected single-byte replacement, got U+FFFD in %q", out)
}
// Valid bytes (\x10, \x3D '=', \x5C '\\', \x7F) must be preserved; only
// \x98 and \x8A — the standalone continuation bytes — get replaced.
if !strings.HasSuffix(out, ".trashinfo.9a51454f.partial") {
t.Fatalf("trailing valid bytes were dropped: %q", out)
}
}
func TestSanitizeFuseName_PassThroughValidUTF8(t *testing.T) {
// An already-valid UTF-8 string must be returned unchanged — no heap
// allocation, no alteration of byte content. Preserving identity matters
// because the overwhelming hot path is valid input.
for _, s := range []string{
"plain.txt",
"日本語.txt",
"🦑 squid",
"",
strings.Repeat("a", 255),
} {
if got := sanitizeFuseName(s); got != s {
t.Fatalf("sanitizeFuseName(%q) = %q, want unchanged", s, got)
}
}
}
// TestCheckName_SanitizesBeforeLengthCheck verifies the caller contract: the
// returned name is always safe to put into a proto string field, and the
// length guard still fires on over-long inputs.
func TestCheckName_SanitizesBeforeLengthCheck(t *testing.T) {
bad := "foo\x80bar"
got, s := checkName(bad)
if s != fuse.OK {
t.Fatalf("checkName(%q) status = %v, want OK", bad, s)
}
if !utf8.ValidString(got) {
t.Fatalf("checkName did not sanitize: %q", got)
}
tooLong := strings.Repeat("x", 300)
if _, s := checkName(tooLong); s != fuse.Status(syscall.ENAMETOOLONG) {
t.Fatalf("checkName length guard lost: status=%v", s)
}
}