default ftyp detection to mp4; fix #562

There are too many ftyp codes to look for (about 100 registered + the unregistered ones.) Previously all ftyps where on the same level in the detection tree. Now mp4 is parent to all other ftyps.
gabriel-vasile · Aug 6, 2024 · 71a146e · 71a146e
1 parent b36b70f
commit 71a146e
Show file tree

Hide file tree

Showing 5 changed files with 40 additions and 33 deletions.
diff --git a/internal/magic/ftyp.go b/internal/magic/ftyp.go
@@ -1,22 +1,14 @@
 package magic
 
-import "bytes"
+import (
+	"bytes"
+)
 
 var (
 	// AVIF matches an AV1 Image File Format still or animated.
 	// Wikipedia page seems outdated listing image/avif-sequence for animations.
 	// https://github.com/AOMediaCodec/av1-avif/issues/59
 	AVIF = ftyp([]byte("avif"), []byte("avis"))
-	// Mp4 matches an MP4 file.
-	Mp4 = ftyp(
-		[]byte("avc1"), []byte("dash"), []byte("iso2"), []byte("iso3"),
-		[]byte("iso4"), []byte("iso5"), []byte("iso6"), []byte("isom"),
-		[]byte("mmp4"), []byte("mp41"), []byte("mp42"), []byte("mp4v"),
-		[]byte("mp71"), []byte("MSNV"), []byte("NDAS"), []byte("NDSC"),
-		[]byte("NSDC"), []byte("NSDH"), []byte("NDSM"), []byte("NDSP"),
-		[]byte("NDSS"), []byte("NDXC"), []byte("NDXH"), []byte("NDXM"),
-		[]byte("NDXP"), []byte("NDXS"), []byte("F4V "), []byte("F4P "),
-	)
 	// ThreeGP matches a 3GPP file.
 	ThreeGP = ftyp(
 		[]byte("3gp1"), []byte("3gp2"), []byte("3gp3"), []byte("3gp4"),
@@ -86,3 +78,21 @@ func QuickTime(raw []byte, _ uint32) bool {
 	}
 	return bytes.Equal(raw[:8], []byte("\x00\x00\x00\x08wide"))
 }
+
+// Mp4 detects an .mp4 file. Mp4 detections only does a basic ftyp check.
+// Mp4 has many registered and unregistered code points so it's hard to keep track
+// of all. Detection will default on video/mp4 for all ftyp files.
+// ISO_IEC_14496-12 is the specification for the iso container.
+func Mp4(raw []byte, _ uint32) bool {
+	if len(raw) < 12 {
+		return false
+	}
+	// ftyps are made out of boxes. The first 4 bytes of the box represent
+	// its size in big-endian uint32. First box is the ftyp box and it is small
+	// in size. Check most significant byte is 0 to filter out false positive
+	// text files that happen to contain the string "ftyp" at index 4.
+	if raw[0] != 0 {
+		return false
+	}
+	return bytes.Equal(raw[4:8], []byte("ftyp"))
+}
diff --git a/internal/magic/magic.go b/internal/magic/magic.go
@@ -153,9 +153,6 @@ func ftyp(sigs ...[]byte) Detector {
 		if len(raw) < 12 {
 			return false
 		}
-		if !bytes.Equal(raw[4:8], []byte("ftyp")) {
-			return false
-		}
 		for _, s := range sigs {
 			if bytes.Equal(raw[8:12], s) {
 				return true

diff --git a/mimetype_test.go b/mimetype_test.go
@@ -564,6 +564,7 @@ func BenchmarkCommon(b *testing.B) {
 		"xls.xls",
 		"webm.webm",
 		"csv.csv",
+		"mp4.mp4",
 	}
 	for _, file := range commonFiles {
 		f, err := os.ReadFile(filepath.Join(testDataDir, file))

diff --git a/supported_mimes.md b/supported_mimes.md
@@ -75,21 +75,26 @@ Extension | MIME type | Aliases
 **.au** | audio/basic | -
 **.mpeg** | video/mpeg | -
 **.mov** | video/quicktime | -
-**.mqv** | video/quicktime | -
 **.mp4** | video/mp4 | -
-**.webm** | video/webm | audio/webm
+**.avif** | image/avif | -
 **.3gp** | video/3gpp | video/3gp, audio/3gpp
 **.3g2** | video/3gpp2 | video/3g2, audio/3gpp2
+**.mp4** | audio/mp4 | audio/x-m4a, audio/x-mp4a
+**.mqv** | video/quicktime | -
+**.m4a** | audio/x-m4a | -
+**.m4v** | video/x-m4v | -
+**.heic** | image/heic | -
+**.heic** | image/heic-sequence | -
+**.heif** | image/heif | -
+**.heif** | image/heif-sequence | -
+**.webm** | video/webm | audio/webm
 **.avi** | video/x-msvideo | video/avi, video/msvideo
 **.flv** | video/x-flv | -
 **.mkv** | video/x-matroska | -
 **.asf** | video/x-ms-asf | video/asf, video/x-ms-wmv
 **.aac** | audio/aac | -
 **.voc** | audio/x-unknown | -
-**.mp4** | audio/mp4 | audio/x-m4a, audio/x-mp4a
-**.m4a** | audio/x-m4a | -
 **.m3u** | application/vnd.apple.mpegurl | audio/mpegurl
-**.m4v** | video/x-m4v | -
 **.rmvb** | application/vnd.rn-realmedia-vbr | -
 **.gz** | application/gzip | application/x-gzip, application/x-gunzip, application/gzipped, application/gzip-compressed, application/x-gzip-compressed, gzip/document
 **.class** | application/x-java-applet | -
@@ -118,10 +123,6 @@ Extension | MIME type | Aliases
 **.macho** | application/x-mach-binary | -
 **.qcp** | audio/qcelp | -
 **.icns** | image/x-icns | -
-**.heic** | image/heic | -
-**.heic** | image/heic-sequence | -
-**.heif** | image/heif | -
-**.heif** | image/heif-sequence | -
 **.hdr** | image/vnd.radiance | -
 **.mrc** | application/marc | -
 **.mdb** | application/x-msaccess | -
@@ -138,7 +139,6 @@ Extension | MIME type | Aliases
 **.pat** | image/x-gimp-pat | -
 **.gbr** | image/x-gimp-gbr | -
 **.glb** | model/gltf-binary | -
-**.avif** | image/avif | -
 **.cab** | application/x-installshield | -
 **.jxr** | image/jxr | image/vnd.ms-photo
 **.txt** | text/plain | -

diff --git a/tree.go b/tree.go
@@ -18,14 +18,13 @@ import (
 var root = newMIME("application/octet-stream", "",
 	func([]byte, uint32) bool { return true },
 	xpm, sevenZ, zip, pdf, fdf, ole, ps, psd, p7s, ogg, png, jpg, jxl, jp2, jpx,
-	jpm, jxs, gif, webp, exe, elf, ar, tar, xar, bz2, fits, tiff, bmp, ico, mp3, flac,
-	midi, ape, musePack, amr, wav, aiff, au, mpeg, quickTime, mqv, mp4, webM,
-	threeGP, threeG2, avi, flv, mkv, asf, aac, voc, aMp4, m4a, m3u, m4v, rmvb,
-	gzip, class, swf, crx, ttf, woff, woff2, otf, ttc, eot, wasm, shx, dbf, dcm, rar,
-	djvu, mobi, lit, bpg, sqlite3, dwg, nes, lnk, macho, qcp, icns, heic,
-	heicSeq, heif, heifSeq, hdr, mrc, mdb, accdb, zstd, cab, rpm, xz, lzip,
-	torrent, cpio, tzif, xcf, pat, gbr, glb, avif, cabIS, jxr,
-	// Keep text last because it is the slowest check
+	jpm, jxs, gif, webp, exe, elf, ar, tar, xar, bz2, fits, tiff, bmp, ico, mp3,
+	flac, midi, ape, musePack, amr, wav, aiff, au, mpeg, quickTime, mp4, webM,
+	avi, flv, mkv, asf, aac, voc, m3u, rmvb, gzip, class, swf, crx, ttf, woff,
+	woff2, otf, ttc, eot, wasm, shx, dbf, dcm, rar, djvu, mobi, lit, bpg,
+	sqlite3, dwg, nes, lnk, macho, qcp, icns, hdr, mrc, mdb, accdb, zstd, cab,
+	rpm, xz, lzip, torrent, cpio, tzif, xcf, pat, gbr, glb, cabIS, jxr,
+	// Keep text last because it is the slowest check.
 	text,
 )
 
@@ -161,7 +160,7 @@ var (
 	m3u = newMIME("application/vnd.apple.mpegurl", ".m3u", magic.M3u).
 		alias("audio/mpegurl")
 	m4v  = newMIME("video/x-m4v", ".m4v", magic.M4v)
-	mp4  = newMIME("video/mp4", ".mp4", magic.Mp4)
+	mp4  = newMIME("video/mp4", ".mp4", magic.Mp4, avif, threeGP, threeG2, aMp4, mqv, m4a, m4v, heic, heicSeq, heif, heifSeq)
 	webM = newMIME("video/webm", ".webm", magic.WebM).
 		alias("audio/webm")
 	mpeg      = newMIME("video/mpeg", ".mpeg", magic.Mpeg)