1
// Copyright 2011 The Go Authors. All rights reserved.
2
// Use of this source code is governed by a BSD-style
3
// license that can be found in the LICENSE file.
12
// The algorithm uses at most sniffLen bytes to make its decision.
15
// DetectContentType implements the algorithm described
16
// at https://mimesniff.spec.whatwg.org/ to determine the
17
// Content-Type of the given data. It considers at most the
18
// first 512 bytes of data. DetectContentType always returns
19
// a valid MIME type: if it cannot determine a more specific one, it
20
// returns "application/octet-stream".
21
func DetectContentType(data []byte) string {
22
if len(data) > sniffLen {
23
data = data[:sniffLen]
26
// Index of the first non-whitespace byte in data.
28
for ; firstNonWS < len(data) && isWS(data[firstNonWS]); firstNonWS++ {
31
for _, sig := range sniffSignatures {
32
if ct := sig.match(data, firstNonWS); ct != "" {
37
return "application/octet-stream" // fallback
40
func isWS(b byte) bool {
42
case '\t', '\n', '\x0c', '\r', ' ':
48
type sniffSig interface {
49
// match returns the MIME type of the data, or "" if unknown.
50
match(data []byte, firstNonWS int) string
53
// Data matching the table in section 6.
54
var sniffSignatures = []sniffSig{
55
htmlSig("<!DOCTYPE HTML"),
73
&maskedSig{mask: []byte("\xFF\xFF\xFF\xFF\xFF"), pat: []byte("<?xml"), skipWS: true, ct: "text/xml; charset=utf-8"},
75
&exactSig{[]byte("%PDF-"), "application/pdf"},
76
&exactSig{[]byte("%!PS-Adobe-"), "application/postscript"},
79
&maskedSig{mask: []byte("\xFF\xFF\x00\x00"), pat: []byte("\xFE\xFF\x00\x00"), ct: "text/plain; charset=utf-16be"},
80
&maskedSig{mask: []byte("\xFF\xFF\x00\x00"), pat: []byte("\xFF\xFE\x00\x00"), ct: "text/plain; charset=utf-16le"},
81
&maskedSig{mask: []byte("\xFF\xFF\xFF\x00"), pat: []byte("\xEF\xBB\xBF\x00"), ct: "text/plain; charset=utf-8"},
83
&exactSig{[]byte("GIF87a"), "image/gif"},
84
&exactSig{[]byte("GIF89a"), "image/gif"},
85
&exactSig{[]byte("\x89\x50\x4E\x47\x0D\x0A\x1A\x0A"), "image/png"},
86
&exactSig{[]byte("\xFF\xD8\xFF"), "image/jpeg"},
87
&exactSig{[]byte("BM"), "image/bmp"},
89
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF\xFF\xFF"),
90
pat: []byte("RIFF\x00\x00\x00\x00WEBPVP"),
93
&exactSig{[]byte("\x00\x00\x01\x00"), "image/vnd.microsoft.icon"},
96
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
97
pat: []byte("RIFF\x00\x00\x00\x00WAVE"),
101
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
102
pat: []byte("FORM\x00\x00\x00\x00AIFF"),
106
mask: []byte("\xFF\xFF\xFF\xFF"),
111
mask: []byte("\xFF\xFF\xFF\xFF\xFF"),
112
pat: []byte("OggS\x00"),
113
ct: "application/ogg",
116
mask: []byte("\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF"),
117
pat: []byte("MThd\x00\x00\x00\x06"),
121
mask: []byte("\xFF\xFF\xFF"),
126
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
127
pat: []byte("RIFF\x00\x00\x00\x00AVI "),
133
// 34 NULL bytes followed by the string "LP"
134
pat: []byte("\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x4C\x50"),
135
// 34 NULL bytes followed by \xF\xF
136
mask: []byte("\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\x00\xFF\xFF"),
137
ct: "application/vnd.ms-fontobject",
139
&exactSig{[]byte("\x00\x01\x00\x00"), "font/ttf"},
140
&exactSig{[]byte("OTTO"), "font/otf"},
141
&exactSig{[]byte("ttcf"), "font/collection"},
142
&exactSig{[]byte("wOFF"), "font/woff"},
143
&exactSig{[]byte("wOF2"), "font/woff2"},
145
&exactSig{[]byte("\x1A\x45\xDF\xA3"), "video/webm"},
146
&exactSig{[]byte("\x52\x61\x72\x20\x1A\x07\x00"), "application/x-rar-compressed"},
147
&exactSig{[]byte("\x50\x4B\x03\x04"), "application/zip"},
148
&exactSig{[]byte("\x1F\x8B\x08"), "application/x-gzip"},
150
&exactSig{[]byte("\x00\x61\x73\x6D"), "application/wasm"},
154
textSig{}, // should be last
157
type exactSig struct {
162
func (e *exactSig) match(data []byte, firstNonWS int) string {
163
if bytes.HasPrefix(data, e.sig) {
169
type maskedSig struct {
175
func (m *maskedSig) match(data []byte, firstNonWS int) string {
176
// pattern matching algorithm section 6
177
// https://mimesniff.spec.whatwg.org/#pattern-matching-algorithm
180
data = data[firstNonWS:]
182
if len(m.pat) != len(m.mask) {
185
if len(data) < len(m.mask) {
188
for i, mask := range m.mask {
199
func (h htmlSig) match(data []byte, firstNonWS int) string {
200
data = data[firstNonWS:]
201
if len(data) < len(h)+1 {
204
for i, b := range h {
206
if 'A' <= b && b <= 'Z' {
213
// Next byte must be space or right angle bracket.
214
if db := data[len(h)]; db != ' ' && db != '>' {
217
return "text/html; charset=utf-8"
220
var mp4ftype = []byte("ftyp")
221
var mp4 = []byte("mp4")
225
func (mp4Sig) match(data []byte, firstNonWS int) string {
226
// https://mimesniff.spec.whatwg.org/#signature-for-mp4
227
// c.f. section 6.2.1
231
boxSize := int(binary.BigEndian.Uint32(data[:4]))
232
if boxSize%4 != 0 || len(data) < boxSize {
235
if !bytes.Equal(data[4:8], mp4ftype) {
238
for st := 8; st < boxSize; st += 4 {
240
// minor version number
243
if bytes.Equal(data[st:st+3], mp4) {
252
func (textSig) match(data []byte, firstNonWS int) string {
253
// c.f. section 5, step 4.
254
for _, b := range data[firstNonWS:] {
258
0x0E <= b && b <= 0x1A,
259
0x1C <= b && b <= 0x1F:
263
return "text/plain; charset=utf-8"