Golang 通过读取文件头，识别文件类型
package main
import (
"fmt"
"net/http"
"os"
)
func main() {
// Open File
f, err := os.Open("heilong.mp3")
if err != nil {
panic(err)
}
defer f.Close()
// Get the content
contentType, err := GetFileContentType(f)
if err != nil {
panic(err)
}
fmt.Println("Content Type: " + contentType)
}
func GetFileContentType(out *os.File) (string, error) {
// Only the first 512 bytes are used to sniff the content type.
buffer := make([]byte, 512)
_, err := out.Read(buffer)
if err != nil {
return "", err
}
// Output: [512/512]0xc0000f8000
println(buffer)
// Use the net/http package's handy DectectContentType function. Always returns a valid
// content-type by returning "application/octet-stream" if no others seemed to match.
contentType := http.DetectContentType(buffer)
return contentType, nil
}
文件头映射表
// Data matching the table in section 6.
var sniffSignatures = []sniffSig{
htmlSig("<!DOCTYPE HTML"),
htmlSig("<HTML"),
htmlSig("<HEAD"),
htmlSig("<SCRIPT"),
htmlSig("<IFRAME"),
htmlSig("<H1"),
htmlSig("<DIV"),
htmlSig("<FONT"),
htmlSig("<TABLE"),
htmlSig("<A"),
htmlSig("<STYLE"),
htmlSig("<TITLE"),
htmlSig("<B"),
htmlSig("<BODY"),
htmlSig("<BR"),
htmlSig("<P"),
htmlSig("<!--"),
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\xFF"),
pat: []byte("<?xml"),
skipWS: true,
ct: "text/xml; charset=utf-8"},
&exactSig{[]byte("%PDF-"), "application/pdf"},
&exactSig{[]byte("%!PS-Adobe-"), "application/postscript"},
// UTF BOMs.
&maskedSig{
mask: []byte("\xFF\xFF\x00\x00"),
pat: []byte("\xFE\xFF\x00\x00"),
ct: "text/plain; charset=utf-16be",
},
&maskedSig{
mask: []byte("\xFF\xFF\x00\x00"),
pat: []byte("\xFF\xFE\x00\x00"),
ct: "text/plain; charset=utf-16le",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\x00"),
pat: []byte("\xEF\xBB\xBF\x00"),
ct: "text/plain; charset=utf-8",
},
// Image types
// For posterity, we originally returned "image/vnd.microsoft.icon" from
// https://tools.ietf.org/html/draft-ietf-websec-mime-sniff-03#section-7
// https://codereview.appspot.com/4746042
// but that has since been replaced with "image/x-icon" in Section 6.2
// of https://mimesniff.spec.whatwg.org/#matching-an-image-type-pattern
&exactSig{[]byte("\x00\x00\x01\x00"), "image/x-icon"},
&exactSig{[]byte("\x00\x00\x02\x00"), "image/x-icon"},
&exactSig{[]byte("BM"), "image/bmp"},
&exactSig{[]byte("GIF87a"), "image/gif"},
&exactSig{[]byte("GIF89a"), "image/gif"},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF\xFF\xFF"),
pat: []byte("RIFF\x00\x00\x00\x00WEBPVP"),
ct: "image/webp",
},
&exactSig{[]byte("\x89PNG\x0D\x0A\x1A\x0A"), "image/png"},
&exactSig{[]byte("\xFF\xD8\xFF"), "image/jpeg"},
// Audio and Video types
// Enforce the pattern match ordering as prescribed in
// https://mimesniff.spec.whatwg.org/#matching-an-audio-or-video-type-pattern
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF"),
pat: []byte(".snd"),
ct: "audio/basic",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
pat: []byte("FORM\x00\x00\x00\x00AIFF"),
ct: "audio/aiff",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF"),
pat: []byte("ID3"),
ct: "audio/mpeg",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\xFF"),
pat: []byte("OggS\x00"),
ct: "application/ogg",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\xFF\xFF\xFF\xFF"),
pat: []byte("MThd\x00\x00\x00\x06"),
ct: "audio/midi",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
pat: []byte("RIFF\x00\x00\x00\x00AVI "),
ct: "video/avi",
},
&maskedSig{
mask: []byte("\xFF\xFF\xFF\xFF\x00\x00\x00\x00\xFF\xFF\xFF\xFF"),
pat: []byte("RIFF\x00\x00\x00\x00WAVE"),
ct: "audio/wave",
},
// 6.2.0.2. video/mp4
mp4Sig{},
// 6.2.0.3. video/webm
&exactSig{[]byte("\x1A\x45\xDF\xA3"), "video/webm"},
// Font types
&maskedSig{
// 34 NULL bytes followed by the string "LP"
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\x00LP"),
// 34 NULL bytes followed by \xF\xF
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"),
ct: "application/vnd.ms-fontobject",
},
&exactSig{[]byte("\x00\x01\x00\x00"), "font/ttf"},
&exactSig{[]byte("OTTO"), "font/otf"},
&exactSig{[]byte("ttcf"), "font/collection"},
&exactSig{[]byte("wOFF"), "font/woff"},
&exactSig{[]byte("wOF2"), "font/woff2"},
// Archive types
&exactSig{[]byte("\x1F\x8B\x08"), "application/x-gzip"},
&exactSig{[]byte("PK\x03\x04"), "application/zip"},
// RAR's signatures are incorrectly defined by the MIME spec as per
// https://github.com/whatwg/mimesniff/issues/63
// However, RAR Labs correctly defines it at:
// https://www.rarlab.com/technote.htm#rarsign
// so we use the definition from RAR Labs.
// TODO: do whatever the spec ends up doing.
&exactSig{[]byte("Rar!\x1A\x07\x00"), "application/x-rar-compressed"}, // RAR v1.5-v4.0
&exactSig{[]byte("Rar!\x1A\x07\x01\x00"), "application/x-rar-compressed"}, // RAR v5+
&exactSig{[]byte("\x00\x61\x73\x6D"), "application/wasm"},
textSig{}, // should be last
}
发表评论