Close #17108
This PR uses a trick (removing the ID3 tag) to detect the content again
to to see whether the content is text type.
---------
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
}
}
+ if strings.HasPrefix(ct, "audio/") && bytes.HasPrefix(data, []byte("ID3")) {
+ // The MP3 detection is quite inaccurate, any content with "ID3" prefix will result in "audio/mpeg".
+ // So remove the "ID3" prefix and detect again, if result is text, then it must be text content.
+ // This works especially because audio files contain many unprintable/invalid characters like `0x00`
+ ct2 := http.DetectContentType(data[3:])
+ if strings.HasPrefix(ct2, "text/") {
+ ct = ct2
+ }
+ }
+
return SniffedType{ct}
}
mp3, _ := base64.StdEncoding.DecodeString("SUQzBAAAAAABAFRYWFgAAAASAAADbWFqb3JfYnJhbmQAbXA0MgBUWFhYAAAAEQAAA21pbm9yX3Zl")
assert.True(t, DetectContentType(mp3).IsAudio())
assert.False(t, DetectContentType([]byte("plain text")).IsAudio())
+
+ assert.True(t, DetectContentType([]byte("ID3Toy\000")).IsAudio())
+ assert.True(t, DetectContentType([]byte("ID3Toy\n====\t* hi 🌞, ...")).IsText()) // test ID3 tag for plain text
+ assert.True(t, DetectContentType([]byte("ID3Toy\n====\t* hi 🌞, ..."+"🌛"[0:2])).IsText()) // test ID3 tag with incomplete UTF8 char
}
func TestDetectContentTypeFromReader(t *testing.T) {