diff options
| author | Giteabot <teabot@gitea.io> | 2023-03-08 03:03:05 -0500 |
|---|---|---|
| committer | GitHub <noreply@github.com> | 2023-03-08 16:03:05 +0800 |
| commit | cf80f829b49b35fd91854798f9ead913145782f1 (patch) | |
| tree | b77d5eedd32a9dae77cdfa2a8476c2d495f971b0 /modules | |
| parent | ed25e094abfc93e83c6fecb5d82ce64d0e220717 (diff) | |
| download | gitea-cf80f829b49b35fd91854798f9ead913145782f1.tar.gz gitea-cf80f829b49b35fd91854798f9ead913145782f1.zip | |
Do not recognize text files as audio (#23355) (#23368)
Backport #23355
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: wxiaoguang <wxiaoguang@gmail.com>
Co-authored-by: delvh <dev.lh@web.de>
Co-authored-by: techknowlogick <techknowlogick@gitea.io>
Co-authored-by: Lunny Xiao <xiaolunwen@gmail.com>
Diffstat (limited to 'modules')
| -rw-r--r-- | modules/typesniffer/typesniffer.go | 10 | ||||
| -rw-r--r-- | modules/typesniffer/typesniffer_test.go | 4 |
2 files changed, 14 insertions, 0 deletions
diff --git a/modules/typesniffer/typesniffer.go b/modules/typesniffer/typesniffer.go index 5b215496b8..7887fd42b7 100644 --- a/modules/typesniffer/typesniffer.go +++ b/modules/typesniffer/typesniffer.go @@ -106,6 +106,16 @@ func DetectContentType(data []byte) SniffedType { } } + 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} } diff --git a/modules/typesniffer/typesniffer_test.go b/modules/typesniffer/typesniffer_test.go index 2bafdffd14..6c6da34aa0 100644 --- a/modules/typesniffer/typesniffer_test.go +++ b/modules/typesniffer/typesniffer_test.go @@ -109,6 +109,10 @@ func TestIsAudio(t *testing.T) { 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) { |