Preserve BOM in web editor (#28935) (#28959)

Backport #28935 by @silverwind

The `ToUTF8*` functions were stripping BOM, while BOM is actually valid
in UTF8, so the stripping must be optional depending on use case. This
does:

- Add a options struct to all `ToUTF8*` functions, that by default will
strip BOM to preserve existing behaviour
- Remove `ToUTF8` function, it was dead code
- Rename `ToUTF8WithErr` to `ToUTF8`
- Preserve BOM in Monaco Editor
- Remove a unnecessary newline in the textarea value. Browsers did
ignore it, it seems but it's better not to rely on this behaviour.

Fixes: https://github.com/go-gitea/gitea/issues/28743
Related: https://github.com/go-gitea/gitea/issues/6716 which seems to
have once introduced a mechanism that strips and re-adds the BOM, but
from what I can tell, this mechanism was removed at some point after
that PR.

Co-authored-by: silverwind <me@silverwind.io>
(cherry picked from commit b8e6cffd31)
This commit is contained in:
Giteabot 2024-01-28 06:23:31 +08:00 committed by Earl Warren
parent afdce6bd45
commit b272224918
No known key found for this signature in database
GPG key ID: 0579CB2928A78A00
13 changed files with 69 additions and 134 deletions

View file

@ -22,17 +22,21 @@ import (
// UTF8BOM is the utf-8 byte-order marker // UTF8BOM is the utf-8 byte-order marker
var UTF8BOM = []byte{'\xef', '\xbb', '\xbf'} var UTF8BOM = []byte{'\xef', '\xbb', '\xbf'}
type ConvertOpts struct {
KeepBOM bool
}
// ToUTF8WithFallbackReader detects the encoding of content and converts to UTF-8 reader if possible // ToUTF8WithFallbackReader detects the encoding of content and converts to UTF-8 reader if possible
func ToUTF8WithFallbackReader(rd io.Reader) io.Reader { func ToUTF8WithFallbackReader(rd io.Reader, opts ConvertOpts) io.Reader {
buf := make([]byte, 2048) buf := make([]byte, 2048)
n, err := util.ReadAtMost(rd, buf) n, err := util.ReadAtMost(rd, buf)
if err != nil { if err != nil {
return io.MultiReader(bytes.NewReader(RemoveBOMIfPresent(buf[:n])), rd) return io.MultiReader(bytes.NewReader(MaybeRemoveBOM(buf[:n], opts)), rd)
} }
charsetLabel, err := DetectEncoding(buf[:n]) charsetLabel, err := DetectEncoding(buf[:n])
if err != nil || charsetLabel == "UTF-8" { if err != nil || charsetLabel == "UTF-8" {
return io.MultiReader(bytes.NewReader(RemoveBOMIfPresent(buf[:n])), rd) return io.MultiReader(bytes.NewReader(MaybeRemoveBOM(buf[:n], opts)), rd)
} }
encoding, _ := charset.Lookup(charsetLabel) encoding, _ := charset.Lookup(charsetLabel)
@ -42,20 +46,20 @@ func ToUTF8WithFallbackReader(rd io.Reader) io.Reader {
return transform.NewReader( return transform.NewReader(
io.MultiReader( io.MultiReader(
bytes.NewReader(RemoveBOMIfPresent(buf[:n])), bytes.NewReader(MaybeRemoveBOM(buf[:n], opts)),
rd, rd,
), ),
encoding.NewDecoder(), encoding.NewDecoder(),
) )
} }
// ToUTF8WithErr converts content to UTF8 encoding // ToUTF8 converts content to UTF8 encoding
func ToUTF8WithErr(content []byte) (string, error) { func ToUTF8(content []byte, opts ConvertOpts) (string, error) {
charsetLabel, err := DetectEncoding(content) charsetLabel, err := DetectEncoding(content)
if err != nil { if err != nil {
return "", err return "", err
} else if charsetLabel == "UTF-8" { } else if charsetLabel == "UTF-8" {
return string(RemoveBOMIfPresent(content)), nil return string(MaybeRemoveBOM(content, opts)), nil
} }
encoding, _ := charset.Lookup(charsetLabel) encoding, _ := charset.Lookup(charsetLabel)
@ -70,28 +74,22 @@ func ToUTF8WithErr(content []byte) (string, error) {
result = append(result, content[n:]...) result = append(result, content[n:]...)
} }
result = RemoveBOMIfPresent(result) result = MaybeRemoveBOM(result, opts)
return string(result), err return string(result), err
} }
// ToUTF8WithFallback detects the encoding of content and converts to UTF-8 if possible // ToUTF8WithFallback detects the encoding of content and converts to UTF-8 if possible
func ToUTF8WithFallback(content []byte) []byte { func ToUTF8WithFallback(content []byte, opts ConvertOpts) []byte {
bs, _ := io.ReadAll(ToUTF8WithFallbackReader(bytes.NewReader(content))) bs, _ := io.ReadAll(ToUTF8WithFallbackReader(bytes.NewReader(content), opts))
return bs return bs
} }
// ToUTF8 converts content to UTF8 encoding and ignore error
func ToUTF8(content string) string {
res, _ := ToUTF8WithErr([]byte(content))
return res
}
// ToUTF8DropErrors makes sure the return string is valid utf-8; attempts conversion if possible // ToUTF8DropErrors makes sure the return string is valid utf-8; attempts conversion if possible
func ToUTF8DropErrors(content []byte) []byte { func ToUTF8DropErrors(content []byte, opts ConvertOpts) []byte {
charsetLabel, err := DetectEncoding(content) charsetLabel, err := DetectEncoding(content)
if err != nil || charsetLabel == "UTF-8" { if err != nil || charsetLabel == "UTF-8" {
return RemoveBOMIfPresent(content) return MaybeRemoveBOM(content, opts)
} }
encoding, _ := charset.Lookup(charsetLabel) encoding, _ := charset.Lookup(charsetLabel)
@ -117,11 +115,14 @@ func ToUTF8DropErrors(content []byte) []byte {
} }
} }
return RemoveBOMIfPresent(decoded) return MaybeRemoveBOM(decoded, opts)
} }
// RemoveBOMIfPresent removes a UTF-8 BOM from a []byte // MaybeRemoveBOM removes a UTF-8 BOM from a []byte when opts.KeepBOM is false
func RemoveBOMIfPresent(content []byte) []byte { func MaybeRemoveBOM(content []byte, opts ConvertOpts) []byte {
if opts.KeepBOM {
return content
}
if len(content) > 2 && bytes.Equal(content[0:3], UTF8BOM) { if len(content) > 2 && bytes.Equal(content[0:3], UTF8BOM) {
return content[3:] return content[3:]
} }

View file

@ -30,15 +30,15 @@ func resetDefaultCharsetsOrder() {
} }
} }
func TestRemoveBOMIfPresent(t *testing.T) { func TestMaybeRemoveBOM(t *testing.T) {
res := RemoveBOMIfPresent([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res := MaybeRemoveBOM([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
res = RemoveBOMIfPresent([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res = MaybeRemoveBOM([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
} }
func TestToUTF8WithErr(t *testing.T) { func TestToUTF8(t *testing.T) {
resetDefaultCharsetsOrder() resetDefaultCharsetsOrder()
var res string var res string
var err error var err error
@ -47,53 +47,53 @@ func TestToUTF8WithErr(t *testing.T) {
// locale, so some conversions might behave differently. For that reason, we don't // locale, so some conversions might behave differently. For that reason, we don't
// depend on particular conversions but in expected behaviors. // depend on particular conversions but in expected behaviors.
res, err = ToUTF8WithErr([]byte{0x41, 0x42, 0x43}) res, err = ToUTF8([]byte{0x41, 0x42, 0x43}, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
assert.Equal(t, "ABC", res) assert.Equal(t, "ABC", res)
// "áéíóú" // "áéíóú"
res, err = ToUTF8WithErr([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res, err = ToUTF8([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res)) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res))
// "áéíóú" // "áéíóú"
res, err = ToUTF8WithErr([]byte{ res, err = ToUTF8([]byte{
0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3,
0xc3, 0xba, 0xc3, 0xba,
}) }, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res)) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res))
res, err = ToUTF8WithErr([]byte{ res, err = ToUTF8([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e, 0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e,
}) }, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
stringMustStartWith(t, "Hola,", res) stringMustStartWith(t, "Hola,", res)
stringMustEndWith(t, "AAA.", res) stringMustEndWith(t, "AAA.", res)
res, err = ToUTF8WithErr([]byte{ res, err = ToUTF8([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e, 0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e,
}) }, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
stringMustStartWith(t, "Hola,", res) stringMustStartWith(t, "Hola,", res)
stringMustEndWith(t, "AAA.", res) stringMustEndWith(t, "AAA.", res)
res, err = ToUTF8WithErr([]byte{ res, err = ToUTF8([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e, 0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73, 0x41, 0x41, 0x41, 0x2e,
}) }, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
stringMustStartWith(t, "Hola,", res) stringMustStartWith(t, "Hola,", res)
stringMustEndWith(t, "AAA.", res) stringMustEndWith(t, "AAA.", res)
// Japanese (Shift-JIS) // Japanese (Shift-JIS)
// 日属秘ぞしちゅ。 // 日属秘ぞしちゅ。
res, err = ToUTF8WithErr([]byte{ res, err = ToUTF8([]byte{
0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82, 0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82,
0xBF, 0x82, 0xE3, 0x81, 0x42, 0xBF, 0x82, 0xE3, 0x81, 0x42,
}) }, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
assert.Equal(t, []byte{ assert.Equal(t, []byte{
0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3, 0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3,
@ -101,7 +101,7 @@ func TestToUTF8WithErr(t *testing.T) {
}, },
[]byte(res)) []byte(res))
res, err = ToUTF8WithErr([]byte{0x00, 0x00, 0x00, 0x00}) res, err = ToUTF8([]byte{0x00, 0x00, 0x00, 0x00}, ConvertOpts{})
assert.NoError(t, err) assert.NoError(t, err)
assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, []byte(res)) assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, []byte(res))
} }
@ -109,22 +109,22 @@ func TestToUTF8WithErr(t *testing.T) {
func TestToUTF8WithFallback(t *testing.T) { func TestToUTF8WithFallback(t *testing.T) {
resetDefaultCharsetsOrder() resetDefaultCharsetsOrder()
// "ABC" // "ABC"
res := ToUTF8WithFallback([]byte{0x41, 0x42, 0x43}) res := ToUTF8WithFallback([]byte{0x41, 0x42, 0x43}, ConvertOpts{})
assert.Equal(t, []byte{0x41, 0x42, 0x43}, res) assert.Equal(t, []byte{0x41, 0x42, 0x43}, res)
// "áéíóú" // "áéíóú"
res = ToUTF8WithFallback([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res = ToUTF8WithFallback([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
// UTF8 BOM + "áéíóú" // UTF8 BOM + "áéíóú"
res = ToUTF8WithFallback([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res = ToUTF8WithFallback([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
// "Hola, así cómo ños" // "Hola, así cómo ños"
res = ToUTF8WithFallback([]byte{ res = ToUTF8WithFallback([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73, 0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73,
}) }, ConvertOpts{})
assert.Equal(t, []byte{ assert.Equal(t, []byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63, 0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63,
0xC3, 0xB3, 0x6D, 0x6F, 0x20, 0xC3, 0xB1, 0x6F, 0x73, 0xC3, 0xB3, 0x6D, 0x6F, 0x20, 0xC3, 0xB1, 0x6F, 0x73,
@ -133,126 +133,65 @@ func TestToUTF8WithFallback(t *testing.T) {
// "Hola, así cómo " // "Hola, así cómo "
minmatch := []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63, 0xC3, 0xB3, 0x6D, 0x6F, 0x20} minmatch := []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63, 0xC3, 0xB3, 0x6D, 0x6F, 0x20}
res = ToUTF8WithFallback([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73}) res = ToUTF8WithFallback([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73}, ConvertOpts{})
// Do not fail for differences in invalid cases, as the library might change the conversion criteria for those // Do not fail for differences in invalid cases, as the library might change the conversion criteria for those
assert.Equal(t, minmatch, res[0:len(minmatch)]) assert.Equal(t, minmatch, res[0:len(minmatch)])
res = ToUTF8WithFallback([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73}) res = ToUTF8WithFallback([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73}, ConvertOpts{})
// Do not fail for differences in invalid cases, as the library might change the conversion criteria for those // Do not fail for differences in invalid cases, as the library might change the conversion criteria for those
assert.Equal(t, minmatch, res[0:len(minmatch)]) assert.Equal(t, minmatch, res[0:len(minmatch)])
// Japanese (Shift-JIS) // Japanese (Shift-JIS)
// "日属秘ぞしちゅ。" // "日属秘ぞしちゅ。"
res = ToUTF8WithFallback([]byte{0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82, 0xBF, 0x82, 0xE3, 0x81, 0x42}) res = ToUTF8WithFallback([]byte{0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82, 0xBF, 0x82, 0xE3, 0x81, 0x42}, ConvertOpts{})
assert.Equal(t, []byte{ assert.Equal(t, []byte{
0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3, 0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3,
0x81, 0x9E, 0xE3, 0x81, 0x97, 0xE3, 0x81, 0xA1, 0xE3, 0x82, 0x85, 0xE3, 0x80, 0x82, 0x81, 0x9E, 0xE3, 0x81, 0x97, 0xE3, 0x81, 0xA1, 0xE3, 0x82, 0x85, 0xE3, 0x80, 0x82,
}, res) }, res)
res = ToUTF8WithFallback([]byte{0x00, 0x00, 0x00, 0x00}) res = ToUTF8WithFallback([]byte{0x00, 0x00, 0x00, 0x00}, ConvertOpts{})
assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, res) assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, res)
} }
func TestToUTF8(t *testing.T) {
resetDefaultCharsetsOrder()
// Note: golang compiler seems so behave differently depending on the current
// locale, so some conversions might behave differently. For that reason, we don't
// depend on particular conversions but in expected behaviors.
res := ToUTF8(string([]byte{0x41, 0x42, 0x43}))
assert.Equal(t, "ABC", res)
// "áéíóú"
res = ToUTF8(string([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}))
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res))
// BOM + "áéíóú"
res = ToUTF8(string([]byte{
0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3,
0xc3, 0xba,
}))
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, []byte(res))
// Latin1
// Hola, así cómo ños
res = ToUTF8(string([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73,
}))
assert.Equal(t, []byte{
0x48, 0x6f, 0x6c, 0x61, 0x2c, 0x20, 0x61, 0x73, 0xc3, 0xad, 0x20, 0x63,
0xc3, 0xb3, 0x6d, 0x6f, 0x20, 0xc3, 0xb1, 0x6f, 0x73,
}, []byte(res))
// Latin1
// Hola, así cómo \x07ños
res = ToUTF8(string([]byte{
0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63,
0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73,
}))
// Hola,
bytesMustStartWith(t, []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C}, []byte(res))
// This test FAILS
// res = ToUTF8("Hola, así cómo \x81ños")
// Do not fail for differences in invalid cases, as the library might change the conversion criteria for those
// assert.Regexp(t, "^Hola, así cómo", res)
// Japanese (Shift-JIS)
// 日属秘ぞしちゅ。
res = ToUTF8(string([]byte{
0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82,
0xBF, 0x82, 0xE3, 0x81, 0x42,
}))
assert.Equal(t, []byte{
0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3,
0x81, 0x9E, 0xE3, 0x81, 0x97, 0xE3, 0x81, 0xA1, 0xE3, 0x82, 0x85, 0xE3, 0x80, 0x82,
},
[]byte(res))
res = ToUTF8("\x00\x00\x00\x00")
assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, []byte(res))
}
func TestToUTF8DropErrors(t *testing.T) { func TestToUTF8DropErrors(t *testing.T) {
resetDefaultCharsetsOrder() resetDefaultCharsetsOrder()
// "ABC" // "ABC"
res := ToUTF8DropErrors([]byte{0x41, 0x42, 0x43}) res := ToUTF8DropErrors([]byte{0x41, 0x42, 0x43}, ConvertOpts{})
assert.Equal(t, []byte{0x41, 0x42, 0x43}, res) assert.Equal(t, []byte{0x41, 0x42, 0x43}, res)
// "áéíóú" // "áéíóú"
res = ToUTF8DropErrors([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res = ToUTF8DropErrors([]byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
// UTF8 BOM + "áéíóú" // UTF8 BOM + "áéíóú"
res = ToUTF8DropErrors([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}) res = ToUTF8DropErrors([]byte{0xef, 0xbb, 0xbf, 0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, ConvertOpts{})
assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res) assert.Equal(t, []byte{0xc3, 0xa1, 0xc3, 0xa9, 0xc3, 0xad, 0xc3, 0xb3, 0xc3, 0xba}, res)
// "Hola, así cómo ños" // "Hola, así cómo ños"
res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73}) res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0xF1, 0x6F, 0x73}, ConvertOpts{})
assert.Equal(t, []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73}, res[:8]) assert.Equal(t, []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73}, res[:8])
assert.Equal(t, []byte{0x73}, res[len(res)-1:]) assert.Equal(t, []byte{0x73}, res[len(res)-1:])
// "Hola, así cómo " // "Hola, así cómo "
minmatch := []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63, 0xC3, 0xB3, 0x6D, 0x6F, 0x20} minmatch := []byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xC3, 0xAD, 0x20, 0x63, 0xC3, 0xB3, 0x6D, 0x6F, 0x20}
res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73}) res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x07, 0xA4, 0x6F, 0x73}, ConvertOpts{})
// Do not fail for differences in invalid cases, as the library might change the conversion criteria for those // Do not fail for differences in invalid cases, as the library might change the conversion criteria for those
assert.Equal(t, minmatch, res[0:len(minmatch)]) assert.Equal(t, minmatch, res[0:len(minmatch)])
res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73}) res = ToUTF8DropErrors([]byte{0x48, 0x6F, 0x6C, 0x61, 0x2C, 0x20, 0x61, 0x73, 0xED, 0x20, 0x63, 0xF3, 0x6D, 0x6F, 0x20, 0x81, 0xA4, 0x6F, 0x73}, ConvertOpts{})
// Do not fail for differences in invalid cases, as the library might change the conversion criteria for those // Do not fail for differences in invalid cases, as the library might change the conversion criteria for those
assert.Equal(t, minmatch, res[0:len(minmatch)]) assert.Equal(t, minmatch, res[0:len(minmatch)])
// Japanese (Shift-JIS) // Japanese (Shift-JIS)
// "日属秘ぞしちゅ。" // "日属秘ぞしちゅ。"
res = ToUTF8DropErrors([]byte{0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82, 0xBF, 0x82, 0xE3, 0x81, 0x42}) res = ToUTF8DropErrors([]byte{0x93, 0xFA, 0x91, 0xAE, 0x94, 0xE9, 0x82, 0xBC, 0x82, 0xB5, 0x82, 0xBF, 0x82, 0xE3, 0x81, 0x42}, ConvertOpts{})
assert.Equal(t, []byte{ assert.Equal(t, []byte{
0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3, 0xE6, 0x97, 0xA5, 0xE5, 0xB1, 0x9E, 0xE7, 0xA7, 0x98, 0xE3,
0x81, 0x9E, 0xE3, 0x81, 0x97, 0xE3, 0x81, 0xA1, 0xE3, 0x82, 0x85, 0xE3, 0x80, 0x82, 0x81, 0x9E, 0xE3, 0x81, 0x97, 0xE3, 0x81, 0xA1, 0xE3, 0x82, 0x85, 0xE3, 0x80, 0x82,
}, res) }, res)
res = ToUTF8DropErrors([]byte{0x00, 0x00, 0x00, 0x00}) res = ToUTF8DropErrors([]byte{0x00, 0x00, 0x00, 0x00}, ConvertOpts{})
assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, res) assert.Equal(t, []byte{0x00, 0x00, 0x00, 0x00}, res)
} }
@ -302,10 +241,6 @@ func stringMustEndWith(t *testing.T, expected, value string) {
assert.Equal(t, expected, value[len(value)-len(expected):]) assert.Equal(t, expected, value[len(value)-len(expected):])
} }
func bytesMustStartWith(t *testing.T, expected, value []byte) {
assert.Equal(t, expected, value[:len(expected)])
}
func TestToUTF8WithFallbackReader(t *testing.T) { func TestToUTF8WithFallbackReader(t *testing.T) {
resetDefaultCharsetsOrder() resetDefaultCharsetsOrder()
@ -317,7 +252,7 @@ func TestToUTF8WithFallbackReader(t *testing.T) {
} }
input = input[:testLen] input = input[:testLen]
input += "// Выключаем" input += "// Выключаем"
rd := ToUTF8WithFallbackReader(bytes.NewReader([]byte(input))) rd := ToUTF8WithFallbackReader(bytes.NewReader([]byte(input)), ConvertOpts{})
r, _ := io.ReadAll(rd) r, _ := io.ReadAll(rd)
assert.EqualValuesf(t, input, string(r), "testing string len=%d", testLen) assert.EqualValuesf(t, input, string(r), "testing string len=%d", testLen)
} }

View file

@ -174,7 +174,7 @@ func (b *Indexer) addUpdate(ctx context.Context, batchWriter git.WriteCloserErro
return batch.Index(id, &RepoIndexerData{ return batch.Index(id, &RepoIndexerData{
RepoID: repo.ID, RepoID: repo.ID,
CommitID: commitSha, CommitID: commitSha,
Content: string(charset.ToUTF8DropErrors(fileContents)), Content: string(charset.ToUTF8DropErrors(fileContents, charset.ConvertOpts{})),
Language: analyze.GetCodeLanguage(update.Filename, fileContents), Language: analyze.GetCodeLanguage(update.Filename, fileContents),
UpdatedAt: time.Now().UTC(), UpdatedAt: time.Now().UTC(),
}) })

View file

@ -135,7 +135,7 @@ func (b *Indexer) addUpdate(ctx context.Context, batchWriter git.WriteCloserErro
Id(id). Id(id).
Doc(map[string]any{ Doc(map[string]any{
"repo_id": repo.ID, "repo_id": repo.ID,
"content": string(charset.ToUTF8DropErrors(fileContents)), "content": string(charset.ToUTF8DropErrors(fileContents, charset.ConvertOpts{})),
"commit_id": sha, "commit_id": sha,
"language": analyze.GetCodeLanguage(update.Filename, fileContents), "language": analyze.GetCodeLanguage(update.Filename, fileContents),
"updated_at": timeutil.TimeStampNow(), "updated_at": timeutil.TimeStampNow(),

View file

@ -400,7 +400,7 @@ func Diff(ctx *context.Context) {
Metas: ctx.Repo.Repository.ComposeMetas(), Metas: ctx.Repo.Repository.ComposeMetas(),
GitRepo: ctx.Repo.GitRepo, GitRepo: ctx.Repo.GitRepo,
Ctx: ctx, Ctx: ctx,
}, template.HTMLEscapeString(string(charset.ToUTF8WithFallback(note.Message)))) }, template.HTMLEscapeString(string(charset.ToUTF8WithFallback(note.Message, charset.ConvertOpts{}))))
if err != nil { if err != nil {
ctx.ServerError("RenderCommitMessage", err) ctx.ServerError("RenderCommitMessage", err)
return return

View file

@ -142,7 +142,7 @@ func setCsvCompareContext(ctx *context.Context) {
return nil, nil, err return nil, nil, err
} }
csvReader, err := csv_module.CreateReaderAndDetermineDelimiter(ctx, charset.ToUTF8WithFallbackReader(reader)) csvReader, err := csv_module.CreateReaderAndDetermineDelimiter(ctx, charset.ToUTF8WithFallbackReader(reader, charset.ConvertOpts{}))
return csvReader, reader, err return csvReader, reader, err
} }

View file

@ -166,8 +166,8 @@ func editFile(ctx *context.Context, isNewFile bool) {
} }
buf = append(buf, d...) buf = append(buf, d...)
if content, err := charset.ToUTF8WithErr(buf); err != nil { if content, err := charset.ToUTF8(buf, charset.ConvertOpts{KeepBOM: true}); err != nil {
log.Error("ToUTF8WithErr: %v", err) log.Error("ToUTF8: %v", err)
ctx.Data["FileContent"] = string(buf) ctx.Data["FileContent"] = string(buf)
} else { } else {
ctx.Data["FileContent"] = content ctx.Data["FileContent"] = content

View file

@ -43,7 +43,7 @@ func RenderFile(ctx *context.Context) {
st := typesniffer.DetectContentType(buf) st := typesniffer.DetectContentType(buf)
isTextFile := st.IsText() isTextFile := st.IsText()
rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc)) rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc), charset.ConvertOpts{})
if markupType := markup.Type(blob.Name()); markupType == "" { if markupType := markup.Type(blob.Name()); markupType == "" {
if isTextFile { if isTextFile {

View file

@ -303,7 +303,7 @@ func LFSFileGet(ctx *context.Context) {
break break
} }
rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc)) rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc), charset.ConvertOpts{})
// Building code view blocks with line number on server side. // Building code view blocks with line number on server side.
escapedContent := &bytes.Buffer{} escapedContent := &bytes.Buffer{}

View file

@ -303,7 +303,7 @@ func renderReadmeFile(ctx *context.Context, subfolder string, readmeFile *git.Tr
return return
} }
rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc)) rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc), charset.ConvertOpts{})
if markupType := markup.Type(readmeFile.Name()); markupType != "" { if markupType := markup.Type(readmeFile.Name()); markupType != "" {
ctx.Data["IsMarkup"] = true ctx.Data["IsMarkup"] = true
@ -449,7 +449,7 @@ func renderFile(ctx *context.Context, entry *git.TreeEntry) {
break break
} }
rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc)) rd := charset.ToUTF8WithFallbackReader(io.MultiReader(bytes.NewReader(buf), dataRc), charset.ConvertOpts{})
shouldRenderSource := ctx.FormString("display") == "source" shouldRenderSource := ctx.FormString("display") == "source"
readmeExist := util.IsReadmeFileName(blob.Name()) readmeExist := util.IsReadmeFileName(blob.Name())

View file

@ -38,8 +38,7 @@
data-url="{{.Repository.Link}}/markup" data-url="{{.Repository.Link}}/markup"
data-context="{{.RepoLink}}" data-context="{{.RepoLink}}"
data-previewable-extensions="{{.PreviewableExtensions}}" data-previewable-extensions="{{.PreviewableExtensions}}"
data-line-wrap-extensions="{{.LineWrapExtensions}}"> data-line-wrap-extensions="{{.LineWrapExtensions}}">{{.FileContent}}</textarea>
{{.FileContent}}</textarea>
<div class="editor-loading is-loading"></div> <div class="editor-loading is-loading"></div>
</div> </div>
<div class="ui bottom attached tab segment markup" data-tab="preview"> <div class="ui bottom attached tab segment markup" data-tab="preview">

View file

@ -144,7 +144,7 @@ func readSQLFromFile(version string) (string, error) {
if err != nil { if err != nil {
return "", err return "", err
} }
return string(charset.RemoveBOMIfPresent(bytes)), nil return string(charset.MaybeRemoveBOM(bytes, charset.ConvertOpts{})), nil
} }
func restoreOldDB(t *testing.T, version string) bool { func restoreOldDB(t *testing.T, version string) bool {

View file

@ -114,7 +114,7 @@ export async function createMonaco(textarea, filename, editorOpts) {
const model = editor.getModel(); const model = editor.getModel();
model.onDidChangeContent(() => { model.onDidChangeContent(() => {
textarea.value = editor.getValue(); textarea.value = editor.getValue({preserveBOM: true});
textarea.dispatchEvent(new Event('change')); // seems to be needed for jquery-are-you-sure textarea.dispatchEvent(new Event('change')); // seems to be needed for jquery-are-you-sure
}); });