zlog/internal/json/string.go

181 lines
4.7 KiB
Go

package json
import "unicode/utf8"
const hex = "0123456789abcdef"
func AppendStrings(dst []byte, vals []string) []byte {
if len(vals) == 0 {
return append(dst, '[', ']')
}
dst = append(dst, '[')
dst = AppendString(dst, vals[0])
if len(vals) > 1 {
for _, val := range vals[1:] {
dst = AppendString(append(dst, ','), val)
}
}
dst = append(dst, ']')
return dst
}
// AppendString encodes the input string to json and appends
// the encoded string to the input byte slice.
//
// The operation loops though each byte in the string looking
// for characters that need json or utf8 encoding. If the string
// does not need encoding, then the string is appended in it's
// entirety to the byte slice.
// If we encounter a byte that does need encoding, switch up
// the operation and perform a byte-by-byte read-encode-append.
func AppendString(dst []byte, s string) []byte {
// Start with a double quote.
dst = append(dst, '"')
// Loop through each character in the string.
for i := 0; i < len(s); i++ {
// Check if the character needs encoding. Control characters, slashes,
// and the double quote need json encoding. Bytes above the ascii
// boundary needs utf8 encoding.
if s[i] < 0x20 || s[i] > 0x7e || s[i] == '\\' || s[i] == '"' {
// We encountered a character that needs to be encoded. Switch
// to complex version of the algorithm.
dst = appendStringComplex(dst, s, i)
return append(dst, '"')
}
}
// The string has no need for encoding an therefore is directly
// appended to the byte slice.
dst = append(dst, s...)
// End with a double quote
return append(dst, '"')
}
// appendStringComplex is used by appendString to take over an in
// progress JSON string encoding that encountered a character that needs
// to be encoded.
func appendStringComplex(dst []byte, s string, i int) []byte {
start := 0
for i < len(s) {
b := s[i]
if b >= utf8.RuneSelf {
r, size := utf8.DecodeRuneInString(s[i:])
if r == utf8.RuneError && size == 1 {
// In case of error, first append previous simple characters to
// the byte slice if any and append a remplacement character code
// in place of the invalid sequence.
if start < i {
dst = append(dst, s[start:i]...)
}
dst = append(dst, `\ufffd`...)
i += size
start = i
continue
}
i += size
continue
}
if b >= 0x20 && b <= 0x7e && b != '\\' && b != '"' {
i++
continue
}
// We encountered a character that needs to be encoded.
// Let's append the previous simple characters to the byte slice
// and switch our operation to read and encode the remainder
// characters byte-by-byte.
if start < i {
dst = append(dst, s[start:i]...)
}
switch b {
case '"', '\\':
dst = append(dst, '\\', b)
case '\b':
dst = append(dst, '\\', 'b')
case '\f':
dst = append(dst, '\\', 'f')
case '\n':
dst = append(dst, '\\', 'n')
case '\r':
dst = append(dst, '\\', 'r')
case '\t':
dst = append(dst, '\\', 't')
default:
dst = append(dst, '\\', 'u', '0', '0', hex[b>>4], hex[b&0xF])
}
i++
start = i
}
if start < len(s) {
dst = append(dst, s[start:]...)
}
return dst
}
// AppendBytes is a mirror of appendString with []byte arg
func AppendBytes(dst, s []byte) []byte {
dst = append(dst, '"')
for i := 0; i < len(s); i++ {
if s[i] < 0x20 || s[i] > 0x7e || s[i] == '\\' || s[i] == '"' {
dst = appendBytesComplex(dst, s, i)
return append(dst, '"')
}
}
dst = append(dst, s...)
return append(dst, '"')
}
// appendBytesComplex is a mirror of the appendStringComplex
// with []byte arg
func appendBytesComplex(dst, s []byte, i int) []byte {
start := 0
for i < len(s) {
b := s[i]
if b >= utf8.RuneSelf {
r, size := utf8.DecodeRune(s[i:])
if r == utf8.RuneError && size == 1 {
if start < i {
dst = append(dst, s[start:i]...)
}
dst = append(dst, `\ufffd`...)
i += size
start = i
continue
}
i += size
continue
}
if b >= 0x20 && b <= 0x7e && b != '\\' && b != '"' {
i++
continue
}
// We encountered a character that needs to be encoded.
// Let's append the previous simple characters to the byte slice
// and switch our operation to read and encode the remainder
// characters byte-by-byte.
if start < i {
dst = append(dst, s[start:i]...)
}
switch b {
case '"', '\\':
dst = append(dst, '\\', b)
case '\b':
dst = append(dst, '\\', 'b')
case '\f':
dst = append(dst, '\\', 'f')
case '\n':
dst = append(dst, '\\', 'n')
case '\r':
dst = append(dst, '\\', 'r')
case '\t':
dst = append(dst, '\\', 't')
default:
dst = append(dst, '\\', 'u', '0', '0', hex[b>>4], hex[b&0xF])
}
i++
start = i
}
if start < len(s) {
dst = append(dst, s[start:]...)
}
return dst
}