lib/encoder: add EncodeRaw
This commit is contained in:
URenko
Nick Craig-Wood
parent
846c1aeed0
commit
1b9217bc78
3 changed files with 46 additions and 36 deletions
|
|
@ -35,34 +35,35 @@ const (
|
|||
|
||||
// Possible flags for the MultiEncoder
|
||||
const (
|
||||
EncodeZero MultiEncoder = 0 // NUL(0x00)
|
||||
EncodeSlash MultiEncoder = 1 << iota // /
|
||||
EncodeLtGt // <>
|
||||
EncodeDoubleQuote // "
|
||||
EncodeSingleQuote // '
|
||||
EncodeBackQuote // `
|
||||
EncodeDollar // $
|
||||
EncodeColon // :
|
||||
EncodeQuestion // ?
|
||||
EncodeAsterisk // *
|
||||
EncodePipe // |
|
||||
EncodeHash // #
|
||||
EncodePercent // %
|
||||
EncodeBackSlash // \
|
||||
EncodeCrLf // CR(0x0D), LF(0x0A)
|
||||
EncodeDel // DEL(0x7F)
|
||||
EncodeCtl // CTRL(0x01-0x1F)
|
||||
EncodeLeftSpace // Leading SPACE
|
||||
EncodeLeftPeriod // Leading .
|
||||
EncodeLeftTilde // Leading ~
|
||||
EncodeLeftCrLfHtVt // Leading CR LF HT VT
|
||||
EncodeRightSpace // Trailing SPACE
|
||||
EncodeRightPeriod // Trailing .
|
||||
EncodeRightCrLfHtVt // Trailing CR LF HT VT
|
||||
EncodeInvalidUtf8 // Invalid UTF-8 bytes
|
||||
EncodeDot // . and .. names
|
||||
EncodeSquareBracket // []
|
||||
EncodeSemicolon // ;
|
||||
EncodeZero MultiEncoder = 0 // NUL(0x00)
|
||||
EncodeRaw MultiEncoder = 1 << (iota - 1)
|
||||
EncodeSlash // /
|
||||
EncodeLtGt // <>
|
||||
EncodeDoubleQuote // "
|
||||
EncodeSingleQuote // '
|
||||
EncodeBackQuote // `
|
||||
EncodeDollar // $
|
||||
EncodeColon // :
|
||||
EncodeQuestion // ?
|
||||
EncodeAsterisk // *
|
||||
EncodePipe // |
|
||||
EncodeHash // #
|
||||
EncodePercent // %
|
||||
EncodeBackSlash // \
|
||||
EncodeCrLf // CR(0x0D), LF(0x0A)
|
||||
EncodeDel // DEL(0x7F)
|
||||
EncodeCtl // CTRL(0x01-0x1F)
|
||||
EncodeLeftSpace // Leading SPACE
|
||||
EncodeLeftPeriod // Leading .
|
||||
EncodeLeftTilde // Leading ~
|
||||
EncodeLeftCrLfHtVt // Leading CR LF HT VT
|
||||
EncodeRightSpace // Trailing SPACE
|
||||
EncodeRightPeriod // Trailing .
|
||||
EncodeRightCrLfHtVt // Trailing CR LF HT VT
|
||||
EncodeInvalidUtf8 // Invalid UTF-8 bytes
|
||||
EncodeDot // . and .. names
|
||||
EncodeSquareBracket // []
|
||||
EncodeSemicolon // ;
|
||||
|
||||
// Synthetic
|
||||
EncodeWin = EncodeColon | EncodeQuestion | EncodeDoubleQuote | EncodeAsterisk | EncodeLtGt | EncodePipe // :?"*<>|
|
||||
|
|
@ -117,6 +118,7 @@ func alias(name string, mask MultiEncoder) {
|
|||
}
|
||||
|
||||
func init() {
|
||||
alias("Raw", EncodeRaw)
|
||||
alias("None", EncodeZero)
|
||||
alias("Slash", EncodeSlash)
|
||||
alias("LtGt", EncodeLtGt)
|
||||
|
|
@ -214,6 +216,10 @@ func (mask *MultiEncoder) Scan(s fmt.ScanState, ch rune) error {
|
|||
// Encode takes a raw name and substitutes any reserved characters and
|
||||
// patterns in it
|
||||
func (mask MultiEncoder) Encode(in string) string {
|
||||
if mask == EncodeRaw {
|
||||
return in
|
||||
}
|
||||
|
||||
if in == "" {
|
||||
return ""
|
||||
}
|
||||
|
|
@ -671,6 +677,10 @@ func (mask MultiEncoder) Encode(in string) string {
|
|||
|
||||
// Decode takes a name and undoes any substitutions made by Encode
|
||||
func (mask MultiEncoder) Decode(in string) string {
|
||||
if mask == EncodeRaw {
|
||||
return in
|
||||
}
|
||||
|
||||
if mask.Has(EncodeDot) {
|
||||
switch in {
|
||||
case "．":
|
||||
|
|
|
|||
|
|
@ -22,7 +22,7 @@ func TestEncodeString(t *testing.T) {
|
|||
mask MultiEncoder
|
||||
want string
|
||||
}{
|
||||
{0, "None"},
|
||||
{EncodeRaw, "Raw"},
|
||||
{EncodeZero, "None"},
|
||||
{EncodeDoubleQuote, "DoubleQuote"},
|
||||
{EncodeDot, "Dot"},
|
||||
|
|
@ -44,7 +44,7 @@ func TestEncodeSet(t *testing.T) {
|
|||
wantErr bool
|
||||
}{
|
||||
{"", 0, true},
|
||||
{"None", 0, false},
|
||||
{"Raw", EncodeRaw, false},
|
||||
{"None", EncodeZero, false},
|
||||
{"DoubleQuote", EncodeDoubleQuote, false},
|
||||
{"Dot", EncodeDot, false},
|
||||
|
|
@ -178,7 +178,7 @@ func TestEncodeInvalidUnicode(t *testing.T) {
|
|||
func TestEncodeDot(t *testing.T) {
|
||||
for i, tc := range []testCase{
|
||||
{
|
||||
mask: 0,
|
||||
mask: EncodeZero,
|
||||
in: ".",
|
||||
out: ".",
|
||||
}, {
|
||||
|
|
@ -186,7 +186,7 @@ func TestEncodeDot(t *testing.T) {
|
|||
in: ".",
|
||||
out: "．",
|
||||
}, {
|
||||
mask: 0,
|
||||
mask: EncodeZero,
|
||||
in: "..",
|
||||
out: "..",
|
||||
}, {
|
||||
|
|
@ -224,7 +224,7 @@ func TestDecodeHalf(t *testing.T) {
|
|||
in: "‛",
|
||||
out: "‛",
|
||||
}, {
|
||||
mask: 0,
|
||||
mask: EncodeZero,
|
||||
in: "‛‛",
|
||||
out: "‛",
|
||||
}, {
|
||||
|
|
|
|||
|
|
@ -230,7 +230,7 @@ func main() {
|
|||
}
|
||||
in, out := buildTestString(
|
||||
[]mapping{getMapping(m.mask)}, // pick
|
||||
[]mapping{getMapping(0)}, // quote
|
||||
[]mapping{getMapping(encoder.EncodeZero)}, // quote
|
||||
printables, fullwidthPrintables, encodables, encoded, greek) // fill
|
||||
fatalW(fmt.Fprintf(fd, `{ // %d
|
||||
mask: %s,
|
||||
|
|
@ -262,7 +262,7 @@ var testCasesSingleEdge = []testCase{
|
|||
for idx, orig := range e.orig {
|
||||
replace := e.replace[idx]
|
||||
pairs := buildEdgeTestString(
|
||||
[]edge{e}, []mapping{getMapping(0), getMapping(m.mask)}, // quote
|
||||
[]edge{e}, []mapping{getMapping(encoder.EncodeZero), getMapping(m.mask)}, // quote
|
||||
[][]rune{printables, fullwidthPrintables, encodables, encoded, greek}, // fill
|
||||
func(rIn, rOut []rune, quoteOut []bool, testMappings []mapping) (out []stringPair) {
|
||||
testL := len(rIn)
|
||||
|
|
@ -386,7 +386,7 @@ var testCasesDoubleEdge = []testCase{
|
|||
orig, replace := e1.orig[0], e1.replace[0]
|
||||
edges := []edge{e1, e2}
|
||||
pairs := buildEdgeTestString(
|
||||
edges, []mapping{getMapping(0), getMapping(m.mask)}, // quote
|
||||
edges, []mapping{getMapping(encoder.EncodeZero), getMapping(m.mask)}, // quote
|
||||
[][]rune{printables, fullwidthPrintables, encodables, encoded, greek}, // fill
|
||||
func(rIn, rOut []rune, quoteOut []bool, testMappings []mapping) (out []stringPair) {
|
||||
testL := len(rIn)
|
||||
|
|
|
|||
Loading…
Reference in a new issue