package bytes import ( "internal/bytealg" "unicode" "unicode/utf8" ) // It splits the slice s around each instance of one or more consecutive white space // characters, as defined by [unicode.IsSpace], returning a slice of subslices of s or an // empty slice if s contains only white space. Every element of the returned slice is // non-empty. Unlike [Split], leading and trailing runs of white space characters // are discarded. func Fields(s []byte) [][]byte { asciiSpace := _asciiSpace() n := 0 wasSpace := 1 setBits := uint8(0) for i := 0; i < len(s); i++ { r := s[i] setBits |= r isSpace := int32(asciiSpace[r]) n += wasSpace & ^isSpace wasSpace = isSpace } if setBits >= utf8.RuneSelf { t := unicode.NewTables() return FieldsFunc(s, t.IsSpace) } a := [][]byte{:n} na := 0 fieldStart := 0 i := 0 for i < len(s) && asciiSpace[s[i]] != 0 { i++ } fieldStart = i for i < len(s) { if asciiSpace[s[i]] == 0 { i++ continue } a[na] = s[fieldStart:i:i] na++ i++ for i < len(s) && asciiSpace[s[i]] != 0 { i++ } fieldStart = i } if fieldStart < len(s) { // Last field might end at EOF. a[na] = s[fieldStart:len(s):len(s)] } return a } // FieldsFunc interprets s as a sequence of UTF-8-encoded code points. // It splits the slice s at each run of code points c satisfying f(c) and // returns a slice of subslices of s. If all code points in s satisfy f(c), or // len(s) == 0, an empty slice is returned. Every element of the returned slice is // non-empty. Unlike [SplitFunc], leading and trailing runs of code points // satisfying f(c) are discarded. // // FieldsFunc makes no guarantees about the order in which it calls f(c) // and assumes that f always returns the same value for a given c. func FieldsFunc(s []byte, f func(rune) bool) [][]byte { // A span is used to record a slice of s of the form s[start:end]. // The start index is inclusive and the end index is exclusive. type span struct { start int32 end int32 } spans := []span{:0:32} // Find the field start and end indices. // Doing this in a separate pass (rather than slicing the string s // and collecting the result substrings right away) is significantly // more efficient, possibly due to cache effects. start := -1 // valid span start if >= 0 for i := 0; i < len(s); { size := 1 r := rune(s[i]) if r >= utf8.RuneSelf { r, size = utf8.DecodeRune(s[i:]) } if f(r) { if start >= 0 { spans = append(spans, span{start, i}) start = -1 } } else { if start < 0 { start = i } } i += size } // Last field might end at EOF. if start >= 0 { spans = append(spans, span{start, len(s)}) } // Create subslices from recorded field indices. a := [][]byte{:len(spans)} for i, span := range spans { a[i] = s[span.start:span.end:span.end] } return a } // Join concatenates the elements of s to create a new byte slice. The separator // sep is placed between elements in the resulting slice. func Join(s [][]byte, sep []byte) []byte { if len(s) == 0 { return []byte{} } if len(s) == 1 { // Just return a copy. return append([]byte(nil), s[0]...) } var n int32 if len(sep) > 0 { if len(sep) >= maxInt/(len(s)-1) { panic("bytes: Join output length overflow") } n += len(sep) * (len(s) - 1) } for _, v := range s { if len(v) > maxInt-n { panic("bytes: Join output length overflow") } n += len(v) } b := bytealg.MakeNoZero(n)[:n:n] bp := copy(b, s[0]) for _, v := range s[1:] { bp += copy(b[bp:], sep) bp += copy(b[bp:], v) } return b } // HasPrefix reports whether the byte slice s begins with prefix. func HasPrefix(s, prefix []byte) bool { return len(s) >= len(prefix) && Equal(s[:len(prefix)], prefix) } // HasSuffix reports whether the byte slice s ends with suffix. func HasSuffix(s, suffix []byte) bool { return len(s) >= len(suffix) && Equal(s[len(s)-len(suffix):], suffix) } // Map returns a copy of the byte slice s with all its characters modified // according to the mapping function. If mapping returns a negative value, the character is // dropped from the byte slice with no replacement. The characters in s and the // output are interpreted as UTF-8-encoded code points. func Map(mapping func(r rune) rune, s []byte) []byte { // In the worst case, the slice can grow when mapped, making // things unpleasant. But it's so rare we barge in assuming it's // fine. It could also shrink but that falls out naturally. b := []byte{:0:len(s)} for i := 0; i < len(s); { wid := 1 r := rune(s[i]) if r >= utf8.RuneSelf { r, wid = utf8.DecodeRune(s[i:]) } r = mapping(r) if r >= 0 { b = utf8.AppendRune(b, r) } i += wid } return b } // Despite being an exported symbol, // Repeat is linknamed by widely used packages. // Notable members of the hall of shame include: // - gitee.com/quant1x/num // // Do not remove or change the type signature. // See go.dev/issue/67401. // // Note that this comment is not part of the doc comment. // // linkname Repeat (stripped for moxie) // ToUpper returns a copy of the byte slice s with all Unicode letters mapped to // their upper case. func ToUpper(s []byte) []byte { isASCII, hasLower := true, false for i := 0; i < len(s); i++ { c := s[i] if c >= utf8.RuneSelf { isASCII = false break } hasLower = hasLower || ('a' <= c && c <= 'z') } if isASCII { // optimize for ASCII-only byte slices. if !hasLower { // Just return a copy. return append([]byte(""), s...) } b := bytealg.MakeNoZero(len(s))[:len(s):len(s)] for i := 0; i < len(s); i++ { c := s[i] if 'a' <= c && c <= 'z' { c -= 'a' - 'A' } b[i] = c } return b } t := unicode.NewTables() return Map(t.ToUpper, s) } // ToLower returns a copy of the byte slice s with all Unicode letters mapped to // their lower case. func ToLower(s []byte) []byte { isASCII, hasUpper := true, false for i := 0; i < len(s); i++ { c := s[i] if c >= utf8.RuneSelf { isASCII = false break } hasUpper = hasUpper || ('A' <= c && c <= 'Z') } if isASCII { // optimize for ASCII-only byte slices. if !hasUpper { return append([]byte(""), s...) } b := bytealg.MakeNoZero(len(s))[:len(s):len(s)] for i := 0; i < len(s); i++ { c := s[i] if 'A' <= c && c <= 'Z' { c += 'a' - 'A' } b[i] = c } return b } t := unicode.NewTables() return Map(t.ToLower, s) } // ToTitle treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their title case. func ToTitle(s []byte) []byte { t := unicode.NewTables() return Map(t.ToTitle, s) } // ToUpperSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their // upper case, giving priority to the special casing rules. func ToUpperSpecial(c unicode.SpecialCase, s []byte) []byte { return Map(c.ToUpper, s) } // ToLowerSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their // lower case, giving priority to the special casing rules. func ToLowerSpecial(c unicode.SpecialCase, s []byte) []byte { return Map(c.ToLower, s) } // ToTitleSpecial treats s as UTF-8-encoded bytes and returns a copy with all the Unicode letters mapped to their // title case, giving priority to the special casing rules. func ToTitleSpecial(c unicode.SpecialCase, s []byte) []byte { return Map(c.ToTitle, s) } // ToValidUTF8 treats s as UTF-8-encoded bytes and returns a copy with each run of bytes // representing invalid UTF-8 replaced with the bytes in replacement, which may be empty. func ToValidUTF8(s, replacement []byte) []byte { b := []byte{:0:len(s)+len(replacement)} invalid := false // previous byte was from an invalid UTF-8 sequence for i := 0; i < len(s); { c := s[i] if c < utf8.RuneSelf { i++ invalid = false b = append(b, c) continue } _, wid := utf8.DecodeRune(s[i:]) if wid == 1 { i++ if !invalid { invalid = true b = append(b, replacement...) } continue } invalid = false b = append(b, s[i:i+wid]...) i += wid } return b } // isSeparator reports whether the rune could mark a word boundary. // TODO: update when package unicode captures more of the properties. func isSeparator(r rune) bool { // ASCII alphanumerics and underscore are not separators if r <= 0x7F { switch { case '0' <= r && r <= '9': return false case 'a' <= r && r <= 'z': return false case 'A' <= r && r <= 'Z': return false case r == '_': return false } return true } // Letters and digits are not separators t := unicode.NewTables() if t.IsLetter(r) || t.IsDigit(r) { return false } // Otherwise, all we can do for now is treat spaces as separators. return t.IsSpace(r) } // Title treats s as UTF-8-encoded bytes and returns a copy with all Unicode letters that begin // words mapped to their title case. // // Deprecated: The rule Title uses for word boundaries does not handle Unicode // punctuation properly. Use golang.org/x/text/cases instead. func Title(s []byte) []byte { // Use a closure here to remember state. // Hackish but effective. Depends on Map scanning in order and calling // the closure once per rune. t := unicode.NewTables() prev := ' ' return Map( func(r rune) rune { if isSeparator(prev) { prev = r return t.ToTitle(r) } prev = r return r }, s) } // TrimLeftFunc treats s as UTF-8-encoded bytes and returns a subslice of s by slicing off // all leading UTF-8-encoded code points c that satisfy f(c). func TrimLeftFunc(s []byte, f func(r rune) bool) []byte { i := indexFunc(s, f, false) if i == -1 { return nil } return s[i:] } // TrimRightFunc returns a subslice of s by slicing off all trailing // UTF-8-encoded code points c that satisfy f(c). func TrimRightFunc(s []byte, f func(r rune) bool) []byte { i := lastIndexFunc(s, f, false) if i >= 0 && s[i] >= utf8.RuneSelf { _, wid := utf8.DecodeRune(s[i:]) i += wid } else { i++ } return s[0:i] } // TrimFunc returns a subslice of s by slicing off all leading and trailing // UTF-8-encoded code points c that satisfy f(c). func TrimFunc(s []byte, f func(r rune) bool) []byte { return TrimRightFunc(TrimLeftFunc(s, f), f) } // TrimPrefix returns s without the provided leading prefix string. // If s doesn't start with prefix, s is returned unchanged. func TrimPrefix(s, prefix []byte) []byte { if HasPrefix(s, prefix) { return s[len(prefix):] } return s } // TrimSuffix returns s without the provided trailing suffix string. // If s doesn't end with suffix, s is returned unchanged. func TrimSuffix(s, suffix []byte) []byte { if HasSuffix(s, suffix) { return s[:len(s)-len(suffix)] } return s } // IndexFunc interprets s as a sequence of UTF-8-encoded code points. // It returns the byte index in s of the first Unicode // code point satisfying f(c), or -1 if none do. func IndexFunc(s []byte, f func(r rune) bool) int32 { return indexFunc(s, f, true) } // LastIndexFunc interprets s as a sequence of UTF-8-encoded code points. // It returns the byte index in s of the last Unicode // code point satisfying f(c), or -1 if none do. func LastIndexFunc(s []byte, f func(r rune) bool) int32 { return lastIndexFunc(s, f, true) } // indexFunc is the same as IndexFunc except that if // truth==false, the sense of the predicate function is // inverted. func indexFunc(s []byte, f func(r rune) bool, truth bool) int32 { start := 0 for start < len(s) { wid := 1 r := rune(s[start]) if r >= utf8.RuneSelf { r, wid = utf8.DecodeRune(s[start:]) } if f(r) == truth { return start } start += wid } return -1 } // lastIndexFunc is the same as LastIndexFunc except that if // truth==false, the sense of the predicate function is // inverted. func lastIndexFunc(s []byte, f func(r rune) bool, truth bool) int32 { for i := len(s); i > 0; { r, size := rune(s[i-1]), 1 if r >= utf8.RuneSelf { r, size = utf8.DecodeLastRune(s[0:i]) } i -= size if f(r) == truth { return i } } return -1 } // asciiSet is a 32-byte value, where each bit represents the presence of a // given ASCII character in the set. The 128-bits of the lower 16 bytes, // starting with the least-significant bit of the lowest word to the // most-significant bit of the highest word, map to the full range of all // 128 ASCII characters. The 128-bits of the upper 16 bytes will be zeroed, // ensuring that any non-ASCII character will be reported as not in the set. // This allocates a total of 32 bytes even though the upper half // is unused to avoid bounds checks in asciiSet.contains. type asciiSet [8]uint32 // makeASCIISet creates a set of ASCII characters and reports whether all // characters in chars are ASCII. func makeASCIISet(chars []byte) (as asciiSet, ok bool) { for i := 0; i < len(chars); i++ { c := chars[i] if c >= utf8.RuneSelf { return as, false } as[c/32] |= 1 << (c % 32) } return as, true } // contains reports whether c is inside the set. func (as *asciiSet) contains(c byte) bool { return (as[c/32] & (1 << (c % 32))) != 0 } // containsRune is a simplified version of strings.ContainsRune // to avoid importing the strings package. // We avoid bytes.ContainsRune to avoid allocating a temporary copy of s. func containsRune(s []byte, r rune) bool { for i := 0; i < len(s); { c, w := utf8.DecodeRune(s[i:]) if c == r { return true } i += w } return false } // Trim returns a subslice of s by slicing off all leading and // trailing UTF-8-encoded code points contained in cutset. func Trim(s []byte, cutset []byte) []byte { if len(s) == 0 { // This is what we've historically done. return nil } if cutset == "" { return s } if len(cutset) == 1 && cutset[0] < utf8.RuneSelf { return trimLeftByte(trimRightByte(s, cutset[0]), cutset[0]) } if as, ok := makeASCIISet(cutset); ok { return trimLeftASCII(trimRightASCII(s, &as), &as) } return trimLeftUnicode(trimRightUnicode(s, cutset), cutset) } // TrimLeft returns a subslice of s by slicing off all leading // UTF-8-encoded code points contained in cutset. func TrimLeft(s []byte, cutset []byte) []byte { if len(s) == 0 { // This is what we've historically done. return nil } if cutset == "" { return s } if len(cutset) == 1 && cutset[0] < utf8.RuneSelf { return trimLeftByte(s, cutset[0]) } if as, ok := makeASCIISet(cutset); ok { return trimLeftASCII(s, &as) } return trimLeftUnicode(s, cutset) } func trimLeftByte(s []byte, c byte) []byte { for len(s) > 0 && s[0] == c { s = s[1:] } if len(s) == 0 { // This is what we've historically done. return nil } return s } func trimLeftASCII(s []byte, as *asciiSet) []byte { for len(s) > 0 { if !as.contains(s[0]) { break } s = s[1:] } if len(s) == 0 { // This is what we've historically done. return nil } return s } func trimLeftUnicode(s []byte, cutset []byte) []byte { for len(s) > 0 { r, n := rune(s[0]), 1 if r >= utf8.RuneSelf { r, n = utf8.DecodeRune(s) } if !containsRune(cutset, r) { break } s = s[n:] } if len(s) == 0 { // This is what we've historically done. return nil } return s } // TrimRight returns a subslice of s by slicing off all trailing // UTF-8-encoded code points that are contained in cutset. func TrimRight(s []byte, cutset []byte) []byte { if len(s) == 0 || cutset == "" { return s } if len(cutset) == 1 && cutset[0] < utf8.RuneSelf { return trimRightByte(s, cutset[0]) } if as, ok := makeASCIISet(cutset); ok { return trimRightASCII(s, &as) } return trimRightUnicode(s, cutset) } func trimRightByte(s []byte, c byte) []byte { for len(s) > 0 && s[len(s)-1] == c { s = s[:len(s)-1] } return s } func trimRightASCII(s []byte, as *asciiSet) []byte { for len(s) > 0 { if !as.contains(s[len(s)-1]) { break } s = s[:len(s)-1] } return s } func trimRightUnicode(s []byte, cutset []byte) []byte { for len(s) > 0 { r, n := rune(s[len(s)-1]), 1 if r >= utf8.RuneSelf { r, n = utf8.DecodeLastRune(s) } if !containsRune(cutset, r) { break } s = s[:len(s)-n] } return s } // TrimSpace returns a subslice of s by slicing off all leading and // trailing white space, as defined by Unicode. func TrimSpace(s []byte) []byte { t := unicode.NewTables() asciiSpace := _asciiSpace() start := 0 for ; start < len(s); start++ { c := s[start] if c >= utf8.RuneSelf { return TrimFunc(s[start:], t.IsSpace) } if asciiSpace[c] == 0 { break } } stop := len(s) for ; stop > start; stop-- { c := s[stop-1] if c >= utf8.RuneSelf { return TrimFunc(s[start:stop], t.IsSpace) } if asciiSpace[c] == 0 { break } } // At this point s[start:stop] starts and ends with an ASCII // non-space bytes, so we're done. Non-ASCII cases have already // been handled above. if start == stop { // Special case to preserve previous TrimLeftFunc behavior, // returning nil instead of empty slice if all spaces. return nil } return s[start:stop] } // Runes interprets s as a sequence of UTF-8-encoded code points. // It returns a slice of runes (Unicode code points) equivalent to s. func Runes(s []byte) []rune { t := []rune{:utf8.RuneCount(s)} i := 0 for len(s) > 0 { r, l := utf8.DecodeRune(s) t[i] = r i++ s = s[l:] } return t } // Replace returns a copy of the slice s with the first n // non-overlapping instances of old replaced by new. // If old is empty, it matches at the beginning of the slice // and after each UTF-8 sequence, yielding up to k+1 replacements // for a k-rune slice. // If n < 0, there is no limit on the number of replacements. func Replace(s, old, new []byte, n int32) []byte { m := 0 if n != 0 { // Compute number of replacements. m = Count(s, old) } if m == 0 { // Just return a copy. return append([]byte(nil), s...) } if n < 0 || m < n { n = m } // Apply replacements to buffer. t := []byte{:len(s)+n*(len(new)-len(old))} w := 0 start := 0 if len(old) > 0 { for range n { j := start + Index(s[start:], old) w += copy(t[w:], s[start:j]) w += copy(t[w:], new) start = j + len(old) } } else { // len(old) == 0 w += copy(t[w:], new) for range n - 1 { _, wid := utf8.DecodeRune(s[start:]) j := start + wid w += copy(t[w:], s[start:j]) w += copy(t[w:], new) start = j } } w += copy(t[w:], s[start:]) return t[0:w] } // ReplaceAll returns a copy of the slice s with all // non-overlapping instances of old replaced by new. // If old is empty, it matches at the beginning of the slice // and after each UTF-8 sequence, yielding up to k+1 replacements // for a k-rune slice. func ReplaceAll(s, old, new []byte) []byte { return Replace(s, old, new, -1) } // EqualFold reports whether s and t, interpreted as UTF-8 strings, // are equal under simple Unicode case-folding, which is a more general // form of case-insensitivity. func EqualFold(s, t []byte) bool { // ASCII fast path i := 0 for n := min(len(s), len(t)); i < n; i++ { sr := s[i] tr := t[i] if sr|tr >= utf8.RuneSelf { goto hasUnicode } // Easy case. if tr == sr { continue } // Make sr < tr to simplify what follows. if tr < sr { tr, sr = sr, tr } // ASCII only, sr/tr must be upper/lower case if 'A' <= sr && sr <= 'Z' && tr == sr+'a'-'A' { continue } return false } // Check if we've exhausted both strings. return len(s) == len(t) hasUnicode: s = s[i:] t = t[i:] for len(s) != 0 && len(t) != 0 { // Extract first rune from each. var sr, tr rune if s[0] < utf8.RuneSelf { sr, s = rune(s[0]), s[1:] } else { r, size := utf8.DecodeRune(s) sr, s = r, s[size:] } if t[0] < utf8.RuneSelf { tr, t = rune(t[0]), t[1:] } else { r, size := utf8.DecodeRune(t) tr, t = r, t[size:] } // If they match, keep going; if not, return false. // Easy case. if tr == sr { continue } // Make sr < tr to simplify what follows. if tr < sr { tr, sr = sr, tr } // Fast check for ASCII. if tr < utf8.RuneSelf { // ASCII only, sr/tr must be upper/lower case if 'A' <= sr && sr <= 'Z' && tr == sr+'a'-'A' { continue } return false } // General case. SimpleFold(x) returns the next equivalent rune > x // or wraps around to smaller values. tb := unicode.NewTables() r := tb.SimpleFold(sr) for r != sr && r < tr { r = tb.SimpleFold(r) } if r == tr { continue } return false } // One string is empty. Are both? return len(s) == len(t) } // Index returns the index of the first instance of sep in s, or -1 if sep is not present in s. func Index(s, sep []byte) int32 { n := len(sep) if n == 0 { return 0 } if n == 1 { return IndexByte(s, sep[0]) } if n == len(s) { if Equal(sep, s) { return 0 } return -1 } if n > len(s) { return -1 } return indexBrute(s, sep, n) } func indexBrute(s, sep []byte, n int32) int32 { c0 := sep[0] c1 := sep[1] i := int32(0) fails := int32(0) t := len(s) - n + 1 for i < t { if s[i] != c0 { o := IndexByte(s[i+1:t], c0) if o < 0 { break } i += o + 1 } if s[i+1] == c1 && Equal(s[i:i+n], sep) { return i } i++ fails++ if fails >= 4+i>>4 && i < t { j := bytealg.IndexRabinKarp(s[i:], sep) if j < 0 { return -1 } return i + j } } return -1 } // Cut slices s around the first instance of sep, // returning the text before and after sep. // The found result reports whether sep appears in s. // If sep does not appear in s, cut returns s, nil, false. // // Cut returns slices of the original slice s, not copies. func Cut(s, sep []byte) (before, after []byte, found bool) { if i := Index(s, sep); i >= 0 { return s[:i], s[i+len(sep):], true } return s, nil, false } // Clone returns a copy of b[:len(b)]. // The result may have additional unused capacity. // Clone(nil) returns nil. func Clone(b []byte) []byte { if b == nil { return nil } return append([]byte{}, b...) } // CutPrefix returns s without the provided leading prefix byte slice // and reports whether it found the prefix. // If s doesn't start with prefix, CutPrefix returns s, false. // If prefix is the empty byte slice, CutPrefix returns s, true. // // CutPrefix returns slices of the original slice s, not copies. func CutPrefix(s, prefix []byte) (after []byte, found bool) { if !HasPrefix(s, prefix) { return s, false } return s[len(prefix):], true } // CutSuffix returns s without the provided ending suffix byte slice // and reports whether it found the suffix. // If s doesn't end with suffix, CutSuffix returns s, false. // If suffix is the empty byte slice, CutSuffix returns s, true. // // CutSuffix returns slices of the original slice s, not copies. func CutSuffix(s, suffix []byte) (before []byte, found bool) { if !HasSuffix(s, suffix) { return s, false } return s[:len(s)-len(suffix)], true }