strings.go

     1  // Copyright 2024 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  // Package stringslite implements a subset of strings,
     6  // only using packages that may be imported by "os".
     7  //
     8  // Most tests for these functions are in the strings package.
     9  package stringslite
    10  
    11  import (
    12  	"internal/bytealg"
    13  	"unsafe"
    14  )
    15  
    16  func HasPrefix(s, prefix string) bool {
    17  	return len(s) >= len(prefix) && s[:len(prefix)] == prefix
    18  }
    19  
    20  func HasSuffix(s, suffix string) bool {
    21  	return len(s) >= len(suffix) && s[len(s)-len(suffix):] == suffix
    22  }
    23  
    24  func IndexByte(s string, c byte) int {
    25  	return bytealg.IndexByteString(s, c)
    26  }
    27  
    28  func Index(s, substr string) int {
    29  	n := len(substr)
    30  	switch {
    31  	case n == 0 || substr == s:
    32  		return 0
    33  	case n == 1:
    34  		return IndexByte(s, substr[0])
    35  	case n >= len(s):
    36  		return -1
    37  	case n <= bytealg.MaxLen && len(s) <= bytealg.MaxBruteForce:
    38  		// Use brute force when s and substr both are small
    39  		return bytealg.IndexString(s, substr)
    40  	}
    41  	c0 := substr[0]
    42  	c1 := substr[1]
    43  	i := 0
    44  	t := len(s) - n + 1
    45  	fails := 0
    46  	for i < t {
    47  		if s[i] != c0 {
    48  			// IndexByte is faster than bytealg.IndexString, so use it as long as
    49  			// we're not getting lots of false positives.
    50  			o := IndexByte(s[i+1:t], c0)
    51  			if o < 0 {
    52  				return -1
    53  			}
    54  			i += o + 1
    55  		}
    56  		if s[i+1] == c1 && s[i:i+n] == substr {
    57  			return i
    58  		}
    59  		i++
    60  		fails++
    61  		if n <= bytealg.MaxLen && fails > bytealg.Cutover(i) {
    62  			// Switch to bytealg.IndexString when IndexByte produces too many false positives.
    63  			r := bytealg.IndexString(s[i:], substr)
    64  			if r >= 0 {
    65  				return r + i
    66  			}
    67  			return -1
    68  		} else if n > bytealg.MaxLen && fails >= 4+i>>4 && i < t {
    69  			// See comment in ../bytes/bytes.go.
    70  			j := bytealg.IndexRabinKarp(s[i:], substr)
    71  			if j < 0 {
    72  				return -1
    73  			}
    74  			return i + j
    75  		}
    76  	}
    77  	return -1
    78  }
    79  
    80  func Cut(s, sep string) (before, after string, found bool) {
    81  	if i := Index(s, sep); i >= 0 {
    82  		return s[:i], s[i+len(sep):], true
    83  	}
    84  	return s, "", false
    85  }
    86  
    87  func CutPrefix(s, prefix string) (after string, found bool) {
    88  	if !HasPrefix(s, prefix) {
    89  		return s, false
    90  	}
    91  	return s[len(prefix):], true
    92  }
    93  
    94  func CutSuffix(s, suffix string) (before string, found bool) {
    95  	if !HasSuffix(s, suffix) {
    96  		return s, false
    97  	}
    98  	return s[:len(s)-len(suffix)], true
    99  }
   100  
   101  func TrimPrefix(s, prefix string) string {
   102  	if HasPrefix(s, prefix) {
   103  		return s[len(prefix):]
   104  	}
   105  	return s
   106  }
   107  
   108  func TrimSuffix(s, suffix string) string {
   109  	if HasSuffix(s, suffix) {
   110  		return s[:len(s)-len(suffix)]
   111  	}
   112  	return s
   113  }
   114  
   115  // IsSpace reports whether r is a space character as defined by
   116  // Unicode's White Space property.
   117  func IsSpace(r rune) bool {
   118  	switch r {
   119  	case '\t', '\n', '\v', '\f', '\r', ' ',
   120  		0x0085, // NEXT LINE
   121  		0x00A0, // NO-BREAK SPACE
   122  		0x1680, // OGHAM SPACE MARK
   123  		0x2000, // EN QUAD
   124  		0x2001, // EM QUAD
   125  		0x2002, // EN SPACE
   126  		0x2003, // EM SPACE
   127  		0x2004, // THREE-PER-EM SPACE
   128  		0x2005, // FOUR-PER-EM SPACE
   129  		0x2006, // SIX-PER-EM SPACE
   130  		0x2007, // FIGURE SPACE
   131  		0x2008, // PUNCTUATION SPACE
   132  		0x2009, // THIN SPACE
   133  		0x200A, // HAIR SPACE
   134  		0x2028, // LINE SEPARATOR
   135  		0x2029, // PARAGRAPH SEPARATOR
   136  		0x202F, // NARROW NO-BREAK SPACE
   137  		0x205F, // MEDIUM MATHEMATICAL SPACE
   138  		0x3000: // IDEOGRAPHIC SPACE
   139  		return true
   140  	}
   141  	return false
   142  }
   143  
   144  func Clone(s string) string {
   145  	if len(s) == 0 {
   146  		return ""
   147  	}
   148  	b := make([]byte, len(s))
   149  	copy(b, s)
   150  	return unsafe.String(&b[0], len(b))
   151  }
   152
View as plain text