Home | History | Annotate | Download | only in bytes 
      1 // Copyright 2016 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 bytes
      6 
      7 //go:noescape
      8 
      9 // indexShortStr returns the index of the first instance of sep in s,
     10 // or -1 if sep is not present in s.
     11 // indexShortStr requires 2 <= len(sep) <= shortStringLen
     12 func indexShortStr(s, c []byte) int // ../runtime/asm_s390x.s
     13 
     14 // supportsVX reports whether the vector facility is available.
     15 // indexShortStr must not be called if the vector facility is not
     16 // available.
     17 func supportsVX() bool // ../runtime/asm_s390x.s
     18 
     19 var shortStringLen = -1
     20 
     21 func init() {
     22 	if supportsVX() {
     23 		shortStringLen = 64
     24 	}
     25 }
     26 
     27 // Index returns the index of the first instance of sep in s, or -1 if sep is not present in s.
     28 func Index(s, sep []byte) int {
     29 	n := len(sep)
     30 	switch {
     31 	case n == 0:
     32 		return 0
     33 	case n == 1:
     34 		return IndexByte(s, sep[0])
     35 	case n == len(s):
     36 		if Equal(sep, s) {
     37 			return 0
     38 		}
     39 		return -1
     40 	case n > len(s):
     41 		return -1
     42 	case n <= shortStringLen:
     43 		// Use brute force when s and sep both are small
     44 		if len(s) <= 64 {
     45 			return indexShortStr(s, sep)
     46 		}
     47 		c := sep[0]
     48 		i := 0
     49 		t := s[:len(s)-n+1]
     50 		fails := 0
     51 		for i < len(t) {
     52 			if t[i] != c {
     53 				// IndexByte skips 16/32 bytes per iteration,
     54 				// so it's faster than indexShortStr.
     55 				o := IndexByte(t[i:], c)
     56 				if o < 0 {
     57 					return -1
     58 				}
     59 				i += o
     60 			}
     61 			if Equal(s[i:i+n], sep) {
     62 				return i
     63 			}
     64 			fails++
     65 			i++
     66 			// Switch to indexShortStr when IndexByte produces too many false positives.
     67 			// Too many means more that 1 error per 8 characters.
     68 			// Allow some errors in the beginning.
     69 			if fails > (i+16)/8 {
     70 				r := indexShortStr(s[i:], sep)
     71 				if r >= 0 {
     72 					return r + i
     73 				}
     74 				return -1
     75 			}
     76 		}
     77 		return -1
     78 	}
     79 	// Rabin-Karp search
     80 	hashsep, pow := hashStr(sep)
     81 	var h uint32
     82 	for i := 0; i < n; i++ {
     83 		h = h*primeRK + uint32(s[i])
     84 	}
     85 	if h == hashsep && Equal(s[:n], sep) {
     86 		return 0
     87 	}
     88 	for i := n; i < len(s); {
     89 		h *= primeRK
     90 		h += uint32(s[i])
     91 		h -= pow * uint32(s[i-n])
     92 		i++
     93 		if h == hashsep && Equal(s[i-n:i], sep) {
     94 			return i - n
     95 		}
     96 	}
     97 	return -1
     98 }
     99 
    100 // primeRK is the prime base used in Rabin-Karp algorithm.
    101 const primeRK = 16777619
    102 
    103 // hashStr returns the hash and the appropriate multiplicative
    104 // factor for use in Rabin-Karp algorithm.
    105 func hashStr(sep []byte) (uint32, uint32) {
    106 	hash := uint32(0)
    107 	for i := 0; i < len(sep); i++ {
    108 		hash = hash*primeRK + uint32(sep[i])
    109 	}
    110 	var pow, sq uint32 = 1, primeRK
    111 	for i := len(sep); i > 0; i >>= 1 {
    112 		if i&1 != 0 {
    113 			pow *= sq
    114 		}
    115 		sq *= sq
    116 	}
    117 	return hash, pow
    118 }
    119