example_iter_test.go

Documentation: golang.org/x/text/unicode/norm

     1  // Copyright 2012 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 norm_test
     6  
     7  import (
     8  	"bytes"
     9  	"fmt"
    10  	"io"
    11  	"unicode/utf8"
    12  
    13  	"golang.org/x/text/unicode/norm"
    14  )
    15  
    16  // EqualSimple uses a norm.Iter to compare two non-normalized
    17  // strings for equivalence.
    18  func EqualSimple(a, b string) bool {
    19  	var ia, ib norm.Iter
    20  	ia.InitString(norm.NFKD, a)
    21  	ib.InitString(norm.NFKD, b)
    22  	for !ia.Done() && !ib.Done() {
    23  		if !bytes.Equal(ia.Next(), ib.Next()) {
    24  			return false
    25  		}
    26  	}
    27  	return ia.Done() && ib.Done()
    28  }
    29  
    30  // FindPrefix finds the longest common prefix of ASCII characters
    31  // of a and b.
    32  func FindPrefix(a, b string) int {
    33  	i := 0
    34  	for ; i < len(a) && i < len(b) && a[i] < utf8.RuneSelf && a[i] == b[i]; i++ {
    35  	}
    36  	return i
    37  }
    38  
    39  // EqualOpt is like EqualSimple, but optimizes the special
    40  // case for ASCII characters.
    41  func EqualOpt(a, b string) bool {
    42  	n := FindPrefix(a, b)
    43  	a, b = a[n:], b[n:]
    44  	var ia, ib norm.Iter
    45  	ia.InitString(norm.NFKD, a)
    46  	ib.InitString(norm.NFKD, b)
    47  	for !ia.Done() && !ib.Done() {
    48  		if !bytes.Equal(ia.Next(), ib.Next()) {
    49  			return false
    50  		}
    51  		if n := int64(FindPrefix(a[ia.Pos():], b[ib.Pos():])); n != 0 {
    52  			ia.Seek(n, io.SeekCurrent)
    53  			ib.Seek(n, io.SeekCurrent)
    54  		}
    55  	}
    56  	return ia.Done() && ib.Done()
    57  }
    58  
    59  var compareTests = []struct{ a, b string }{
    60  	{"aaa", "aaa"},
    61  	{"aaa", "aab"},
    62  	{"a\u0300a", "\u00E0a"},
    63  	{"a\u0300\u0320b", "a\u0320\u0300b"},
    64  	{"\u1E0A\u0323", "\x44\u0323\u0307"},
    65  	// A character that decomposes into multiple segments
    66  	// spans several iterations.
    67  	{"\u3304", "\u30A4\u30CB\u30F3\u30AF\u3099"},
    68  }
    69  
    70  func ExampleIter() {
    71  	for i, t := range compareTests {
    72  		r0 := EqualSimple(t.a, t.b)
    73  		r1 := EqualOpt(t.a, t.b)
    74  		fmt.Printf("%d: %v %v\n", i, r0, r1)
    75  	}
    76  	// Output:
    77  	// 0: true true
    78  	// 1: false false
    79  	// 2: true true
    80  	// 3: true true
    81  	// 4: true true
    82  	// 5: true true
    83  }
    84
View as plain text