example_test.go

Documentation: golang.org/x/text/internal/ucd

     1  // Copyright 2014 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 ucd_test
     6  
     7  import (
     8  	"fmt"
     9  	"strings"
    10  
    11  	"golang.org/x/text/internal/ucd"
    12  )
    13  
    14  func Example() {
    15  	// Read rune-by-rune from UnicodeData.
    16  	var count int
    17  	p := ucd.New(strings.NewReader(unicodeData))
    18  	for p.Next() {
    19  		count++
    20  		if lower := p.Runes(ucd.SimpleLowercaseMapping); lower != nil {
    21  			fmt.Printf("lower(%U) -> %U\n", p.Rune(0), lower[0])
    22  		}
    23  	}
    24  	if err := p.Err(); err != nil {
    25  		fmt.Println(err)
    26  	}
    27  	fmt.Println("Number of runes visited:", count)
    28  
    29  	// Read raw ranges from Scripts.
    30  	p = ucd.New(strings.NewReader(scripts), ucd.KeepRanges)
    31  	for p.Next() {
    32  		start, end := p.Range(0)
    33  		fmt.Printf("%04X..%04X: %s\n", start, end, p.String(1))
    34  	}
    35  	if err := p.Err(); err != nil {
    36  		fmt.Println(err)
    37  	}
    38  
    39  	// Output:
    40  	// lower(U+00C0) -> U+00E0
    41  	// lower(U+00C1) -> U+00E1
    42  	// lower(U+00C2) -> U+00E2
    43  	// lower(U+00C3) -> U+00E3
    44  	// lower(U+00C4) -> U+00E4
    45  	// Number of runes visited: 6594
    46  	// 0000..001F: Common
    47  	// 0020..0020: Common
    48  	// 0021..0023: Common
    49  	// 0024..0024: Common
    50  }
    51  
    52  // Excerpt from UnicodeData.txt
    53  const unicodeData = `
    54  00B9;SUPERSCRIPT ONE;No;0;EN;<super> 0031;;1;1;N;SUPERSCRIPT DIGIT ONE;;;;
    55  00BA;MASCULINE ORDINAL INDICATOR;Lo;0;L;<super> 006F;;;;N;;;;;
    56  00BB;RIGHT-POINTING DOUBLE ANGLE QUOTATION MARK;Pf;0;ON;;;;;Y;RIGHT POINTING GUILLEMET;;;;
    57  00BC;VULGAR FRACTION ONE QUARTER;No;0;ON;<fraction> 0031 2044 0034;;;1/4;N;FRACTION ONE QUARTER;;;;
    58  00BD;VULGAR FRACTION ONE HALF;No;0;ON;<fraction> 0031 2044 0032;;;1/2;N;FRACTION ONE HALF;;;;
    59  00BE;VULGAR FRACTION THREE QUARTERS;No;0;ON;<fraction> 0033 2044 0034;;;3/4;N;FRACTION THREE QUARTERS;;;;
    60  00BF;INVERTED QUESTION MARK;Po;0;ON;;;;;N;;;;;
    61  00C0;LATIN CAPITAL LETTER A WITH GRAVE;Lu;0;L;0041 0300;;;;N;LATIN CAPITAL LETTER A GRAVE;;;00E0;
    62  00C1;LATIN CAPITAL LETTER A WITH ACUTE;Lu;0;L;0041 0301;;;;N;LATIN CAPITAL LETTER A ACUTE;;;00E1;
    63  00C2;LATIN CAPITAL LETTER A WITH CIRCUMFLEX;Lu;0;L;0041 0302;;;;N;LATIN CAPITAL LETTER A CIRCUMFLEX;;;00E2;
    64  00C3;LATIN CAPITAL LETTER A WITH TILDE;Lu;0;L;0041 0303;;;;N;LATIN CAPITAL LETTER A TILDE;;;00E3;
    65  00C4;LATIN CAPITAL LETTER A WITH DIAERESIS;Lu;0;L;0041 0308;;;;N;LATIN CAPITAL LETTER A DIAERESIS;;;00E4;
    66  
    67  # A legacy rune range.
    68  3400;<CJK Ideograph Extension A, First>;Lo;0;L;;;;;N;;;;;
    69  4DB5;<CJK Ideograph Extension A, Last>;Lo;0;L;;;;;N;;;;;
    70  `
    71  
    72  // Excerpt from Scripts.txt
    73  const scripts = `
    74  # Property:	Script
    75  # ================================================
    76  
    77  0000..001F    ; Common # Cc  [32] <control-0000>..<control-001F>
    78  0020          ; Common # Zs       SPACE
    79  0021..0023    ; Common # Po   [3] EXCLAMATION MARK..NUMBER SIGN
    80  0024          ; Common # Sc       DOLLAR SIGN
    81  `
    82
View as plain text