vendor: Update everything

GitHub-Pull-Request: https://github.com/syncthing/syncthing/pull/4620
This commit is contained in:
Jakob Borg
2017-12-29 11:38:00 +00:00
parent 1296a22069
commit c24bf7ea55
1070 changed files with 294926 additions and 488191 deletions
+27
View File
@@ -0,0 +1,27 @@
Copyright (c) 2009 The Go Authors. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
* Neither the name of Google Inc. nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+41
View File
@@ -0,0 +1,41 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package format contains types for defining language-specific formatting of
// values.
//
// This package is internal now, but will eventually be exposed after the API
// settles.
package format // import "golang.org/x/text/internal/format"
import (
"fmt"
"golang.org/x/text/language"
)
// State represents the printer state passed to custom formatters. It provides
// access to the fmt.State interface and the sentence and language-related
// context.
type State interface {
fmt.State
// Language reports the requested language in which to render a message.
Language() language.Tag
// TODO: consider this and removing rune from the Format method in the
// Formatter interface.
//
// Verb returns the format variant to render, analogous to the types used
// in fmt. Use 'v' for the default or only variant.
// Verb() rune
// TODO: more info:
// - sentence context such as linguistic features passed by the translator.
}
// Formatter is analogous to fmt.Formatter.
type Formatter interface {
Format(state State, verb rune)
}
+357
View File
@@ -0,0 +1,357 @@
// Copyright 2017 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
package format
import (
"reflect"
"unicode/utf8"
)
// A Parser parses a format string. The result from the parse are set in the
// struct fields.
type Parser struct {
Verb rune
WidthPresent bool
PrecPresent bool
Minus bool
Plus bool
Sharp bool
Space bool
Zero bool
// For the formats %+v %#v, we set the plusV/sharpV flags
// and clear the plus/sharp flags since %+v and %#v are in effect
// different, flagless formats set at the top level.
PlusV bool
SharpV bool
HasIndex bool
Width int
Prec int // precision
// retain arguments across calls.
Args []interface{}
// retain current argument number across calls
ArgNum int
// reordered records whether the format string used argument reordering.
Reordered bool
// goodArgNum records whether the most recent reordering directive was valid.
goodArgNum bool
// position info
format string
startPos int
endPos int
Status Status
}
// Reset initializes a parser to scan format strings for the given args.
func (p *Parser) Reset(args []interface{}) {
p.Args = args
p.ArgNum = 0
p.startPos = 0
p.Reordered = false
}
// Text returns the part of the format string that was parsed by the last call
// to Scan. It returns the original substitution clause if the current scan
// parsed a substitution.
func (p *Parser) Text() string { return p.format[p.startPos:p.endPos] }
// SetFormat sets a new format string to parse. It does not reset the argument
// count.
func (p *Parser) SetFormat(format string) {
p.format = format
p.startPos = 0
p.endPos = 0
}
// Status indicates the result type of a call to Scan.
type Status int
const (
StatusText Status = iota
StatusSubstitution
StatusBadWidthSubstitution
StatusBadPrecSubstitution
StatusNoVerb
StatusBadArgNum
StatusMissingArg
)
// ClearFlags reset the parser to default behavior.
func (p *Parser) ClearFlags() {
p.WidthPresent = false
p.PrecPresent = false
p.Minus = false
p.Plus = false
p.Sharp = false
p.Space = false
p.Zero = false
p.PlusV = false
p.SharpV = false
p.HasIndex = false
}
// Scan scans the next part of the format string and sets the status to
// indicate whether it scanned a string literal, substitution or error.
func (p *Parser) Scan() bool {
p.Status = StatusText
format := p.format
end := len(format)
if p.endPos >= end {
return false
}
afterIndex := false // previous item in format was an index like [3].
p.startPos = p.endPos
p.goodArgNum = true
i := p.startPos
for i < end && format[i] != '%' {
i++
}
if i > p.startPos {
p.endPos = i
return true
}
// Process one verb
i++
p.Status = StatusSubstitution
// Do we have flags?
p.ClearFlags()
simpleFormat:
for ; i < end; i++ {
c := p.format[i]
switch c {
case '#':
p.Sharp = true
case '0':
p.Zero = !p.Minus // Only allow zero padding to the left.
case '+':
p.Plus = true
case '-':
p.Minus = true
p.Zero = false // Do not pad with zeros to the right.
case ' ':
p.Space = true
default:
// Fast path for common case of ascii lower case simple verbs
// without precision or width or argument indices.
if 'a' <= c && c <= 'z' && p.ArgNum < len(p.Args) {
if c == 'v' {
// Go syntax
p.SharpV = p.Sharp
p.Sharp = false
// Struct-field syntax
p.PlusV = p.Plus
p.Plus = false
}
p.Verb = rune(c)
p.ArgNum++
p.endPos = i + 1
return true
}
// Format is more complex than simple flags and a verb or is malformed.
break simpleFormat
}
}
// Do we have an explicit argument index?
i, afterIndex = p.updateArgNumber(format, i)
// Do we have width?
if i < end && format[i] == '*' {
i++
p.Width, p.WidthPresent = p.intFromArg()
if !p.WidthPresent {
p.Status = StatusBadWidthSubstitution
}
// We have a negative width, so take its value and ensure
// that the minus flag is set
if p.Width < 0 {
p.Width = -p.Width
p.Minus = true
p.Zero = false // Do not pad with zeros to the right.
}
afterIndex = false
} else {
p.Width, p.WidthPresent, i = parsenum(format, i, end)
if afterIndex && p.WidthPresent { // "%[3]2d"
p.goodArgNum = false
}
}
// Do we have precision?
if i+1 < end && format[i] == '.' {
i++
if afterIndex { // "%[3].2d"
p.goodArgNum = false
}
i, afterIndex = p.updateArgNumber(format, i)
if i < end && format[i] == '*' {
i++
p.Prec, p.PrecPresent = p.intFromArg()
// Negative precision arguments don't make sense
if p.Prec < 0 {
p.Prec = 0
p.PrecPresent = false
}
if !p.PrecPresent {
p.Status = StatusBadPrecSubstitution
}
afterIndex = false
} else {
p.Prec, p.PrecPresent, i = parsenum(format, i, end)
if !p.PrecPresent {
p.Prec = 0
p.PrecPresent = true
}
}
}
if !afterIndex {
i, afterIndex = p.updateArgNumber(format, i)
}
p.HasIndex = afterIndex
if i >= end {
p.endPos = i
p.Status = StatusNoVerb
return true
}
verb, w := utf8.DecodeRuneInString(format[i:])
p.endPos = i + w
p.Verb = verb
switch {
case verb == '%': // Percent does not absorb operands and ignores f.wid and f.prec.
p.startPos = p.endPos - 1
p.Status = StatusText
case !p.goodArgNum:
p.Status = StatusBadArgNum
case p.ArgNum >= len(p.Args): // No argument left over to print for the current verb.
p.Status = StatusMissingArg
case verb == 'v':
// Go syntax
p.SharpV = p.Sharp
p.Sharp = false
// Struct-field syntax
p.PlusV = p.Plus
p.Plus = false
fallthrough
default:
p.ArgNum++
}
return true
}
// intFromArg gets the ArgNumth element of Args. On return, isInt reports
// whether the argument has integer type.
func (p *Parser) intFromArg() (num int, isInt bool) {
if p.ArgNum < len(p.Args) {
arg := p.Args[p.ArgNum]
num, isInt = arg.(int) // Almost always OK.
if !isInt {
// Work harder.
switch v := reflect.ValueOf(arg); v.Kind() {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
n := v.Int()
if int64(int(n)) == n {
num = int(n)
isInt = true
}
case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
n := v.Uint()
if int64(n) >= 0 && uint64(int(n)) == n {
num = int(n)
isInt = true
}
default:
// Already 0, false.
}
}
p.ArgNum++
if tooLarge(num) {
num = 0
isInt = false
}
}
return
}
// parseArgNumber returns the value of the bracketed number, minus 1
// (explicit argument numbers are one-indexed but we want zero-indexed).
// The opening bracket is known to be present at format[0].
// The returned values are the index, the number of bytes to consume
// up to the closing paren, if present, and whether the number parsed
// ok. The bytes to consume will be 1 if no closing paren is present.
func parseArgNumber(format string) (index int, wid int, ok bool) {
// There must be at least 3 bytes: [n].
if len(format) < 3 {
return 0, 1, false
}
// Find closing bracket.
for i := 1; i < len(format); i++ {
if format[i] == ']' {
width, ok, newi := parsenum(format, 1, i)
if !ok || newi != i {
return 0, i + 1, false
}
return width - 1, i + 1, true // arg numbers are one-indexed and skip paren.
}
}
return 0, 1, false
}
// updateArgNumber returns the next argument to evaluate, which is either the value of the passed-in
// argNum or the value of the bracketed integer that begins format[i:]. It also returns
// the new value of i, that is, the index of the next byte of the format to process.
func (p *Parser) updateArgNumber(format string, i int) (newi int, found bool) {
if len(format) <= i || format[i] != '[' {
return i, false
}
p.Reordered = true
index, wid, ok := parseArgNumber(format[i:])
if ok && 0 <= index && index < len(p.Args) {
p.ArgNum = index
return i + wid, true
}
p.goodArgNum = false
return i + wid, ok
}
// tooLarge reports whether the magnitude of the integer is
// too large to be used as a formatting width or precision.
func tooLarge(x int) bool {
const max int = 1e6
return x > max || x < -max
}
// parsenum converts ASCII to integer. num is 0 (and isnum is false) if no number present.
func parsenum(s string, start, end int) (num int, isnum bool, newi int) {
if start >= end {
return 0, false, end
}
for newi = start; newi < end && '0' <= s[newi] && s[newi] <= '9'; newi++ {
if tooLarge(num) {
return 0, false, end // Overflow; crazy long number most likely.
}
num = num*10 + int(s[newi]-'0')
isnum = true
}
return
}
+21 -3
View File
@@ -55,18 +55,36 @@ func (w *CodeWriter) WriteGoFile(filename, pkg string) {
log.Fatalf("Could not create file %s: %v", filename, err)
}
defer f.Close()
if _, err = w.WriteGo(f, pkg); err != nil {
if _, err = w.WriteGo(f, pkg, ""); err != nil {
log.Fatalf("Error writing file %s: %v", filename, err)
}
}
// WriteVersionedGoFile appends the buffer with the total size of all created
// structures and writes it as a Go file to the the given file with the given
// package name and build tags for the current Unicode version,
func (w *CodeWriter) WriteVersionedGoFile(filename, pkg string) {
tags := buildTags()
if tags != "" {
filename = insertVersion(filename, UnicodeVersion())
}
f, err := os.Create(filename)
if err != nil {
log.Fatalf("Could not create file %s: %v", filename, err)
}
defer f.Close()
if _, err = w.WriteGo(f, pkg, tags); err != nil {
log.Fatalf("Error writing file %s: %v", filename, err)
}
}
// WriteGo appends the buffer with the total size of all created structures and
// writes it as a Go file to the the given writer with the given package name.
func (w *CodeWriter) WriteGo(out io.Writer, pkg string) (n int, err error) {
func (w *CodeWriter) WriteGo(out io.Writer, pkg, tags string) (n int, err error) {
sz := w.Size
w.WriteComment("Total table size %d bytes (%dKiB); checksum: %X\n", sz, sz/1024, w.Hash.Sum32())
defer w.buf.Reset()
return WriteGo(out, pkg, w.buf.Bytes())
return WriteGo(out, pkg, tags, w.buf.Bytes())
}
func (w *CodeWriter) printf(f string, x ...interface{}) {
+58 -6
View File
@@ -31,6 +31,7 @@ import (
"os"
"path"
"path/filepath"
"strings"
"sync"
"unicode"
@@ -69,8 +70,6 @@ func Init() {
const header = `// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
package %s
`
// UnicodeVersion reports the requested Unicode version.
@@ -78,11 +77,33 @@ func UnicodeVersion() string {
return *unicodeVersion
}
// UnicodeVersion reports the requested CLDR version.
// CLDRVersion reports the requested CLDR version.
func CLDRVersion() string {
return *cldrVersion
}
var tags = []struct{ version, buildTags string }{
{"10.0.0", "go1.10"},
{"", "!go1.10"},
}
// buildTags reports the build tags used for the current Unicode version.
func buildTags() string {
v := UnicodeVersion()
for _, x := range tags {
// We should do a numeric comparison, but including the collate package
// would create an import cycle. We approximate it by assuming that
// longer version strings are later.
if len(x.version) <= len(v) {
return x.buildTags
}
if len(x.version) == len(v) && x.version <= v {
return x.buildTags
}
}
return tags[0].buildTags
}
// IsLocal reports whether data files are available locally.
func IsLocal() bool {
dir, err := localReadmeFile()
@@ -243,15 +264,46 @@ func WriteGoFile(filename, pkg string, b []byte) {
log.Fatalf("Could not create file %s: %v", filename, err)
}
defer w.Close()
if _, err = WriteGo(w, pkg, b); err != nil {
if _, err = WriteGo(w, pkg, "", b); err != nil {
log.Fatalf("Error writing file %s: %v", filename, err)
}
}
func insertVersion(filename, version string) string {
suffix := ".go"
if strings.HasSuffix(filename, "_test.go") {
suffix = "_test.go"
}
return fmt.Sprint(filename[:len(filename)-len(suffix)], version, suffix)
}
// WriteVersionedGoFile prepends a standard file comment, adds build tags to
// version the file for the current Unicode version, and package statement to
// the given bytes, applies gofmt, and writes them to a file with the given
// name. It will call log.Fatal if there are any errors.
func WriteVersionedGoFile(filename, pkg string, b []byte) {
tags := buildTags()
if tags != "" {
filename = insertVersion(filename, UnicodeVersion())
}
w, err := os.Create(filename)
if err != nil {
log.Fatalf("Could not create file %s: %v", filename, err)
}
defer w.Close()
if _, err = WriteGo(w, pkg, tags, b); err != nil {
log.Fatalf("Error writing file %s: %v", filename, err)
}
}
// WriteGo prepends a standard file comment and package statement to the given
// bytes, applies gofmt, and writes them to w.
func WriteGo(w io.Writer, pkg string, b []byte) (n int, err error) {
src := []byte(fmt.Sprintf(header, pkg))
func WriteGo(w io.Writer, pkg, tags string, b []byte) (n int, err error) {
src := []byte(header)
if tags != "" {
src = append(src, fmt.Sprintf("// +build %s\n\n", tags)...)
}
src = append(src, fmt.Sprintf("package %s\n\n", pkg)...)
src = append(src, b...)
formatted, err := format.Source(src)
if err != nil {
+27
View File
@@ -0,0 +1,27 @@
Copyright (c) 2009 The Go Authors. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
* Neither the name of Google Inc. nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+100
View File
@@ -0,0 +1,100 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package tag contains functionality handling tags and related data.
package tag // import "golang.org/x/text/internal/tag"
import "sort"
// An Index converts tags to a compact numeric value.
//
// All elements are of size 4. Tags may be up to 4 bytes long. Excess bytes can
// be used to store additional information about the tag.
type Index string
// Elem returns the element data at the given index.
func (s Index) Elem(x int) string {
return string(s[x*4 : x*4+4])
}
// Index reports the index of the given key or -1 if it could not be found.
// Only the first len(key) bytes from the start of the 4-byte entries will be
// considered for the search and the first match in Index will be returned.
func (s Index) Index(key []byte) int {
n := len(key)
// search the index of the first entry with an equal or higher value than
// key in s.
index := sort.Search(len(s)/4, func(i int) bool {
return cmp(s[i*4:i*4+n], key) != -1
})
i := index * 4
if cmp(s[i:i+len(key)], key) != 0 {
return -1
}
return index
}
// Next finds the next occurrence of key after index x, which must have been
// obtained from a call to Index using the same key. It returns x+1 or -1.
func (s Index) Next(key []byte, x int) int {
if x++; x*4 < len(s) && cmp(s[x*4:x*4+len(key)], key) == 0 {
return x
}
return -1
}
// cmp returns an integer comparing a and b lexicographically.
func cmp(a Index, b []byte) int {
n := len(a)
if len(b) < n {
n = len(b)
}
for i, c := range b[:n] {
switch {
case a[i] > c:
return 1
case a[i] < c:
return -1
}
}
switch {
case len(a) < len(b):
return -1
case len(a) > len(b):
return 1
}
return 0
}
// Compare returns an integer comparing a and b lexicographically.
func Compare(a string, b []byte) int {
return cmp(Index(a), b)
}
// FixCase reformats b to the same pattern of cases as form.
// If returns false if string b is malformed.
func FixCase(form string, b []byte) bool {
if len(form) != len(b) {
return false
}
for i, c := range b {
if form[i] <= 'Z' {
if c >= 'a' {
c -= 'z' - 'Z'
}
if c < 'A' || 'Z' < c {
return false
}
} else {
if c <= 'Z' {
c += 'z' - 'Z'
}
if c < 'a' || 'z' < c {
return false
}
}
b[i] = c
}
return true
}
+55 -60
View File
@@ -11,8 +11,8 @@ package ucd // import "golang.org/x/text/internal/ucd"
import (
"bufio"
"bytes"
"errors"
"fmt"
"io"
"log"
"regexp"
@@ -92,10 +92,11 @@ type Parser struct {
keepRanges bool // Don't expand rune ranges in field 0.
err error
comment []byte
field [][]byte
comment string
field []string
// parsedRange is needed in case Range(0) is called more than once for one
// field. In some cases this requires scanning ahead.
line int
parsedRange bool
rangeStart, rangeEnd rune
@@ -103,15 +104,19 @@ type Parser struct {
commentHandler func(s string)
}
func (p *Parser) setError(err error) {
if p.err == nil {
p.err = err
func (p *Parser) setError(err error, msg string) {
if p.err == nil && err != nil {
if msg == "" {
p.err = fmt.Errorf("ucd:line:%d: %v", p.line, err)
} else {
p.err = fmt.Errorf("ucd:line:%d:%s: %v", p.line, msg, err)
}
}
}
func (p *Parser) getField(i int) []byte {
func (p *Parser) getField(i int) string {
if i >= len(p.field) {
return nil
return ""
}
return p.field[i]
}
@@ -139,65 +144,66 @@ func (p *Parser) Next() bool {
p.rangeStart++
return true
}
p.comment = nil
p.comment = ""
p.field = p.field[:0]
p.parsedRange = false
for p.scanner.Scan() {
b := p.scanner.Bytes()
if len(b) == 0 {
for p.scanner.Scan() && p.err == nil {
p.line++
s := p.scanner.Text()
if s == "" {
continue
}
if b[0] == '#' {
if s[0] == '#' {
if p.commentHandler != nil {
p.commentHandler(strings.TrimSpace(string(b[1:])))
p.commentHandler(strings.TrimSpace(s[1:]))
}
continue
}
// Parse line
if i := bytes.IndexByte(b, '#'); i != -1 {
p.comment = bytes.TrimSpace(b[i+1:])
b = b[:i]
if i := strings.IndexByte(s, '#'); i != -1 {
p.comment = strings.TrimSpace(s[i+1:])
s = s[:i]
}
if b[0] == '@' {
if s[0] == '@' {
if p.partHandler != nil {
p.field = append(p.field, bytes.TrimSpace(b[1:]))
p.field = append(p.field, strings.TrimSpace(s[1:]))
p.partHandler(p)
p.field = p.field[:0]
}
p.comment = nil
p.comment = ""
continue
}
for {
i := bytes.IndexByte(b, ';')
i := strings.IndexByte(s, ';')
if i == -1 {
p.field = append(p.field, bytes.TrimSpace(b))
p.field = append(p.field, strings.TrimSpace(s))
break
}
p.field = append(p.field, bytes.TrimSpace(b[:i]))
b = b[i+1:]
p.field = append(p.field, strings.TrimSpace(s[:i]))
s = s[i+1:]
}
if !p.keepRanges {
p.rangeStart, p.rangeEnd = p.getRange(0)
}
return true
}
p.setError(p.scanner.Err())
p.setError(p.scanner.Err(), "scanner failed")
return false
}
func parseRune(b []byte) (rune, error) {
func parseRune(b string) (rune, error) {
if len(b) > 2 && b[0] == 'U' && b[1] == '+' {
b = b[2:]
}
x, err := strconv.ParseUint(string(b), 16, 32)
x, err := strconv.ParseUint(b, 16, 32)
return rune(x), err
}
func (p *Parser) parseRune(b []byte) rune {
x, err := parseRune(b)
p.setError(err)
func (p *Parser) parseRune(s string) rune {
x, err := parseRune(s)
p.setError(err, "failed to parse rune")
return x
}
@@ -211,13 +217,13 @@ func (p *Parser) Rune(i int) rune {
// Runes interprets and returns field i as a sequence of runes.
func (p *Parser) Runes(i int) (runes []rune) {
add := func(b []byte) {
if b = bytes.TrimSpace(b); len(b) > 0 {
runes = append(runes, p.parseRune(b))
add := func(s string) {
if s = strings.TrimSpace(s); len(s) > 0 {
runes = append(runes, p.parseRune(s))
}
}
for b := p.getField(i); ; {
i := bytes.IndexByte(b, ' ')
i := strings.IndexByte(b, ' ')
if i == -1 {
add(b)
break
@@ -247,7 +253,7 @@ func (p *Parser) Range(i int) (first, last rune) {
func (p *Parser) getRange(i int) (first, last rune) {
b := p.getField(i)
if k := bytes.Index(b, []byte("..")); k != -1 {
if k := strings.Index(b, ".."); k != -1 {
return p.parseRune(b[:k]), p.parseRune(b[k+2:])
}
// The first field may not be a rune, in which case we may ignore any error
@@ -260,23 +266,24 @@ func (p *Parser) getRange(i int) (first, last rune) {
p.keepRanges = true
}
// Special case for UnicodeData that was retained for backwards compatibility.
if i == 0 && len(p.field) > 1 && bytes.HasSuffix(p.field[1], []byte("First>")) {
if i == 0 && len(p.field) > 1 && strings.HasSuffix(p.field[1], "First>") {
if p.parsedRange {
return p.rangeStart, p.rangeEnd
}
mf := reRange.FindStringSubmatch(p.scanner.Text())
p.line++
if mf == nil || !p.scanner.Scan() {
p.setError(errIncorrectLegacyRange)
p.setError(errIncorrectLegacyRange, "")
return x, x
}
// Using Bytes would be more efficient here, but Text is a lot easier
// and this is not a frequent case.
ml := reRange.FindStringSubmatch(p.scanner.Text())
if ml == nil || mf[2] != ml[2] || ml[3] != "Last" || mf[4] != ml[4] {
p.setError(errIncorrectLegacyRange)
p.setError(errIncorrectLegacyRange, "")
return x, x
}
p.rangeStart, p.rangeEnd = x, p.parseRune(p.scanner.Bytes()[:len(ml[1])])
p.rangeStart, p.rangeEnd = x, p.parseRune(p.scanner.Text()[:len(ml[1])])
p.parsedRange = true
return p.rangeStart, p.rangeEnd
}
@@ -298,34 +305,34 @@ var bools = map[string]bool{
// Bool parses and returns field i as a boolean value.
func (p *Parser) Bool(i int) bool {
b := p.getField(i)
f := p.getField(i)
for s, v := range bools {
if bstrEq(b, s) {
if f == s {
return v
}
}
p.setError(strconv.ErrSyntax)
p.setError(strconv.ErrSyntax, "error parsing bool")
return false
}
// Int parses and returns field i as an integer value.
func (p *Parser) Int(i int) int {
x, err := strconv.ParseInt(string(p.getField(i)), 10, 64)
p.setError(err)
p.setError(err, "error parsing int")
return int(x)
}
// Uint parses and returns field i as an unsigned integer value.
func (p *Parser) Uint(i int) uint {
x, err := strconv.ParseUint(string(p.getField(i)), 10, 64)
p.setError(err)
p.setError(err, "error parsing uint")
return uint(x)
}
// Float parses and returns field i as a decimal value.
func (p *Parser) Float(i int) float64 {
x, err := strconv.ParseFloat(string(p.getField(i)), 64)
p.setError(err)
p.setError(err, "error parsing float")
return x
}
@@ -353,24 +360,12 @@ var errUndefinedEnum = errors.New("ucd: undefined enum value")
// Enum interprets and returns field i as a value that must be one of the values
// in enum.
func (p *Parser) Enum(i int, enum ...string) string {
b := p.getField(i)
f := p.getField(i)
for _, s := range enum {
if bstrEq(b, s) {
if f == s {
return s
}
}
p.setError(errUndefinedEnum)
p.setError(errUndefinedEnum, "error parsing enum")
return ""
}
func bstrEq(b []byte, s string) bool {
if len(b) != len(s) {
return false
}
for i, c := range b {
if c != s[i] {
return false
}
}
return true
}
+27
View File
@@ -0,0 +1,27 @@
Copyright (c) 2009 The Go Authors. All rights reserved.
Redistribution and use in source and binary forms, with or without
modification, are permitted provided that the following conditions are
met:
* Redistributions of source code must retain the above copyright
notice, this list of conditions and the following disclaimer.
* Redistributions in binary form must reproduce the above
copyright notice, this list of conditions and the following disclaimer
in the documentation and/or other materials provided with the
distribution.
* Neither the name of Google Inc. nor the names of its
contributors may be used to endorse or promote products derived from
this software without specific prior written permission.
THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
"AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR
A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT
OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL,
SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT
LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE,
DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY
THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
(INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+87
View File
@@ -0,0 +1,87 @@
// Copyright 2015 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package utf8internal contains low-level utf8-related constants, tables, etc.
// that are used internally by the text package.
package utf8internal
// The default lowest and highest continuation byte.
const (
LoCB = 0x80 // 1000 0000
HiCB = 0xBF // 1011 1111
)
// Constants related to getting information of first bytes of UTF-8 sequences.
const (
// ASCII identifies a UTF-8 byte as ASCII.
ASCII = as
// FirstInvalid indicates a byte is invalid as a first byte of a UTF-8
// sequence.
FirstInvalid = xx
// SizeMask is a mask for the size bits. Use use x&SizeMask to get the size.
SizeMask = 7
// AcceptShift is the right-shift count for the first byte info byte to get
// the index into the AcceptRanges table. See AcceptRanges.
AcceptShift = 4
// The names of these constants are chosen to give nice alignment in the
// table below. The first nibble is an index into acceptRanges or F for
// special one-byte cases. The second nibble is the Rune length or the
// Status for the special one-byte case.
xx = 0xF1 // invalid: size 1
as = 0xF0 // ASCII: size 1
s1 = 0x02 // accept 0, size 2
s2 = 0x13 // accept 1, size 3
s3 = 0x03 // accept 0, size 3
s4 = 0x23 // accept 2, size 3
s5 = 0x34 // accept 3, size 4
s6 = 0x04 // accept 0, size 4
s7 = 0x44 // accept 4, size 4
)
// First is information about the first byte in a UTF-8 sequence.
var First = [256]uint8{
// 1 2 3 4 5 6 7 8 9 A B C D E F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x00-0x0F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x10-0x1F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x20-0x2F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x30-0x3F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x40-0x4F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x50-0x5F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x60-0x6F
as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, as, // 0x70-0x7F
// 1 2 3 4 5 6 7 8 9 A B C D E F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x80-0x8F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0x90-0x9F
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xA0-0xAF
xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xB0-0xBF
xx, xx, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xC0-0xCF
s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, s1, // 0xD0-0xDF
s2, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s3, s4, s3, s3, // 0xE0-0xEF
s5, s6, s6, s6, s7, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, xx, // 0xF0-0xFF
}
// AcceptRange gives the range of valid values for the second byte in a UTF-8
// sequence for any value for First that is not ASCII or FirstInvalid.
type AcceptRange struct {
Lo uint8 // lowest value for second byte.
Hi uint8 // highest value for second byte.
}
// AcceptRanges is a slice of AcceptRange values. For a given byte sequence b
//
// AcceptRanges[First[b[0]]>>AcceptShift]
//
// will give the value of AcceptRange for the multi-byte UTF-8 sequence starting
// at b[0].
var AcceptRanges = [...]AcceptRange{
0: {LoCB, HiCB},
1: {0xA0, HiCB},
2: {LoCB, 0x9F},
3: {0x90, HiCB},
4: {LoCB, 0x8F},
}