diff options
Diffstat (limited to 'vendor/golang.org/x/text/internal')
-rw-r--r-- | vendor/golang.org/x/text/internal/gen/code.go | 339 | ||||
-rw-r--r-- | vendor/golang.org/x/text/internal/gen/gen.go | 281 | ||||
-rw-r--r-- | vendor/golang.org/x/text/internal/tag/tag.go | 100 | ||||
-rw-r--r-- | vendor/golang.org/x/text/internal/utf8internal/utf8internal.go | 87 |
4 files changed, 807 insertions, 0 deletions
diff --git a/vendor/golang.org/x/text/internal/gen/code.go b/vendor/golang.org/x/text/internal/gen/code.go new file mode 100644 index 000000000..2202c9f64 --- /dev/null +++ b/vendor/golang.org/x/text/internal/gen/code.go @@ -0,0 +1,339 @@ +// 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 gen + +import ( + "bytes" + "encoding/gob" + "fmt" + "hash" + "hash/fnv" + "io" + "log" + "os" + "reflect" + "strings" + "unicode" + "unicode/utf8" +) + +// This file contains utilities for generating code. + +// TODO: other write methods like: +// - slices, maps, types, etc. + +// CodeWriter is a utility for writing structured code. It computes the content +// hash and size of written content. It ensures there are newlines between +// written code blocks. +type CodeWriter struct { + buf bytes.Buffer + Size int + Hash hash.Hash32 // content hash + gob *gob.Encoder + // For comments we skip the usual one-line separator if they are followed by + // a code block. + skipSep bool +} + +func (w *CodeWriter) Write(p []byte) (n int, err error) { + return w.buf.Write(p) +} + +// NewCodeWriter returns a new CodeWriter. +func NewCodeWriter() *CodeWriter { + h := fnv.New32() + return &CodeWriter{Hash: h, gob: gob.NewEncoder(h)} +} + +// WriteGoFile 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. +func (w *CodeWriter) WriteGoFile(filename, pkg string) { + 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); 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) { + 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()) +} + +func (w *CodeWriter) printf(f string, x ...interface{}) { + fmt.Fprintf(w, f, x...) +} + +func (w *CodeWriter) insertSep() { + if w.skipSep { + w.skipSep = false + return + } + // Use at least two newlines to ensure a blank space between the previous + // block. WriteGoFile will remove extraneous newlines. + w.printf("\n\n") +} + +// WriteComment writes a comment block. All line starts are prefixed with "//". +// Initial empty lines are gobbled. The indentation for the first line is +// stripped from consecutive lines. +func (w *CodeWriter) WriteComment(comment string, args ...interface{}) { + s := fmt.Sprintf(comment, args...) + s = strings.Trim(s, "\n") + + // Use at least two newlines to ensure a blank space between the previous + // block. WriteGoFile will remove extraneous newlines. + w.printf("\n\n// ") + w.skipSep = true + + // strip first indent level. + sep := "\n" + for ; len(s) > 0 && (s[0] == '\t' || s[0] == ' '); s = s[1:] { + sep += s[:1] + } + + strings.NewReplacer(sep, "\n// ", "\n", "\n// ").WriteString(w, s) + + w.printf("\n") +} + +func (w *CodeWriter) writeSizeInfo(size int) { + w.printf("// Size: %d bytes\n", size) +} + +// WriteConst writes a constant of the given name and value. +func (w *CodeWriter) WriteConst(name string, x interface{}) { + w.insertSep() + v := reflect.ValueOf(x) + + switch v.Type().Kind() { + case reflect.String: + // See golang.org/issue/13145. + const arbitraryCutoff = 16 + if v.Len() > arbitraryCutoff { + w.printf("var %s %s = ", name, typeName(x)) + } else { + w.printf("const %s %s = ", name, typeName(x)) + } + w.WriteString(v.String()) + w.printf("\n") + default: + w.printf("const %s = %#v\n", name, x) + } +} + +// WriteVar writes a variable of the given name and value. +func (w *CodeWriter) WriteVar(name string, x interface{}) { + w.insertSep() + v := reflect.ValueOf(x) + oldSize := w.Size + sz := int(v.Type().Size()) + w.Size += sz + + switch v.Type().Kind() { + case reflect.String: + w.printf("var %s %s = ", name, typeName(x)) + w.WriteString(v.String()) + case reflect.Struct: + w.gob.Encode(x) + fallthrough + case reflect.Slice, reflect.Array: + w.printf("var %s = ", name) + w.writeValue(v) + w.writeSizeInfo(w.Size - oldSize) + default: + w.printf("var %s %s = ", name, typeName(x)) + w.gob.Encode(x) + w.writeValue(v) + w.writeSizeInfo(w.Size - oldSize) + } + w.printf("\n") +} + +func (w *CodeWriter) writeValue(v reflect.Value) { + x := v.Interface() + switch v.Kind() { + case reflect.String: + w.WriteString(v.String()) + case reflect.Array: + // Don't double count: callers of WriteArray count on the size being + // added, so we need to discount it here. + w.Size -= int(v.Type().Size()) + w.writeSlice(x, true) + case reflect.Slice: + w.writeSlice(x, false) + case reflect.Struct: + w.printf("%s{\n", typeName(v.Interface())) + t := v.Type() + for i := 0; i < v.NumField(); i++ { + w.printf("%s: ", t.Field(i).Name) + w.writeValue(v.Field(i)) + w.printf(",\n") + } + w.printf("}") + default: + w.printf("%#v", x) + } +} + +// WriteString writes a string literal. +func (w *CodeWriter) WriteString(s string) { + s = strings.Replace(s, `\`, `\\`, -1) + io.WriteString(w.Hash, s) // content hash + w.Size += len(s) + + const maxInline = 40 + if len(s) <= maxInline { + w.printf("%q", s) + return + } + + // We will render the string as a multi-line string. + const maxWidth = 80 - 4 - len(`"`) - len(`" +`) + + // When starting on its own line, go fmt indents line 2+ an extra level. + n, max := maxWidth, maxWidth-4 + + // Print "" +\n, if a string does not start on its own line. + b := w.buf.Bytes() + if p := len(bytes.TrimRight(b, " \t")); p > 0 && b[p-1] != '\n' { + w.printf("\"\" + // Size: %d bytes\n", len(s)) + n, max = maxWidth, maxWidth + } + + w.printf(`"`) + + for sz, p := 0, 0; p < len(s); { + var r rune + r, sz = utf8.DecodeRuneInString(s[p:]) + out := s[p : p+sz] + chars := 1 + if !unicode.IsPrint(r) || r == utf8.RuneError || r == '"' { + switch sz { + case 1: + out = fmt.Sprintf("\\x%02x", s[p]) + case 2, 3: + out = fmt.Sprintf("\\u%04x", r) + case 4: + out = fmt.Sprintf("\\U%08x", r) + } + chars = len(out) + } + if n -= chars; n < 0 { + w.printf("\" +\n\"") + n = max - len(out) + } + w.printf("%s", out) + p += sz + } + w.printf(`"`) +} + +// WriteSlice writes a slice value. +func (w *CodeWriter) WriteSlice(x interface{}) { + w.writeSlice(x, false) +} + +// WriteArray writes an array value. +func (w *CodeWriter) WriteArray(x interface{}) { + w.writeSlice(x, true) +} + +func (w *CodeWriter) writeSlice(x interface{}, isArray bool) { + v := reflect.ValueOf(x) + w.gob.Encode(v.Len()) + w.Size += v.Len() * int(v.Type().Elem().Size()) + name := typeName(x) + if isArray { + name = fmt.Sprintf("[%d]%s", v.Len(), name[strings.Index(name, "]")+1:]) + } + if isArray { + w.printf("%s{\n", name) + } else { + w.printf("%s{ // %d elements\n", name, v.Len()) + } + + switch kind := v.Type().Elem().Kind(); kind { + case reflect.String: + for _, s := range x.([]string) { + w.WriteString(s) + w.printf(",\n") + } + case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64, + reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64: + // nLine and nBlock are the number of elements per line and block. + nLine, nBlock, format := 8, 64, "%d," + switch kind { + case reflect.Uint8: + format = "%#02x," + case reflect.Uint16: + format = "%#04x," + case reflect.Uint32: + nLine, nBlock, format = 4, 32, "%#08x," + case reflect.Uint, reflect.Uint64: + nLine, nBlock, format = 4, 32, "%#016x," + case reflect.Int8: + nLine = 16 + } + n := nLine + for i := 0; i < v.Len(); i++ { + if i%nBlock == 0 && v.Len() > nBlock { + w.printf("// Entry %X - %X\n", i, i+nBlock-1) + } + x := v.Index(i).Interface() + w.gob.Encode(x) + w.printf(format, x) + if n--; n == 0 { + n = nLine + w.printf("\n") + } + } + w.printf("\n") + case reflect.Struct: + zero := reflect.Zero(v.Type().Elem()).Interface() + for i := 0; i < v.Len(); i++ { + x := v.Index(i).Interface() + w.gob.EncodeValue(v) + if !reflect.DeepEqual(zero, x) { + line := fmt.Sprintf("%#v,\n", x) + line = line[strings.IndexByte(line, '{'):] + w.printf("%d: ", i) + w.printf(line) + } + } + case reflect.Array: + for i := 0; i < v.Len(); i++ { + w.printf("%d: %#v,\n", i, v.Index(i).Interface()) + } + default: + panic("gen: slice elem type not supported") + } + w.printf("}") +} + +// WriteType writes a definition of the type of the given value and returns the +// type name. +func (w *CodeWriter) WriteType(x interface{}) string { + t := reflect.TypeOf(x) + w.printf("type %s struct {\n", t.Name()) + for i := 0; i < t.NumField(); i++ { + w.printf("\t%s %s\n", t.Field(i).Name, t.Field(i).Type) + } + w.printf("}\n") + return t.Name() +} + +// typeName returns the name of the go type of x. +func typeName(x interface{}) string { + t := reflect.ValueOf(x).Type() + return strings.Replace(fmt.Sprint(t), "main.", "", 1) +} diff --git a/vendor/golang.org/x/text/internal/gen/gen.go b/vendor/golang.org/x/text/internal/gen/gen.go new file mode 100644 index 000000000..84c699faa --- /dev/null +++ b/vendor/golang.org/x/text/internal/gen/gen.go @@ -0,0 +1,281 @@ +// 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 gen contains common code for the various code generation tools in the +// text repository. Its usage ensures consistency between tools. +// +// This package defines command line flags that are common to most generation +// tools. The flags allow for specifying specific Unicode and CLDR versions +// in the public Unicode data repository (http://www.unicode.org/Public). +// +// A local Unicode data mirror can be set through the flag -local or the +// environment variable UNICODE_DIR. The former takes precedence. The local +// directory should follow the same structure as the public repository. +// +// IANA data can also optionally be mirrored by putting it in the iana directory +// rooted at the top of the local mirror. Beware, though, that IANA data is not +// versioned. So it is up to the developer to use the right version. +package gen // import "golang.org/x/text/internal/gen" + +import ( + "bytes" + "flag" + "fmt" + "go/build" + "go/format" + "io" + "io/ioutil" + "log" + "net/http" + "os" + "path" + "path/filepath" + "sync" + "unicode" + + "golang.org/x/text/unicode/cldr" +) + +var ( + url = flag.String("url", + "http://www.unicode.org/Public", + "URL of Unicode database directory") + iana = flag.String("iana", + "http://www.iana.org", + "URL of the IANA repository") + unicodeVersion = flag.String("unicode", + getEnv("UNICODE_VERSION", unicode.Version), + "unicode version to use") + cldrVersion = flag.String("cldr", + getEnv("CLDR_VERSION", cldr.Version), + "cldr version to use") +) + +func getEnv(name, def string) string { + if v := os.Getenv(name); v != "" { + return v + } + return def +} + +// Init performs common initialization for a gen command. It parses the flags +// and sets up the standard logging parameters. +func Init() { + log.SetPrefix("") + log.SetFlags(log.Lshortfile) + flag.Parse() +} + +const header = `// This file was generated by go generate; DO NOT EDIT + +package %s + +` + +// UnicodeVersion reports the requested Unicode version. +func UnicodeVersion() string { + return *unicodeVersion +} + +// UnicodeVersion reports the requested CLDR version. +func CLDRVersion() string { + return *cldrVersion +} + +// IsLocal reports whether data files are available locally. +func IsLocal() bool { + dir, err := localReadmeFile() + if err != nil { + return false + } + if _, err = os.Stat(dir); err != nil { + return false + } + return true +} + +// OpenUCDFile opens the requested UCD file. The file is specified relative to +// the public Unicode root directory. It will call log.Fatal if there are any +// errors. +func OpenUCDFile(file string) io.ReadCloser { + return openUnicode(path.Join(*unicodeVersion, "ucd", file)) +} + +// OpenCLDRCoreZip opens the CLDR core zip file. It will call log.Fatal if there +// are any errors. +func OpenCLDRCoreZip() io.ReadCloser { + return OpenUnicodeFile("cldr", *cldrVersion, "core.zip") +} + +// OpenUnicodeFile opens the requested file of the requested category from the +// root of the Unicode data archive. The file is specified relative to the +// public Unicode root directory. If version is "", it will use the default +// Unicode version. It will call log.Fatal if there are any errors. +func OpenUnicodeFile(category, version, file string) io.ReadCloser { + if version == "" { + version = UnicodeVersion() + } + return openUnicode(path.Join(category, version, file)) +} + +// OpenIANAFile opens the requested IANA file. The file is specified relative +// to the IANA root, which is typically either http://www.iana.org or the +// iana directory in the local mirror. It will call log.Fatal if there are any +// errors. +func OpenIANAFile(path string) io.ReadCloser { + return Open(*iana, "iana", path) +} + +var ( + dirMutex sync.Mutex + localDir string +) + +const permissions = 0755 + +func localReadmeFile() (string, error) { + p, err := build.Import("golang.org/x/text", "", build.FindOnly) + if err != nil { + return "", fmt.Errorf("Could not locate package: %v", err) + } + return filepath.Join(p.Dir, "DATA", "README"), nil +} + +func getLocalDir() string { + dirMutex.Lock() + defer dirMutex.Unlock() + + readme, err := localReadmeFile() + if err != nil { + log.Fatal(err) + } + dir := filepath.Dir(readme) + if _, err := os.Stat(readme); err != nil { + if err := os.MkdirAll(dir, permissions); err != nil { + log.Fatalf("Could not create directory: %v", err) + } + ioutil.WriteFile(readme, []byte(readmeTxt), permissions) + } + return dir +} + +const readmeTxt = `Generated by golang.org/x/text/internal/gen. DO NOT EDIT. + +This directory contains downloaded files used to generate the various tables +in the golang.org/x/text subrepo. + +Note that the language subtag repo (iana/assignments/language-subtag-registry) +and all other times in the iana subdirectory are not versioned and will need +to be periodically manually updated. The easiest way to do this is to remove +the entire iana directory. This is mostly of concern when updating the language +package. +` + +// Open opens subdir/path if a local directory is specified and the file exists, +// where subdir is a directory relative to the local root, or fetches it from +// urlRoot/path otherwise. It will call log.Fatal if there are any errors. +func Open(urlRoot, subdir, path string) io.ReadCloser { + file := filepath.Join(getLocalDir(), subdir, filepath.FromSlash(path)) + return open(file, urlRoot, path) +} + +func openUnicode(path string) io.ReadCloser { + file := filepath.Join(getLocalDir(), filepath.FromSlash(path)) + return open(file, *url, path) +} + +// TODO: automatically periodically update non-versioned files. + +func open(file, urlRoot, path string) io.ReadCloser { + if f, err := os.Open(file); err == nil { + return f + } + r := get(urlRoot, path) + defer r.Close() + b, err := ioutil.ReadAll(r) + if err != nil { + log.Fatalf("Could not download file: %v", err) + } + os.MkdirAll(filepath.Dir(file), permissions) + if err := ioutil.WriteFile(file, b, permissions); err != nil { + log.Fatalf("Could not create file: %v", err) + } + return ioutil.NopCloser(bytes.NewReader(b)) +} + +func get(root, path string) io.ReadCloser { + url := root + "/" + path + fmt.Printf("Fetching %s...", url) + defer fmt.Println(" done.") + resp, err := http.Get(url) + if err != nil { + log.Fatalf("HTTP GET: %v", err) + } + if resp.StatusCode != 200 { + log.Fatalf("Bad GET status for %q: %q", url, resp.Status) + } + return resp.Body +} + +// TODO: use Write*Version in all applicable packages. + +// WriteUnicodeVersion writes a constant for the Unicode version from which the +// tables are generated. +func WriteUnicodeVersion(w io.Writer) { + fmt.Fprintf(w, "// UnicodeVersion is the Unicode version from which the tables in this package are derived.\n") + fmt.Fprintf(w, "const UnicodeVersion = %q\n\n", UnicodeVersion()) +} + +// WriteCLDRVersion writes a constant for the CLDR version from which the +// tables are generated. +func WriteCLDRVersion(w io.Writer) { + fmt.Fprintf(w, "// CLDRVersion is the CLDR version from which the tables in this package are derived.\n") + fmt.Fprintf(w, "const CLDRVersion = %q\n\n", CLDRVersion()) +} + +// WriteGoFile prepends a standard file comment 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 WriteGoFile(filename, pkg string, b []byte) { + 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, 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)) + src = append(src, b...) + formatted, err := format.Source(src) + if err != nil { + // Print the generated code even in case of an error so that the + // returned error can be meaningfully interpreted. + n, _ = w.Write(src) + return n, err + } + return w.Write(formatted) +} + +// Repackage rewrites a Go file from belonging to package main to belonging to +// the given package. +func Repackage(inFile, outFile, pkg string) { + src, err := ioutil.ReadFile(inFile) + if err != nil { + log.Fatalf("reading %s: %v", inFile, err) + } + const toDelete = "package main\n\n" + i := bytes.Index(src, []byte(toDelete)) + if i < 0 { + log.Fatalf("Could not find %q in %s.", toDelete, inFile) + } + w := &bytes.Buffer{} + w.Write(src[i+len(toDelete):]) + WriteGoFile(outFile, pkg, w.Bytes()) +} diff --git a/vendor/golang.org/x/text/internal/tag/tag.go b/vendor/golang.org/x/text/internal/tag/tag.go new file mode 100644 index 000000000..b5d348891 --- /dev/null +++ b/vendor/golang.org/x/text/internal/tag/tag.go @@ -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 +} diff --git a/vendor/golang.org/x/text/internal/utf8internal/utf8internal.go b/vendor/golang.org/x/text/internal/utf8internal/utf8internal.go new file mode 100644 index 000000000..575cea870 --- /dev/null +++ b/vendor/golang.org/x/text/internal/utf8internal/utf8internal.go @@ -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}, +} |