func canonicalCase(field string) string {
upper := strings.ToUpper(field)
switch upper {
case "SHA1", "SHA256", "SHA512":
return upper
case "MD5SUM":
return "MD5Sum"
case "NOTAUTOMATIC":
return "NotAutomatic"
case "BUTAUTOMATICUPGRADES":
return "ButAutomaticUpgrades"
}
startOfWord := true
return strings.Map(func(r rune) rune {
if startOfWord {
startOfWord = false
return unicode.ToUpper(r)
}
if r == '-' {
startOfWord = true
}
return unicode.ToLower(r)
}, field)
}
func standardizedLocale(s string) string {
localeParts := strings.Split(s, "-")
language := localeParts[0]
regionOrScript := localeParts[1]
switch len(s) {
case 5:
newRegion := ""
for _, v := range regionOrScript {
newRegion += string(unicode.ToUpper(v))
}
return language + "_" + newRegion
case 7:
newScript := ""
for i, v := range regionOrScript {
if i == 0 {
newScript += string(unicode.ToUpper(v))
} else {
newScript += string(v)
}
}
return language + "-" + newScript
}
return s
}
// ToUpperCamelCase returns a copy of the string s with all Unicode letters mapped to their camel case.
// It will convert to upper case previous letter of '_' and first letter, and remove letter of '_'.
func ToUpperCamelCase(s string) string {
if s == "" {
return ""
}
upper := true
start := 0
result := make([]byte, 0, len(s))
var runeBuf [utf8.UTFMax]byte
var initialism []byte
for _, c := range s {
if c == '_' {
upper = true
candidate := string(result[start:])
initialism = initialism[:0]
for _, r := range candidate {
if r < utf8.RuneSelf {
initialism = append(initialism, toUpperASCII(byte(r)))
} else {
n := utf8.EncodeRune(runeBuf[:], unicode.ToUpper(r))
initialism = append(initialism, runeBuf[:n]...)
}
}
if length := commonInitialism.LookupByBytes(initialism); length > 0 {
result = append(result[:start], initialism...)
}
start = len(result)
continue
}
if upper {
if c < utf8.RuneSelf {
result = append(result, toUpperASCII(byte(c)))
} else {
n := utf8.EncodeRune(runeBuf[:], unicode.ToUpper(c))
result = append(result, runeBuf[:n]...)
}
upper = false
continue
}
if c < utf8.RuneSelf {
result = append(result, byte(c))
} else {
n := utf8.EncodeRune(runeBuf[:], c)
result = append(result, runeBuf[:n]...)
}
}
candidate := string(result[start:])
initialism = initialism[:0]
for _, r := range candidate {
if r < utf8.RuneSelf {
initialism = append(initialism, toUpperASCII(byte(r)))
} else {
n := utf8.EncodeRune(runeBuf[:], unicode.ToUpper(r))
initialism = append(initialism, runeBuf[:n]...)
}
}
if length := commonInitialism.LookupByBytes(initialism); length > 0 {
result = append(result[:start], initialism...)
}
return string(result)
}
func (g Garbler) uppercase(p string, numUppercase int) string {
if numUppercase <= 0 {
return p
}
b := []byte(p)
numsDone := 0
for i := 0; i < len(b); i++ {
//play nice with environ sequence,
//just uppercase 1st and 2nd consonants,
//which should make the whole thing more readable
if i%5 == 0 || i%5 == 2 {
b[i] = byte(unicode.ToUpper(rune(b[i])))
numsDone++
if numsDone >= numUppercase {
return string(b)
}
}
}
//playing nice didn't work out so do the other letters too
//in no particular order
for i := 0; i < len(b); i++ {
if !(i%5 == 0 || i%5 == 2) {
b[i] = byte(unicode.ToUpper(rune(b[i])))
numsDone++
if numsDone >= numUppercase {
return string(b)
}
}
}
//still here? then numUppercase was too large, panic
panic("ouch")
}
// This is what generates the actual markov chain
func markov(channel string, conn *irc.Conn) {
markovData.mutex.RLock()
var markovchain string
messageLength := random(50) + 10
for i := 0; i < messageLength; i++ {
splitchain := strings.Split(markovchain, " ")
if len(splitchain) < 2 {
s := []rune(markovData.keys[random(len(markovData.keys))])
s[0] = unicode.ToUpper(s[0])
markovchain = string(s)
continue
}
chainlength := len(splitchain)
searchfor := strings.ToLower(splitchain[chainlength-2] + " " + splitchain[chainlength-1])
if len(markovData.bigmap[searchfor]) == 0 || strings.LastIndex(markovchain, ".") < len(markovchain)-50 {
s := []rune(markovData.keys[random(len(markovData.keys))])
s[0] = unicode.ToUpper(s[0])
markovchain = markovchain + ". " + string(s)
continue
}
randnext := random(len(markovData.bigmap[searchfor]))
markovchain = markovchain + " " + markovData.bigmap[searchfor][randnext]
}
conn.Privmsg(channel, markovchain+".")
markovData.mutex.RUnlock()
}
func upperWordLetterPairs(runes []rune) ([]runeBigram, int) {
limit := len(runes) - 1
if limit < 1 {
return make([]runeBigram, 0), 0
}
bigrams := make([]runeBigram, limit)
var a rune
var b rune
numPairs := 0
for i := 0; i < limit; i++ {
a = runes[i]
b = runes[i+1]
if unicode.IsSpace(b) {
i++
continue
}
if unicode.IsSpace(a) {
continue
}
bigrams[numPairs] = runeBigram{rA: unicode.ToUpper(a), rB: unicode.ToUpper(b)}
numPairs++
}
bigrams = bigrams[0:numPairs]
return bigrams, numPairs
}
func defaultFileName(lang string, pkg *types.Package) string {
switch lang {
case "java":
if pkg == nil {
return "Universe.java"
}
firstRune, size := utf8.DecodeRuneInString(pkg.Name())
className := string(unicode.ToUpper(firstRune)) + pkg.Name()[size:]
return className + ".java"
case "go":
if pkg == nil {
return "go_universe.go"
}
return "go_" + pkg.Name() + ".go"
case "objc":
if pkg == nil {
return "GoUniverse.m"
}
firstRune, size := utf8.DecodeRuneInString(pkg.Name())
className := string(unicode.ToUpper(firstRune)) + pkg.Name()[size:]
return "Go" + className + ".m"
}
errorf("unknown target language: %q", lang)
os.Exit(exitStatus)
return ""
}
// Goify makes a valid Go identifier out of any string.
// It does that by removing any non letter and non digit character and by making sure the first
// character is a letter or "_".
// Goify produces a "CamelCase" version of the string, if firstUpper is true the first character
// of the identifier is uppercase otherwise it's lowercase.
func Goify(str string, firstUpper bool) string {
runes := []rune(str)
w, i := 0, 0 // index of start of word, scan
for i+1 <= len(runes) {
eow := false // whether we hit the end of a word
if i+1 == len(runes) {
eow = true
} else if !validIdentifier(runes[i]) {
// get rid of it
runes = append(runes[:i], runes[i+1:]...)
} else if runes[i+1] == '_' {
// underscore; shift the remainder forward over any run of underscores
eow = true
n := 1
for i+n+1 < len(runes) && runes[i+n+1] == '_' {
n++
}
copy(runes[i+1:], runes[i+n+1:])
runes = runes[:len(runes)-n]
} else if unicode.IsLower(runes[i]) && !unicode.IsLower(runes[i+1]) {
// lower->non-lower
eow = true
}
i++
if !eow {
continue
}
// [w,i] is a word.
word := string(runes[w:i])
// is it one of our initialisms?
if u := strings.ToUpper(word); commonInitialisms[u] {
if firstUpper {
u = strings.ToUpper(u)
}
// All the common initialisms are ASCII,
// so we can replace the bytes exactly.
copy(runes[w:], []rune(u))
} else if w > 0 && strings.ToLower(word) == word {
// already all lowercase, and not the first word, so uppercase the first character.
runes[w] = unicode.ToUpper(runes[w])
} else if w == 0 && strings.ToLower(word) == word && firstUpper {
runes[w] = unicode.ToUpper(runes[w])
}
if w == 0 && !firstUpper {
runes[w] = unicode.ToLower(runes[w])
}
//advance to next word
w = i
}
res := string(runes)
if _, ok := reserved[res]; ok {
res += "_"
}
return res
}
func skinkefy(word string) string {
isNoun := foundWord(Nouns, word)
isVerb := foundWord(Verbs, word)
isAdjective := foundWord(Adjectives, word)
if UseWordList && !isNoun && !isVerb && !isAdjective {
return word
}
camelCaseCheckR, _ := regexp.Compile("[\\p{Ll}][\\p{Lu}][\\p{L}]+")
camelCaseR, _ := regexp.Compile("([\\p{Lu}][\\p{Ll}]+|[\\p{Lu}]+)")
if camelCaseCheckR.MatchString(word) {
word = camelCaseR.ReplaceAllStringFunc(word, skinkefy)
} else {
if (rand.Intn(100) <= skinkeFactor &&
!IkkeSkinker.Contains(word) &&
len(word) >= skinkeLength) ||
Skinker.Contains(word) {
// Yes, we got a skinke, let's add the word.
Skinker.Append(word)
var skinkeCondition int
first, _ := utf8.DecodeRuneInString(word)
switch {
case strings.ToUpper(word) == word:
skinkeCondition = scFullUpper
case unicode.ToUpper(first) == first:
skinkeCondition = scCapitalised
default:
skinkeCondition = scRegular
}
// Get the last two letters of the word.
end, size := utf8.DecodeLastRuneInString(word)
end2, _ := utf8.DecodeLastRuneInString(word[:len(word)-size])
end = unicode.ToUpper(end)
end2 = unicode.ToUpper(end2)
// If it ends on R, it's plural. That's our rule!
// If it ends on ET/EN, it's definitive.
if isNoun {
word = oneSkinkeNoun(skinkeCondition, end == 'R', end2 == 'E' && (end == 'T' || end == 'N'))
}
if isVerb {
word = oneSkinkeVerb(skinkeCondition, end == 'R', end2 == 'D' && end == 'E')
}
if isAdjective {
word = oneSkinkeAdjective(skinkeCondition)
}
} else {
// Not a skinke word. If it appears again, it should remain
// not skinke.
IkkeSkinker.Append(word)
}
}
return word
}
func singleLiteralToCharClass(rx *syntax.Regexp) {
if rx.Op == syntax.OpLiteral && len(rx.Rune) == 1 {
char := rx.Rune[0]
if rx.Flags&syntax.FoldCase != 0 && unicode.ToLower(char) != unicode.ToUpper(char) {
l, h := unicode.ToLower(char), unicode.ToUpper(char)
rx.Rune = []rune{h, h, l, l}
rx.Rune0 = [...]rune{h, h}
} else {
rx.Rune = []rune{char, char}
rx.Rune0 = [...]rune{char, char}
}
rx.Op = syntax.OpCharClass
}
}
// Goify makes a valid Go identifier out of any string.
// It does that by removing any non letter and non digit character and by making sure the first
// character is a letter or "_".
// Goify produces a "CamelCase" version of the string, if firstUpper is true the first character
// of the identifier is uppercase otherwise it's lowercase.
func Goify(str string, firstUpper bool) string {
if firstUpper {
if val, ok := acros[str]; ok {
return val
}
for a, u := range acros {
p := a + "_"
if strings.HasPrefix(str, p) {
rest := strings.TrimPrefix(str, p)
res := Goify(rest, true)
return u + res
}
}
}
if str == "ok" && firstUpper {
return "OK"
} else if str == "id" && firstUpper {
return "ID"
}
var b bytes.Buffer
var firstWritten, nextUpper bool
for i := 0; i < len(str); i++ {
r := rune(str[i])
if r == '_' {
nextUpper = true
} else if unicode.IsLetter(r) || unicode.IsDigit(r) {
if !firstWritten {
if firstUpper {
r = unicode.ToUpper(r)
} else {
r = unicode.ToLower(r)
}
firstWritten = true
nextUpper = false
} else if nextUpper {
r = unicode.ToUpper(r)
nextUpper = false
}
b.WriteRune(r)
}
}
if b.Len() == 0 {
return "_v" // you have a better idea?
}
res := b.String()
if _, ok := reserved[res]; ok {
res += "_"
}
return res
}
func upperFirst(s string) string {
if s == "" {
return ""
}
r, n := utf8.DecodeRuneInString(s)
return string(unicode.ToUpper(r)) + s[n:]
}
func GenSentence(wordCnt int) string {
initWords()
words := genWordSlice(wordCnt)
firstWord := []rune(words[0])
words[0] = string(unicode.ToUpper(firstWord[0])) + string(firstWord[1:])
return strings.Join(words, " ") + "."
}
func New(file string, args []string) (*Script, error) {
if len(args)&1 == 1 {
return nil, errors.New("number of arguments for template script must be even")
}
sc := &Script{}
if len(args) != 0 {
sc.args = make(map[string]string)
for i := 0; i < len(args); i += 2 {
if len(args[i]) < 2 || args[i][0] != '-' {
return nil, errors.New("invalid flag name: " + args[i])
}
r, n := utf8.DecodeRuneInString(args[i][1:])
if r == utf8.RuneError {
return nil, errors.New("invalid flag name: " + args[i])
}
sc.args[string(unicode.ToUpper(r))+args[i][1+n:]] = args[i+1]
}
}
tmpl, err := template.New(filepath.Base(file)).Funcs(sc.funcs()).ParseFiles(file)
if err != nil {
return nil, err
}
sc.tmpl = tmpl
return sc, nil
}
func helpDocumentation(path string) {
w := new(bytes.Buffer)
w.WriteString(documentationHeader)
w.WriteString("\n/*\n")
if err := usageTmpl.Execute(w, commands); err != nil {
log.Fatal(err)
}
for _, cmd := range commands {
r, rlen := utf8.DecodeRuneInString(cmd.Short)
w.WriteString("\n\n")
w.WriteRune(unicode.ToUpper(r))
w.WriteString(cmd.Short[rlen:])
w.WriteString("\n\nUsage:\n\n\tgomobile " + cmd.Name)
if cmd.Usage != "" {
w.WriteRune(' ')
w.WriteString(cmd.Usage)
}
w.WriteRune('\n')
w.WriteString(cmd.Long)
}
w.WriteString("*/\npackage main // import \"golang.org/x/mobile/cmd/gomobile\"\n")
if err := ioutil.WriteFile(path, w.Bytes(), 0666); err != nil {
log.Fatal(err)
}
}
func toCamelCase(x string) string {
if len(x) == 0 {
return ""
}
output := make([]byte, 0)
uppercase := true
for len(x) > 0 {
v, size := utf8.DecodeRuneInString(x)
// If underscore, append and keep going.
if v == '_' {
uppercase = true
} else if unicode.IsLetter(v) {
if uppercase {
uppercase = false
buf := make([]byte, size)
utf8.EncodeRune(buf, unicode.ToUpper(v))
output = bytes.Join([][]byte{output, buf}, nil)
} else if unicode.IsUpper(v) {
buf := make([]byte, size)
utf8.EncodeRune(buf, v)
output = bytes.Join([][]byte{output, buf}, []byte("_"))
}
}
x = x[size:]
}
return string(output)
}
func isRegister(ident string) bool {
if len(ident) != 1 {
return false
}
ch := unicode.ToUpper(rune(ident[0]))
return (ch >= 'A' && ch <= 'C' || ch >= 'X' && ch <= 'Z' || ch == 'I' || ch == 'J')
}
func UpperFirstCharacter(s string) string {
r := []rune(s)
r[0] = unicode.ToUpper(r[0])
return string(r)
}
func camelize(str string) string {
runes := []rune(str)
w, i := 0, 0
for i+1 <= len(runes) {
eow := false
if i+1 == len(runes) {
eow = true
} else if !validIdentifier(runes[i]) {
runes = append(runes[:i], runes[i+1:]...)
} else if spacer(runes[i+1]) {
eow = true
n := 1
for i+n+1 < len(runes) && spacer(runes[i+n+1]) {
n++
}
copy(runes[i+1:], runes[i+n+1:])
runes = runes[:len(runes)-n]
} else if unicode.IsLower(runes[i]) && !unicode.IsLower(runes[i+1]) {
eow = true
}
i++
if !eow {
continue
}
runes[w] = unicode.ToUpper(runes[w])
w = i
}
return string(runes)
}
// capitalize changes the first letter of a string to upper case.
func capitalize(s string) string {
if len(s) == 0 {
return s
}
x, i := utf8.DecodeRuneInString(s)
return string(unicode.ToUpper(x)) + string(s[i:])
}
// ToLatin returns a string transliterated using provided mapping table.
// Runes that cannot be transliterated inserted as-is.
func ToLatin(s string, table map[int]string) string {
runes := []int(s)
out := make([]int, 0, len(s))
for i, rune := range runes {
if tr, ok := table[unicode.ToLower(rune)]; ok {
if tr == "" {
continue
}
if unicode.IsUpper(rune) {
// Correctly translate case of successive characters:
// ЩИ -> SCHI
// Щи -> Schi
if i+1 < len(runes) && !unicode.IsUpper(runes[i+1]) {
t := []int(tr)
t[0] = unicode.ToUpper(t[0])
out = append(out, t...)
continue
}
out = append(out, []int(strings.ToUpper(tr))...)
continue
}
out = append(out, []int(tr)...)
} else {
out = append(out, rune)
}
}
return string(out)
}
// SwapCase will swap characters case from upper to lower or lower to upper.
func SwapCase(str string) string {
var r rune
var size int
buf := &bytes.Buffer{}
for len(str) > 0 {
r, size = utf8.DecodeRuneInString(str)
switch {
case unicode.IsUpper(r):
buf.WriteRune(unicode.ToLower(r))
case unicode.IsLower(r):
buf.WriteRune(unicode.ToUpper(r))
default:
buf.WriteRune(r)
}
str = str[size:]
}
return buf.String()
}
func (t *Textonym) ReadFrom(r io.Reader) (n int64, err error) {
t.numberMap = make(map[string][]string)
buf := make([]byte, 0, 32)
sc := bufio.NewScanner(r)
sc.Split(bufio.ScanWords)
scan:
for sc.Scan() {
buf = buf[:0]
word := sc.Text()
// XXX we only bother approximating the number of bytes
// consumed. This isn't used in the calling code and was
// only included to match the io.ReaderFrom interface.
n += int64(len(word)) + 1
for _, r := range word {
d, ok := t.letterMap[unicode.ToUpper(r)]
if !ok {
//log.Printf("ignoring %q\n", word)
continue scan
}
buf = append(buf, d)
}
//log.Printf("scanned %q\n", word)
num := string(buf)
t.numberMap[num] = append(t.numberMap[num], word)
t.count++
if len(t.numberMap[num]) == 2 {
t.textonyms++
}
//log.Printf("%q → %v\t→ %v\n", word, num, t.numberMap[num])
}
return n, sc.Err()
}
// pascalCase combines the given words using PascalCase.
//
// If allowAllCaps is true, when an all-caps word that is not a known
// abbreviation is encountered, it is left unchanged. Otherwise, it is
// Titlecased.
func pascalCase(allowAllCaps bool, words ...string) string {
for i, chunk := range words {
if len(chunk) == 0 {
// foo__bar
continue
}
// known initalism
init := strings.ToUpper(chunk)
if _, ok := commonInitialisms[init]; ok {
words[i] = init
continue
}
// Was SCREAMING_SNAKE_CASE and not a known initialism so Titlecase it.
if isAllCaps(chunk) && !allowAllCaps {
// A single ALLCAPS word does not count as SCREAMING_SNAKE_CASE.
// There must be at least one underscore.
words[i] = strings.Title(strings.ToLower(chunk))
continue
}
// Just another word, but could already be camelCased somehow, so just
// change the first letter.
head, headIndex := utf8.DecodeRuneInString(chunk)
words[i] = string(unicode.ToUpper(head)) + string(chunk[headIndex:])
}
return strings.Join(words, "")
}
func Capitalize(s string) string {
if s == "" {
return ""
}
r, n := utf8.DecodeRuneInString(s)
return string(unicode.ToUpper(r)) + strings.ToLower(s[n:])
}
// Converts a string (e.g. a line from a file) into an array of tokens.
func TokenizeLine(line string, lowerCase bool, returnChars bool) (tokens []string) {
// Lower case everything if required.
if lowerCase {
line = strings.ToLower(line)
}
// Split line into characters.
if returnChars {
// Create a map of acceptable characters.
var okChars = make(map[rune]bool)
for _, rn := range "abcdefghijklmnopqrstuvwxyz0123456789 ,;:." {
okChars[rn] = true
okChars[unicode.ToUpper(rn)] = true
}
// Add rune to tokens if it is acceptable.
for _, rn := range line {
if okChars[rn] {
tokens = append(tokens, string(rn))
} else {
tokens = append(tokens, "XXX")
}
}
// Or else split line into "words" by splitting on space.
} else {
// Insert spaces around break character, remove repeated spaces, and then split on individual spaces.
splitchars, _ := regexp.Compile(`\b`)
multiplespaces, _ := regexp.Compile(` +`)
line = splitchars.ReplaceAllString(line, " $1 ")
line = multiplespaces.ReplaceAllString(line, " ")
tokens = strings.Split(line, " ")
}
return
}
// lintName returns a different name if it should be different.
func lintName(name string) (should string) {
// Fast path for simple cases: "_" and all lowercase.
if name == "_" {
return name
}
allLower := true
for _, r := range name {
if !unicode.IsLower(r) {
allLower = false
break
}
}
if allLower {
return name
}
// Split camelCase at any lower->upper transition, and split on underscores.
// Check each word for common initialisms.
runes := []rune(name)
w, i := 0, 0 // index of start of word, scan
for i+1 <= len(runes) {
eow := false // whether we hit the end of a word
if i+1 == len(runes) {
eow = true
} else if runes[i+1] == '_' {
// underscore; shift the remainder forward over any run of underscores
eow = true
n := 1
for i+n+1 < len(runes) && runes[i+n+1] == '_' {
n++
}
copy(runes[i+1:], runes[i+n+1:])
runes = runes[:len(runes)-n]
} else if unicode.IsLower(runes[i]) && !unicode.IsLower(runes[i+1]) {
// lower->non-lower
eow = true
}
i++
if !eow {
continue
}
// [w,i) is a word.
word := string(runes[w:i])
if u := strings.ToUpper(word); commonInitialisms[u] {
// Keep consistent case, which is lowercase only at the start.
if w == 0 && unicode.IsLower(runes[w]) {
u = strings.ToLower(u)
}
// All the common initialisms are ASCII,
// so we can replace the bytes exactly.
copy(runes[w:], []rune(u))
} else if w > 0 && strings.ToLower(word) == word {
// already all lowercase, and not the first word, so uppercase the first character.
runes[w] = unicode.ToUpper(runes[w])
}
w = i
}
return string(runes)
}
func removeInvalidCharacters(target string) string {
var ret bytes.Buffer
var channel chan rune
var previousInvalid bool
channel = make(chan rune)
previousInvalid = false
go getCharacterChannel(target, channel)
for character := range channel {
if previousInvalid {
character = unicode.ToUpper(character)
}
previousInvalid = !unicode.IsLetter(character) && !unicode.IsDigit(character)
if !previousInvalid {
ret.WriteRune(character)
}
}
return ret.String()
}
// LookupString attempts to convert a (modifiers, keycode) to an english string.
// It essentially implements the rules described at http://goo.gl/qum9q
// Namely, the bulleted list that describes how key syms should be interpreted
// when various modifiers are pressed.
// Note that we ignore the logic that asks us to check if particular key codes
// are mapped to particular modifiers (i.e., "XK_Caps_Lock" to "Lock" modifier).
// We just check if the modifiers are activated. That's good enough for me.
// XXX: We ignore num lock stuff.
// XXX: We ignore MODE SWITCH stuff. (i.e., we don't use group 2 key syms.)
func LookupString(xu *xgbutil.XUtil, mods uint16,
keycode xproto.Keycode) string {
k1, k2, _, _ := interpretSymList(xu, keycode)
shift := mods&xproto.ModMaskShift > 0
lock := mods&xproto.ModMaskLock > 0
switch {
case !shift && !lock:
return k1
case !shift && lock:
if len(k1) == 1 && unicode.IsLower(rune(k1[0])) {
return k2
} else {
return k1
}
case shift && lock:
if len(k2) == 1 && unicode.IsLower(rune(k2[0])) {
return string(unicode.ToUpper(rune(k2[0])))
} else {
return k2
}
case shift:
return k2
}
return ""
}
// Extract a sensible template name from the file path
func templateNameFromFilePath(p string) string {
name := path.Base(p)
if ext := path.Ext(name); ext != "" {
name = name[:len(name)-len(ext)]
}
return string(unicode.ToUpper(rune(name[0]))) + name[1:]
}
