1
0
Fork 0
mirror of https://github.com/miniflux/v2.git synced 2025-09-15 18:57:04 +00:00

Update dependencies

This commit is contained in:
Frédéric Guillot 2019-09-05 21:39:32 -07:00 committed by Frédéric Guillot
parent 456ebaf423
commit b94160df72
350 changed files with 119448 additions and 25865 deletions

View file

@ -1,712 +0,0 @@
// Copyright 2012 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.
// +build ignore
//go:generate go run gen.go
//go:generate go run gen.go -test
package main
import (
"bytes"
"flag"
"fmt"
"go/format"
"io/ioutil"
"math/rand"
"os"
"sort"
"strings"
)
// identifier converts s to a Go exported identifier.
// It converts "div" to "Div" and "accept-charset" to "AcceptCharset".
func identifier(s string) string {
b := make([]byte, 0, len(s))
cap := true
for _, c := range s {
if c == '-' {
cap = true
continue
}
if cap && 'a' <= c && c <= 'z' {
c -= 'a' - 'A'
}
cap = false
b = append(b, byte(c))
}
return string(b)
}
var test = flag.Bool("test", false, "generate table_test.go")
func genFile(name string, buf *bytes.Buffer) {
b, err := format.Source(buf.Bytes())
if err != nil {
fmt.Fprintln(os.Stderr, err)
os.Exit(1)
}
if err := ioutil.WriteFile(name, b, 0644); err != nil {
fmt.Fprintln(os.Stderr, err)
os.Exit(1)
}
}
func main() {
flag.Parse()
var all []string
all = append(all, elements...)
all = append(all, attributes...)
all = append(all, eventHandlers...)
all = append(all, extra...)
sort.Strings(all)
// uniq - lists have dups
w := 0
for _, s := range all {
if w == 0 || all[w-1] != s {
all[w] = s
w++
}
}
all = all[:w]
if *test {
var buf bytes.Buffer
fmt.Fprintln(&buf, "// Code generated by go generate gen.go; DO NOT EDIT.\n")
fmt.Fprintln(&buf, "//go:generate go run gen.go -test\n")
fmt.Fprintln(&buf, "package atom\n")
fmt.Fprintln(&buf, "var testAtomList = []string{")
for _, s := range all {
fmt.Fprintf(&buf, "\t%q,\n", s)
}
fmt.Fprintln(&buf, "}")
genFile("table_test.go", &buf)
return
}
// Find hash that minimizes table size.
var best *table
for i := 0; i < 1000000; i++ {
if best != nil && 1<<(best.k-1) < len(all) {
break
}
h := rand.Uint32()
for k := uint(0); k <= 16; k++ {
if best != nil && k >= best.k {
break
}
var t table
if t.init(h, k, all) {
best = &t
break
}
}
}
if best == nil {
fmt.Fprintf(os.Stderr, "failed to construct string table\n")
os.Exit(1)
}
// Lay out strings, using overlaps when possible.
layout := append([]string{}, all...)
// Remove strings that are substrings of other strings
for changed := true; changed; {
changed = false
for i, s := range layout {
if s == "" {
continue
}
for j, t := range layout {
if i != j && t != "" && strings.Contains(s, t) {
changed = true
layout[j] = ""
}
}
}
}
// Join strings where one suffix matches another prefix.
for {
// Find best i, j, k such that layout[i][len-k:] == layout[j][:k],
// maximizing overlap length k.
besti := -1
bestj := -1
bestk := 0
for i, s := range layout {
if s == "" {
continue
}
for j, t := range layout {
if i == j {
continue
}
for k := bestk + 1; k <= len(s) && k <= len(t); k++ {
if s[len(s)-k:] == t[:k] {
besti = i
bestj = j
bestk = k
}
}
}
}
if bestk > 0 {
layout[besti] += layout[bestj][bestk:]
layout[bestj] = ""
continue
}
break
}
text := strings.Join(layout, "")
atom := map[string]uint32{}
for _, s := range all {
off := strings.Index(text, s)
if off < 0 {
panic("lost string " + s)
}
atom[s] = uint32(off<<8 | len(s))
}
var buf bytes.Buffer
// Generate the Go code.
fmt.Fprintln(&buf, "// Code generated by go generate gen.go; DO NOT EDIT.\n")
fmt.Fprintln(&buf, "//go:generate go run gen.go\n")
fmt.Fprintln(&buf, "package atom\n\nconst (")
// compute max len
maxLen := 0
for _, s := range all {
if maxLen < len(s) {
maxLen = len(s)
}
fmt.Fprintf(&buf, "\t%s Atom = %#x\n", identifier(s), atom[s])
}
fmt.Fprintln(&buf, ")\n")
fmt.Fprintf(&buf, "const hash0 = %#x\n\n", best.h0)
fmt.Fprintf(&buf, "const maxAtomLen = %d\n\n", maxLen)
fmt.Fprintf(&buf, "var table = [1<<%d]Atom{\n", best.k)
for i, s := range best.tab {
if s == "" {
continue
}
fmt.Fprintf(&buf, "\t%#x: %#x, // %s\n", i, atom[s], s)
}
fmt.Fprintf(&buf, "}\n")
datasize := (1 << best.k) * 4
fmt.Fprintln(&buf, "const atomText =")
textsize := len(text)
for len(text) > 60 {
fmt.Fprintf(&buf, "\t%q +\n", text[:60])
text = text[60:]
}
fmt.Fprintf(&buf, "\t%q\n\n", text)
genFile("table.go", &buf)
fmt.Fprintf(os.Stdout, "%d atoms; %d string bytes + %d tables = %d total data\n", len(all), textsize, datasize, textsize+datasize)
}
type byLen []string
func (x byLen) Less(i, j int) bool { return len(x[i]) > len(x[j]) }
func (x byLen) Swap(i, j int) { x[i], x[j] = x[j], x[i] }
func (x byLen) Len() int { return len(x) }
// fnv computes the FNV hash with an arbitrary starting value h.
func fnv(h uint32, s string) uint32 {
for i := 0; i < len(s); i++ {
h ^= uint32(s[i])
h *= 16777619
}
return h
}
// A table represents an attempt at constructing the lookup table.
// The lookup table uses cuckoo hashing, meaning that each string
// can be found in one of two positions.
type table struct {
h0 uint32
k uint
mask uint32
tab []string
}
// hash returns the two hashes for s.
func (t *table) hash(s string) (h1, h2 uint32) {
h := fnv(t.h0, s)
h1 = h & t.mask
h2 = (h >> 16) & t.mask
return
}
// init initializes the table with the given parameters.
// h0 is the initial hash value,
// k is the number of bits of hash value to use, and
// x is the list of strings to store in the table.
// init returns false if the table cannot be constructed.
func (t *table) init(h0 uint32, k uint, x []string) bool {
t.h0 = h0
t.k = k
t.tab = make([]string, 1<<k)
t.mask = 1<<k - 1
for _, s := range x {
if !t.insert(s) {
return false
}
}
return true
}
// insert inserts s in the table.
func (t *table) insert(s string) bool {
h1, h2 := t.hash(s)
if t.tab[h1] == "" {
t.tab[h1] = s
return true
}
if t.tab[h2] == "" {
t.tab[h2] = s
return true
}
if t.push(h1, 0) {
t.tab[h1] = s
return true
}
if t.push(h2, 0) {
t.tab[h2] = s
return true
}
return false
}
// push attempts to push aside the entry in slot i.
func (t *table) push(i uint32, depth int) bool {
if depth > len(t.tab) {
return false
}
s := t.tab[i]
h1, h2 := t.hash(s)
j := h1 + h2 - i
if t.tab[j] != "" && !t.push(j, depth+1) {
return false
}
t.tab[j] = s
return true
}
// The lists of element names and attribute keys were taken from
// https://html.spec.whatwg.org/multipage/indices.html#index
// as of the "HTML Living Standard - Last Updated 16 April 2018" version.
// "command", "keygen" and "menuitem" have been removed from the spec,
// but are kept here for backwards compatibility.
var elements = []string{
"a",
"abbr",
"address",
"area",
"article",
"aside",
"audio",
"b",
"base",
"bdi",
"bdo",
"blockquote",
"body",
"br",
"button",
"canvas",
"caption",
"cite",
"code",
"col",
"colgroup",
"command",
"data",
"datalist",
"dd",
"del",
"details",
"dfn",
"dialog",
"div",
"dl",
"dt",
"em",
"embed",
"fieldset",
"figcaption",
"figure",
"footer",
"form",
"h1",
"h2",
"h3",
"h4",
"h5",
"h6",
"head",
"header",
"hgroup",
"hr",
"html",
"i",
"iframe",
"img",
"input",
"ins",
"kbd",
"keygen",
"label",
"legend",
"li",
"link",
"main",
"map",
"mark",
"menu",
"menuitem",
"meta",
"meter",
"nav",
"noscript",
"object",
"ol",
"optgroup",
"option",
"output",
"p",
"param",
"picture",
"pre",
"progress",
"q",
"rp",
"rt",
"ruby",
"s",
"samp",
"script",
"section",
"select",
"slot",
"small",
"source",
"span",
"strong",
"style",
"sub",
"summary",
"sup",
"table",
"tbody",
"td",
"template",
"textarea",
"tfoot",
"th",
"thead",
"time",
"title",
"tr",
"track",
"u",
"ul",
"var",
"video",
"wbr",
}
// https://html.spec.whatwg.org/multipage/indices.html#attributes-3
//
// "challenge", "command", "contextmenu", "dropzone", "icon", "keytype", "mediagroup",
// "radiogroup", "spellcheck", "scoped", "seamless", "sortable" and "sorted" have been removed from the spec,
// but are kept here for backwards compatibility.
var attributes = []string{
"abbr",
"accept",
"accept-charset",
"accesskey",
"action",
"allowfullscreen",
"allowpaymentrequest",
"allowusermedia",
"alt",
"as",
"async",
"autocomplete",
"autofocus",
"autoplay",
"challenge",
"charset",
"checked",
"cite",
"class",
"color",
"cols",
"colspan",
"command",
"content",
"contenteditable",
"contextmenu",
"controls",
"coords",
"crossorigin",
"data",
"datetime",
"default",
"defer",
"dir",
"dirname",
"disabled",
"download",
"draggable",
"dropzone",
"enctype",
"for",
"form",
"formaction",
"formenctype",
"formmethod",
"formnovalidate",
"formtarget",
"headers",
"height",
"hidden",
"high",
"href",
"hreflang",
"http-equiv",
"icon",
"id",
"inputmode",
"integrity",
"is",
"ismap",
"itemid",
"itemprop",
"itemref",
"itemscope",
"itemtype",
"keytype",
"kind",
"label",
"lang",
"list",
"loop",
"low",
"manifest",
"max",
"maxlength",
"media",
"mediagroup",
"method",
"min",
"minlength",
"multiple",
"muted",
"name",
"nomodule",
"nonce",
"novalidate",
"open",
"optimum",
"pattern",
"ping",
"placeholder",
"playsinline",
"poster",
"preload",
"radiogroup",
"readonly",
"referrerpolicy",
"rel",
"required",
"reversed",
"rows",
"rowspan",
"sandbox",
"spellcheck",
"scope",
"scoped",
"seamless",
"selected",
"shape",
"size",
"sizes",
"sortable",
"sorted",
"slot",
"span",
"spellcheck",
"src",
"srcdoc",
"srclang",
"srcset",
"start",
"step",
"style",
"tabindex",
"target",
"title",
"translate",
"type",
"typemustmatch",
"updateviacache",
"usemap",
"value",
"width",
"workertype",
"wrap",
}
// "onautocomplete", "onautocompleteerror", "onmousewheel",
// "onshow" and "onsort" have been removed from the spec,
// but are kept here for backwards compatibility.
var eventHandlers = []string{
"onabort",
"onautocomplete",
"onautocompleteerror",
"onauxclick",
"onafterprint",
"onbeforeprint",
"onbeforeunload",
"onblur",
"oncancel",
"oncanplay",
"oncanplaythrough",
"onchange",
"onclick",
"onclose",
"oncontextmenu",
"oncopy",
"oncuechange",
"oncut",
"ondblclick",
"ondrag",
"ondragend",
"ondragenter",
"ondragexit",
"ondragleave",
"ondragover",
"ondragstart",
"ondrop",
"ondurationchange",
"onemptied",
"onended",
"onerror",
"onfocus",
"onhashchange",
"oninput",
"oninvalid",
"onkeydown",
"onkeypress",
"onkeyup",
"onlanguagechange",
"onload",
"onloadeddata",
"onloadedmetadata",
"onloadend",
"onloadstart",
"onmessage",
"onmessageerror",
"onmousedown",
"onmouseenter",
"onmouseleave",
"onmousemove",
"onmouseout",
"onmouseover",
"onmouseup",
"onmousewheel",
"onwheel",
"onoffline",
"ononline",
"onpagehide",
"onpageshow",
"onpaste",
"onpause",
"onplay",
"onplaying",
"onpopstate",
"onprogress",
"onratechange",
"onreset",
"onresize",
"onrejectionhandled",
"onscroll",
"onsecuritypolicyviolation",
"onseeked",
"onseeking",
"onselect",
"onshow",
"onsort",
"onstalled",
"onstorage",
"onsubmit",
"onsuspend",
"ontimeupdate",
"ontoggle",
"onunhandledrejection",
"onunload",
"onvolumechange",
"onwaiting",
}
// extra are ad-hoc values not covered by any of the lists above.
var extra = []string{
"acronym",
"align",
"annotation",
"annotation-xml",
"applet",
"basefont",
"bgsound",
"big",
"blink",
"center",
"color",
"desc",
"face",
"font",
"foreignObject", // HTML is case-insensitive, but SVG-embedded-in-HTML is case-sensitive.
"foreignobject",
"frame",
"frameset",
"image",
"isindex",
"listing",
"malignmark",
"marquee",
"math",
"mglyph",
"mi",
"mn",
"mo",
"ms",
"mtext",
"nobr",
"noembed",
"noframes",
"plaintext",
"prompt",
"public",
"rb",
"rtc",
"spacer",
"strike",
"svg",
"system",
"tt",
"xmp",
}

127
vendor/golang.org/x/net/html/parse.go generated vendored
View file

@ -630,7 +630,16 @@ func inHeadIM(p *parser) bool {
p.oe.pop()
p.acknowledgeSelfClosingTag()
return true
case a.Script, a.Title, a.Noscript, a.Noframes, a.Style:
case a.Noscript:
p.addElement()
if p.scripting {
p.setOriginalIM()
p.im = textIM
} else {
p.im = inHeadNoscriptIM
}
return true
case a.Script, a.Title, a.Noframes, a.Style:
p.addElement()
p.setOriginalIM()
p.im = textIM
@ -692,6 +701,49 @@ func inHeadIM(p *parser) bool {
return false
}
// 12.2.6.4.5.
func inHeadNoscriptIM(p *parser) bool {
switch p.tok.Type {
case DoctypeToken:
// Ignore the token.
return true
case StartTagToken:
switch p.tok.DataAtom {
case a.Html:
return inBodyIM(p)
case a.Basefont, a.Bgsound, a.Link, a.Meta, a.Noframes, a.Style:
return inHeadIM(p)
case a.Head, a.Noscript:
// Ignore the token.
return true
}
case EndTagToken:
switch p.tok.DataAtom {
case a.Noscript, a.Br:
default:
// Ignore the token.
return true
}
case TextToken:
s := strings.TrimLeft(p.tok.Data, whitespace)
if len(s) == 0 {
// It was all whitespace.
return inHeadIM(p)
}
case CommentToken:
return inHeadIM(p)
}
p.oe.pop()
if p.top().DataAtom != a.Head {
panic("html: the new current node will be a head element.")
}
p.im = inHeadIM
if p.tok.DataAtom == a.Noscript {
return true
}
return false
}
// Section 12.2.6.4.6.
func afterHeadIM(p *parser) bool {
switch p.tok.Type {
@ -901,7 +953,7 @@ func inBodyIM(p *parser) bool {
case a.A:
for i := len(p.afe) - 1; i >= 0 && p.afe[i].Type != scopeMarkerNode; i-- {
if n := p.afe[i]; n.Type == ElementNode && n.DataAtom == a.A {
p.inBodyEndTagFormatting(a.A)
p.inBodyEndTagFormatting(a.A, "a")
p.oe.remove(n)
p.afe.remove(n)
break
@ -915,7 +967,7 @@ func inBodyIM(p *parser) bool {
case a.Nobr:
p.reconstructActiveFormattingElements()
if p.elementInScope(defaultScope, a.Nobr) {
p.inBodyEndTagFormatting(a.Nobr)
p.inBodyEndTagFormatting(a.Nobr, "nobr")
p.reconstructActiveFormattingElements()
}
p.addFormattingElement()
@ -1123,7 +1175,7 @@ func inBodyIM(p *parser) bool {
case a.H1, a.H2, a.H3, a.H4, a.H5, a.H6:
p.popUntil(defaultScope, a.H1, a.H2, a.H3, a.H4, a.H5, a.H6)
case a.A, a.B, a.Big, a.Code, a.Em, a.Font, a.I, a.Nobr, a.S, a.Small, a.Strike, a.Strong, a.Tt, a.U:
p.inBodyEndTagFormatting(p.tok.DataAtom)
p.inBodyEndTagFormatting(p.tok.DataAtom, p.tok.Data)
case a.Applet, a.Marquee, a.Object:
if p.popUntil(defaultScope, p.tok.DataAtom) {
p.clearActiveFormattingElements()
@ -1134,7 +1186,7 @@ func inBodyIM(p *parser) bool {
case a.Template:
return inHeadIM(p)
default:
p.inBodyEndTagOther(p.tok.DataAtom)
p.inBodyEndTagOther(p.tok.DataAtom, p.tok.Data)
}
case CommentToken:
p.addChild(&Node{
@ -1161,7 +1213,7 @@ func inBodyIM(p *parser) bool {
return true
}
func (p *parser) inBodyEndTagFormatting(tagAtom a.Atom) {
func (p *parser) inBodyEndTagFormatting(tagAtom a.Atom, tagName string) {
// This is the "adoption agency" algorithm, described at
// https://html.spec.whatwg.org/multipage/syntax.html#adoptionAgency
@ -1183,7 +1235,7 @@ func (p *parser) inBodyEndTagFormatting(tagAtom a.Atom) {
}
}
if formattingElement == nil {
p.inBodyEndTagOther(tagAtom)
p.inBodyEndTagOther(tagAtom, tagName)
return
}
feIndex := p.oe.index(formattingElement)
@ -1288,9 +1340,17 @@ func (p *parser) inBodyEndTagFormatting(tagAtom a.Atom) {
// inBodyEndTagOther performs the "any other end tag" algorithm for inBodyIM.
// "Any other end tag" handling from 12.2.6.5 The rules for parsing tokens in foreign content
// https://html.spec.whatwg.org/multipage/syntax.html#parsing-main-inforeign
func (p *parser) inBodyEndTagOther(tagAtom a.Atom) {
func (p *parser) inBodyEndTagOther(tagAtom a.Atom, tagName string) {
for i := len(p.oe) - 1; i >= 0; i-- {
if p.oe[i].DataAtom == tagAtom {
// Two element nodes have the same tag if they have the same Data (a
// string-typed field). As an optimization, for common HTML tags, each
// Data string is assigned a unique, non-zero DataAtom (a uint32-typed
// field), since integer comparison is faster than string comparison.
// Uncommon (custom) tags get a zero DataAtom.
//
// The if condition here is equivalent to (p.oe[i].Data == tagName).
if (p.oe[i].DataAtom == tagAtom) &&
((tagAtom != 0) || (p.oe[i].Data == tagName)) {
p.oe = p.oe[:i]
break
}
@ -1684,8 +1744,9 @@ func inCellIM(p *parser) bool {
return true
}
// Close the cell and reprocess.
p.popUntil(tableScope, a.Td, a.Th)
p.clearActiveFormattingElements()
if p.popUntil(tableScope, a.Td, a.Th) {
p.clearActiveFormattingElements()
}
p.im = inRowIM
return false
}
@ -2239,6 +2300,33 @@ func (p *parser) parse() error {
//
// The input is assumed to be UTF-8 encoded.
func Parse(r io.Reader) (*Node, error) {
return ParseWithOptions(r)
}
// ParseFragment parses a fragment of HTML and returns the nodes that were
// found. If the fragment is the InnerHTML for an existing element, pass that
// element in context.
//
// It has the same intricacies as Parse.
func ParseFragment(r io.Reader, context *Node) ([]*Node, error) {
return ParseFragmentWithOptions(r, context)
}
// ParseOption configures a parser.
type ParseOption func(p *parser)
// ParseOptionEnableScripting configures the scripting flag.
// https://html.spec.whatwg.org/multipage/webappapis.html#enabling-and-disabling-scripting
//
// By default, scripting is enabled.
func ParseOptionEnableScripting(enable bool) ParseOption {
return func(p *parser) {
p.scripting = enable
}
}
// ParseWithOptions is like Parse, with options.
func ParseWithOptions(r io.Reader, opts ...ParseOption) (*Node, error) {
p := &parser{
tokenizer: NewTokenizer(r),
doc: &Node{
@ -2248,6 +2336,11 @@ func Parse(r io.Reader) (*Node, error) {
framesetOK: true,
im: initialIM,
}
for _, f := range opts {
f(p)
}
err := p.parse()
if err != nil {
return nil, err
@ -2255,12 +2348,8 @@ func Parse(r io.Reader) (*Node, error) {
return p.doc, nil
}
// ParseFragment parses a fragment of HTML and returns the nodes that were
// found. If the fragment is the InnerHTML for an existing element, pass that
// element in context.
//
// It has the same intricacies as Parse.
func ParseFragment(r io.Reader, context *Node) ([]*Node, error) {
// ParseFragmentWithOptions is like ParseFragment, with options.
func ParseFragmentWithOptions(r io.Reader, context *Node, opts ...ParseOption) ([]*Node, error) {
contextTag := ""
if context != nil {
if context.Type != ElementNode {
@ -2284,6 +2373,10 @@ func ParseFragment(r io.Reader, context *Node) ([]*Node, error) {
context: context,
}
for _, f := range opts {
f(p)
}
root := &Node{
Type: ElementNode,
DataAtom: a.Html,

734
vendor/golang.org/x/net/idna/idna10.0.0.go generated vendored Normal file
View file

@ -0,0 +1,734 @@
// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
// Copyright 2016 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.
// +build go1.10
// Package idna implements IDNA2008 using the compatibility processing
// defined by UTS (Unicode Technical Standard) #46, which defines a standard to
// deal with the transition from IDNA2003.
//
// IDNA2008 (Internationalized Domain Names for Applications), is defined in RFC
// 5890, RFC 5891, RFC 5892, RFC 5893 and RFC 5894.
// UTS #46 is defined in https://www.unicode.org/reports/tr46.
// See https://unicode.org/cldr/utility/idna.jsp for a visualization of the
// differences between these two standards.
package idna // import "golang.org/x/net/idna"
import (
"fmt"
"strings"
"unicode/utf8"
"golang.org/x/text/secure/bidirule"
"golang.org/x/text/unicode/bidi"
"golang.org/x/text/unicode/norm"
)
// NOTE: Unlike common practice in Go APIs, the functions will return a
// sanitized domain name in case of errors. Browsers sometimes use a partially
// evaluated string as lookup.
// TODO: the current error handling is, in my opinion, the least opinionated.
// Other strategies are also viable, though:
// Option 1) Return an empty string in case of error, but allow the user to
// specify explicitly which errors to ignore.
// Option 2) Return the partially evaluated string if it is itself a valid
// string, otherwise return the empty string in case of error.
// Option 3) Option 1 and 2.
// Option 4) Always return an empty string for now and implement Option 1 as
// needed, and document that the return string may not be empty in case of
// error in the future.
// I think Option 1 is best, but it is quite opinionated.
// ToASCII is a wrapper for Punycode.ToASCII.
func ToASCII(s string) (string, error) {
return Punycode.process(s, true)
}
// ToUnicode is a wrapper for Punycode.ToUnicode.
func ToUnicode(s string) (string, error) {
return Punycode.process(s, false)
}
// An Option configures a Profile at creation time.
type Option func(*options)
// Transitional sets a Profile to use the Transitional mapping as defined in UTS
// #46. This will cause, for example, "ß" to be mapped to "ss". Using the
// transitional mapping provides a compromise between IDNA2003 and IDNA2008
// compatibility. It is used by most browsers when resolving domain names. This
// option is only meaningful if combined with MapForLookup.
func Transitional(transitional bool) Option {
return func(o *options) { o.transitional = true }
}
// VerifyDNSLength sets whether a Profile should fail if any of the IDN parts
// are longer than allowed by the RFC.
func VerifyDNSLength(verify bool) Option {
return func(o *options) { o.verifyDNSLength = verify }
}
// RemoveLeadingDots removes leading label separators. Leading runes that map to
// dots, such as U+3002 IDEOGRAPHIC FULL STOP, are removed as well.
//
// This is the behavior suggested by the UTS #46 and is adopted by some
// browsers.
func RemoveLeadingDots(remove bool) Option {
return func(o *options) { o.removeLeadingDots = remove }
}
// ValidateLabels sets whether to check the mandatory label validation criteria
// as defined in Section 5.4 of RFC 5891. This includes testing for correct use
// of hyphens ('-'), normalization, validity of runes, and the context rules.
func ValidateLabels(enable bool) Option {
return func(o *options) {
// Don't override existing mappings, but set one that at least checks
// normalization if it is not set.
if o.mapping == nil && enable {
o.mapping = normalize
}
o.trie = trie
o.validateLabels = enable
o.fromPuny = validateFromPunycode
}
}
// StrictDomainName limits the set of permissible ASCII characters to those
// allowed in domain names as defined in RFC 1034 (A-Z, a-z, 0-9 and the
// hyphen). This is set by default for MapForLookup and ValidateForRegistration.
//
// This option is useful, for instance, for browsers that allow characters
// outside this range, for example a '_' (U+005F LOW LINE). See
// http://www.rfc-editor.org/std/std3.txt for more details This option
// corresponds to the UseSTD3ASCIIRules option in UTS #46.
func StrictDomainName(use bool) Option {
return func(o *options) {
o.trie = trie
o.useSTD3Rules = use
o.fromPuny = validateFromPunycode
}
}
// NOTE: the following options pull in tables. The tables should not be linked
// in as long as the options are not used.
// BidiRule enables the Bidi rule as defined in RFC 5893. Any application
// that relies on proper validation of labels should include this rule.
func BidiRule() Option {
return func(o *options) { o.bidirule = bidirule.ValidString }
}
// ValidateForRegistration sets validation options to verify that a given IDN is
// properly formatted for registration as defined by Section 4 of RFC 5891.
func ValidateForRegistration() Option {
return func(o *options) {
o.mapping = validateRegistration
StrictDomainName(true)(o)
ValidateLabels(true)(o)
VerifyDNSLength(true)(o)
BidiRule()(o)
}
}
// MapForLookup sets validation and mapping options such that a given IDN is
// transformed for domain name lookup according to the requirements set out in
// Section 5 of RFC 5891. The mappings follow the recommendations of RFC 5894,
// RFC 5895 and UTS 46. It does not add the Bidi Rule. Use the BidiRule option
// to add this check.
//
// The mappings include normalization and mapping case, width and other
// compatibility mappings.
func MapForLookup() Option {
return func(o *options) {
o.mapping = validateAndMap
StrictDomainName(true)(o)
ValidateLabels(true)(o)
}
}
type options struct {
transitional bool
useSTD3Rules bool
validateLabels bool
verifyDNSLength bool
removeLeadingDots bool
trie *idnaTrie
// fromPuny calls validation rules when converting A-labels to U-labels.
fromPuny func(p *Profile, s string) error
// mapping implements a validation and mapping step as defined in RFC 5895
// or UTS 46, tailored to, for example, domain registration or lookup.
mapping func(p *Profile, s string) (mapped string, isBidi bool, err error)
// bidirule, if specified, checks whether s conforms to the Bidi Rule
// defined in RFC 5893.
bidirule func(s string) bool
}
// A Profile defines the configuration of an IDNA mapper.
type Profile struct {
options
}
func apply(o *options, opts []Option) {
for _, f := range opts {
f(o)
}
}
// New creates a new Profile.
//
// With no options, the returned Profile is the most permissive and equals the
// Punycode Profile. Options can be passed to further restrict the Profile. The
// MapForLookup and ValidateForRegistration options set a collection of options,
// for lookup and registration purposes respectively, which can be tailored by
// adding more fine-grained options, where later options override earlier
// options.
func New(o ...Option) *Profile {
p := &Profile{}
apply(&p.options, o)
return p
}
// ToASCII converts a domain or domain label to its ASCII form. For example,
// ToASCII("bücher.example.com") is "xn--bcher-kva.example.com", and
// ToASCII("golang") is "golang". If an error is encountered it will return
// an error and a (partially) processed result.
func (p *Profile) ToASCII(s string) (string, error) {
return p.process(s, true)
}
// ToUnicode converts a domain or domain label to its Unicode form. For example,
// ToUnicode("xn--bcher-kva.example.com") is "bücher.example.com", and
// ToUnicode("golang") is "golang". If an error is encountered it will return
// an error and a (partially) processed result.
func (p *Profile) ToUnicode(s string) (string, error) {
pp := *p
pp.transitional = false
return pp.process(s, false)
}
// String reports a string with a description of the profile for debugging
// purposes. The string format may change with different versions.
func (p *Profile) String() string {
s := ""
if p.transitional {
s = "Transitional"
} else {
s = "NonTransitional"
}
if p.useSTD3Rules {
s += ":UseSTD3Rules"
}
if p.validateLabels {
s += ":ValidateLabels"
}
if p.verifyDNSLength {
s += ":VerifyDNSLength"
}
return s
}
var (
// Punycode is a Profile that does raw punycode processing with a minimum
// of validation.
Punycode *Profile = punycode
// Lookup is the recommended profile for looking up domain names, according
// to Section 5 of RFC 5891. The exact configuration of this profile may
// change over time.
Lookup *Profile = lookup
// Display is the recommended profile for displaying domain names.
// The configuration of this profile may change over time.
Display *Profile = display
// Registration is the recommended profile for checking whether a given
// IDN is valid for registration, according to Section 4 of RFC 5891.
Registration *Profile = registration
punycode = &Profile{}
lookup = &Profile{options{
transitional: true,
useSTD3Rules: true,
validateLabels: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateAndMap,
bidirule: bidirule.ValidString,
}}
display = &Profile{options{
useSTD3Rules: true,
validateLabels: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateAndMap,
bidirule: bidirule.ValidString,
}}
registration = &Profile{options{
useSTD3Rules: true,
validateLabels: true,
verifyDNSLength: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateRegistration,
bidirule: bidirule.ValidString,
}}
// TODO: profiles
// Register: recommended for approving domain names: don't do any mappings
// but rather reject on invalid input. Bundle or block deviation characters.
)
type labelError struct{ label, code_ string }
func (e labelError) code() string { return e.code_ }
func (e labelError) Error() string {
return fmt.Sprintf("idna: invalid label %q", e.label)
}
type runeError rune
func (e runeError) code() string { return "P1" }
func (e runeError) Error() string {
return fmt.Sprintf("idna: disallowed rune %U", e)
}
// process implements the algorithm described in section 4 of UTS #46,
// see https://www.unicode.org/reports/tr46.
func (p *Profile) process(s string, toASCII bool) (string, error) {
var err error
var isBidi bool
if p.mapping != nil {
s, isBidi, err = p.mapping(p, s)
}
// Remove leading empty labels.
if p.removeLeadingDots {
for ; len(s) > 0 && s[0] == '.'; s = s[1:] {
}
}
// TODO: allow for a quick check of the tables data.
// It seems like we should only create this error on ToASCII, but the
// UTS 46 conformance tests suggests we should always check this.
if err == nil && p.verifyDNSLength && s == "" {
err = &labelError{s, "A4"}
}
labels := labelIter{orig: s}
for ; !labels.done(); labels.next() {
label := labels.label()
if label == "" {
// Empty labels are not okay. The label iterator skips the last
// label if it is empty.
if err == nil && p.verifyDNSLength {
err = &labelError{s, "A4"}
}
continue
}
if strings.HasPrefix(label, acePrefix) {
u, err2 := decode(label[len(acePrefix):])
if err2 != nil {
if err == nil {
err = err2
}
// Spec says keep the old label.
continue
}
isBidi = isBidi || bidirule.DirectionString(u) != bidi.LeftToRight
labels.set(u)
if err == nil && p.validateLabels {
err = p.fromPuny(p, u)
}
if err == nil {
// This should be called on NonTransitional, according to the
// spec, but that currently does not have any effect. Use the
// original profile to preserve options.
err = p.validateLabel(u)
}
} else if err == nil {
err = p.validateLabel(label)
}
}
if isBidi && p.bidirule != nil && err == nil {
for labels.reset(); !labels.done(); labels.next() {
if !p.bidirule(labels.label()) {
err = &labelError{s, "B"}
break
}
}
}
if toASCII {
for labels.reset(); !labels.done(); labels.next() {
label := labels.label()
if !ascii(label) {
a, err2 := encode(acePrefix, label)
if err == nil {
err = err2
}
label = a
labels.set(a)
}
n := len(label)
if p.verifyDNSLength && err == nil && (n == 0 || n > 63) {
err = &labelError{label, "A4"}
}
}
}
s = labels.result()
if toASCII && p.verifyDNSLength && err == nil {
// Compute the length of the domain name minus the root label and its dot.
n := len(s)
if n > 0 && s[n-1] == '.' {
n--
}
if len(s) < 1 || n > 253 {
err = &labelError{s, "A4"}
}
}
return s, err
}
func normalize(p *Profile, s string) (mapped string, isBidi bool, err error) {
// TODO: consider first doing a quick check to see if any of these checks
// need to be done. This will make it slower in the general case, but
// faster in the common case.
mapped = norm.NFC.String(s)
isBidi = bidirule.DirectionString(mapped) == bidi.RightToLeft
return mapped, isBidi, nil
}
func validateRegistration(p *Profile, s string) (idem string, bidi bool, err error) {
// TODO: filter need for normalization in loop below.
if !norm.NFC.IsNormalString(s) {
return s, false, &labelError{s, "V1"}
}
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
if sz == 0 {
return s, bidi, runeError(utf8.RuneError)
}
bidi = bidi || info(v).isBidi(s[i:])
// Copy bytes not copied so far.
switch p.simplify(info(v).category()) {
// TODO: handle the NV8 defined in the Unicode idna data set to allow
// for strict conformance to IDNA2008.
case valid, deviation:
case disallowed, mapped, unknown, ignored:
r, _ := utf8.DecodeRuneInString(s[i:])
return s, bidi, runeError(r)
}
i += sz
}
return s, bidi, nil
}
func (c info) isBidi(s string) bool {
if !c.isMapped() {
return c&attributesMask == rtl
}
// TODO: also store bidi info for mapped data. This is possible, but a bit
// cumbersome and not for the common case.
p, _ := bidi.LookupString(s)
switch p.Class() {
case bidi.R, bidi.AL, bidi.AN:
return true
}
return false
}
func validateAndMap(p *Profile, s string) (vm string, bidi bool, err error) {
var (
b []byte
k int
)
// combinedInfoBits contains the or-ed bits of all runes. We use this
// to derive the mayNeedNorm bit later. This may trigger normalization
// overeagerly, but it will not do so in the common case. The end result
// is another 10% saving on BenchmarkProfile for the common case.
var combinedInfoBits info
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
if sz == 0 {
b = append(b, s[k:i]...)
b = append(b, "\ufffd"...)
k = len(s)
if err == nil {
err = runeError(utf8.RuneError)
}
break
}
combinedInfoBits |= info(v)
bidi = bidi || info(v).isBidi(s[i:])
start := i
i += sz
// Copy bytes not copied so far.
switch p.simplify(info(v).category()) {
case valid:
continue
case disallowed:
if err == nil {
r, _ := utf8.DecodeRuneInString(s[start:])
err = runeError(r)
}
continue
case mapped, deviation:
b = append(b, s[k:start]...)
b = info(v).appendMapping(b, s[start:i])
case ignored:
b = append(b, s[k:start]...)
// drop the rune
case unknown:
b = append(b, s[k:start]...)
b = append(b, "\ufffd"...)
}
k = i
}
if k == 0 {
// No changes so far.
if combinedInfoBits&mayNeedNorm != 0 {
s = norm.NFC.String(s)
}
} else {
b = append(b, s[k:]...)
if norm.NFC.QuickSpan(b) != len(b) {
b = norm.NFC.Bytes(b)
}
// TODO: the punycode converters require strings as input.
s = string(b)
}
return s, bidi, err
}
// A labelIter allows iterating over domain name labels.
type labelIter struct {
orig string
slice []string
curStart int
curEnd int
i int
}
func (l *labelIter) reset() {
l.curStart = 0
l.curEnd = 0
l.i = 0
}
func (l *labelIter) done() bool {
return l.curStart >= len(l.orig)
}
func (l *labelIter) result() string {
if l.slice != nil {
return strings.Join(l.slice, ".")
}
return l.orig
}
func (l *labelIter) label() string {
if l.slice != nil {
return l.slice[l.i]
}
p := strings.IndexByte(l.orig[l.curStart:], '.')
l.curEnd = l.curStart + p
if p == -1 {
l.curEnd = len(l.orig)
}
return l.orig[l.curStart:l.curEnd]
}
// next sets the value to the next label. It skips the last label if it is empty.
func (l *labelIter) next() {
l.i++
if l.slice != nil {
if l.i >= len(l.slice) || l.i == len(l.slice)-1 && l.slice[l.i] == "" {
l.curStart = len(l.orig)
}
} else {
l.curStart = l.curEnd + 1
if l.curStart == len(l.orig)-1 && l.orig[l.curStart] == '.' {
l.curStart = len(l.orig)
}
}
}
func (l *labelIter) set(s string) {
if l.slice == nil {
l.slice = strings.Split(l.orig, ".")
}
l.slice[l.i] = s
}
// acePrefix is the ASCII Compatible Encoding prefix.
const acePrefix = "xn--"
func (p *Profile) simplify(cat category) category {
switch cat {
case disallowedSTD3Mapped:
if p.useSTD3Rules {
cat = disallowed
} else {
cat = mapped
}
case disallowedSTD3Valid:
if p.useSTD3Rules {
cat = disallowed
} else {
cat = valid
}
case deviation:
if !p.transitional {
cat = valid
}
case validNV8, validXV8:
// TODO: handle V2008
cat = valid
}
return cat
}
func validateFromPunycode(p *Profile, s string) error {
if !norm.NFC.IsNormalString(s) {
return &labelError{s, "V1"}
}
// TODO: detect whether string may have to be normalized in the following
// loop.
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
if sz == 0 {
return runeError(utf8.RuneError)
}
if c := p.simplify(info(v).category()); c != valid && c != deviation {
return &labelError{s, "V6"}
}
i += sz
}
return nil
}
const (
zwnj = "\u200c"
zwj = "\u200d"
)
type joinState int8
const (
stateStart joinState = iota
stateVirama
stateBefore
stateBeforeVirama
stateAfter
stateFAIL
)
var joinStates = [][numJoinTypes]joinState{
stateStart: {
joiningL: stateBefore,
joiningD: stateBefore,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateVirama,
},
stateVirama: {
joiningL: stateBefore,
joiningD: stateBefore,
},
stateBefore: {
joiningL: stateBefore,
joiningD: stateBefore,
joiningT: stateBefore,
joinZWNJ: stateAfter,
joinZWJ: stateFAIL,
joinVirama: stateBeforeVirama,
},
stateBeforeVirama: {
joiningL: stateBefore,
joiningD: stateBefore,
joiningT: stateBefore,
},
stateAfter: {
joiningL: stateFAIL,
joiningD: stateBefore,
joiningT: stateAfter,
joiningR: stateStart,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateAfter, // no-op as we can't accept joiners here
},
stateFAIL: {
0: stateFAIL,
joiningL: stateFAIL,
joiningD: stateFAIL,
joiningT: stateFAIL,
joiningR: stateFAIL,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateFAIL,
},
}
// validateLabel validates the criteria from Section 4.1. Item 1, 4, and 6 are
// already implicitly satisfied by the overall implementation.
func (p *Profile) validateLabel(s string) (err error) {
if s == "" {
if p.verifyDNSLength {
return &labelError{s, "A4"}
}
return nil
}
if !p.validateLabels {
return nil
}
trie := p.trie // p.validateLabels is only set if trie is set.
if len(s) > 4 && s[2] == '-' && s[3] == '-' {
return &labelError{s, "V2"}
}
if s[0] == '-' || s[len(s)-1] == '-' {
return &labelError{s, "V3"}
}
// TODO: merge the use of this in the trie.
v, sz := trie.lookupString(s)
x := info(v)
if x.isModifier() {
return &labelError{s, "V5"}
}
// Quickly return in the absence of zero-width (non) joiners.
if strings.Index(s, zwj) == -1 && strings.Index(s, zwnj) == -1 {
return nil
}
st := stateStart
for i := 0; ; {
jt := x.joinType()
if s[i:i+sz] == zwj {
jt = joinZWJ
} else if s[i:i+sz] == zwnj {
jt = joinZWNJ
}
st = joinStates[st][jt]
if x.isViramaModifier() {
st = joinStates[st][joinVirama]
}
if i += sz; i == len(s) {
break
}
v, sz = trie.lookupString(s[i:])
x = info(v)
}
if st == stateFAIL || st == stateAfter {
return &labelError{s, "C"}
}
return nil
}
func ascii(s string) bool {
for i := 0; i < len(s); i++ {
if s[i] >= utf8.RuneSelf {
return false
}
}
return true
}

682
vendor/golang.org/x/net/idna/idna9.0.0.go generated vendored Normal file
View file

@ -0,0 +1,682 @@
// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
// Copyright 2016 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.
// +build !go1.10
// Package idna implements IDNA2008 using the compatibility processing
// defined by UTS (Unicode Technical Standard) #46, which defines a standard to
// deal with the transition from IDNA2003.
//
// IDNA2008 (Internationalized Domain Names for Applications), is defined in RFC
// 5890, RFC 5891, RFC 5892, RFC 5893 and RFC 5894.
// UTS #46 is defined in https://www.unicode.org/reports/tr46.
// See https://unicode.org/cldr/utility/idna.jsp for a visualization of the
// differences between these two standards.
package idna // import "golang.org/x/net/idna"
import (
"fmt"
"strings"
"unicode/utf8"
"golang.org/x/text/secure/bidirule"
"golang.org/x/text/unicode/norm"
)
// NOTE: Unlike common practice in Go APIs, the functions will return a
// sanitized domain name in case of errors. Browsers sometimes use a partially
// evaluated string as lookup.
// TODO: the current error handling is, in my opinion, the least opinionated.
// Other strategies are also viable, though:
// Option 1) Return an empty string in case of error, but allow the user to
// specify explicitly which errors to ignore.
// Option 2) Return the partially evaluated string if it is itself a valid
// string, otherwise return the empty string in case of error.
// Option 3) Option 1 and 2.
// Option 4) Always return an empty string for now and implement Option 1 as
// needed, and document that the return string may not be empty in case of
// error in the future.
// I think Option 1 is best, but it is quite opinionated.
// ToASCII is a wrapper for Punycode.ToASCII.
func ToASCII(s string) (string, error) {
return Punycode.process(s, true)
}
// ToUnicode is a wrapper for Punycode.ToUnicode.
func ToUnicode(s string) (string, error) {
return Punycode.process(s, false)
}
// An Option configures a Profile at creation time.
type Option func(*options)
// Transitional sets a Profile to use the Transitional mapping as defined in UTS
// #46. This will cause, for example, "ß" to be mapped to "ss". Using the
// transitional mapping provides a compromise between IDNA2003 and IDNA2008
// compatibility. It is used by most browsers when resolving domain names. This
// option is only meaningful if combined with MapForLookup.
func Transitional(transitional bool) Option {
return func(o *options) { o.transitional = true }
}
// VerifyDNSLength sets whether a Profile should fail if any of the IDN parts
// are longer than allowed by the RFC.
func VerifyDNSLength(verify bool) Option {
return func(o *options) { o.verifyDNSLength = verify }
}
// RemoveLeadingDots removes leading label separators. Leading runes that map to
// dots, such as U+3002 IDEOGRAPHIC FULL STOP, are removed as well.
//
// This is the behavior suggested by the UTS #46 and is adopted by some
// browsers.
func RemoveLeadingDots(remove bool) Option {
return func(o *options) { o.removeLeadingDots = remove }
}
// ValidateLabels sets whether to check the mandatory label validation criteria
// as defined in Section 5.4 of RFC 5891. This includes testing for correct use
// of hyphens ('-'), normalization, validity of runes, and the context rules.
func ValidateLabels(enable bool) Option {
return func(o *options) {
// Don't override existing mappings, but set one that at least checks
// normalization if it is not set.
if o.mapping == nil && enable {
o.mapping = normalize
}
o.trie = trie
o.validateLabels = enable
o.fromPuny = validateFromPunycode
}
}
// StrictDomainName limits the set of permissable ASCII characters to those
// allowed in domain names as defined in RFC 1034 (A-Z, a-z, 0-9 and the
// hyphen). This is set by default for MapForLookup and ValidateForRegistration.
//
// This option is useful, for instance, for browsers that allow characters
// outside this range, for example a '_' (U+005F LOW LINE). See
// http://www.rfc-editor.org/std/std3.txt for more details This option
// corresponds to the UseSTD3ASCIIRules option in UTS #46.
func StrictDomainName(use bool) Option {
return func(o *options) {
o.trie = trie
o.useSTD3Rules = use
o.fromPuny = validateFromPunycode
}
}
// NOTE: the following options pull in tables. The tables should not be linked
// in as long as the options are not used.
// BidiRule enables the Bidi rule as defined in RFC 5893. Any application
// that relies on proper validation of labels should include this rule.
func BidiRule() Option {
return func(o *options) { o.bidirule = bidirule.ValidString }
}
// ValidateForRegistration sets validation options to verify that a given IDN is
// properly formatted for registration as defined by Section 4 of RFC 5891.
func ValidateForRegistration() Option {
return func(o *options) {
o.mapping = validateRegistration
StrictDomainName(true)(o)
ValidateLabels(true)(o)
VerifyDNSLength(true)(o)
BidiRule()(o)
}
}
// MapForLookup sets validation and mapping options such that a given IDN is
// transformed for domain name lookup according to the requirements set out in
// Section 5 of RFC 5891. The mappings follow the recommendations of RFC 5894,
// RFC 5895 and UTS 46. It does not add the Bidi Rule. Use the BidiRule option
// to add this check.
//
// The mappings include normalization and mapping case, width and other
// compatibility mappings.
func MapForLookup() Option {
return func(o *options) {
o.mapping = validateAndMap
StrictDomainName(true)(o)
ValidateLabels(true)(o)
RemoveLeadingDots(true)(o)
}
}
type options struct {
transitional bool
useSTD3Rules bool
validateLabels bool
verifyDNSLength bool
removeLeadingDots bool
trie *idnaTrie
// fromPuny calls validation rules when converting A-labels to U-labels.
fromPuny func(p *Profile, s string) error
// mapping implements a validation and mapping step as defined in RFC 5895
// or UTS 46, tailored to, for example, domain registration or lookup.
mapping func(p *Profile, s string) (string, error)
// bidirule, if specified, checks whether s conforms to the Bidi Rule
// defined in RFC 5893.
bidirule func(s string) bool
}
// A Profile defines the configuration of a IDNA mapper.
type Profile struct {
options
}
func apply(o *options, opts []Option) {
for _, f := range opts {
f(o)
}
}
// New creates a new Profile.
//
// With no options, the returned Profile is the most permissive and equals the
// Punycode Profile. Options can be passed to further restrict the Profile. The
// MapForLookup and ValidateForRegistration options set a collection of options,
// for lookup and registration purposes respectively, which can be tailored by
// adding more fine-grained options, where later options override earlier
// options.
func New(o ...Option) *Profile {
p := &Profile{}
apply(&p.options, o)
return p
}
// ToASCII converts a domain or domain label to its ASCII form. For example,
// ToASCII("bücher.example.com") is "xn--bcher-kva.example.com", and
// ToASCII("golang") is "golang". If an error is encountered it will return
// an error and a (partially) processed result.
func (p *Profile) ToASCII(s string) (string, error) {
return p.process(s, true)
}
// ToUnicode converts a domain or domain label to its Unicode form. For example,
// ToUnicode("xn--bcher-kva.example.com") is "bücher.example.com", and
// ToUnicode("golang") is "golang". If an error is encountered it will return
// an error and a (partially) processed result.
func (p *Profile) ToUnicode(s string) (string, error) {
pp := *p
pp.transitional = false
return pp.process(s, false)
}
// String reports a string with a description of the profile for debugging
// purposes. The string format may change with different versions.
func (p *Profile) String() string {
s := ""
if p.transitional {
s = "Transitional"
} else {
s = "NonTransitional"
}
if p.useSTD3Rules {
s += ":UseSTD3Rules"
}
if p.validateLabels {
s += ":ValidateLabels"
}
if p.verifyDNSLength {
s += ":VerifyDNSLength"
}
return s
}
var (
// Punycode is a Profile that does raw punycode processing with a minimum
// of validation.
Punycode *Profile = punycode
// Lookup is the recommended profile for looking up domain names, according
// to Section 5 of RFC 5891. The exact configuration of this profile may
// change over time.
Lookup *Profile = lookup
// Display is the recommended profile for displaying domain names.
// The configuration of this profile may change over time.
Display *Profile = display
// Registration is the recommended profile for checking whether a given
// IDN is valid for registration, according to Section 4 of RFC 5891.
Registration *Profile = registration
punycode = &Profile{}
lookup = &Profile{options{
transitional: true,
useSTD3Rules: true,
validateLabels: true,
removeLeadingDots: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateAndMap,
bidirule: bidirule.ValidString,
}}
display = &Profile{options{
useSTD3Rules: true,
validateLabels: true,
removeLeadingDots: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateAndMap,
bidirule: bidirule.ValidString,
}}
registration = &Profile{options{
useSTD3Rules: true,
validateLabels: true,
verifyDNSLength: true,
trie: trie,
fromPuny: validateFromPunycode,
mapping: validateRegistration,
bidirule: bidirule.ValidString,
}}
// TODO: profiles
// Register: recommended for approving domain names: don't do any mappings
// but rather reject on invalid input. Bundle or block deviation characters.
)
type labelError struct{ label, code_ string }
func (e labelError) code() string { return e.code_ }
func (e labelError) Error() string {
return fmt.Sprintf("idna: invalid label %q", e.label)
}
type runeError rune
func (e runeError) code() string { return "P1" }
func (e runeError) Error() string {
return fmt.Sprintf("idna: disallowed rune %U", e)
}
// process implements the algorithm described in section 4 of UTS #46,
// see https://www.unicode.org/reports/tr46.
func (p *Profile) process(s string, toASCII bool) (string, error) {
var err error
if p.mapping != nil {
s, err = p.mapping(p, s)
}
// Remove leading empty labels.
if p.removeLeadingDots {
for ; len(s) > 0 && s[0] == '.'; s = s[1:] {
}
}
// It seems like we should only create this error on ToASCII, but the
// UTS 46 conformance tests suggests we should always check this.
if err == nil && p.verifyDNSLength && s == "" {
err = &labelError{s, "A4"}
}
labels := labelIter{orig: s}
for ; !labels.done(); labels.next() {
label := labels.label()
if label == "" {
// Empty labels are not okay. The label iterator skips the last
// label if it is empty.
if err == nil && p.verifyDNSLength {
err = &labelError{s, "A4"}
}
continue
}
if strings.HasPrefix(label, acePrefix) {
u, err2 := decode(label[len(acePrefix):])
if err2 != nil {
if err == nil {
err = err2
}
// Spec says keep the old label.
continue
}
labels.set(u)
if err == nil && p.validateLabels {
err = p.fromPuny(p, u)
}
if err == nil {
// This should be called on NonTransitional, according to the
// spec, but that currently does not have any effect. Use the
// original profile to preserve options.
err = p.validateLabel(u)
}
} else if err == nil {
err = p.validateLabel(label)
}
}
if toASCII {
for labels.reset(); !labels.done(); labels.next() {
label := labels.label()
if !ascii(label) {
a, err2 := encode(acePrefix, label)
if err == nil {
err = err2
}
label = a
labels.set(a)
}
n := len(label)
if p.verifyDNSLength && err == nil && (n == 0 || n > 63) {
err = &labelError{label, "A4"}
}
}
}
s = labels.result()
if toASCII && p.verifyDNSLength && err == nil {
// Compute the length of the domain name minus the root label and its dot.
n := len(s)
if n > 0 && s[n-1] == '.' {
n--
}
if len(s) < 1 || n > 253 {
err = &labelError{s, "A4"}
}
}
return s, err
}
func normalize(p *Profile, s string) (string, error) {
return norm.NFC.String(s), nil
}
func validateRegistration(p *Profile, s string) (string, error) {
if !norm.NFC.IsNormalString(s) {
return s, &labelError{s, "V1"}
}
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
// Copy bytes not copied so far.
switch p.simplify(info(v).category()) {
// TODO: handle the NV8 defined in the Unicode idna data set to allow
// for strict conformance to IDNA2008.
case valid, deviation:
case disallowed, mapped, unknown, ignored:
r, _ := utf8.DecodeRuneInString(s[i:])
return s, runeError(r)
}
i += sz
}
return s, nil
}
func validateAndMap(p *Profile, s string) (string, error) {
var (
err error
b []byte
k int
)
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
start := i
i += sz
// Copy bytes not copied so far.
switch p.simplify(info(v).category()) {
case valid:
continue
case disallowed:
if err == nil {
r, _ := utf8.DecodeRuneInString(s[start:])
err = runeError(r)
}
continue
case mapped, deviation:
b = append(b, s[k:start]...)
b = info(v).appendMapping(b, s[start:i])
case ignored:
b = append(b, s[k:start]...)
// drop the rune
case unknown:
b = append(b, s[k:start]...)
b = append(b, "\ufffd"...)
}
k = i
}
if k == 0 {
// No changes so far.
s = norm.NFC.String(s)
} else {
b = append(b, s[k:]...)
if norm.NFC.QuickSpan(b) != len(b) {
b = norm.NFC.Bytes(b)
}
// TODO: the punycode converters require strings as input.
s = string(b)
}
return s, err
}
// A labelIter allows iterating over domain name labels.
type labelIter struct {
orig string
slice []string
curStart int
curEnd int
i int
}
func (l *labelIter) reset() {
l.curStart = 0
l.curEnd = 0
l.i = 0
}
func (l *labelIter) done() bool {
return l.curStart >= len(l.orig)
}
func (l *labelIter) result() string {
if l.slice != nil {
return strings.Join(l.slice, ".")
}
return l.orig
}
func (l *labelIter) label() string {
if l.slice != nil {
return l.slice[l.i]
}
p := strings.IndexByte(l.orig[l.curStart:], '.')
l.curEnd = l.curStart + p
if p == -1 {
l.curEnd = len(l.orig)
}
return l.orig[l.curStart:l.curEnd]
}
// next sets the value to the next label. It skips the last label if it is empty.
func (l *labelIter) next() {
l.i++
if l.slice != nil {
if l.i >= len(l.slice) || l.i == len(l.slice)-1 && l.slice[l.i] == "" {
l.curStart = len(l.orig)
}
} else {
l.curStart = l.curEnd + 1
if l.curStart == len(l.orig)-1 && l.orig[l.curStart] == '.' {
l.curStart = len(l.orig)
}
}
}
func (l *labelIter) set(s string) {
if l.slice == nil {
l.slice = strings.Split(l.orig, ".")
}
l.slice[l.i] = s
}
// acePrefix is the ASCII Compatible Encoding prefix.
const acePrefix = "xn--"
func (p *Profile) simplify(cat category) category {
switch cat {
case disallowedSTD3Mapped:
if p.useSTD3Rules {
cat = disallowed
} else {
cat = mapped
}
case disallowedSTD3Valid:
if p.useSTD3Rules {
cat = disallowed
} else {
cat = valid
}
case deviation:
if !p.transitional {
cat = valid
}
case validNV8, validXV8:
// TODO: handle V2008
cat = valid
}
return cat
}
func validateFromPunycode(p *Profile, s string) error {
if !norm.NFC.IsNormalString(s) {
return &labelError{s, "V1"}
}
for i := 0; i < len(s); {
v, sz := trie.lookupString(s[i:])
if c := p.simplify(info(v).category()); c != valid && c != deviation {
return &labelError{s, "V6"}
}
i += sz
}
return nil
}
const (
zwnj = "\u200c"
zwj = "\u200d"
)
type joinState int8
const (
stateStart joinState = iota
stateVirama
stateBefore
stateBeforeVirama
stateAfter
stateFAIL
)
var joinStates = [][numJoinTypes]joinState{
stateStart: {
joiningL: stateBefore,
joiningD: stateBefore,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateVirama,
},
stateVirama: {
joiningL: stateBefore,
joiningD: stateBefore,
},
stateBefore: {
joiningL: stateBefore,
joiningD: stateBefore,
joiningT: stateBefore,
joinZWNJ: stateAfter,
joinZWJ: stateFAIL,
joinVirama: stateBeforeVirama,
},
stateBeforeVirama: {
joiningL: stateBefore,
joiningD: stateBefore,
joiningT: stateBefore,
},
stateAfter: {
joiningL: stateFAIL,
joiningD: stateBefore,
joiningT: stateAfter,
joiningR: stateStart,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateAfter, // no-op as we can't accept joiners here
},
stateFAIL: {
0: stateFAIL,
joiningL: stateFAIL,
joiningD: stateFAIL,
joiningT: stateFAIL,
joiningR: stateFAIL,
joinZWNJ: stateFAIL,
joinZWJ: stateFAIL,
joinVirama: stateFAIL,
},
}
// validateLabel validates the criteria from Section 4.1. Item 1, 4, and 6 are
// already implicitly satisfied by the overall implementation.
func (p *Profile) validateLabel(s string) error {
if s == "" {
if p.verifyDNSLength {
return &labelError{s, "A4"}
}
return nil
}
if p.bidirule != nil && !p.bidirule(s) {
return &labelError{s, "B"}
}
if !p.validateLabels {
return nil
}
trie := p.trie // p.validateLabels is only set if trie is set.
if len(s) > 4 && s[2] == '-' && s[3] == '-' {
return &labelError{s, "V2"}
}
if s[0] == '-' || s[len(s)-1] == '-' {
return &labelError{s, "V3"}
}
// TODO: merge the use of this in the trie.
v, sz := trie.lookupString(s)
x := info(v)
if x.isModifier() {
return &labelError{s, "V5"}
}
// Quickly return in the absence of zero-width (non) joiners.
if strings.Index(s, zwj) == -1 && strings.Index(s, zwnj) == -1 {
return nil
}
st := stateStart
for i := 0; ; {
jt := x.joinType()
if s[i:i+sz] == zwj {
jt = joinZWJ
} else if s[i:i+sz] == zwnj {
jt = joinZWNJ
}
st = joinStates[st][jt]
if x.isViramaModifier() {
st = joinStates[st][joinVirama]
}
if i += sz; i == len(s) {
break
}
v, sz = trie.lookupString(s[i:])
x = info(v)
}
if st == stateFAIL || st == stateAfter {
return &labelError{s, "C"}
}
return nil
}
func ascii(s string) bool {
for i := 0; i < len(s); i++ {
if s[i] >= utf8.RuneSelf {
return false
}
}
return true
}

203
vendor/golang.org/x/net/idna/punycode.go generated vendored Normal file
View file

@ -0,0 +1,203 @@
// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
// Copyright 2016 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 idna
// This file implements the Punycode algorithm from RFC 3492.
import (
"math"
"strings"
"unicode/utf8"
)
// These parameter values are specified in section 5.
//
// All computation is done with int32s, so that overflow behavior is identical
// regardless of whether int is 32-bit or 64-bit.
const (
base int32 = 36
damp int32 = 700
initialBias int32 = 72
initialN int32 = 128
skew int32 = 38
tmax int32 = 26
tmin int32 = 1
)
func punyError(s string) error { return &labelError{s, "A3"} }
// decode decodes a string as specified in section 6.2.
func decode(encoded string) (string, error) {
if encoded == "" {
return "", nil
}
pos := 1 + strings.LastIndex(encoded, "-")
if pos == 1 {
return "", punyError(encoded)
}
if pos == len(encoded) {
return encoded[:len(encoded)-1], nil
}
output := make([]rune, 0, len(encoded))
if pos != 0 {
for _, r := range encoded[:pos-1] {
output = append(output, r)
}
}
i, n, bias := int32(0), initialN, initialBias
for pos < len(encoded) {
oldI, w := i, int32(1)
for k := base; ; k += base {
if pos == len(encoded) {
return "", punyError(encoded)
}
digit, ok := decodeDigit(encoded[pos])
if !ok {
return "", punyError(encoded)
}
pos++
i += digit * w
if i < 0 {
return "", punyError(encoded)
}
t := k - bias
if t < tmin {
t = tmin
} else if t > tmax {
t = tmax
}
if digit < t {
break
}
w *= base - t
if w >= math.MaxInt32/base {
return "", punyError(encoded)
}
}
x := int32(len(output) + 1)
bias = adapt(i-oldI, x, oldI == 0)
n += i / x
i %= x
if n > utf8.MaxRune || len(output) >= 1024 {
return "", punyError(encoded)
}
output = append(output, 0)
copy(output[i+1:], output[i:])
output[i] = n
i++
}
return string(output), nil
}
// encode encodes a string as specified in section 6.3 and prepends prefix to
// the result.
//
// The "while h < length(input)" line in the specification becomes "for
// remaining != 0" in the Go code, because len(s) in Go is in bytes, not runes.
func encode(prefix, s string) (string, error) {
output := make([]byte, len(prefix), len(prefix)+1+2*len(s))
copy(output, prefix)
delta, n, bias := int32(0), initialN, initialBias
b, remaining := int32(0), int32(0)
for _, r := range s {
if r < 0x80 {
b++
output = append(output, byte(r))
} else {
remaining++
}
}
h := b
if b > 0 {
output = append(output, '-')
}
for remaining != 0 {
m := int32(0x7fffffff)
for _, r := range s {
if m > r && r >= n {
m = r
}
}
delta += (m - n) * (h + 1)
if delta < 0 {
return "", punyError(s)
}
n = m
for _, r := range s {
if r < n {
delta++
if delta < 0 {
return "", punyError(s)
}
continue
}
if r > n {
continue
}
q := delta
for k := base; ; k += base {
t := k - bias
if t < tmin {
t = tmin
} else if t > tmax {
t = tmax
}
if q < t {
break
}
output = append(output, encodeDigit(t+(q-t)%(base-t)))
q = (q - t) / (base - t)
}
output = append(output, encodeDigit(q))
bias = adapt(delta, h+1, h == b)
delta = 0
h++
remaining--
}
delta++
n++
}
return string(output), nil
}
func decodeDigit(x byte) (digit int32, ok bool) {
switch {
case '0' <= x && x <= '9':
return int32(x - ('0' - 26)), true
case 'A' <= x && x <= 'Z':
return int32(x - 'A'), true
case 'a' <= x && x <= 'z':
return int32(x - 'a'), true
}
return 0, false
}
func encodeDigit(digit int32) byte {
switch {
case 0 <= digit && digit < 26:
return byte(digit + 'a')
case 26 <= digit && digit < 36:
return byte(digit + ('0' - 26))
}
panic("idna: internal error in punycode encoding")
}
// adapt is the bias adaptation function specified in section 6.1.
func adapt(delta, numPoints int32, firstTime bool) int32 {
if firstTime {
delta /= damp
} else {
delta /= 2
}
delta += delta / numPoints
k := int32(0)
for delta > ((base-tmin)*tmax)/2 {
delta /= base - tmin
k += base
}
return k + (base-tmin+1)*delta/(delta+skew)
}

4559
vendor/golang.org/x/net/idna/tables10.0.0.go generated vendored Normal file

File diff suppressed because it is too large Load diff

4653
vendor/golang.org/x/net/idna/tables11.0.0.go generated vendored Normal file

File diff suppressed because it is too large Load diff

4486
vendor/golang.org/x/net/idna/tables9.0.0.go generated vendored Normal file

File diff suppressed because it is too large Load diff

72
vendor/golang.org/x/net/idna/trie.go generated vendored Normal file
View file

@ -0,0 +1,72 @@
// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
// Copyright 2016 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 idna
// appendMapping appends the mapping for the respective rune. isMapped must be
// true. A mapping is a categorization of a rune as defined in UTS #46.
func (c info) appendMapping(b []byte, s string) []byte {
index := int(c >> indexShift)
if c&xorBit == 0 {
s := mappings[index:]
return append(b, s[1:s[0]+1]...)
}
b = append(b, s...)
if c&inlineXOR == inlineXOR {
// TODO: support and handle two-byte inline masks
b[len(b)-1] ^= byte(index)
} else {
for p := len(b) - int(xorData[index]); p < len(b); p++ {
index++
b[p] ^= xorData[index]
}
}
return b
}
// Sparse block handling code.
type valueRange struct {
value uint16 // header: value:stride
lo, hi byte // header: lo:n
}
type sparseBlocks struct {
values []valueRange
offset []uint16
}
var idnaSparse = sparseBlocks{
values: idnaSparseValues[:],
offset: idnaSparseOffset[:],
}
// Don't use newIdnaTrie to avoid unconditional linking in of the table.
var trie = &idnaTrie{}
// lookup determines the type of block n and looks up the value for b.
// For n < t.cutoff, the block is a simple lookup table. Otherwise, the block
// is a list of ranges with an accompanying value. Given a matching range r,
// the value for b is by r.value + (b - r.lo) * stride.
func (t *sparseBlocks) lookup(n uint32, b byte) uint16 {
offset := t.offset[n]
header := t.values[offset]
lo := offset + 1
hi := lo + uint16(header.lo)
for lo < hi {
m := lo + (hi-lo)/2
r := t.values[m]
if r.lo <= b && b <= r.hi {
return r.value + uint16(b-r.lo)*header.value
}
if b < r.lo {
hi = m
} else {
lo = m + 1
}
}
return 0
}

119
vendor/golang.org/x/net/idna/trieval.go generated vendored Normal file
View file

@ -0,0 +1,119 @@
// Code generated by running "go generate" in golang.org/x/text. DO NOT EDIT.
package idna
// This file contains definitions for interpreting the trie value of the idna
// trie generated by "go run gen*.go". It is shared by both the generator
// program and the resultant package. Sharing is achieved by the generator
// copying gen_trieval.go to trieval.go and changing what's above this comment.
// info holds information from the IDNA mapping table for a single rune. It is
// the value returned by a trie lookup. In most cases, all information fits in
// a 16-bit value. For mappings, this value may contain an index into a slice
// with the mapped string. Such mappings can consist of the actual mapped value
// or an XOR pattern to be applied to the bytes of the UTF8 encoding of the
// input rune. This technique is used by the cases packages and reduces the
// table size significantly.
//
// The per-rune values have the following format:
//
// if mapped {
// if inlinedXOR {
// 15..13 inline XOR marker
// 12..11 unused
// 10..3 inline XOR mask
// } else {
// 15..3 index into xor or mapping table
// }
// } else {
// 15..14 unused
// 13 mayNeedNorm
// 12..11 attributes
// 10..8 joining type
// 7..3 category type
// }
// 2 use xor pattern
// 1..0 mapped category
//
// See the definitions below for a more detailed description of the various
// bits.
type info uint16
const (
catSmallMask = 0x3
catBigMask = 0xF8
indexShift = 3
xorBit = 0x4 // interpret the index as an xor pattern
inlineXOR = 0xE000 // These bits are set if the XOR pattern is inlined.
joinShift = 8
joinMask = 0x07
// Attributes
attributesMask = 0x1800
viramaModifier = 0x1800
modifier = 0x1000
rtl = 0x0800
mayNeedNorm = 0x2000
)
// A category corresponds to a category defined in the IDNA mapping table.
type category uint16
const (
unknown category = 0 // not currently defined in unicode.
mapped category = 1
disallowedSTD3Mapped category = 2
deviation category = 3
)
const (
valid category = 0x08
validNV8 category = 0x18
validXV8 category = 0x28
disallowed category = 0x40
disallowedSTD3Valid category = 0x80
ignored category = 0xC0
)
// join types and additional rune information
const (
joiningL = (iota + 1)
joiningD
joiningT
joiningR
//the following types are derived during processing
joinZWJ
joinZWNJ
joinVirama
numJoinTypes
)
func (c info) isMapped() bool {
return c&0x3 != 0
}
func (c info) category() category {
small := c & catSmallMask
if small != 0 {
return category(small)
}
return category(c & catBigMask)
}
func (c info) joinType() info {
if c.isMapped() {
return 0
}
return (c >> joinShift) & joinMask
}
func (c info) isModifier() bool {
return c&(modifier|catSmallMask) == modifier
}
func (c info) isViramaModifier() bool {
return c&(attributesMask|catSmallMask) == viramaModifier
}