Change markdown rendering from blackfriday to goldmark (#9533)

* Move to goldmark

Markdown rendering moved from blackfriday to the goldmark.

Multiple subtle changes required to the goldmark extensions to keep
current rendering and defaults.

Can go further with goldmark linkify and have this work within markdown
rendering making the link processor unnecessary.

Need to think about how to go about allowing extensions - at present it
seems that these would be hard to do without recompilation.

* linter fixes

Co-authored-by: Lauris BH <lauris@nix.lv>
lunny/display_deleted_branch2
zeripath 4 years ago committed by Lauris BH
parent 0c07f1de5b
commit 27757714d0

@ -79,11 +79,9 @@ require (
github.com/prometheus/procfs v0.0.4 // indirect
github.com/quasoft/websspi v1.0.0
github.com/remyoudompheng/bigfft v0.0.0-20190321074620-2f0d2b0e0001 // indirect
github.com/russross/blackfriday/v2 v2.0.1
github.com/satori/go.uuid v1.2.0
github.com/sergi/go-diff v1.0.0
github.com/shurcooL/httpfs v0.0.0-20190527155220-6a4d4a70508b // indirect
github.com/shurcooL/sanitized_anchor_name v1.0.0 // indirect
github.com/shurcooL/vfsgen v0.0.0-20181202132449-6a9ea43bcacd
github.com/steveyen/gtreap v0.0.0-20150807155958-0abe01ef9be2 // indirect
github.com/stretchr/testify v1.4.0
@ -95,6 +93,7 @@ require (
github.com/unknwon/paginater v0.0.0-20151104151617-7748a72e0141
github.com/urfave/cli v1.20.0
github.com/yohcop/openid-go v0.0.0-20160914080427-2c050d2dae53
github.com/yuin/goldmark v1.1.19
go.etcd.io/bbolt v1.3.3 // indirect
golang.org/x/crypto v0.0.0-20191227163750-53104e6ec876
golang.org/x/net v0.0.0-20191101175033-0deb6923b6d9

@ -462,16 +462,12 @@ github.com/remyoudompheng/bigfft v0.0.0-20190321074620-2f0d2b0e0001/go.mod h1:qq
github.com/rogpeppe/fastuuid v0.0.0-20150106093220-6724a57986af/go.mod h1:XWv6SoW27p1b0cqNHllgS5HIMJraePCO15w5zCzIWYg=
github.com/russross/blackfriday v1.5.2 h1:HyvC0ARfnZBqnXwABFeSZHpKvJHJJfPz81GNueLj0oo=
github.com/russross/blackfriday v1.5.2/go.mod h1:JO/DiYxRf+HjHt06OyowR9PTA263kcR/rfWxYHBV53g=
github.com/russross/blackfriday/v2 v2.0.1 h1:lPqVAte+HuHNfhJ/0LC98ESWRz8afy9tM/0RK8m9o+Q=
github.com/russross/blackfriday/v2 v2.0.1/go.mod h1:+Rmxgy9KzJVeS9/2gXHxylqXiyQDYRxCVz55jmeOWTM=
github.com/satori/go.uuid v1.2.0 h1:0uYX9dsZ2yD7q2RtLRtPSdGDWzjeM3TbMJP9utgA0ww=
github.com/satori/go.uuid v1.2.0/go.mod h1:dA0hQrYB0VpLJoorglMZABFdXlWrHn1NEOzdhQKdks0=
github.com/sergi/go-diff v1.0.0 h1:Kpca3qRNrduNnOQeazBd0ysaKrUJiIuISHxogkT9RPQ=
github.com/sergi/go-diff v1.0.0/go.mod h1:0CfEIISq7TuYL3j771MWULgwwjU+GofnZX9QAmXWZgo=
github.com/shurcooL/httpfs v0.0.0-20190527155220-6a4d4a70508b h1:4kg1wyftSKxLtnPAvcRWakIPpokB9w780/KwrNLnfPA=
github.com/shurcooL/httpfs v0.0.0-20190527155220-6a4d4a70508b/go.mod h1:ZY1cvUeJuFPAdZ/B6v7RHavJWZn2YPVFQ1OSXhCGOkg=
github.com/shurcooL/sanitized_anchor_name v1.0.0 h1:PdmoCO6wvbs+7yrJyMORt4/BmY5IYyJwS/kOiWx8mHo=
github.com/shurcooL/sanitized_anchor_name v1.0.0/go.mod h1:1NzhyTcUVG4SuEtjjoZeVRXNmyL/1OwPU0+IJeTBvfc=
github.com/shurcooL/vfsgen v0.0.0-20181202132449-6a9ea43bcacd h1:ug7PpSOB5RBPK1Kg6qskGBoP3Vnj/aNYFTznWvlkGo0=
github.com/shurcooL/vfsgen v0.0.0-20181202132449-6a9ea43bcacd/go.mod h1:TrYk7fJVaAttu97ZZKrO9UbRa8izdowaMIZcxYMbVaw=
github.com/siddontang/go v0.0.0-20180604090527-bdc77568d726/go.mod h1:3yhqj7WBBfRhbBlzyOC3gUxftwsU0u8gqevxwIHQpMw=
@ -550,6 +546,8 @@ github.com/xiang90/probing v0.0.0-20190116061207-43a291ad63a2/go.mod h1:UETIi67q
github.com/xordataexchange/crypt v0.0.3-0.20170626215501-b2862e3d0a77/go.mod h1:aYKd//L2LvnjZzWKhF00oedf4jCCReLcmhLdhm1A27Q=
github.com/yohcop/openid-go v0.0.0-20160914080427-2c050d2dae53 h1:HsIQ6yAjfjQ3IxPGrTusxp6Qxn92gNVq2x5CbvQvx3w=
github.com/yohcop/openid-go v0.0.0-20160914080427-2c050d2dae53/go.mod h1:f6elajwZV+xceiaqgRL090YzLEDGSbqr3poGL3ZgXYo=
github.com/yuin/goldmark v1.1.19 h1:0s2/60x0XsFCXHeFut+F3azDVAAyIMyUfJRbRexiTYs=
github.com/yuin/goldmark v1.1.19/go.mod h1:3hX8gzYuyVAZsxl0MRgGTJEmQBFcNTphYh9decYSb74=
github.com/ziutek/mymysql v1.5.4 h1:GB0qdRGsTwQSBVYuVShFBKaXSnSnYYC2d9knnE1LHFs=
github.com/ziutek/mymysql v1.5.4/go.mod h1:LMSpPZ6DbqWFxNCHW77HeMg9I646SAhApZ/wKdgO/C0=
go.etcd.io/bbolt v1.3.2/go.mod h1:IbVyRI1SCnLcuJnV2u8VeU0CEYM7e686BmAb1XKL+uU=

@ -0,0 +1,507 @@
// Copyright 2019 Yusuke Inuzuka
// Copyright 2019 The Gitea Authors. All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
// Most of what follows is a subtly changed version of github.com/yuin/goldmark/extension/footnote.go
package common
import (
"bytes"
"fmt"
"os"
"strconv"
"unicode"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// CleanValue will clean a value to make it safe to be an id
// This function is quite different from the original goldmark function
// and more closely matches the output from the shurcooL sanitizer
// In particular Unicode letters and numbers are a lot more than a-zA-Z0-9...
func CleanValue(value []byte) []byte {
value = bytes.TrimSpace(value)
rs := bytes.Runes(value)
result := make([]rune, 0, len(rs))
needsDash := false
for _, r := range rs {
switch {
case unicode.IsLetter(r) || unicode.IsNumber(r):
if needsDash && len(result) > 0 {
result = append(result, '-')
}
needsDash = false
result = append(result, unicode.ToLower(r))
default:
needsDash = true
}
}
return []byte(string(result))
}
// Most of what follows is a subtly changed version of github.com/yuin/goldmark/extension/footnote.go
// A FootnoteLink struct represents a link to a footnote of Markdown
// (PHP Markdown Extra) text.
type FootnoteLink struct {
ast.BaseInline
Index int
Name []byte
}
// Dump implements Node.Dump.
func (n *FootnoteLink) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
m["Name"] = fmt.Sprintf("%v", n.Name)
ast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteLink is a NodeKind of the FootnoteLink node.
var KindFootnoteLink = ast.NewNodeKind("GiteaFootnoteLink")
// Kind implements Node.Kind.
func (n *FootnoteLink) Kind() ast.NodeKind {
return KindFootnoteLink
}
// NewFootnoteLink returns a new FootnoteLink node.
func NewFootnoteLink(index int, name []byte) *FootnoteLink {
return &FootnoteLink{
Index: index,
Name: name,
}
}
// A FootnoteBackLink struct represents a link to a footnote of Markdown
// (PHP Markdown Extra) text.
type FootnoteBackLink struct {
ast.BaseInline
Index int
Name []byte
}
// Dump implements Node.Dump.
func (n *FootnoteBackLink) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
m["Name"] = fmt.Sprintf("%v", n.Name)
ast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteBackLink is a NodeKind of the FootnoteBackLink node.
var KindFootnoteBackLink = ast.NewNodeKind("GiteaFootnoteBackLink")
// Kind implements Node.Kind.
func (n *FootnoteBackLink) Kind() ast.NodeKind {
return KindFootnoteBackLink
}
// NewFootnoteBackLink returns a new FootnoteBackLink node.
func NewFootnoteBackLink(index int, name []byte) *FootnoteBackLink {
return &FootnoteBackLink{
Index: index,
Name: name,
}
}
// A Footnote struct represents a footnote of Markdown
// (PHP Markdown Extra) text.
type Footnote struct {
ast.BaseBlock
Ref []byte
Index int
Name []byte
}
// Dump implements Node.Dump.
func (n *Footnote) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
m["Ref"] = fmt.Sprintf("%s", n.Ref)
m["Name"] = fmt.Sprintf("%v", n.Name)
ast.DumpHelper(n, source, level, m, nil)
}
// KindFootnote is a NodeKind of the Footnote node.
var KindFootnote = ast.NewNodeKind("GiteaFootnote")
// Kind implements Node.Kind.
func (n *Footnote) Kind() ast.NodeKind {
return KindFootnote
}
// NewFootnote returns a new Footnote node.
func NewFootnote(ref []byte) *Footnote {
return &Footnote{
Ref: ref,
Index: -1,
Name: ref,
}
}
// A FootnoteList struct represents footnotes of Markdown
// (PHP Markdown Extra) text.
type FootnoteList struct {
ast.BaseBlock
Count int
}
// Dump implements Node.Dump.
func (n *FootnoteList) Dump(source []byte, level int) {
m := map[string]string{}
m["Count"] = fmt.Sprintf("%v", n.Count)
ast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteList is a NodeKind of the FootnoteList node.
var KindFootnoteList = ast.NewNodeKind("GiteaFootnoteList")
// Kind implements Node.Kind.
func (n *FootnoteList) Kind() ast.NodeKind {
return KindFootnoteList
}
// NewFootnoteList returns a new FootnoteList node.
func NewFootnoteList() *FootnoteList {
return &FootnoteList{
Count: 0,
}
}
var footnoteListKey = parser.NewContextKey()
type footnoteBlockParser struct {
}
var defaultFootnoteBlockParser = &footnoteBlockParser{}
// NewFootnoteBlockParser returns a new parser.BlockParser that can parse
// footnotes of the Markdown(PHP Markdown Extra) text.
func NewFootnoteBlockParser() parser.BlockParser {
return defaultFootnoteBlockParser
}
func (b *footnoteBlockParser) Trigger() []byte {
return []byte{'['}
}
func (b *footnoteBlockParser) Open(parent ast.Node, reader text.Reader, pc parser.Context) (ast.Node, parser.State) {
line, segment := reader.PeekLine()
pos := pc.BlockOffset()
if pos < 0 || line[pos] != '[' {
return nil, parser.NoChildren
}
pos++
if pos > len(line)-1 || line[pos] != '^' {
return nil, parser.NoChildren
}
open := pos + 1
closes := 0
closure := util.FindClosure(line[pos+1:], '[', ']', false, false)
closes = pos + 1 + closure
next := closes + 1
if closure > -1 {
if next >= len(line) || line[next] != ':' {
return nil, parser.NoChildren
}
} else {
return nil, parser.NoChildren
}
padding := segment.Padding
label := reader.Value(text.NewSegment(segment.Start+open-padding, segment.Start+closes-padding))
if util.IsBlank(label) {
return nil, parser.NoChildren
}
item := NewFootnote(label)
pos = next + 1 - padding
if pos >= len(line) {
reader.Advance(pos)
return item, parser.NoChildren
}
reader.AdvanceAndSetPadding(pos, padding)
return item, parser.HasChildren
}
func (b *footnoteBlockParser) Continue(node ast.Node, reader text.Reader, pc parser.Context) parser.State {
line, _ := reader.PeekLine()
if util.IsBlank(line) {
return parser.Continue | parser.HasChildren
}
childpos, padding := util.IndentPosition(line, reader.LineOffset(), 4)
if childpos < 0 {
return parser.Close
}
reader.AdvanceAndSetPadding(childpos, padding)
return parser.Continue | parser.HasChildren
}
func (b *footnoteBlockParser) Close(node ast.Node, reader text.Reader, pc parser.Context) {
var list *FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*FootnoteList)
} else {
list = NewFootnoteList()
pc.Set(footnoteListKey, list)
node.Parent().InsertBefore(node.Parent(), node, list)
}
node.Parent().RemoveChild(node.Parent(), node)
list.AppendChild(list, node)
}
func (b *footnoteBlockParser) CanInterruptParagraph() bool {
return true
}
func (b *footnoteBlockParser) CanAcceptIndentedLine() bool {
return false
}
type footnoteParser struct {
}
var defaultFootnoteParser = &footnoteParser{}
// NewFootnoteParser returns a new parser.InlineParser that can parse
// footnote links of the Markdown(PHP Markdown Extra) text.
func NewFootnoteParser() parser.InlineParser {
return defaultFootnoteParser
}
func (s *footnoteParser) Trigger() []byte {
// footnote syntax probably conflict with the image syntax.
// So we need trigger this parser with '!'.
return []byte{'!', '['}
}
func (s *footnoteParser) Parse(parent ast.Node, block text.Reader, pc parser.Context) ast.Node {
line, segment := block.PeekLine()
pos := 1
if len(line) > 0 && line[0] == '!' {
pos++
}
if pos >= len(line) || line[pos] != '^' {
return nil
}
pos++
if pos >= len(line) {
return nil
}
open := pos
closure := util.FindClosure(line[pos:], '[', ']', false, false)
if closure < 0 {
return nil
}
closes := pos + closure
value := block.Value(text.NewSegment(segment.Start+open, segment.Start+closes))
block.Advance(closes + 1)
var list *FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*FootnoteList)
}
if list == nil {
return nil
}
index := 0
name := []byte{}
for def := list.FirstChild(); def != nil; def = def.NextSibling() {
d := def.(*Footnote)
if bytes.Equal(d.Ref, value) {
if d.Index < 0 {
list.Count++
d.Index = list.Count
val := CleanValue(d.Name)
if len(val) == 0 {
val = []byte(strconv.Itoa(d.Index))
}
d.Name = pc.IDs().Generate(val, KindFootnote)
}
index = d.Index
name = d.Name
break
}
}
if index == 0 {
return nil
}
return NewFootnoteLink(index, name)
}
type footnoteASTTransformer struct {
}
var defaultFootnoteASTTransformer = &footnoteASTTransformer{}
// NewFootnoteASTTransformer returns a new parser.ASTTransformer that
// insert a footnote list to the last of the document.
func NewFootnoteASTTransformer() parser.ASTTransformer {
return defaultFootnoteASTTransformer
}
func (a *footnoteASTTransformer) Transform(node *ast.Document, reader text.Reader, pc parser.Context) {
var list *FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*FootnoteList)
} else {
return
}
pc.Set(footnoteListKey, nil)
for footnote := list.FirstChild(); footnote != nil; {
var container ast.Node = footnote
next := footnote.NextSibling()
if fc := container.LastChild(); fc != nil && ast.IsParagraph(fc) {
container = fc
}
footnoteNode := footnote.(*Footnote)
index := footnoteNode.Index
name := footnoteNode.Name
if index < 0 {
list.RemoveChild(list, footnote)
} else {
container.AppendChild(container, NewFootnoteBackLink(index, name))
}
footnote = next
}
list.SortChildren(func(n1, n2 ast.Node) int {
if n1.(*Footnote).Index < n2.(*Footnote).Index {
return -1
}
return 1
})
if list.Count <= 0 {
list.Parent().RemoveChild(list.Parent(), list)
return
}
node.AppendChild(node, list)
}
// FootnoteHTMLRenderer is a renderer.NodeRenderer implementation that
// renders FootnoteLink nodes.
type FootnoteHTMLRenderer struct {
html.Config
}
// NewFootnoteHTMLRenderer returns a new FootnoteHTMLRenderer.
func NewFootnoteHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &FootnoteHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *FootnoteHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(KindFootnoteLink, r.renderFootnoteLink)
reg.Register(KindFootnoteBackLink, r.renderFootnoteBackLink)
reg.Register(KindFootnote, r.renderFootnote)
reg.Register(KindFootnoteList, r.renderFootnoteList)
}
func (r *FootnoteHTMLRenderer) renderFootnoteLink(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
n := node.(*FootnoteLink)
n.Dump(source, 0)
is := strconv.Itoa(n.Index)
_, _ = w.WriteString(`<sup id="fnref:`)
_, _ = w.Write(n.Name)
_, _ = w.WriteString(`"><a href="#fn:`)
_, _ = w.Write(n.Name)
_, _ = w.WriteString(`" class="footnote-ref" role="doc-noteref">`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`</a></sup>`)
}
return ast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnoteBackLink(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
n := node.(*FootnoteBackLink)
fmt.Fprintf(os.Stdout, "source:\n%s\n", string(n.Text(source)))
_, _ = w.WriteString(` <a href="#fnref:`)
_, _ = w.Write(n.Name)
_, _ = w.WriteString(`" class="footnote-backref" role="doc-backlink">`)
_, _ = w.WriteString("&#x21a9;&#xfe0e;")
_, _ = w.WriteString(`</a>`)
}
return ast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnote(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*Footnote)
if entering {
fmt.Fprintf(os.Stdout, "source:\n%s\n", string(n.Text(source)))
_, _ = w.WriteString(`<li id="fn:`)
_, _ = w.Write(n.Name)
_, _ = w.WriteString(`" role="doc-endnote"`)
if node.Attributes() != nil {
html.RenderAttributes(w, node, html.ListItemAttributeFilter)
}
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("</li>\n")
}
return ast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnoteList(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
tag := "div"
if entering {
_, _ = w.WriteString("<")
_, _ = w.WriteString(tag)
_, _ = w.WriteString(` class="footnotes" role="doc-endnotes"`)
if node.Attributes() != nil {
html.RenderAttributes(w, node, html.GlobalAttributeFilter)
}
_ = w.WriteByte('>')
if r.Config.XHTML {
_, _ = w.WriteString("\n<hr />\n")
} else {
_, _ = w.WriteString("\n<hr>\n")
}
_, _ = w.WriteString("<ol>\n")
} else {
_, _ = w.WriteString("</ol>\n")
_, _ = w.WriteString("</")
_, _ = w.WriteString(tag)
_, _ = w.WriteString(">\n")
}
return ast.WalkContinue, nil
}
type footnoteExtension struct{}
// FootnoteExtension represents the Gitea Footnote
var FootnoteExtension = &footnoteExtension{}
// Extend extends the markdown converter with the Gitea Footnote parser
func (e *footnoteExtension) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(
parser.WithBlockParsers(
util.Prioritized(NewFootnoteBlockParser(), 999),
),
parser.WithInlineParsers(
util.Prioritized(NewFootnoteParser(), 101),
),
parser.WithASTTransformers(
util.Prioritized(NewFootnoteASTTransformer(), 999),
),
)
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewFootnoteHTMLRenderer(), 500),
))
}

@ -0,0 +1,19 @@
// Copyright 2019 The Gitea Authors. All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
package common
import (
"mvdan.cc/xurls/v2"
)
var (
// NOTE: All below regex matching do not perform any extra validation.
// Thus a link is produced even if the linked entity does not exist.
// While fast, this is also incorrect and lead to false positives.
// TODO: fix invalid linking issue
// LinkRegex is a regexp matching a valid link
LinkRegex, _ = xurls.StrictMatchingScheme("https?://")
)

@ -0,0 +1,156 @@
// Copyright 2019 Yusuke Inuzuka
// Copyright 2019 The Gitea Authors. All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
// Most of this file is a subtly changed version of github.com/yuin/goldmark/extension/linkify.go
package common
import (
"bytes"
"regexp"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var wwwURLRegxp = regexp.MustCompile(`^www\.[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}((?:/|[#?])[-a-zA-Z0-9@:%_\+.~#!?&//=\(\);,'">\^{}\[\]` + "`" + `]*)?`)
type linkifyParser struct {
}
var defaultLinkifyParser = &linkifyParser{}
// NewLinkifyParser return a new InlineParser can parse
// text that seems like a URL.
func NewLinkifyParser() parser.InlineParser {
return defaultLinkifyParser
}
func (s *linkifyParser) Trigger() []byte {
// ' ' indicates any white spaces and a line head
return []byte{' ', '*', '_', '~', '('}
}
var protoHTTP = []byte("http:")
var protoHTTPS = []byte("https:")
var protoFTP = []byte("ftp:")
var domainWWW = []byte("www.")
func (s *linkifyParser) Parse(parent ast.Node, block text.Reader, pc parser.Context) ast.Node {
if pc.IsInLinkLabel() {
return nil
}
line, segment := block.PeekLine()
consumes := 0
start := segment.Start
c := line[0]
// advance if current position is not a line head.
if c == ' ' || c == '*' || c == '_' || c == '~' || c == '(' {
consumes++
start++
line = line[1:]
}
var m []int
var protocol []byte
var typ ast.AutoLinkType = ast.AutoLinkURL
if bytes.HasPrefix(line, protoHTTP) || bytes.HasPrefix(line, protoHTTPS) || bytes.HasPrefix(line, protoFTP) {
m = LinkRegex.FindSubmatchIndex(line)
}
if m == nil && bytes.HasPrefix(line, domainWWW) {
m = wwwURLRegxp.FindSubmatchIndex(line)
protocol = []byte("http")
}
if m != nil {
lastChar := line[m[1]-1]
if lastChar == '.' {
m[1]--
} else if lastChar == ')' {
closing := 0
for i := m[1] - 1; i >= m[0]; i-- {
if line[i] == ')' {
closing++
} else if line[i] == '(' {
closing--
}
}
if closing > 0 {
m[1] -= closing
}
} else if lastChar == ';' {
i := m[1] - 2
for ; i >= m[0]; i-- {
if util.IsAlphaNumeric(line[i]) {
continue
}
break
}
if i != m[1]-2 {
if line[i] == '&' {
m[1] -= m[1] - i
}
}
}
}
if m == nil {
if len(line) > 0 && util.IsPunct(line[0]) {
return nil
}
typ = ast.AutoLinkEmail
stop := util.FindEmailIndex(line)
if stop < 0 {
return nil
}
at := bytes.IndexByte(line, '@')
m = []int{0, stop, at, stop - 1}
if m == nil || bytes.IndexByte(line[m[2]:m[3]], '.') < 0 {
return nil
}
lastChar := line[m[1]-1]
if lastChar == '.' {
m[1]--
}
if m[1] < len(line) {
nextChar := line[m[1]]
if nextChar == '-' || nextChar == '_' {
return nil
}
}
}
if m == nil {
return nil
}
if consumes != 0 {
s := segment.WithStop(segment.Start + 1)
ast.MergeOrAppendTextSegment(parent, s)
}
consumes += m[1]
block.Advance(consumes)
n := ast.NewTextSegment(text.NewSegment(start, start+m[1]))
link := ast.NewAutoLink(typ, n)
link.Protocol = protocol
return link
}
func (s *linkifyParser) CloseBlock(parent ast.Node, pc parser.Context) {
// nothing to do
}
type linkify struct {
}
// Linkify is an extension that allow you to parse text that seems like a URL.
var Linkify = &linkify{}
func (e *linkify) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(
parser.WithInlineParsers(
util.Prioritized(NewLinkifyParser(), 999),
),
)
}

@ -15,6 +15,7 @@ import (
"code.gitea.io/gitea/modules/base"
"code.gitea.io/gitea/modules/git"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/markup/common"
"code.gitea.io/gitea/modules/references"
"code.gitea.io/gitea/modules/setting"
"code.gitea.io/gitea/modules/util"
@ -57,8 +58,6 @@ var (
// https://html.spec.whatwg.org/multipage/input.html#e-mail-state-(type%3Demail)
emailRegex = regexp.MustCompile("(?:\\s|^|\\(|\\[)([a-zA-Z0-9.!#$%&'*+\\/=?^_`{|}~-]+@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\\.[a-zA-Z0-9]{2,}(?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)+)(?:\\s|$|\\)|\\]|\\.(\\s|$))")
linkRegex, _ = xurls.StrictMatchingScheme("https?://")
// blackfriday extensions create IDs like fn:user-content-footnote
blackfridayExtRegex = regexp.MustCompile(`[^:]*:user-content-`)
)
@ -118,7 +117,7 @@ func CustomLinkURLSchemes(schemes []string) {
}
withAuth = append(withAuth, s)
}
linkRegex, _ = xurls.StrictMatchingScheme(strings.Join(withAuth, "|"))
common.LinkRegex, _ = xurls.StrictMatchingScheme(strings.Join(withAuth, "|"))
}
// IsSameDomain checks if given url string has the same hostname as current Gitea instance
@ -509,6 +508,12 @@ func shortLinkProcessorFull(ctx *postProcessCtx, node *html.Node, noLink bool) {
(strings.HasPrefix(val, "") && strings.HasSuffix(val, "")) {
const lenQuote = len("")
val = val[lenQuote : len(val)-lenQuote]
} else if (strings.HasPrefix(val, "\"") && strings.HasSuffix(val, "\"")) ||
(strings.HasPrefix(val, "'") && strings.HasSuffix(val, "'")) {
val = val[1 : len(val)-1]
} else if strings.HasPrefix(val, "'") && strings.HasSuffix(val, "") {
const lenQuote = len("")
val = val[1 : len(val)-lenQuote]
}
props[key] = val
}
@ -803,7 +808,7 @@ func emailAddressProcessor(ctx *postProcessCtx, node *html.Node) {
// linkProcessor creates links for any HTTP or HTTPS URL not captured by
// markdown.
func linkProcessor(ctx *postProcessCtx, node *html.Node) {
m := linkRegex.FindStringIndex(node.Data)
m := common.LinkRegex.FindStringIndex(node.Data)
if m == nil {
return
}
@ -832,7 +837,7 @@ func genDefaultLinkProcessor(defaultLink string) processor {
// descriptionLinkProcessor creates links for DescriptionHTML
func descriptionLinkProcessor(ctx *postProcessCtx, node *html.Node) {
m := linkRegex.FindStringIndex(node.Data)
m := common.LinkRegex.FindStringIndex(node.Data)
if m == nil {
return
}

@ -323,6 +323,6 @@ func TestRender_ShortLinks(t *testing.T) {
`<p><a href="`+notencodedImgurlWiki+`" rel="nofollow"><img src="`+notencodedImgurlWiki+`"/></a></p>`)
test(
"<p><a href=\"https://example.org\">[[foobar]]</a></p>",
`<p></p><p><a href="https://example.org" rel="nofollow">[[foobar]]</a></p><p></p>`,
`<p></p><p><a href="https://example.org" rel="nofollow">[[foobar]]</a></p><p></p>`)
`<p><a href="https://example.org" rel="nofollow">[[foobar]]</a></p>`,
`<p><a href="https://example.org" rel="nofollow">[[foobar]]</a></p>`)
}

@ -0,0 +1,178 @@
// Copyright 2019 The Gitea Authors. All rights reserved.
// Use of this source code is governed by a MIT-style
// license that can be found in the LICENSE file.
package markdown
import (
"bytes"
"fmt"
"strings"
"code.gitea.io/gitea/modules/markup"
"code.gitea.io/gitea/modules/markup/common"
giteautil "code.gitea.io/gitea/modules/util"
"github.com/yuin/goldmark/ast"
east "github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var byteMailto = []byte("mailto:")
// GiteaASTTransformer is a default transformer of the goldmark tree.
type GiteaASTTransformer struct{}
// Transform transforms the given AST tree.
func (g *GiteaASTTransformer) Transform(node *ast.Document, reader text.Reader, pc parser.Context) {
_ = ast.Walk(node, func(n ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
switch v := n.(type) {
case *ast.Image:
// Images need two things:
//
// 1. Their src needs to munged to be a real value
// 2. If they're not wrapped with a link they need a link wrapper
// Check if the destination is a real link
link := v.Destination
if len(link) > 0 && !markup.IsLink(link) {
prefix := pc.Get(urlPrefixKey).(string)
if pc.Get(isWikiKey).(bool) {
prefix = giteautil.URLJoin(prefix, "wiki", "raw")
}
prefix = strings.Replace(prefix, "/src/", "/media/", 1)
lnk := string(link)
lnk = giteautil.URLJoin(prefix, lnk)
lnk = strings.Replace(lnk, " ", "+", -1)
link = []byte(lnk)
}
v.Destination = link
parent := n.Parent()
// Create a link around image only if parent is not already a link
if _, ok := parent.(*ast.Link); !ok && parent != nil {
wrap := ast.NewLink()
wrap.Destination = link
wrap.Title = v.Title
parent.ReplaceChild(parent, n, wrap)
wrap.AppendChild(wrap, n)
}
case *ast.Link:
// Links need their href to munged to be a real value
link := v.Destination
if len(link) > 0 && !markup.IsLink(link) &&
link[0] != '#' && !bytes.HasPrefix(link, byteMailto) {
// special case: this is not a link, a hash link or a mailto:, so it's a
// relative URL
lnk := string(link)
if pc.Get(isWikiKey).(bool) {
lnk = giteautil.URLJoin("wiki", lnk)
}
link = []byte(giteautil.URLJoin(pc.Get(urlPrefixKey).(string), lnk))
}
v.Destination = link
}
return ast.WalkContinue, nil
})
}
type prefixedIDs struct {
values map[string]bool
}
// Generate generates a new element id.
func (p *prefixedIDs) Generate(value []byte, kind ast.NodeKind) []byte {
dft := []byte("id")
if kind == ast.KindHeading {
dft = []byte("heading")
}
return p.GenerateWithDefault(value, dft)
}
// Generate generates a new element id.
func (p *prefixedIDs) GenerateWithDefault(value []byte, dft []byte) []byte {
result := common.CleanValue(value)
if len(result) == 0 {
result = dft
}
if !bytes.HasPrefix(result, []byte("user-content-")) {
result = append([]byte("user-content-"), result...)
}
if _, ok := p.values[util.BytesToReadOnlyString(result)]; !ok {
p.values[util.BytesToReadOnlyString(result)] = true
return result
}
for i := 1; ; i++ {
newResult := fmt.Sprintf("%s-%d", result, i)
if _, ok := p.values[newResult]; !ok {
p.values[newResult] = true
return []byte(newResult)
}
}
}
// Put puts a given element id to the used ids table.
func (p *prefixedIDs) Put(value []byte) {
p.values[util.BytesToReadOnlyString(value)] = true
}
func newPrefixedIDs() *prefixedIDs {
return &prefixedIDs{
values: map[string]bool{},
}
}
// NewTaskCheckBoxHTMLRenderer creates a TaskCheckBoxHTMLRenderer to render tasklists
// in the gitea form.
func NewTaskCheckBoxHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &TaskCheckBoxHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// TaskCheckBoxHTMLRenderer is a renderer.NodeRenderer implementation that
// renders checkboxes in list items.
// Overrides the default goldmark one to present the gitea format
type TaskCheckBoxHTMLRenderer struct {
html.Config
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *TaskCheckBoxHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(east.KindTaskCheckBox, r.renderTaskCheckBox)
}
func (r *TaskCheckBoxHTMLRenderer) renderTaskCheckBox(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
n := node.(*east.TaskCheckBox)
end := ">"
if r.XHTML {
end = " />"
}
var err error
if n.IsChecked {
_, err = w.WriteString(`<span class="ui fitted disabled checkbox"><input type="checkbox" disabled="disabled"` + end + `<label` + end + `</span>`)
} else {
_, err = w.WriteString(`<span class="ui checked fitted disabled checkbox"><input type="checkbox" checked="" disabled="disabled"` + end + `<label` + end + `</span>`)
}
if err != nil {
return ast.WalkStop, err
}
return ast.WalkContinue, nil
}

@ -7,161 +7,83 @@ package markdown
import (
"bytes"
"io"
"strings"
"sync"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/markup"
"code.gitea.io/gitea/modules/markup/common"
"code.gitea.io/gitea/modules/setting"
"code.gitea.io/gitea/modules/util"
"github.com/russross/blackfriday/v2"
giteautil "code.gitea.io/gitea/modules/util"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/extension"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/util"
)
// Renderer is a extended version of underlying render object.
type Renderer struct {
blackfriday.Renderer
URLPrefix string
IsWiki bool
}
var converter goldmark.Markdown
var once = sync.Once{}
var byteMailto = []byte("mailto:")
var urlPrefixKey = parser.NewContextKey()
var isWikiKey = parser.NewContextKey()
var htmlEscaper = [256][]byte{
'&': []byte("&amp;"),
'<': []byte("&lt;"),
'>': []byte("&gt;"),
'"': []byte("&quot;"),
// NewGiteaParseContext creates a parser.Context with the gitea context set
func NewGiteaParseContext(urlPrefix string, isWiki bool) parser.Context {
pc := parser.NewContext(parser.WithIDs(newPrefixedIDs()))
pc.Set(urlPrefixKey, urlPrefix)
pc.Set(isWikiKey, isWiki)
return pc
}
func escapeHTML(w io.Writer, s []byte) {
var start, end int
for end < len(s) {
escSeq := htmlEscaper[s[end]]
if escSeq != nil {
_, _ = w.Write(s[start:end])
_, _ = w.Write(escSeq)
start = end + 1
}
end++
}
if start < len(s) && end <= len(s) {
_, _ = w.Write(s[start:end])
}
}
// RenderNode is a default renderer of a single node of a syntax tree. For
// block nodes it will be called twice: first time with entering=true, second
// time with entering=false, so that it could know when it's working on an open
// tag and when on close. It writes the result to w.
//
// The return value is a way to tell the calling walker to adjust its walk
// pattern: e.g. it can terminate the traversal by returning Terminate. Or it
// can ask the walker to skip a subtree of this node by returning SkipChildren.
// The typical behavior is to return GoToNext, which asks for the usual
// traversal to the next node.
func (r *Renderer) RenderNode(w io.Writer, node *blackfriday.Node, entering bool) blackfriday.WalkStatus {
switch node.Type {
case blackfriday.Image:
prefix := r.URLPrefix
if r.IsWiki {
prefix = util.URLJoin(prefix, "wiki", "raw")
}
prefix = strings.Replace(prefix, "/src/", "/media/", 1)
link := node.LinkData.Destination
if len(link) > 0 && !markup.IsLink(link) {
lnk := string(link)
lnk = util.URLJoin(prefix, lnk)
lnk = strings.Replace(lnk, " ", "+", -1)
link = []byte(lnk)
}
node.LinkData.Destination = link
// Render link around image only if parent is not link already
if node.Parent != nil && node.Parent.Type != blackfriday.Link {
if entering {
_, _ = w.Write([]byte(`<a href="`))
escapeHTML(w, link)
_, _ = w.Write([]byte(`">`))
return r.Renderer.RenderNode(w, node, entering)
}
s := r.Renderer.RenderNode(w, node, entering)
_, _ = w.Write([]byte(`</a>`))
return s
}
return r.Renderer.RenderNode(w, node, entering)
case blackfriday.Link:
// special case: this is not a link, a hash link or a mailto:, so it's a
// relative URL
link := node.LinkData.Destination
if len(link) > 0 && !markup.IsLink(link) &&
link[0] != '#' && !bytes.HasPrefix(link, byteMailto) &&
node.LinkData.Footnote == nil {
lnk := string(link)
if r.IsWiki {
lnk = util.URLJoin("wiki", lnk)
}
link = []byte(util.URLJoin(r.URLPrefix, lnk))
}
node.LinkData.Destination = link
return r.Renderer.RenderNode(w, node, entering)
case blackfriday.Text:
isListItem := false
for n := node.Parent; n != nil; n = n.Parent {
if n.Type == blackfriday.Item {
isListItem = true
break
}
}
if isListItem {
text := node.Literal
switch {
case bytes.HasPrefix(text, []byte("[ ] ")):
_, _ = w.Write([]byte(`<span class="ui fitted disabled checkbox"><input type="checkbox" disabled="disabled" /><label /></span>`))
text = text[3:]
case bytes.HasPrefix(text, []byte("[x] ")):
_, _ = w.Write([]byte(`<span class="ui checked fitted disabled checkbox"><input type="checkbox" checked="" disabled="disabled" /><label /></span>`))
text = text[3:]
}
node.Literal = text
}
}
return r.Renderer.RenderNode(w, node, entering)
}
const (
blackfridayExtensions = 0 |
blackfriday.NoIntraEmphasis |
blackfriday.Tables |
blackfriday.FencedCode |
blackfriday.Strikethrough |
blackfriday.NoEmptyLineBeforeBlock |
blackfriday.DefinitionLists |
blackfriday.Footnotes |
blackfriday.HeadingIDs |
blackfriday.AutoHeadingIDs
blackfridayHTMLFlags = 0 |
blackfriday.Smartypants
)
// RenderRaw renders Markdown to HTML without handling special links.
func RenderRaw(body []byte, urlPrefix string, wikiMarkdown bool) []byte {
renderer := &Renderer{
Renderer: blackfriday.NewHTMLRenderer(blackfriday.HTMLRendererParameters{
Flags: blackfridayHTMLFlags,
FootnoteAnchorPrefix: "user-content-",
HeadingIDPrefix: "user-content-",
}),
URLPrefix: urlPrefix,
IsWiki: wikiMarkdown,
}
once.Do(func() {
converter = goldmark.New(
goldmark.WithExtensions(extension.Table,
extension.Strikethrough,
extension.TaskList,
extension.DefinitionList,
common.FootnoteExtension,
extension.NewTypographer(
extension.WithTypographicSubstitutions(extension.TypographicSubstitutions{
extension.EnDash: nil,
extension.EmDash: nil,
}),
),
),
goldmark.WithParserOptions(
parser.WithAttribute(),
parser.WithAutoHeadingID(),
parser.WithASTTransformers(
util.Prioritized(&GiteaASTTransformer{}, 10000),
),
),
goldmark.WithRendererOptions(
html.WithUnsafe(),
),
)
// Override the original Tasklist renderer!
converter.Renderer().AddOptions(
renderer.WithNodeRenderers(
util.Prioritized(NewTaskCheckBoxHTMLRenderer(), 1000),
),
)
if setting.Markdown.EnableHardLineBreak {
converter.Renderer().AddOptions(html.WithHardWraps())
}
})
exts := blackfridayExtensions
if setting.Markdown.EnableHardLineBreak {
exts |= blackfriday.HardLineBreak
pc := NewGiteaParseContext(urlPrefix, wikiMarkdown)
var buf bytes.Buffer
if err := converter.Convert(giteautil.NormalizeEOL(body), &buf, parser.WithContext(pc)); err != nil {
log.Error("Unable to render: %v", err)
}
// Need to normalize EOL to UNIX LF to have consistent results in rendering
body = blackfriday.Run(util.NormalizeEOL(body), blackfriday.WithRenderer(renderer), blackfriday.WithExtensions(exts))
return markup.SanitizeBytes(body)
return markup.SanitizeReader(&buf).Bytes()
}
var (
@ -174,8 +96,7 @@ func init() {
}
// Parser implements markup.Parser
type Parser struct {
}
type Parser struct{}
// Name implements markup.Parser
func (Parser) Name() string {

@ -98,16 +98,12 @@ func TestRender_Images(t *testing.T) {
func testAnswers(baseURLContent, baseURLImages string) []string {
return []string{
`<p>Wiki! Enjoy :)</p>
<ul>
<li><a href="` + baseURLContent + `/Links" rel="nofollow">Links, Language bindings, Engine bindings</a></li>
<li><a href="` + baseURLContent + `/Tips" rel="nofollow">Tips</a></li>
</ul>
<p>See commit <a href="http://localhost:3000/gogits/gogs/commit/65f1bf27bc" rel="nofollow"><code>65f1bf27bc</code></a></p>
<p>Ideas and codes</p>
<ul>
<li>Bezier widget (by <a href="` + AppURL + `r-lyeh" rel="nofollow">@r-lyeh</a>) <a href="http://localhost:3000/ocornut/imgui/issues/786" rel="nofollow">ocornut/imgui#786</a></li>
<li>Bezier widget (by <a href="` + AppURL + `r-lyeh" rel="nofollow">@r-lyeh</a>) <a href="http://localhost:3000/gogits/gogs/issues/786" rel="nofollow">#786</a></li>
@ -117,13 +113,9 @@ func testAnswers(baseURLContent, baseURLImages string) []string {
</ul>
`,
`<h2 id="user-content-what-is-wine-staging">What is Wine Staging?</h2>
<p><strong>Wine Staging</strong> on website <a href="http://wine-staging.com" rel="nofollow">wine-staging.com</a>.</p>
<h2 id="user-content-quick-links">Quick Links</h2>
<p>Here are some links to the most important topics. You can find the full list of pages at the sidebar.</p>
<table>
<thead>
<tr>
@ -131,7 +123,6 @@ func testAnswers(baseURLContent, baseURLImages string) []string {
<th><a href="` + baseURLContent + `/Installation" rel="nofollow">Installation</a></th>
</tr>
</thead>
<tbody>
<tr>
<td><a href="` + baseURLImages + `/images/icon-usage.png" rel="nofollow"><img src="` + baseURLImages + `/images/icon-usage.png" title="icon-usage.png" alt="images/icon-usage.png"/></a></td>
@ -141,20 +132,15 @@ func testAnswers(baseURLContent, baseURLImages string) []string {
</table>
`,
`<p><a href="http://www.excelsiorjet.com/" rel="nofollow">Excelsior JET</a> allows you to create native executables for Windows, Linux and Mac OS X.</p>
<ol>
<li><a href="https://github.com/libgdx/libgdx/wiki/Gradle-on-the-Commandline#packaging-for-the-desktop" rel="nofollow">Package your libGDX application</a>
<a href="` + baseURLImages + `/images/1.png" rel="nofollow"><img src="` + baseURLImages + `/images/1.png" title="1.png" alt="images/1.png"/></a></li>
<li>Perform a test run by hitting the Run! button.
<a href="` + baseURLImages + `/images/2.png" rel="nofollow"><img src="` + baseURLImages + `/images/2.png" title="2.png" alt="images/2.png"/></a></li>
</ol>
<h2 id="user-content-custom-id">More tests</h2>
<p>(from <a href="https://www.markdownguide.org/extended-syntax/" rel="nofollow">https://www.markdownguide.org/extended-syntax/</a>)</p>
<h3 id="user-content-definition-list">Definition list</h3>
<dl>
<dt>First Term</dt>
<dd>This is the definition of the first term.</dd>
@ -162,27 +148,21 @@ func testAnswers(baseURLContent, baseURLImages string) []string {
<dd>This is one definition of the second term.</dd>
<dd>This is another definition of the second term.</dd>
</dl>
<h3 id="user-content-footnotes">Footnotes</h3>
<p>Here is a simple footnote,<sup id="fnref:user-content-1"><a href="#fn:user-content-1" rel="nofollow">1</a></sup> and here is a longer one.<sup id="fnref:user-content-bignote"><a href="#fn:user-content-bignote" rel="nofollow">2</a></sup></p>
<div>
<hr/>
<ol>
<li id="fn:user-content-1">This is the first footnote.</li>
<li id="fn:user-content-bignote"><p>Here is one with multiple paragraphs and code.</p>
<li id="fn:user-content-1">
<p>This is the first footnote. <a href="#fnref:user-content-1" rel="nofollow"></a></p>
</li>
<li id="fn:user-content-bignote">
<p>Here is one with multiple paragraphs and code.</p>
<p>Indent paragraphs to include them in the footnote.</p>
<p><code>{ my code }</code></p>
<p>Add as many paragraphs as you like.</p></li>
<p>Add as many paragraphs as you like. <a href="#fnref:user-content-bignote" rel="nofollow"></a></p>
</li>
</ol>
</div>
`,
}
@ -299,15 +279,15 @@ func TestRender_RenderParagraphs(t *testing.T) {
test := func(t *testing.T, str string, cnt int) {
unix := []byte(str)
res := string(RenderRaw(unix, "", false))
assert.Equal(t, strings.Count(res, "<p"), cnt)
assert.Equal(t, strings.Count(res, "<p"), cnt, "Rendered result for unix should have %d paragraph(s) but has %d:\n%s\n", cnt, strings.Count(res, "<p"), res)
mac := []byte(strings.ReplaceAll(str, "\n", "\r"))
res = string(RenderRaw(mac, "", false))
assert.Equal(t, strings.Count(res, "<p"), cnt)
assert.Equal(t, strings.Count(res, "<p"), cnt, "Rendered result for mac should have %d paragraph(s) but has %d:\n%s\n", cnt, strings.Count(res, "<p"), res)
dos := []byte(strings.ReplaceAll(str, "\n", "\r\n"))
res = string(RenderRaw(dos, "", false))
assert.Equal(t, strings.Count(res, "<p"), cnt)
assert.Equal(t, strings.Count(res, "<p"), cnt, "Rendered result for windows should have %d paragraph(s) but has %d:\n%s\n", cnt, strings.Count(res, "<p"), res)
}
test(t, "\nOne\nTwo\nThree", 1)

@ -6,113 +6,128 @@ package mdstripper
import (
"bytes"
"io"
"sync"
"github.com/russross/blackfriday/v2"
)
"io"
// MarkdownStripper extends blackfriday.Renderer
type MarkdownStripper struct {
links []string
coallesce bool
empty bool
}
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/markup/common"
const (
blackfridayExtensions = 0 |
blackfriday.NoIntraEmphasis |
blackfriday.Tables |
blackfriday.FencedCode |
blackfriday.Strikethrough |
blackfriday.NoEmptyLineBeforeBlock |
blackfriday.DefinitionLists |
blackfriday.Footnotes |
blackfriday.HeadingIDs |
blackfriday.AutoHeadingIDs |
// Not included in modules/markup/markdown/markdown.go;
// required here to process inline links
blackfriday.Autolink
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
)
// StripMarkdown parses markdown content by removing all markup and code blocks
// in order to extract links and other references
func StripMarkdown(rawBytes []byte) (string, []string) {
buf, links := StripMarkdownBytes(rawBytes)
return string(buf), links
type stripRenderer struct {
links []string
empty bool
}
// StripMarkdownBytes parses markdown content by removing all markup and code blocks
// in order to extract links and other references
func StripMarkdownBytes(rawBytes []byte) ([]byte, []string) {
stripper := &MarkdownStripper{
links: make([]string, 0, 10),
empty: true,
}
parser := blackfriday.New(blackfriday.WithRenderer(stripper), blackfriday.WithExtensions(blackfridayExtensions))
ast := parser.Parse(rawBytes)
var buf bytes.Buffer
stripper.RenderHeader(&buf, ast)
ast.Walk(func(node *blackfriday.Node, entering bool) blackfriday.WalkStatus {
return stripper.RenderNode(&buf, node, entering)
func (r *stripRenderer) Render(w io.Writer, source []byte, doc ast.Node) error {
return ast.Walk(doc, func(n ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
switch v := n.(type) {
case *ast.Text:
if !v.IsRaw() {
_, prevSibIsText := n.PreviousSibling().(*ast.Text)
coalesce := prevSibIsText
r.processString(
w,
v.Text(source),
coalesce)
if v.SoftLineBreak() {
r.doubleSpace(w)
}
}
return ast.WalkContinue, nil
case *ast.Link:
r.processLink(w, v.Destination)
return ast.WalkSkipChildren, nil
case *ast.AutoLink:
r.processLink(w, v.URL(source))
return ast.WalkSkipChildren, nil
}
return ast.WalkContinue, nil
})
stripper.RenderFooter(&buf, ast)
return buf.Bytes(), stripper.GetLinks()
}
// RenderNode is the main rendering method. It will be called once for
// every leaf node and twice for every non-leaf node (first with
// entering=true, then with entering=false). The method should write its
// rendition of the node to the supplied writer w.
func (r *MarkdownStripper) RenderNode(w io.Writer, node *blackfriday.Node, entering bool) blackfriday.WalkStatus {
if !entering {
return blackfriday.GoToNext
}
switch node.Type {
case blackfriday.Text:
r.processString(w, node.Literal, node.Parent == nil)
return blackfriday.GoToNext
case blackfriday.Link:
r.processLink(w, node.LinkData.Destination)
r.coallesce = false
return blackfriday.SkipChildren
}
r.coallesce = false
return blackfriday.GoToNext
}
// RenderHeader is a method that allows the renderer to produce some
// content preceding the main body of the output document.
func (r *MarkdownStripper) RenderHeader(w io.Writer, ast *blackfriday.Node) {
}
// RenderFooter is a symmetric counterpart of RenderHeader.
func (r *MarkdownStripper) RenderFooter(w io.Writer, ast *blackfriday.Node) {
}
func (r *MarkdownStripper) doubleSpace(w io.Writer) {
func (r *stripRenderer) doubleSpace(w io.Writer) {
if !r.empty {
_, _ = w.Write([]byte{'\n'})
}
}
func (r *MarkdownStripper) processString(w io.Writer, text []byte, coallesce bool) {
func (r *stripRenderer) processString(w io.Writer, text []byte, coalesce bool) {
// Always break-up words
if !coallesce || !r.coallesce {
if !coalesce {
r.doubleSpace(w)
}
_, _ = w.Write(text)
r.coallesce = coallesce
r.empty = false
}
func (r *MarkdownStripper) processLink(w io.Writer, link []byte) {
func (r *stripRenderer) processLink(w io.Writer, link []byte) {
// Links are processed out of band
r.links = append(r.links, string(link))
r.coallesce = false
}
// GetLinks returns the list of link data collected while parsing
func (r *MarkdownStripper) GetLinks() []string {
func (r *stripRenderer) GetLinks() []string {
return r.links
}
// AddOptions adds given option to this renderer.
func (r *stripRenderer) AddOptions(...renderer.Option) {
// no-op
}
// StripMarkdown parses markdown content by removing all markup and code blocks
// in order to extract links and other references
func StripMarkdown(rawBytes []byte) (string, []string) {
buf, links := StripMarkdownBytes(rawBytes)
return string(buf), links
}
var stripParser parser.Parser
var once = sync.Once{}
// StripMarkdownBytes parses markdown content by removing all markup and code blocks
// in order to extract links and other references
func StripMarkdownBytes(rawBytes []byte) ([]byte, []string) {
once.Do(func() {
gdMarkdown := goldmark.New(
goldmark.WithExtensions(extension.Table,
extension.Strikethrough,
extension.TaskList,
extension.DefinitionList,
common.FootnoteExtension,
common.Linkify,
),
goldmark.WithParserOptions(
parser.WithAttribute(),
parser.WithAutoHeadingID(),
),
goldmark.WithRendererOptions(
html.WithUnsafe(),
),
)
stripParser = gdMarkdown.Parser()
})
stripper := &stripRenderer{
links: make([]string, 0, 10),
empty: true,
}
reader := text.NewReader(rawBytes)
doc := stripParser.Parse(reader)
var buf bytes.Buffer
if err := stripper.Render(&buf, rawBytes, doc); err != nil {
log.Error("Unable to strip: %v", err)
}
return buf.Bytes(), stripper.GetLinks()
}

@ -53,6 +53,20 @@ A HIDDEN ` + "`" + `GHOST` + "`" + ` IN THIS LINE.
[]string{
"link",
}},
{
"Simply closes: #29 yes",
[]string{
"Simply closes: #29 yes",
},
[]string{},
},
{
"Simply closes: !29 yes",
[]string{
"Simply closes: !29 yes",
},
[]string{},
},
}
for _, test := range list {

@ -6,6 +6,8 @@
package markup
import (
"bytes"
"io"
"regexp"
"sync"
@ -67,6 +69,12 @@ func Sanitize(s string) string {
return sanitizer.policy.Sanitize(s)
}
// SanitizeReader sanitizes a Reader
func SanitizeReader(r io.Reader) *bytes.Buffer {
NewSanitizer()
return sanitizer.policy.SanitizeReader(r)
}
// SanitizeBytes takes a []byte slice that contains a HTML fragment or document and applies policy whitelist.
func SanitizeBytes(b []byte) []byte {
if len(b) == 0 {

@ -43,10 +43,6 @@ func TestFindAllIssueReferences(t *testing.T) {
{29, "", "", "29", true, XRefActionCloses, &RefSpan{Start: 15, End: 18}, &RefSpan{Start: 7, End: 13}},
},
},
{
"#123 no, this is a title.",
[]testResult{},
},
{
" #124 yes, this is a reference.",
[]testResult{

@ -69,7 +69,6 @@ func TestAPI_RenderGFM(t *testing.T) {
- Bezier widget (by @r-lyeh) https://github.com/ocornut/imgui/issues/786`,
// rendered
`<p>Wiki! Enjoy :)</p>
<ul>
<li><a href="` + AppSubURL + `wiki/Links" rel="nofollow">Links, Language bindings, Engine bindings</a></li>
<li><a href="` + AppSubURL + `wiki/Tips" rel="nofollow">Tips</a></li>
@ -88,13 +87,9 @@ Here are some links to the most important topics. You can find the full list of
`,
// rendered
`<h2 id="user-content-what-is-wine-staging">What is Wine Staging?</h2>
<p><strong>Wine Staging</strong> on website <a href="http://wine-staging.com" rel="nofollow">wine-staging.com</a>.</p>
<h2 id="user-content-quick-links">Quick Links</h2>
<p>Here are some links to the most important topics. You can find the full list of pages at the sidebar.</p>
<p><a href="` + AppSubURL + `wiki/Configuration" rel="nofollow">Configuration</a>
<a href="` + AppSubURL + `wiki/raw/images/icon-bug.png" rel="nofollow"><img src="` + AppSubURL + `wiki/raw/images/icon-bug.png" title="icon-bug.png" alt="images/icon-bug.png"/></a></p>
`,

@ -1,8 +0,0 @@
*.out
*.swp
*.8
*.6
_obj
_test*
markdown
tags

@ -1,17 +0,0 @@
sudo: false
language: go
go:
- "1.10.x"
- "1.11.x"
- tip
matrix:
fast_finish: true
allow_failures:
- go: tip
install:
- # Do nothing. This is needed to prevent default install action "go get -t -v ./..." from happening here (we want it to happen inside script step).
script:
- go get -t -v ./...
- diff -u <(echo -n) <(gofmt -d -s .)
- go tool vet .
- go test -v ./...

@ -1,29 +0,0 @@
Blackfriday is distributed under the Simplified BSD License:
> Copyright © 2011 Russ Ross
> All rights reserved.
>
> Redistribution and use in source and binary forms, with or without
> modification, are permitted provided that the following conditions
> are met:
>
> 1. Redistributions of source code must retain the above copyright
> notice, this list of conditions and the following disclaimer.
>
> 2. Redistributions in binary form must reproduce the above
> copyright notice, this list of conditions and the following
> disclaimer in the documentation and/or other materials provided with
> the distribution.
>
> THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
> "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
> LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
> FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
> COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
> INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
> BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
> LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
> CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
> LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
> ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
> POSSIBILITY OF SUCH DAMAGE.

@ -1,291 +0,0 @@
Blackfriday [![Build Status](https://travis-ci.org/russross/blackfriday.svg?branch=master)](https://travis-ci.org/russross/blackfriday)
===========
Blackfriday is a [Markdown][1] processor implemented in [Go][2]. It
is paranoid about its input (so you can safely feed it user-supplied
data), it is fast, it supports common extensions (tables, smart
punctuation substitutions, etc.), and it is safe for all utf-8
(unicode) input.
HTML output is currently supported, along with Smartypants
extensions.
It started as a translation from C of [Sundown][3].
Installation
------------
Blackfriday is compatible with any modern Go release. With Go 1.7 and git
installed:
go get gopkg.in/russross/blackfriday.v2
will download, compile, and install the package into your `$GOPATH`
directory hierarchy. Alternatively, you can achieve the same if you
import it into a project:
import "gopkg.in/russross/blackfriday.v2"
and `go get` without parameters.
Versions
--------
Currently maintained and recommended version of Blackfriday is `v2`. It's being
developed on its own branch: https://github.com/russross/blackfriday/tree/v2 and the
documentation is available at
https://godoc.org/gopkg.in/russross/blackfriday.v2.
It is `go get`-able via via [gopkg.in][6] at `gopkg.in/russross/blackfriday.v2`,
but we highly recommend using package management tool like [dep][7] or
[Glide][8] and make use of semantic versioning. With package management you
should import `github.com/russross/blackfriday` and specify that you're using
version 2.0.0.
Version 2 offers a number of improvements over v1:
* Cleaned up API
* A separate call to [`Parse`][4], which produces an abstract syntax tree for
the document
* Latest bug fixes
* Flexibility to easily add your own rendering extensions
Potential drawbacks:
* Our benchmarks show v2 to be slightly slower than v1. Currently in the
ballpark of around 15%.
* API breakage. If you can't afford modifying your code to adhere to the new API
and don't care too much about the new features, v2 is probably not for you.
* Several bug fixes are trailing behind and still need to be forward-ported to
v2. See issue [#348](https://github.com/russross/blackfriday/issues/348) for
tracking.
Usage
-----
For the most sensible markdown processing, it is as simple as getting your input
into a byte slice and calling:
```go
output := blackfriday.Run(input)
```
Your input will be parsed and the output rendered with a set of most popular
extensions enabled. If you want the most basic feature set, corresponding with
the bare Markdown specification, use:
```go
output := blackfriday.Run(input, blackfriday.WithNoExtensions())
```
### Sanitize untrusted content
Blackfriday itself does nothing to protect against malicious content. If you are
dealing with user-supplied markdown, we recommend running Blackfriday's output
through HTML sanitizer such as [Bluemonday][5].
Here's an example of simple usage of Blackfriday together with Bluemonday:
```go
import (
"github.com/microcosm-cc/bluemonday"
"github.com/russross/blackfriday"
)
// ...
unsafe := blackfriday.Run(input)
html := bluemonday.UGCPolicy().SanitizeBytes(unsafe)
```
### Custom options
If you want to customize the set of options, use `blackfriday.WithExtensions`,
`blackfriday.WithRenderer` and `blackfriday.WithRefOverride`.
You can also check out `blackfriday-tool` for a more complete example
of how to use it. Download and install it using:
go get github.com/russross/blackfriday-tool
This is a simple command-line tool that allows you to process a
markdown file using a standalone program. You can also browse the
source directly on github if you are just looking for some example
code:
* <http://github.com/russross/blackfriday-tool>
Note that if you have not already done so, installing
`blackfriday-tool` will be sufficient to download and install
blackfriday in addition to the tool itself. The tool binary will be
installed in `$GOPATH/bin`. This is a statically-linked binary that
can be copied to wherever you need it without worrying about
dependencies and library versions.
Features
--------
All features of Sundown are supported, including:
* **Compatibility**. The Markdown v1.0.3 test suite passes with
the `--tidy` option. Without `--tidy`, the differences are
mostly in whitespace and entity escaping, where blackfriday is
more consistent and cleaner.
* **Common extensions**, including table support, fenced code
blocks, autolinks, strikethroughs, non-strict emphasis, etc.
* **Safety**. Blackfriday is paranoid when parsing, making it safe
to feed untrusted user input without fear of bad things
happening. The test suite stress tests this and there are no
known inputs that make it crash. If you find one, please let me
know and send me the input that does it.
NOTE: "safety" in this context means *runtime safety only*. In order to
protect yourself against JavaScript injection in untrusted content, see
[this example](https://github.com/russross/blackfriday#sanitize-untrusted-content).
* **Fast processing**. It is fast enough to render on-demand in
most web applications without having to cache the output.
* **Thread safety**. You can run multiple parsers in different
goroutines without ill effect. There is no dependence on global
shared state.
* **Minimal dependencies**. Blackfriday only depends on standard
library packages in Go. The source code is pretty
self-contained, so it is easy to add to any project, including
Google App Engine projects.
* **Standards compliant**. Output successfully validates using the
W3C validation tool for HTML 4.01 and XHTML 1.0 Transitional.
Extensions
----------
In addition to the standard markdown syntax, this package
implements the following extensions:
* **Intra-word emphasis supression**. The `_` character is
commonly used inside words when discussing code, so having
markdown interpret it as an emphasis command is usually the
wrong thing. Blackfriday lets you treat all emphasis markers as
normal characters when they occur inside a word.
* **Tables**. Tables can be created by drawing them in the input
using a simple syntax:
```
Name | Age
--------|------
Bob | 27
Alice | 23
```
* **Fenced code blocks**. In addition to the normal 4-space
indentation to mark code blocks, you can explicitly mark them
and supply a language (to make syntax highlighting simple). Just
mark it like this:
```go
func getTrue() bool {
return true
}
```
You can use 3 or more backticks to mark the beginning of the
block, and the same number to mark the end of the block.
* **Definition lists**. A simple definition list is made of a single-line
term followed by a colon and the definition for that term.
Cat
: Fluffy animal everyone likes
Internet
: Vector of transmission for pictures of cats
Terms must be separated from the previous definition by a blank line.
* **Footnotes**. A marker in the text that will become a superscript number;
a footnote definition that will be placed in a list of footnotes at the
end of the document. A footnote looks like this:
This is a footnote.[^1]
[^1]: the footnote text.
* **Autolinking**. Blackfriday can find URLs that have not been
explicitly marked as links and turn them into links.
* **Strikethrough**. Use two tildes (`~~`) to mark text that
should be crossed out.
* **Hard line breaks**. With this extension enabled newlines in the input
translate into line breaks in the output. This extension is off by default.
* **Smart quotes**. Smartypants-style punctuation substitution is
supported, turning normal double- and single-quote marks into
curly quotes, etc.
* **LaTeX-style dash parsing** is an additional option, where `--`
is translated into `&ndash;`, and `---` is translated into
`&mdash;`. This differs from most smartypants processors, which
turn a single hyphen into an ndash and a double hyphen into an
mdash.
* **Smart fractions**, where anything that looks like a fraction
is translated into suitable HTML (instead of just a few special
cases like most smartypant processors). For example, `4/5`
becomes `<sup>4</sup>&frasl;<sub>5</sub>`, which renders as
<sup>4</sup>&frasl;<sub>5</sub>.
Other renderers
---------------
Blackfriday is structured to allow alternative rendering engines. Here
are a few of note:
* [github_flavored_markdown](https://godoc.org/github.com/shurcooL/github_flavored_markdown):
provides a GitHub Flavored Markdown renderer with fenced code block
highlighting, clickable heading anchor links.
It's not customizable, and its goal is to produce HTML output
equivalent to the [GitHub Markdown API endpoint](https://developer.github.com/v3/markdown/#render-a-markdown-document-in-raw-mode),
except the rendering is performed locally.
* [markdownfmt](https://github.com/shurcooL/markdownfmt): like gofmt,
but for markdown.
* [LaTeX output](https://github.com/Ambrevar/Blackfriday-LaTeX):
renders output as LaTeX.
* [Blackfriday-Confluence](https://github.com/kentaro-m/blackfriday-confluence): provides a [Confluence Wiki Markup](https://confluence.atlassian.com/doc/confluence-wiki-markup-251003035.html) renderer.
Todo
----
* More unit testing
* Improve unicode support. It does not understand all unicode
rules (about what constitutes a letter, a punctuation symbol,
etc.), so it may fail to detect word boundaries correctly in
some instances. It is safe on all utf-8 input.
License
-------
[Blackfriday is distributed under the Simplified BSD License](LICENSE.txt)
[1]: https://daringfireball.net/projects/markdown/ "Markdown"
[2]: https://golang.org/ "Go Language"
[3]: https://github.com/vmg/sundown "Sundown"
[4]: https://godoc.org/gopkg.in/russross/blackfriday.v2#Parse "Parse func"
[5]: https://github.com/microcosm-cc/bluemonday "Bluemonday"
[6]: https://labix.org/gopkg.in "gopkg.in"

File diff suppressed because it is too large Load Diff

@ -1,18 +0,0 @@
// Package blackfriday is a markdown processor.
//
// It translates plain text with simple formatting rules into an AST, which can
// then be further processed to HTML (provided by Blackfriday itself) or other
// formats (provided by the community).
//
// The simplest way to invoke Blackfriday is to call the Run function. It will
// take a text input and produce a text output in HTML (or other format).
//
// A slightly more sophisticated way to use Blackfriday is to create a Markdown
// processor and to call Parse, which returns a syntax tree for the input
// document. You can leverage Blackfriday's parsing for content extraction from
// markdown documents. You can assign a custom renderer and set various options
// to the Markdown processor.
//
// If you're interested in calling Blackfriday from command line, see
// https://github.com/russross/blackfriday-tool.
package blackfriday

@ -1,34 +0,0 @@
package blackfriday
import (
"html"
"io"
)
var htmlEscaper = [256][]byte{
'&': []byte("&amp;"),
'<': []byte("&lt;"),
'>': []byte("&gt;"),
'"': []byte("&quot;"),
}
func escapeHTML(w io.Writer, s []byte) {
var start, end int
for end < len(s) {
escSeq := htmlEscaper[s[end]]
if escSeq != nil {
w.Write(s[start:end])
w.Write(escSeq)
start = end + 1
}
end++
}
if start < len(s) && end <= len(s) {
w.Write(s[start:end])
}
}
func escLink(w io.Writer, text []byte) {
unesc := html.UnescapeString(string(text))
escapeHTML(w, []byte(unesc))
}

@ -1 +0,0 @@
module github.com/russross/blackfriday/v2

@ -1,949 +0,0 @@
//
// Blackfriday Markdown Processor
// Available at http://github.com/russross/blackfriday
//
// Copyright © 2011 Russ Ross <russ@russross.com>.
// Distributed under the Simplified BSD License.
// See README.md for details.
//
//
//
// HTML rendering backend
//
//
package blackfriday
import (
"bytes"
"fmt"
"io"
"regexp"
"strings"
)
// HTMLFlags control optional behavior of HTML renderer.
type HTMLFlags int
// HTML renderer configuration options.
const (
HTMLFlagsNone HTMLFlags = 0
SkipHTML HTMLFlags = 1 << iota // Skip preformatted HTML blocks
SkipImages // Skip embedded images
SkipLinks // Skip all links
Safelink // Only link to trusted protocols
NofollowLinks // Only link with rel="nofollow"
NoreferrerLinks // Only link with rel="noreferrer"
NoopenerLinks // Only link with rel="noopener"
HrefTargetBlank // Add a blank target
CompletePage // Generate a complete HTML page
UseXHTML // Generate XHTML output instead of HTML
FootnoteReturnLinks // Generate a link at the end of a footnote to return to the source
Smartypants // Enable smart punctuation substitutions
SmartypantsFractions // Enable smart fractions (with Smartypants)
SmartypantsDashes // Enable smart dashes (with Smartypants)
SmartypantsLatexDashes // Enable LaTeX-style dashes (with Smartypants)
SmartypantsAngledQuotes // Enable angled double quotes (with Smartypants) for double quotes rendering
SmartypantsQuotesNBSP // Enable « French guillemets » (with Smartypants)
TOC // Generate a table of contents
)
var (
htmlTagRe = regexp.MustCompile("(?i)^" + htmlTag)
)
const (
htmlTag = "(?:" + openTag + "|" + closeTag + "|" + htmlComment + "|" +
processingInstruction + "|" + declaration + "|" + cdata + ")"
closeTag = "</" + tagName + "\\s*[>]"
openTag = "<" + tagName + attribute + "*" + "\\s*/?>"
attribute = "(?:" + "\\s+" + attributeName + attributeValueSpec + "?)"
attributeValue = "(?:" + unquotedValue + "|" + singleQuotedValue + "|" + doubleQuotedValue + ")"
attributeValueSpec = "(?:" + "\\s*=" + "\\s*" + attributeValue + ")"
attributeName = "[a-zA-Z_:][a-zA-Z0-9:._-]*"
cdata = "<!\\[CDATA\\[[\\s\\S]*?\\]\\]>"
declaration = "<![A-Z]+" + "\\s+[^>]*>"
doubleQuotedValue = "\"[^\"]*\""
htmlComment = "<!---->|<!--(?:-?[^>-])(?:-?[^-])*-->"
processingInstruction = "[<][?].*?[?][>]"
singleQuotedValue = "'[^']*'"
tagName = "[A-Za-z][A-Za-z0-9-]*"
unquotedValue = "[^\"'=<>`\\x00-\\x20]+"
)
// HTMLRendererParameters is a collection of supplementary parameters tweaking
// the behavior of various parts of HTML renderer.
type HTMLRendererParameters struct {
// Prepend this text to each relative URL.
AbsolutePrefix string
// Add this text to each footnote anchor, to ensure uniqueness.
FootnoteAnchorPrefix string
// Show this text inside the <a> tag for a footnote return link, if the
// HTML_FOOTNOTE_RETURN_LINKS flag is enabled. If blank, the string
// <sup>[return]</sup> is used.
FootnoteReturnLinkContents string
// If set, add this text to the front of each Heading ID, to ensure
// uniqueness.
HeadingIDPrefix string
// If set, add this text to the back of each Heading ID, to ensure uniqueness.
HeadingIDSuffix string
// Increase heading levels: if the offset is 1, <h1> becomes <h2> etc.
// Negative offset is also valid.
// Resulting levels are clipped between 1 and 6.
HeadingLevelOffset int
Title string // Document title (used if CompletePage is set)
CSS string // Optional CSS file URL (used if CompletePage is set)
Icon string // Optional icon file URL (used if CompletePage is set)
Flags HTMLFlags // Flags allow customizing this renderer's behavior
}
// HTMLRenderer is a type that implements the Renderer interface for HTML output.
//
// Do not create this directly, instead use the NewHTMLRenderer function.
type HTMLRenderer struct {
HTMLRendererParameters
closeTag string // how to end singleton tags: either " />" or ">"
// Track heading IDs to prevent ID collision in a single generation.
headingIDs map[string]int
lastOutputLen int
disableTags int
sr *SPRenderer
}
const (
xhtmlClose = " />"
htmlClose = ">"
)
// NewHTMLRenderer creates and configures an HTMLRenderer object, which
// satisfies the Renderer interface.
func NewHTMLRenderer(params HTMLRendererParameters) *HTMLRenderer {
// configure the rendering engine
closeTag := htmlClose
if params.Flags&UseXHTML != 0 {
closeTag = xhtmlClose
}
if params.FootnoteReturnLinkContents == "" {
params.FootnoteReturnLinkContents = `<sup>[return]</sup>`
}
return &HTMLRenderer{
HTMLRendererParameters: params,
closeTag: closeTag,
headingIDs: make(map[string]int),
sr: NewSmartypantsRenderer(params.Flags),
}
}
func isHTMLTag(tag []byte, tagname string) bool {
found, _ := findHTMLTagPos(tag, tagname)
return found
}
// Look for a character, but ignore it when it's in any kind of quotes, it
// might be JavaScript
func skipUntilCharIgnoreQuotes(html []byte, start int, char byte) int {
inSingleQuote := false
inDoubleQuote := false
inGraveQuote := false
i := start
for i < len(html) {
switch {
case html[i] == char && !inSingleQuote && !inDoubleQuote && !inGraveQuote:
return i
case html[i] == '\'':
inSingleQuote = !inSingleQuote
case html[i] == '"':
inDoubleQuote = !inDoubleQuote
case html[i] == '`':
inGraveQuote = !inGraveQuote
}
i++
}
return start
}
func findHTMLTagPos(tag []byte, tagname string) (bool, int) {
i := 0
if i < len(tag) && tag[0] != '<' {
return false, -1
}
i++
i = skipSpace(tag, i)
if i < len(tag) && tag[i] == '/' {
i++
}
i = skipSpace(tag, i)
j := 0
for ; i < len(tag); i, j = i+1, j+1 {
if j >= len(tagname) {
break
}
if strings.ToLower(string(tag[i]))[0] != tagname[j] {
return false, -1
}
}
if i == len(tag) {
return false, -1
}
rightAngle := skipUntilCharIgnoreQuotes(tag, i, '>')
if rightAngle >= i {
return true, rightAngle
}
return false, -1
}
func skipSpace(tag []byte, i int) int {
for i < len(tag) && isspace(tag[i]) {
i++
}
return i
}
func isRelativeLink(link []byte) (yes bool) {
// a tag begin with '#'
if link[0] == '#' {
return true
}
// link begin with '/' but not '//', the second maybe a protocol relative link
if len(link) >= 2 && link[0] == '/' && link[1] != '/' {
return true
}
// only the root '/'
if len(link) == 1 && link[0] == '/' {
return true
}
// current directory : begin with "./"
if bytes.HasPrefix(link, []byte("./")) {
return true
}
// parent directory : begin with "../"
if bytes.HasPrefix(link, []byte("../")) {
return true
}
return false
}
func (r *HTMLRenderer) ensureUniqueHeadingID(id string) string {
for count, found := r.headingIDs[id]; found; count, found = r.headingIDs[id] {
tmp := fmt.Sprintf("%s-%d", id, count+1)
if _, tmpFound := r.headingIDs[tmp]; !tmpFound {
r.headingIDs[id] = count + 1
id = tmp
} else {
id = id + "-1"
}
}
if _, found := r.headingIDs[id]; !found {
r.headingIDs[id] = 0
}
return id
}
func (r *HTMLRenderer) addAbsPrefix(link []byte) []byte {
if r.AbsolutePrefix != "" && isRelativeLink(link) && link[0] != '.' {
newDest := r.AbsolutePrefix
if link[0] != '/' {
newDest += "/"
}
newDest += string(link)
return []byte(newDest)
}
return link
}
func appendLinkAttrs(attrs []string, flags HTMLFlags, link []byte) []string {
if isRelativeLink(link) {
return attrs
}
val := []string{}
if flags&NofollowLinks != 0 {
val = append(val, "nofollow")
}
if flags&NoreferrerLinks != 0 {
val = append(val, "noreferrer")
}
if flags&NoopenerLinks != 0 {
val = append(val, "noopener")
}
if flags&HrefTargetBlank != 0 {
attrs = append(attrs, "target=\"_blank\"")
}
if len(val) == 0 {
return attrs
}
attr := fmt.Sprintf("rel=%q", strings.Join(val, " "))
return append(attrs, attr)
}
func isMailto(link []byte) bool {
return bytes.HasPrefix(link, []byte("mailto:"))
}
func needSkipLink(flags HTMLFlags, dest []byte) bool {
if flags&SkipLinks != 0 {
return true
}
return flags&Safelink != 0 && !isSafeLink(dest) && !isMailto(dest)
}
func isSmartypantable(node *Node) bool {
pt := node.Parent.Type
return pt != Link && pt != CodeBlock && pt != Code
}
func appendLanguageAttr(attrs []string, info []byte) []string {
if len(info) == 0 {
return attrs
}
endOfLang := bytes.IndexAny(info, "\t ")
if endOfLang < 0 {
endOfLang = len(info)
}
return append(attrs, fmt.Sprintf("class=\"language-%s\"", info[:endOfLang]))
}
func (r *HTMLRenderer) tag(w io.Writer, name []byte, attrs []string) {
w.Write(name)
if len(attrs) > 0 {
w.Write(spaceBytes)
w.Write([]byte(strings.Join(attrs, " ")))
}
w.Write(gtBytes)
r.lastOutputLen = 1
}
func footnoteRef(prefix string, node *Node) []byte {
urlFrag := prefix + string(slugify(node.Destination))
anchor := fmt.Sprintf(`<a href="#fn:%s">%d</a>`, urlFrag, node.NoteID)
return []byte(fmt.Sprintf(`<sup class="footnote-ref" id="fnref:%s">%s</sup>`, urlFrag, anchor))
}
func footnoteItem(prefix string, slug []byte) []byte {
return []byte(fmt.Sprintf(`<li id="fn:%s%s">`, prefix, slug))
}
func footnoteReturnLink(prefix, returnLink string, slug []byte) []byte {
const format = ` <a class="footnote-return" href="#fnref:%s%s">%s</a>`
return []byte(fmt.Sprintf(format, prefix, slug, returnLink))
}
func itemOpenCR(node *Node) bool {
if node.Prev == nil {
return false
}
ld := node.Parent.ListData
return !ld.Tight && ld.ListFlags&ListTypeDefinition == 0
}
func skipParagraphTags(node *Node) bool {
grandparent := node.Parent.Parent
if grandparent == nil || grandparent.Type != List {
return false
}
tightOrTerm := grandparent.Tight || node.Parent.ListFlags&ListTypeTerm != 0
return grandparent.Type == List && tightOrTerm
}
func cellAlignment(align CellAlignFlags) string {
switch align {
case TableAlignmentLeft:
return "left"
case TableAlignmentRight:
return "right"
case TableAlignmentCenter:
return "center"
default:
return ""
}
}
func (r *HTMLRenderer) out(w io.Writer, text []byte) {
if r.disableTags > 0 {
w.Write(htmlTagRe.ReplaceAll(text, []byte{}))
} else {
w.Write(text)
}
r.lastOutputLen = len(text)
}
func (r *HTMLRenderer) cr(w io.Writer) {
if r.lastOutputLen > 0 {
r.out(w, nlBytes)
}
}
var (
nlBytes = []byte{'\n'}
gtBytes = []byte{'>'}
spaceBytes = []byte{' '}
)
var (
brTag = []byte("<br>")
brXHTMLTag = []byte("<br />")
emTag = []byte("<em>")
emCloseTag = []byte("</em>")
strongTag = []byte("<strong>")
strongCloseTag = []byte("</strong>")
delTag = []byte("<del>")
delCloseTag = []byte("</del>")
ttTag = []byte("<tt>")
ttCloseTag = []byte("</tt>")
aTag = []byte("<a")
aCloseTag = []byte("</a>")
preTag = []byte("<pre>")
preCloseTag = []byte("</pre>")
codeTag = []byte("<code>")
codeCloseTag = []byte("</code>")
pTag = []byte("<p>")
pCloseTag = []byte("</p>")
blockquoteTag = []byte("<blockquote>")
blockquoteCloseTag = []byte("</blockquote>")
hrTag = []byte("<hr>")
hrXHTMLTag = []byte("<hr />")
ulTag = []byte("<ul>")
ulCloseTag = []byte("</ul>")
olTag = []byte("<ol>")
olCloseTag = []byte("</ol>")
dlTag = []byte("<dl>")
dlCloseTag = []byte("</dl>")
liTag = []byte("<li>")
liCloseTag = []byte("</li>")
ddTag = []byte("<dd>")
ddCloseTag = []byte("</dd>")
dtTag = []byte("<dt>")
dtCloseTag = []byte("</dt>")
tableTag = []byte("<table>")
tableCloseTag = []byte("</table>")
tdTag = []byte("<td")
tdCloseTag = []byte("</td>")
thTag = []byte("<th")
thCloseTag = []byte("</th>")
theadTag = []byte("<thead>")
theadCloseTag = []byte("</thead>")
tbodyTag = []byte("<tbody>")
tbodyCloseTag = []byte("</tbody>")
trTag = []byte("<tr>")
trCloseTag = []byte("</tr>")
h1Tag = []byte("<h1")
h1CloseTag = []byte("</h1>")
h2Tag = []byte("<h2")
h2CloseTag = []byte("</h2>")
h3Tag = []byte("<h3")
h3CloseTag = []byte("</h3>")
h4Tag = []byte("<h4")
h4CloseTag = []byte("</h4>")
h5Tag = []byte("<h5")
h5CloseTag = []byte("</h5>")
h6Tag = []byte("<h6")
h6CloseTag = []byte("</h6>")
footnotesDivBytes = []byte("\n<div class=\"footnotes\">\n\n")
footnotesCloseDivBytes = []byte("\n</div>\n")
)
func headingTagsFromLevel(level int) ([]byte, []byte) {
if level <= 1 {
return h1Tag, h1CloseTag
}
switch level {
case 2:
return h2Tag, h2CloseTag
case 3:
return h3Tag, h3CloseTag
case 4:
return h4Tag, h4CloseTag
case 5:
return h5Tag, h5CloseTag
}
return h6Tag, h6CloseTag
}
func (r *HTMLRenderer) outHRTag(w io.Writer) {
if r.Flags&UseXHTML == 0 {
r.out(w, hrTag)
} else {
r.out(w, hrXHTMLTag)
}
}
// RenderNode is a default renderer of a single node of a syntax tree. For
// block nodes it will be called twice: first time with entering=true, second
// time with entering=false, so that it could know when it's working on an open
// tag and when on close. It writes the result to w.
//
// The return value is a way to tell the calling walker to adjust its walk
// pattern: e.g. it can terminate the traversal by returning Terminate. Or it
// can ask the walker to skip a subtree of this node by returning SkipChildren.
// The typical behavior is to return GoToNext, which asks for the usual
// traversal to the next node.
func (r *HTMLRenderer) RenderNode(w io.Writer, node *Node, entering bool) WalkStatus {
attrs := []string{}
switch node.Type {
case Text:
if r.Flags&Smartypants != 0 {
var tmp bytes.Buffer
escapeHTML(&tmp, node.Literal)
r.sr.Process(w, tmp.Bytes())
} else {
if node.Parent.Type == Link {
escLink(w, node.Literal)
} else {
escapeHTML(w, node.Literal)
}
}
case Softbreak:
r.cr(w)
// TODO: make it configurable via out(renderer.softbreak)
case Hardbreak:
if r.Flags&UseXHTML == 0 {
r.out(w, brTag)
} else {
r.out(w, brXHTMLTag)
}
r.cr(w)
case Emph:
if entering {
r.out(w, emTag)
} else {
r.out(w, emCloseTag)
}
case Strong:
if entering {
r.out(w, strongTag)
} else {
r.out(w, strongCloseTag)
}
case Del:
if entering {
r.out(w, delTag)
} else {
r.out(w, delCloseTag)
}
case HTMLSpan:
if r.Flags&SkipHTML != 0 {
break
}
r.out(w, node.Literal)
case Link:
// mark it but don't link it if it is not a safe link: no smartypants
dest := node.LinkData.Destination
if needSkipLink(r.Flags, dest) {
if entering {
r.out(w, ttTag)
} else {
r.out(w, ttCloseTag)
}
} else {
if entering {
dest = r.addAbsPrefix(dest)
var hrefBuf bytes.Buffer
hrefBuf.WriteString("href=\"")
escLink(&hrefBuf, dest)
hrefBuf.WriteByte('"')
attrs = append(attrs, hrefBuf.String())
if node.NoteID != 0 {
r.out(w, footnoteRef(r.FootnoteAnchorPrefix, node))
break
}
attrs = appendLinkAttrs(attrs, r.Flags, dest)
if len(node.LinkData.Title) > 0 {
var titleBuff bytes.Buffer
titleBuff.WriteString("title=\"")
escapeHTML(&titleBuff, node.LinkData.Title)
titleBuff.WriteByte('"')
attrs = append(attrs, titleBuff.String())
}
r.tag(w, aTag, attrs)
} else {
if node.NoteID != 0 {
break
}
r.out(w, aCloseTag)
}
}
case Image:
if r.Flags&SkipImages != 0 {
return SkipChildren
}
if entering {
dest := node.LinkData.Destination
dest = r.addAbsPrefix(dest)
if r.disableTags == 0 {
//if options.safe && potentiallyUnsafe(dest) {
//out(w, `<img src="" alt="`)
//} else {
r.out(w, []byte(`<img src="`))
escLink(w, dest)
r.out(w, []byte(`" alt="`))
//}
}
r.disableTags++
} else {
r.disableTags--
if r.disableTags == 0 {
if node.LinkData.Title != nil {
r.out(w, []byte(`" title="`))
escapeHTML(w, node.LinkData.Title)
}
r.out(w, []byte(`" />`))
}
}
case Code:
r.out(w, codeTag)
escapeHTML(w, node.Literal)
r.out(w, codeCloseTag)
case Document:
break
case Paragraph:
if skipParagraphTags(node) {
break
}
if entering {
// TODO: untangle this clusterfuck about when the newlines need
// to be added and when not.
if node.Prev != nil {
switch node.Prev.Type {
case HTMLBlock, List, Paragraph, Heading, CodeBlock, BlockQuote, HorizontalRule:
r.cr(w)
}
}
if node.Parent.Type == BlockQuote && node.Prev == nil {
r.cr(w)
}
r.out(w, pTag)
} else {
r.out(w, pCloseTag)
if !(node.Parent.Type == Item && node.Next == nil) {
r.cr(w)
}
}
case BlockQuote:
if entering {
r.cr(w)
r.out(w, blockquoteTag)
} else {
r.out(w, blockquoteCloseTag)
r.cr(w)
}
case HTMLBlock:
if r.Flags&SkipHTML != 0 {
break
}
r.cr(w)
r.out(w, node.Literal)
r.cr(w)
case Heading:
headingLevel := r.HTMLRendererParameters.HeadingLevelOffset + node.Level
openTag, closeTag := headingTagsFromLevel(headingLevel)
if entering {
if node.IsTitleblock {
attrs = append(attrs, `class="title"`)
}
if node.HeadingID != "" {
id := r.ensureUniqueHeadingID(node.HeadingID)
if r.HeadingIDPrefix != "" {
id = r.HeadingIDPrefix + id
}
if r.HeadingIDSuffix != "" {
id = id + r.HeadingIDSuffix
}
attrs = append(attrs, fmt.Sprintf(`id="%s"`, id))
}
r.cr(w)
r.tag(w, openTag, attrs)
} else {
r.out(w, closeTag)
if !(node.Parent.Type == Item && node.Next == nil) {
r.cr(w)
}
}
case HorizontalRule:
r.cr(w)
r.outHRTag(w)
r.cr(w)
case List:
openTag := ulTag
closeTag := ulCloseTag
if node.ListFlags&ListTypeOrdered != 0 {
openTag = olTag
closeTag = olCloseTag
}
if node.ListFlags&ListTypeDefinition != 0 {
openTag = dlTag
closeTag = dlCloseTag
}
if entering {
if node.IsFootnotesList {
r.out(w, footnotesDivBytes)
r.outHRTag(w)
r.cr(w)
}
r.cr(w)
if node.Parent.Type == Item && node.Parent.Parent.Tight {
r.cr(w)
}
r.tag(w, openTag[:len(openTag)-1], attrs)
r.cr(w)
} else {
r.out(w, closeTag)
//cr(w)
//if node.parent.Type != Item {
// cr(w)
//}
if node.Parent.Type == Item && node.Next != nil {
r.cr(w)
}
if node.Parent.Type == Document || node.Parent.Type == BlockQuote {
r.cr(w)
}
if node.IsFootnotesList {
r.out(w, footnotesCloseDivBytes)
}
}
case Item:
openTag := liTag
closeTag := liCloseTag
if node.ListFlags&ListTypeDefinition != 0 {
openTag = ddTag
closeTag = ddCloseTag
}
if node.ListFlags&ListTypeTerm != 0 {
openTag = dtTag
closeTag = dtCloseTag
}
if entering {
if itemOpenCR(node) {
r.cr(w)
}
if node.ListData.RefLink != nil {
slug := slugify(node.ListData.RefLink)
r.out(w, footnoteItem(r.FootnoteAnchorPrefix, slug))
break
}
r.out(w, openTag)
} else {
if node.ListData.RefLink != nil {
slug := slugify(node.ListData.RefLink)
if r.Flags&FootnoteReturnLinks != 0 {
r.out(w, footnoteReturnLink(r.FootnoteAnchorPrefix, r.FootnoteReturnLinkContents, slug))
}
}
r.out(w, closeTag)
r.cr(w)
}
case CodeBlock:
attrs = appendLanguageAttr(attrs, node.Info)
r.cr(w)
r.out(w, preTag)
r.tag(w, codeTag[:len(codeTag)-1], attrs)
escapeHTML(w, node.Literal)
r.out(w, codeCloseTag)
r.out(w, preCloseTag)
if node.Parent.Type != Item {
r.cr(w)
}
case Table:
if entering {
r.cr(w)
r.out(w, tableTag)
} else {
r.out(w, tableCloseTag)
r.cr(w)
}
case TableCell:
openTag := tdTag
closeTag := tdCloseTag
if node.IsHeader {
openTag = thTag
closeTag = thCloseTag
}
if entering {
align := cellAlignment(node.Align)
if align != "" {
attrs = append(attrs, fmt.Sprintf(`align="%s"`, align))
}
if node.Prev == nil {
r.cr(w)
}
r.tag(w, openTag, attrs)
} else {
r.out(w, closeTag)
r.cr(w)
}
case TableHead:
if entering {
r.cr(w)
r.out(w, theadTag)
} else {
r.out(w, theadCloseTag)
r.cr(w)
}
case TableBody:
if entering {
r.cr(w)
r.out(w, tbodyTag)
// XXX: this is to adhere to a rather silly test. Should fix test.
if node.FirstChild == nil {
r.cr(w)
}
} else {
r.out(w, tbodyCloseTag)
r.cr(w)
}
case TableRow:
if entering {
r.cr(w)
r.out(w, trTag)
} else {
r.out(w, trCloseTag)
r.cr(w)
}
default:
panic("Unknown node type " + node.Type.String())
}
return GoToNext
}
// RenderHeader writes HTML document preamble and TOC if requested.
func (r *HTMLRenderer) RenderHeader(w io.Writer, ast *Node) {
r.writeDocumentHeader(w)
if r.Flags&TOC != 0 {
r.writeTOC(w, ast)
}
}
// RenderFooter writes HTML document footer.
func (r *HTMLRenderer) RenderFooter(w io.Writer, ast *Node) {
if r.Flags&CompletePage == 0 {
return
}
io.WriteString(w, "\n</body>\n</html>\n")
}
func (r *HTMLRenderer) writeDocumentHeader(w io.Writer) {
if r.Flags&CompletePage == 0 {
return
}
ending := ""
if r.Flags&UseXHTML != 0 {
io.WriteString(w, "<!DOCTYPE html PUBLIC \"-//W3C//DTD XHTML 1.0 Transitional//EN\" ")
io.WriteString(w, "\"http://www.w3.org/TR/xhtml1/DTD/xhtml1-transitional.dtd\">\n")
io.WriteString(w, "<html xmlns=\"http://www.w3.org/1999/xhtml\">\n")
ending = " /"
} else {
io.WriteString(w, "<!DOCTYPE html>\n")
io.WriteString(w, "<html>\n")
}
io.WriteString(w, "<head>\n")
io.WriteString(w, " <title>")
if r.Flags&Smartypants != 0 {
r.sr.Process(w, []byte(r.Title))
} else {
escapeHTML(w, []byte(r.Title))
}
io.WriteString(w, "</title>\n")
io.WriteString(w, " <meta name=\"GENERATOR\" content=\"Blackfriday Markdown Processor v")
io.WriteString(w, Version)
io.WriteString(w, "\"")
io.WriteString(w, ending)
io.WriteString(w, ">\n")
io.WriteString(w, " <meta charset=\"utf-8\"")
io.WriteString(w, ending)
io.WriteString(w, ">\n")
if r.CSS != "" {
io.WriteString(w, " <link rel=\"stylesheet\" type=\"text/css\" href=\"")
escapeHTML(w, []byte(r.CSS))
io.WriteString(w, "\"")
io.WriteString(w, ending)
io.WriteString(w, ">\n")
}
if r.Icon != "" {
io.WriteString(w, " <link rel=\"icon\" type=\"image/x-icon\" href=\"")
escapeHTML(w, []byte(r.Icon))
io.WriteString(w, "\"")
io.WriteString(w, ending)
io.WriteString(w, ">\n")
}
io.WriteString(w, "</head>\n")
io.WriteString(w, "<body>\n\n")
}
func (r *HTMLRenderer) writeTOC(w io.Writer, ast *Node) {
buf := bytes.Buffer{}
inHeading := false
tocLevel := 0
headingCount := 0
ast.Walk(func(node *Node, entering bool) WalkStatus {
if node.Type == Heading && !node.HeadingData.IsTitleblock {
inHeading = entering
if entering {
node.HeadingID = fmt.Sprintf("toc_%d", headingCount)
if node.Level == tocLevel {
buf.WriteString("</li>\n\n<li>")
} else if node.Level < tocLevel {
for node.Level < tocLevel {
tocLevel--
buf.WriteString("</li>\n</ul>")
}
buf.WriteString("</li>\n\n<li>")
} else {
for node.Level > tocLevel {
tocLevel++
buf.WriteString("\n<ul>\n<li>")
}
}
fmt.Fprintf(&buf, `<a href="#toc_%d">`, headingCount)
headingCount++
} else {
buf.WriteString("</a>")
}
return GoToNext
}
if inHeading {
return r.RenderNode(&buf, node, entering)
}
return GoToNext
})
for ; tocLevel > 0; tocLevel-- {
buf.WriteString("</li>\n</ul>")
}
if buf.Len() > 0 {
io.WriteString(w, "<nav>\n")
w.Write(buf.Bytes())
io.WriteString(w, "\n\n</nav>\n")
}
r.lastOutputLen = buf.Len()
}

File diff suppressed because it is too large Load Diff

@ -1,950 +0,0 @@
// Blackfriday Markdown Processor
// Available at http://github.com/russross/blackfriday
//
// Copyright © 2011 Russ Ross <russ@russross.com>.
// Distributed under the Simplified BSD License.
// See README.md for details.
package blackfriday
import (
"bytes"
"fmt"
"io"
"strings"
"unicode/utf8"
)
//
// Markdown parsing and processing
//
// Version string of the package. Appears in the rendered document when
// CompletePage flag is on.
const Version = "2.0"
// Extensions is a bitwise or'ed collection of enabled Blackfriday's
// extensions.
type Extensions int
// These are the supported markdown parsing extensions.
// OR these values together to select multiple extensions.
const (
NoExtensions Extensions = 0
NoIntraEmphasis Extensions = 1 << iota // Ignore emphasis markers inside words
Tables // Render tables
FencedCode // Render fenced code blocks
Autolink // Detect embedded URLs that are not explicitly marked
Strikethrough // Strikethrough text using ~~test~~
LaxHTMLBlocks // Loosen up HTML block parsing rules
SpaceHeadings // Be strict about prefix heading rules
HardLineBreak // Translate newlines into line breaks
TabSizeEight // Expand tabs to eight spaces instead of four
Footnotes // Pandoc-style footnotes
NoEmptyLineBeforeBlock // No need to insert an empty line to start a (code, quote, ordered list, unordered list) block
HeadingIDs // specify heading IDs with {#id}
Titleblock // Titleblock ala pandoc
AutoHeadingIDs // Create the heading ID from the text
BackslashLineBreak // Translate trailing backslashes into line breaks
DefinitionLists // Render definition lists
CommonHTMLFlags HTMLFlags = UseXHTML | Smartypants |
SmartypantsFractions | SmartypantsDashes | SmartypantsLatexDashes
CommonExtensions Extensions = NoIntraEmphasis | Tables | FencedCode |
Autolink | Strikethrough | SpaceHeadings | HeadingIDs |
BackslashLineBreak | DefinitionLists
)
// ListType contains bitwise or'ed flags for list and list item objects.
type ListType int
// These are the possible flag values for the ListItem renderer.
// Multiple flag values may be ORed together.
// These are mostly of interest if you are writing a new output format.
const (
ListTypeOrdered ListType = 1 << iota
ListTypeDefinition
ListTypeTerm
ListItemContainsBlock
ListItemBeginningOfList // TODO: figure out if this is of any use now
ListItemEndOfList
)
// CellAlignFlags holds a type of alignment in a table cell.
type CellAlignFlags int
// These are the possible flag values for the table cell renderer.
// Only a single one of these values will be used; they are not ORed together.
// These are mostly of interest if you are writing a new output format.
const (
TableAlignmentLeft CellAlignFlags = 1 << iota
TableAlignmentRight
TableAlignmentCenter = (TableAlignmentLeft | TableAlignmentRight)
)
// The size of a tab stop.
const (
TabSizeDefault = 4
TabSizeDouble = 8
)
// blockTags is a set of tags that are recognized as HTML block tags.
// Any of these can be included in markdown text without special escaping.
var blockTags = map[string]struct{}{
"blockquote": {},
"del": {},
"div": {},
"dl": {},
"fieldset": {},
"form": {},
"h1": {},
"h2": {},
"h3": {},
"h4": {},
"h5": {},
"h6": {},
"iframe": {},
"ins": {},
"math": {},
"noscript": {},
"ol": {},
"pre": {},
"p": {},
"script": {},
"style": {},
"table": {},
"ul": {},
// HTML5
"address": {},
"article": {},
"aside": {},
"canvas": {},
"figcaption": {},
"figure": {},
"footer": {},
"header": {},
"hgroup": {},
"main": {},
"nav": {},
"output": {},
"progress": {},
"section": {},
"video": {},
}
// Renderer is the rendering interface. This is mostly of interest if you are
// implementing a new rendering format.
//
// Only an HTML implementation is provided in this repository, see the README
// for external implementations.
type Renderer interface {
// RenderNode is the main rendering method. It will be called once for
// every leaf node and twice for every non-leaf node (first with
// entering=true, then with entering=false). The method should write its
// rendition of the node to the supplied writer w.
RenderNode(w io.Writer, node *Node, entering bool) WalkStatus
// RenderHeader is a method that allows the renderer to produce some
// content preceding the main body of the output document. The header is
// understood in the broad sense here. For example, the default HTML
// renderer will write not only the HTML document preamble, but also the
// table of contents if it was requested.
//
// The method will be passed an entire document tree, in case a particular
// implementation needs to inspect it to produce output.
//
// The output should be written to the supplied writer w. If your
// implementation has no header to write, supply an empty implementation.
RenderHeader(w io.Writer, ast *Node)
// RenderFooter is a symmetric counterpart of RenderHeader.
RenderFooter(w io.Writer, ast *Node)
}
// Callback functions for inline parsing. One such function is defined
// for each character that triggers a response when parsing inline data.
type inlineParser func(p *Markdown, data []byte, offset int) (int, *Node)
// Markdown is a type that holds extensions and the runtime state used by
// Parse, and the renderer. You can not use it directly, construct it with New.
type Markdown struct {
renderer Renderer
referenceOverride ReferenceOverrideFunc
refs map[string]*reference
inlineCallback [256]inlineParser
extensions Extensions
nesting int
maxNesting int
insideLink bool
// Footnotes need to be ordered as well as available to quickly check for
// presence. If a ref is also a footnote, it's stored both in refs and here
// in notes. Slice is nil if footnotes not enabled.
notes []*reference
doc *Node
tip *Node // = doc
oldTip *Node
lastMatchedContainer *Node // = doc
allClosed bool
}
func (p *Markdown) getRef(refid string) (ref *reference, found bool) {
if p.referenceOverride != nil {
r, overridden := p.referenceOverride(refid)
if overridden {
if r == nil {
return nil, false
}
return &reference{
link: []byte(r.Link),
title: []byte(r.Title),
noteID: 0,
hasBlock: false,
text: []byte(r.Text)}, true
}
}
// refs are case insensitive
ref, found = p.refs[strings.ToLower(refid)]
return ref, found
}
func (p *Markdown) finalize(block *Node) {
above := block.Parent
block.open = false
p.tip = above
}
func (p *Markdown) addChild(node NodeType, offset uint32) *Node {
return p.addExistingChild(NewNode(node), offset)
}
func (p *Markdown) addExistingChild(node *Node, offset uint32) *Node {
for !p.tip.canContain(node.Type) {
p.finalize(p.tip)
}
p.tip.AppendChild(node)
p.tip = node
return node
}
func (p *Markdown) closeUnmatchedBlocks() {
if !p.allClosed {
for p.oldTip != p.lastMatchedContainer {
parent := p.oldTip.Parent
p.finalize(p.oldTip)
p.oldTip = parent
}
p.allClosed = true
}
}
//
//
// Public interface
//
//
// Reference represents the details of a link.
// See the documentation in Options for more details on use-case.
type Reference struct {
// Link is usually the URL the reference points to.
Link string
// Title is the alternate text describing the link in more detail.
Title string
// Text is the optional text to override the ref with if the syntax used was
// [refid][]
Text string
}
// ReferenceOverrideFunc is expected to be called with a reference string and
// return either a valid Reference type that the reference string maps to or
// nil. If overridden is false, the default reference logic will be executed.
// See the documentation in Options for more details on use-case.
type ReferenceOverrideFunc func(reference string) (ref *Reference, overridden bool)
// New constructs a Markdown processor. You can use the same With* functions as
// for Run() to customize parser's behavior and the renderer.
func New(opts ...Option) *Markdown {
var p Markdown
for _, opt := range opts {
opt(&p)
}
p.refs = make(map[string]*reference)
p.maxNesting = 16
p.insideLink = false
docNode := NewNode(Document)
p.doc = docNode
p.tip = docNode
p.oldTip = docNode
p.lastMatchedContainer = docNode
p.allClosed = true
// register inline parsers
p.inlineCallback[' '] = maybeLineBreak
p.inlineCallback['*'] = emphasis
p.inlineCallback['_'] = emphasis
if p.extensions&Strikethrough != 0 {
p.inlineCallback['~'] = emphasis
}
p.inlineCallback['`'] = codeSpan
p.inlineCallback['\n'] = lineBreak
p.inlineCallback['['] = link
p.inlineCallback['<'] = leftAngle
p.inlineCallback['\\'] = escape
p.inlineCallback['&'] = entity
p.inlineCallback['!'] = maybeImage
p.inlineCallback['^'] = maybeInlineFootnote
if p.extensions&Autolink != 0 {
p.inlineCallback['h'] = maybeAutoLink
p.inlineCallback['m'] = maybeAutoLink
p.inlineCallback['f'] = maybeAutoLink
p.inlineCallback['H'] = maybeAutoLink
p.inlineCallback['M'] = maybeAutoLink
p.inlineCallback['F'] = maybeAutoLink
}
if p.extensions&Footnotes != 0 {
p.notes = make([]*reference, 0)
}
return &p
}
// Option customizes the Markdown processor's default behavior.
type Option func(*Markdown)
// WithRenderer allows you to override the default renderer.
func WithRenderer(r Renderer) Option {
return func(p *Markdown) {
p.renderer = r
}
}
// WithExtensions allows you to pick some of the many extensions provided by
// Blackfriday. You can bitwise OR them.
func WithExtensions(e Extensions) Option {
return func(p *Markdown) {
p.extensions = e
}
}
// WithNoExtensions turns off all extensions and custom behavior.
func WithNoExtensions() Option {
return func(p *Markdown) {
p.extensions = NoExtensions
p.renderer = NewHTMLRenderer(HTMLRendererParameters{
Flags: HTMLFlagsNone,
})
}
}
// WithRefOverride sets an optional function callback that is called every
// time a reference is resolved.
//
// In Markdown, the link reference syntax can be made to resolve a link to
// a reference instead of an inline URL, in one of the following ways:
//
// * [link text][refid]
// * [refid][]
//
// Usually, the refid is defined at the bottom of the Markdown document. If
// this override function is provided, the refid is passed to the override
// function first, before consulting the defined refids at the bottom. If
// the override function indicates an override did not occur, the refids at
// the bottom will be used to fill in the link details.
func WithRefOverride(o ReferenceOverrideFunc) Option {
return func(p *Markdown) {
p.referenceOverride = o
}
}
// Run is the main entry point to Blackfriday. It parses and renders a
// block of markdown-encoded text.
//
// The simplest invocation of Run takes one argument, input:
// output := Run(input)
// This will parse the input with CommonExtensions enabled and render it with
// the default HTMLRenderer (with CommonHTMLFlags).
//
// Variadic arguments opts can customize the default behavior. Since Markdown
// type does not contain exported fields, you can not use it directly. Instead,
// use the With* functions. For example, this will call the most basic
// functionality, with no extensions:
// output := Run(input, WithNoExtensions())
//
// You can use any number of With* arguments, even contradicting ones. They
// will be applied in order of appearance and the latter will override the
// former:
// output := Run(input, WithNoExtensions(), WithExtensions(exts),
// WithRenderer(yourRenderer))
func Run(input []byte, opts ...Option) []byte {
r := NewHTMLRenderer(HTMLRendererParameters{
Flags: CommonHTMLFlags,
})
optList := []Option{WithRenderer(r), WithExtensions(CommonExtensions)}
optList = append(optList, opts...)
parser := New(optList...)
ast := parser.Parse(input)
var buf bytes.Buffer
parser.renderer.RenderHeader(&buf, ast)
ast.Walk(func(node *Node, entering bool) WalkStatus {
return parser.renderer.RenderNode(&buf, node, entering)
})
parser.renderer.RenderFooter(&buf, ast)
return buf.Bytes()
}
// Parse is an entry point to the parsing part of Blackfriday. It takes an
// input markdown document and produces a syntax tree for its contents. This
// tree can then be rendered with a default or custom renderer, or
// analyzed/transformed by the caller to whatever non-standard needs they have.
// The return value is the root node of the syntax tree.
func (p *Markdown) Parse(input []byte) *Node {
p.block(input)
// Walk the tree and finish up some of unfinished blocks
for p.tip != nil {
p.finalize(p.tip)
}
// Walk the tree again and process inline markdown in each block
p.doc.Walk(func(node *Node, entering bool) WalkStatus {
if node.Type == Paragraph || node.Type == Heading || node.Type == TableCell {
p.inline(node, node.content)
node.content = nil
}
return GoToNext
})
p.parseRefsToAST()
return p.doc
}
func (p *Markdown) parseRefsToAST() {
if p.extensions&Footnotes == 0 || len(p.notes) == 0 {
return
}
p.tip = p.doc
block := p.addBlock(List, nil)
block.IsFootnotesList = true
block.ListFlags = ListTypeOrdered
flags := ListItemBeginningOfList
// Note: this loop is intentionally explicit, not range-form. This is
// because the body of the loop will append nested footnotes to p.notes and
// we need to process those late additions. Range form would only walk over
// the fixed initial set.
for i := 0; i < len(p.notes); i++ {
ref := p.notes[i]
p.addExistingChild(ref.footnote, 0)
block := ref.footnote
block.ListFlags = flags | ListTypeOrdered
block.RefLink = ref.link
if ref.hasBlock {
flags |= ListItemContainsBlock
p.block(ref.title)
} else {
p.inline(block, ref.title)
}
flags &^= ListItemBeginningOfList | ListItemContainsBlock
}
above := block.Parent
finalizeList(block)
p.tip = above
block.Walk(func(node *Node, entering bool) WalkStatus {
if node.Type == Paragraph || node.Type == Heading {
p.inline(node, node.content)
node.content = nil
}
return GoToNext
})
}
//
// Link references
//
// This section implements support for references that (usually) appear
// as footnotes in a document, and can be referenced anywhere in the document.
// The basic format is:
//
// [1]: http://www.google.com/ "Google"
// [2]: http://www.github.com/ "Github"
//
// Anywhere in the document, the reference can be linked by referring to its
// label, i.e., 1 and 2 in this example, as in:
//
// This library is hosted on [Github][2], a git hosting site.
//
// Actual footnotes as specified in Pandoc and supported by some other Markdown
// libraries such as php-markdown are also taken care of. They look like this:
//
// This sentence needs a bit of further explanation.[^note]
//
// [^note]: This is the explanation.
//
// Footnotes should be placed at the end of the document in an ordered list.
// Finally, there are inline footnotes such as:
//
// Inline footnotes^[Also supported.] provide a quick inline explanation,
// but are rendered at the bottom of the document.
//
// reference holds all information necessary for a reference-style links or
// footnotes.
//
// Consider this markdown with reference-style links:
//
// [link][ref]
//
// [ref]: /url/ "tooltip title"
//
// It will be ultimately converted to this HTML:
//
// <p><a href=\"/url/\" title=\"title\">link</a></p>
//
// And a reference structure will be populated as follows:
//
// p.refs["ref"] = &reference{
// link: "/url/",
// title: "tooltip title",
// }
//
// Alternatively, reference can contain information about a footnote. Consider
// this markdown:
//
// Text needing a footnote.[^a]
//
// [^a]: This is the note
//
// A reference structure will be populated as follows:
//
// p.refs["a"] = &reference{
// link: "a",
// title: "This is the note",
// noteID: <some positive int>,
// }
//
// TODO: As you can see, it begs for splitting into two dedicated structures
// for refs and for footnotes.
type reference struct {
link []byte
title []byte
noteID int // 0 if not a footnote ref
hasBlock bool
footnote *Node // a link to the Item node within a list of footnotes
text []byte // only gets populated by refOverride feature with Reference.Text
}
func (r *reference) String() string {
return fmt.Sprintf("{link: %q, title: %q, text: %q, noteID: %d, hasBlock: %v}",
r.link, r.title, r.text, r.noteID, r.hasBlock)
}
// Check whether or not data starts with a reference link.
// If so, it is parsed and stored in the list of references
// (in the render struct).
// Returns the number of bytes to skip to move past it,
// or zero if the first line is not a reference.
func isReference(p *Markdown, data []byte, tabSize int) int {
// up to 3 optional leading spaces
if len(data) < 4 {
return 0
}
i := 0
for i < 3 && data[i] == ' ' {
i++
}
noteID := 0
// id part: anything but a newline between brackets
if data[i] != '[' {
return 0
}
i++
if p.extensions&Footnotes != 0 {
if i < len(data) && data[i] == '^' {
// we can set it to anything here because the proper noteIds will
// be assigned later during the second pass. It just has to be != 0
noteID = 1
i++
}
}
idOffset := i
for i < len(data) && data[i] != '\n' && data[i] != '\r' && data[i] != ']' {
i++
}
if i >= len(data) || data[i] != ']' {
return 0
}
idEnd := i
// footnotes can have empty ID, like this: [^], but a reference can not be
// empty like this: []. Break early if it's not a footnote and there's no ID
if noteID == 0 && idOffset == idEnd {
return 0
}
// spacer: colon (space | tab)* newline? (space | tab)*
i++
if i >= len(data) || data[i] != ':' {
return 0
}
i++
for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
i++
}
if i < len(data) && (data[i] == '\n' || data[i] == '\r') {
i++
if i < len(data) && data[i] == '\n' && data[i-1] == '\r' {
i++
}
}
for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
i++
}
if i >= len(data) {
return 0
}
var (
linkOffset, linkEnd int
titleOffset, titleEnd int
lineEnd int
raw []byte
hasBlock bool
)
if p.extensions&Footnotes != 0 && noteID != 0 {
linkOffset, linkEnd, raw, hasBlock = scanFootnote(p, data, i, tabSize)
lineEnd = linkEnd
} else {
linkOffset, linkEnd, titleOffset, titleEnd, lineEnd = scanLinkRef(p, data, i)
}
if lineEnd == 0 {
return 0
}
// a valid ref has been found
ref := &reference{
noteID: noteID,
hasBlock: hasBlock,
}
if noteID > 0 {
// reusing the link field for the id since footnotes don't have links
ref.link = data[idOffset:idEnd]
// if footnote, it's not really a title, it's the contained text
ref.title = raw
} else {
ref.link = data[linkOffset:linkEnd]
ref.title = data[titleOffset:titleEnd]
}
// id matches are case-insensitive
id := string(bytes.ToLower(data[idOffset:idEnd]))
p.refs[id] = ref
return lineEnd
}
func scanLinkRef(p *Markdown, data []byte, i int) (linkOffset, linkEnd, titleOffset, titleEnd, lineEnd int) {
// link: whitespace-free sequence, optionally between angle brackets
if data[i] == '<' {
i++
}
linkOffset = i
for i < len(data) && data[i] != ' ' && data[i] != '\t' && data[i] != '\n' && data[i] != '\r' {
i++
}
linkEnd = i
if data[linkOffset] == '<' && data[linkEnd-1] == '>' {
linkOffset++
linkEnd--
}
// optional spacer: (space | tab)* (newline | '\'' | '"' | '(' )
for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
i++
}
if i < len(data) && data[i] != '\n' && data[i] != '\r' && data[i] != '\'' && data[i] != '"' && data[i] != '(' {
return
}
// compute end-of-line
if i >= len(data) || data[i] == '\r' || data[i] == '\n' {
lineEnd = i
}
if i+1 < len(data) && data[i] == '\r' && data[i+1] == '\n' {
lineEnd++
}
// optional (space|tab)* spacer after a newline
if lineEnd > 0 {
i = lineEnd + 1
for i < len(data) && (data[i] == ' ' || data[i] == '\t') {
i++
}
}
// optional title: any non-newline sequence enclosed in '"() alone on its line
if i+1 < len(data) && (data[i] == '\'' || data[i] == '"' || data[i] == '(') {
i++
titleOffset = i
// look for EOL
for i < len(data) && data[i] != '\n' && data[i] != '\r' {
i++
}
if i+1 < len(data) && data[i] == '\n' && data[i+1] == '\r' {
titleEnd = i + 1
} else {
titleEnd = i
}
// step back
i--
for i > titleOffset && (data[i] == ' ' || data[i] == '\t') {
i--
}
if i > titleOffset && (data[i] == '\'' || data[i] == '"' || data[i] == ')') {
lineEnd = titleEnd
titleEnd = i
}
}
return
}
// The first bit of this logic is the same as Parser.listItem, but the rest
// is much simpler. This function simply finds the entire block and shifts it
// over by one tab if it is indeed a block (just returns the line if it's not).
// blockEnd is the end of the section in the input buffer, and contents is the
// extracted text that was shifted over one tab. It will need to be rendered at
// the end of the document.
func scanFootnote(p *Markdown, data []byte, i, indentSize int) (blockStart, blockEnd int, contents []byte, hasBlock bool) {
if i == 0 || len(data) == 0 {
return
}
// skip leading whitespace on first line
for i < len(data) && data[i] == ' ' {
i++
}
blockStart = i
// find the end of the line
blockEnd = i
for i < len(data) && data[i-1] != '\n' {
i++
}
// get working buffer
var raw bytes.Buffer
// put the first line into the working buffer
raw.Write(data[blockEnd:i])
blockEnd = i
// process the following lines
containsBlankLine := false
gatherLines:
for blockEnd < len(data) {
i++
// find the end of this line
for i < len(data) && data[i-1] != '\n' {
i++
}
// if it is an empty line, guess that it is part of this item
// and move on to the next line
if p.isEmpty(data[blockEnd:i]) > 0 {
containsBlankLine = true
blockEnd = i
continue
}
n := 0
if n = isIndented(data[blockEnd:i], indentSize); n == 0 {
// this is the end of the block.
// we don't want to include this last line in the index.
break gatherLines
}
// if there were blank lines before this one, insert a new one now
if containsBlankLine {
raw.WriteByte('\n')
containsBlankLine = false
}
// get rid of that first tab, write to buffer
raw.Write(data[blockEnd+n : i])
hasBlock = true
blockEnd = i
}
if data[blockEnd-1] != '\n' {
raw.WriteByte('\n')
}
contents = raw.Bytes()
return
}
//
//
// Miscellaneous helper functions
//
//
// Test if a character is a punctuation symbol.
// Taken from a private function in regexp in the stdlib.
func ispunct(c byte) bool {
for _, r := range []byte("!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~") {
if c == r {
return true
}
}
return false
}
// Test if a character is a whitespace character.
func isspace(c byte) bool {
return ishorizontalspace(c) || isverticalspace(c)
}
// Test if a character is a horizontal whitespace character.
func ishorizontalspace(c byte) bool {
return c == ' ' || c == '\t'
}
// Test if a character is a vertical character.
func isverticalspace(c byte) bool {
return c == '\n' || c == '\r' || c == '\f' || c == '\v'
}
// Test if a character is letter.
func isletter(c byte) bool {
return (c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z')
}
// Test if a character is a letter or a digit.
// TODO: check when this is looking for ASCII alnum and when it should use unicode
func isalnum(c byte) bool {
return (c >= '0' && c <= '9') || isletter(c)
}
// Replace tab characters with spaces, aligning to the next TAB_SIZE column.
// always ends output with a newline
func expandTabs(out *bytes.Buffer, line []byte, tabSize int) {
// first, check for common cases: no tabs, or only tabs at beginning of line
i, prefix := 0, 0
slowcase := false
for i = 0; i < len(line); i++ {
if line[i] == '\t' {
if prefix == i {
prefix++
} else {
slowcase = true
break
}
}
}
// no need to decode runes if all tabs are at the beginning of the line
if !slowcase {
for i = 0; i < prefix*tabSize; i++ {
out.WriteByte(' ')
}
out.Write(line[prefix:])
return
}
// the slow case: we need to count runes to figure out how
// many spaces to insert for each tab
column := 0
i = 0
for i < len(line) {
start := i
for i < len(line) && line[i] != '\t' {
_, size := utf8.DecodeRune(line[i:])
i += size
column++
}
if i > start {
out.Write(line[start:i])
}
if i >= len(line) {
break
}
for {
out.WriteByte(' ')
column++
if column%tabSize == 0 {
break
}
}
i++
}
}
// Find if a line counts as indented or not.
// Returns number of characters the indent is (0 = not indented).
func isIndented(data []byte, indentSize int) int {
if len(data) == 0 {
return 0
}
if data[0] == '\t' {
return 1
}
if len(data) < indentSize {
return 0
}
for i := 0; i < indentSize; i++ {
if data[i] != ' ' {
return 0
}
}
return indentSize
}
// Create a url-safe slug for fragments
func slugify(in []byte) []byte {
if len(in) == 0 {
return in
}
out := make([]byte, 0, len(in))
sym := false
for _, ch := range in {
if isalnum(ch) {
sym = false
out = append(out, ch)
} else if sym {
continue
} else {
out = append(out, '-')
sym = true
}
}
var a, b int
var ch byte
for a, ch = range out {
if ch != '-' {
break
}
}
for b = len(out) - 1; b > 0; b-- {
if out[b] != '-' {
break
}
}
return out[a : b+1]
}

@ -1,354 +0,0 @@
package blackfriday
import (
"bytes"
"fmt"
)
// NodeType specifies a type of a single node of a syntax tree. Usually one
// node (and its type) corresponds to a single markdown feature, e.g. emphasis
// or code block.
type NodeType int
// Constants for identifying different types of nodes. See NodeType.
const (
Document NodeType = iota
BlockQuote
List
Item
Paragraph
Heading
HorizontalRule
Emph
Strong
Del
Link
Image
Text
HTMLBlock
CodeBlock
Softbreak
Hardbreak
Code
HTMLSpan
Table
TableCell
TableHead
TableBody
TableRow
)
var nodeTypeNames = []string{
Document: "Document",
BlockQuote: "BlockQuote",
List: "List",
Item: "Item",
Paragraph: "Paragraph",
Heading: "Heading",
HorizontalRule: "HorizontalRule",
Emph: "Emph",
Strong: "Strong",
Del: "Del",
Link: "Link",
Image: "Image",
Text: "Text",
HTMLBlock: "HTMLBlock",
CodeBlock: "CodeBlock",
Softbreak: "Softbreak",
Hardbreak: "Hardbreak",
Code: "Code",
HTMLSpan: "HTMLSpan",
Table: "Table",
TableCell: "TableCell",
TableHead: "TableHead",
TableBody: "TableBody",
TableRow: "TableRow",
}
func (t NodeType) String() string {
return nodeTypeNames[t]
}
// ListData contains fields relevant to a List and Item node type.
type ListData struct {
ListFlags ListType
Tight bool // Skip <p>s around list item data if true
BulletChar byte // '*', '+' or '-' in bullet lists
Delimiter byte // '.' or ')' after the number in ordered lists
RefLink []byte // If not nil, turns this list item into a footnote item and triggers different rendering
IsFootnotesList bool // This is a list of footnotes
}
// LinkData contains fields relevant to a Link node type.
type LinkData struct {
Destination []byte // Destination is what goes into a href
Title []byte // Title is the tooltip thing that goes in a title attribute
NoteID int // NoteID contains a serial number of a footnote, zero if it's not a footnote
Footnote *Node // If it's a footnote, this is a direct link to the footnote Node. Otherwise nil.
}
// CodeBlockData contains fields relevant to a CodeBlock node type.
type CodeBlockData struct {
IsFenced bool // Specifies whether it's a fenced code block or an indented one
Info []byte // This holds the info string
FenceChar byte
FenceLength int
FenceOffset int
}
// TableCellData contains fields relevant to a TableCell node type.
type TableCellData struct {
IsHeader bool // This tells if it's under the header row
Align CellAlignFlags // This holds the value for align attribute
}
// HeadingData contains fields relevant to a Heading node type.
type HeadingData struct {
Level int // This holds the heading level number
HeadingID string // This might hold heading ID, if present
IsTitleblock bool // Specifies whether it's a title block
}
// Node is a single element in the abstract syntax tree of the parsed document.
// It holds connections to the structurally neighboring nodes and, for certain
// types of nodes, additional information that might be needed when rendering.
type Node struct {
Type NodeType // Determines the type of the node
Parent *Node // Points to the parent
FirstChild *Node // Points to the first child, if any
LastChild *Node // Points to the last child, if any
Prev *Node // Previous sibling; nil if it's the first child
Next *Node // Next sibling; nil if it's the last child
Literal []byte // Text contents of the leaf nodes
HeadingData // Populated if Type is Heading
ListData // Populated if Type is List
CodeBlockData // Populated if Type is CodeBlock
LinkData // Populated if Type is Link
TableCellData // Populated if Type is TableCell
content []byte // Markdown content of the block nodes
open bool // Specifies an open block node that has not been finished to process yet
}
// NewNode allocates a node of a specified type.
func NewNode(typ NodeType) *Node {
return &Node{
Type: typ,
open: true,
}
}
func (n *Node) String() string {
ellipsis := ""
snippet := n.Literal
if len(snippet) > 16 {
snippet = snippet[:16]
ellipsis = "..."
}
return fmt.Sprintf("%s: '%s%s'", n.Type, snippet, ellipsis)
}
// Unlink removes node 'n' from the tree.
// It panics if the node is nil.
func (n *Node) Unlink() {
if n.Prev != nil {
n.Prev.Next = n.Next
} else if n.Parent != nil {
n.Parent.FirstChild = n.Next
}
if n.Next != nil {
n.Next.Prev = n.Prev
} else if n.Parent != nil {
n.Parent.LastChild = n.Prev
}
n.Parent = nil
n.Next = nil
n.Prev = nil
}
// AppendChild adds a node 'child' as a child of 'n'.
// It panics if either node is nil.
func (n *Node) AppendChild(child *Node) {
child.Unlink()
child.Parent = n
if n.LastChild != nil {
n.LastChild.Next = child
child.Prev = n.LastChild
n.LastChild = child
} else {
n.FirstChild = child
n.LastChild = child
}
}
// InsertBefore inserts 'sibling' immediately before 'n'.
// It panics if either node is nil.
func (n *Node) InsertBefore(sibling *Node) {
sibling.Unlink()
sibling.Prev = n.Prev
if sibling.Prev != nil {
sibling.Prev.Next = sibling
}
sibling.Next = n
n.Prev = sibling
sibling.Parent = n.Parent
if sibling.Prev == nil {
sibling.Parent.FirstChild = sibling
}
}
func (n *Node) isContainer() bool {
switch n.Type {
case Document:
fallthrough
case BlockQuote:
fallthrough
case List:
fallthrough
case Item:
fallthrough
case Paragraph:
fallthrough
case Heading:
fallthrough
case Emph:
fallthrough
case Strong:
fallthrough
case Del:
fallthrough
case Link:
fallthrough
case Image:
fallthrough
case Table:
fallthrough
case TableHead:
fallthrough
case TableBody:
fallthrough
case TableRow:
fallthrough
case TableCell:
return true
default:
return false
}
}
func (n *Node) canContain(t NodeType) bool {
if n.Type == List {
return t == Item
}
if n.Type == Document || n.Type == BlockQuote || n.Type == Item {
return t != Item
}
if n.Type == Table {
return t == TableHead || t == TableBody
}
if n.Type == TableHead || n.Type == TableBody {
return t == TableRow
}
if n.Type == TableRow {
return t == TableCell
}
return false
}
// WalkStatus allows NodeVisitor to have some control over the tree traversal.
// It is returned from NodeVisitor and different values allow Node.Walk to
// decide which node to go to next.
type WalkStatus int
const (
// GoToNext is the default traversal of every node.
GoToNext WalkStatus = iota
// SkipChildren tells walker to skip all children of current node.
SkipChildren
// Terminate tells walker to terminate the traversal.
Terminate
)
// NodeVisitor is a callback to be called when traversing the syntax tree.
// Called twice for every node: once with entering=true when the branch is
// first visited, then with entering=false after all the children are done.
type NodeVisitor func(node *Node, entering bool) WalkStatus
// Walk is a convenience method that instantiates a walker and starts a
// traversal of subtree rooted at n.
func (n *Node) Walk(visitor NodeVisitor) {
w := newNodeWalker(n)
for w.current != nil {
status := visitor(w.current, w.entering)
switch status {
case GoToNext:
w.next()
case SkipChildren:
w.entering = false
w.next()
case Terminate:
return
}
}
}
type nodeWalker struct {
current *Node
root *Node
entering bool
}
func newNodeWalker(root *Node) *nodeWalker {
return &nodeWalker{
current: root,
root: root,
entering: true,
}
}
func (nw *nodeWalker) next() {
if (!nw.current.isContainer() || !nw.entering) && nw.current == nw.root {
nw.current = nil
return
}
if nw.entering && nw.current.isContainer() {
if nw.current.FirstChild != nil {
nw.current = nw.current.FirstChild
nw.entering = true
} else {
nw.entering = false
}
} else if nw.current.Next == nil {
nw.current = nw.current.Parent
nw.entering = false
} else {
nw.current = nw.current.Next
nw.entering = true
}
}
func dump(ast *Node) {
fmt.Println(dumpString(ast))
}
func dumpR(ast *Node, depth int) string {
if ast == nil {
return ""
}
indent := bytes.Repeat([]byte("\t"), depth)
content := ast.Literal
if content == nil {
content = ast.content
}
result := fmt.Sprintf("%s%s(%q)\n", indent, ast.Type, content)
for n := ast.FirstChild; n != nil; n = n.Next {
result += dumpR(n, depth+1)
}
return result
}
func dumpString(ast *Node) string {
return dumpR(ast, 0)
}

@ -1,457 +0,0 @@
//
// Blackfriday Markdown Processor
// Available at http://github.com/russross/blackfriday
//
// Copyright © 2011 Russ Ross <russ@russross.com>.
// Distributed under the Simplified BSD License.
// See README.md for details.
//
//
//
// SmartyPants rendering
//
//
package blackfriday
import (
"bytes"
"io"
)
// SPRenderer is a struct containing state of a Smartypants renderer.
type SPRenderer struct {
inSingleQuote bool
inDoubleQuote bool
callbacks [256]smartCallback
}
func wordBoundary(c byte) bool {
return c == 0 || isspace(c) || ispunct(c)
}
func tolower(c byte) byte {
if c >= 'A' && c <= 'Z' {
return c - 'A' + 'a'
}
return c
}
func isdigit(c byte) bool {
return c >= '0' && c <= '9'
}
func smartQuoteHelper(out *bytes.Buffer, previousChar byte, nextChar byte, quote byte, isOpen *bool, addNBSP bool) bool {
// edge of the buffer is likely to be a tag that we don't get to see,
// so we treat it like text sometimes
// enumerate all sixteen possibilities for (previousChar, nextChar)
// each can be one of {0, space, punct, other}
switch {
case previousChar == 0 && nextChar == 0:
// context is not any help here, so toggle
*isOpen = !*isOpen
case isspace(previousChar) && nextChar == 0:
// [ "] might be [ "<code>foo...]
*isOpen = true
case ispunct(previousChar) && nextChar == 0:
// [!"] hmm... could be [Run!"] or [("<code>...]
*isOpen = false
case /* isnormal(previousChar) && */ nextChar == 0:
// [a"] is probably a close
*isOpen = false
case previousChar == 0 && isspace(nextChar):
// [" ] might be [...foo</code>" ]
*isOpen = false
case isspace(previousChar) && isspace(nextChar):
// [ " ] context is not any help here, so toggle
*isOpen = !*isOpen
case ispunct(previousChar) && isspace(nextChar):
// [!" ] is probably a close
*isOpen = false
case /* isnormal(previousChar) && */ isspace(nextChar):
// [a" ] this is one of the easy cases
*isOpen = false
case previousChar == 0 && ispunct(nextChar):
// ["!] hmm... could be ["$1.95] or [</code>"!...]
*isOpen = false
case isspace(previousChar) && ispunct(nextChar):
// [ "!] looks more like [ "$1.95]
*isOpen = true
case ispunct(previousChar) && ispunct(nextChar):
// [!"!] context is not any help here, so toggle
*isOpen = !*isOpen
case /* isnormal(previousChar) && */ ispunct(nextChar):
// [a"!] is probably a close
*isOpen = false
case previousChar == 0 /* && isnormal(nextChar) */ :
// ["a] is probably an open
*isOpen = true
case isspace(previousChar) /* && isnormal(nextChar) */ :
// [ "a] this is one of the easy cases
*isOpen = true
case ispunct(previousChar) /* && isnormal(nextChar) */ :
// [!"a] is probably an open
*isOpen = true
default:
// [a'b] maybe a contraction?
*isOpen = false
}
// Note that with the limited lookahead, this non-breaking
// space will also be appended to single double quotes.
if addNBSP && !*isOpen {
out.WriteString("&nbsp;")
}
out.WriteByte('&')
if *isOpen {
out.WriteByte('l')
} else {
out.WriteByte('r')
}
out.WriteByte(quote)
out.WriteString("quo;")
if addNBSP && *isOpen {
out.WriteString("&nbsp;")
}
return true
}
func (r *SPRenderer) smartSingleQuote(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 2 {
t1 := tolower(text[1])
if t1 == '\'' {
nextChar := byte(0)
if len(text) >= 3 {
nextChar = text[2]
}
if smartQuoteHelper(out, previousChar, nextChar, 'd', &r.inDoubleQuote, false) {
return 1
}
}
if (t1 == 's' || t1 == 't' || t1 == 'm' || t1 == 'd') && (len(text) < 3 || wordBoundary(text[2])) {
out.WriteString("&rsquo;")
return 0
}
if len(text) >= 3 {
t2 := tolower(text[2])
if ((t1 == 'r' && t2 == 'e') || (t1 == 'l' && t2 == 'l') || (t1 == 'v' && t2 == 'e')) &&
(len(text) < 4 || wordBoundary(text[3])) {
out.WriteString("&rsquo;")
return 0
}
}
}
nextChar := byte(0)
if len(text) > 1 {
nextChar = text[1]
}
if smartQuoteHelper(out, previousChar, nextChar, 's', &r.inSingleQuote, false) {
return 0
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartParens(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 3 {
t1 := tolower(text[1])
t2 := tolower(text[2])
if t1 == 'c' && t2 == ')' {
out.WriteString("&copy;")
return 2
}
if t1 == 'r' && t2 == ')' {
out.WriteString("&reg;")
return 2
}
if len(text) >= 4 && t1 == 't' && t2 == 'm' && text[3] == ')' {
out.WriteString("&trade;")
return 3
}
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartDash(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 2 {
if text[1] == '-' {
out.WriteString("&mdash;")
return 1
}
if wordBoundary(previousChar) && wordBoundary(text[1]) {
out.WriteString("&ndash;")
return 0
}
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartDashLatex(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 3 && text[1] == '-' && text[2] == '-' {
out.WriteString("&mdash;")
return 2
}
if len(text) >= 2 && text[1] == '-' {
out.WriteString("&ndash;")
return 1
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartAmpVariant(out *bytes.Buffer, previousChar byte, text []byte, quote byte, addNBSP bool) int {
if bytes.HasPrefix(text, []byte("&quot;")) {
nextChar := byte(0)
if len(text) >= 7 {
nextChar = text[6]
}
if smartQuoteHelper(out, previousChar, nextChar, quote, &r.inDoubleQuote, addNBSP) {
return 5
}
}
if bytes.HasPrefix(text, []byte("&#0;")) {
return 3
}
out.WriteByte('&')
return 0
}
func (r *SPRenderer) smartAmp(angledQuotes, addNBSP bool) func(*bytes.Buffer, byte, []byte) int {
var quote byte = 'd'
if angledQuotes {
quote = 'a'
}
return func(out *bytes.Buffer, previousChar byte, text []byte) int {
return r.smartAmpVariant(out, previousChar, text, quote, addNBSP)
}
}
func (r *SPRenderer) smartPeriod(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 3 && text[1] == '.' && text[2] == '.' {
out.WriteString("&hellip;")
return 2
}
if len(text) >= 5 && text[1] == ' ' && text[2] == '.' && text[3] == ' ' && text[4] == '.' {
out.WriteString("&hellip;")
return 4
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartBacktick(out *bytes.Buffer, previousChar byte, text []byte) int {
if len(text) >= 2 && text[1] == '`' {
nextChar := byte(0)
if len(text) >= 3 {
nextChar = text[2]
}
if smartQuoteHelper(out, previousChar, nextChar, 'd', &r.inDoubleQuote, false) {
return 1
}
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartNumberGeneric(out *bytes.Buffer, previousChar byte, text []byte) int {
if wordBoundary(previousChar) && previousChar != '/' && len(text) >= 3 {
// is it of the form digits/digits(word boundary)?, i.e., \d+/\d+\b
// note: check for regular slash (/) or fraction slash (, 0x2044, or 0xe2 81 84 in utf-8)
// and avoid changing dates like 1/23/2005 into fractions.
numEnd := 0
for len(text) > numEnd && isdigit(text[numEnd]) {
numEnd++
}
if numEnd == 0 {
out.WriteByte(text[0])
return 0
}
denStart := numEnd + 1
if len(text) > numEnd+3 && text[numEnd] == 0xe2 && text[numEnd+1] == 0x81 && text[numEnd+2] == 0x84 {
denStart = numEnd + 3
} else if len(text) < numEnd+2 || text[numEnd] != '/' {
out.WriteByte(text[0])
return 0
}
denEnd := denStart
for len(text) > denEnd && isdigit(text[denEnd]) {
denEnd++
}
if denEnd == denStart {
out.WriteByte(text[0])
return 0
}
if len(text) == denEnd || wordBoundary(text[denEnd]) && text[denEnd] != '/' {
out.WriteString("<sup>")
out.Write(text[:numEnd])
out.WriteString("</sup>&frasl;<sub>")
out.Write(text[denStart:denEnd])
out.WriteString("</sub>")
return denEnd - 1
}
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartNumber(out *bytes.Buffer, previousChar byte, text []byte) int {
if wordBoundary(previousChar) && previousChar != '/' && len(text) >= 3 {
if text[0] == '1' && text[1] == '/' && text[2] == '2' {
if len(text) < 4 || wordBoundary(text[3]) && text[3] != '/' {
out.WriteString("&frac12;")
return 2
}
}
if text[0] == '1' && text[1] == '/' && text[2] == '4' {
if len(text) < 4 || wordBoundary(text[3]) && text[3] != '/' || (len(text) >= 5 && tolower(text[3]) == 't' && tolower(text[4]) == 'h') {
out.WriteString("&frac14;")
return 2
}
}
if text[0] == '3' && text[1] == '/' && text[2] == '4' {
if len(text) < 4 || wordBoundary(text[3]) && text[3] != '/' || (len(text) >= 6 && tolower(text[3]) == 't' && tolower(text[4]) == 'h' && tolower(text[5]) == 's') {
out.WriteString("&frac34;")
return 2
}
}
}
out.WriteByte(text[0])
return 0
}
func (r *SPRenderer) smartDoubleQuoteVariant(out *bytes.Buffer, previousChar byte, text []byte, quote byte) int {
nextChar := byte(0)
if len(text) > 1 {
nextChar = text[1]
}
if !smartQuoteHelper(out, previousChar, nextChar, quote, &r.inDoubleQuote, false) {
out.WriteString("&quot;")
}
return 0
}
func (r *SPRenderer) smartDoubleQuote(out *bytes.Buffer, previousChar byte, text []byte) int {
return r.smartDoubleQuoteVariant(out, previousChar, text, 'd')
}
func (r *SPRenderer) smartAngledDoubleQuote(out *bytes.Buffer, previousChar byte, text []byte) int {
return r.smartDoubleQuoteVariant(out, previousChar, text, 'a')
}
func (r *SPRenderer) smartLeftAngle(out *bytes.Buffer, previousChar byte, text []byte) int {
i := 0
for i < len(text) && text[i] != '>' {
i++
}
out.Write(text[:i+1])
return i
}
type smartCallback func(out *bytes.Buffer, previousChar byte, text []byte) int
// NewSmartypantsRenderer constructs a Smartypants renderer object.
func NewSmartypantsRenderer(flags HTMLFlags) *SPRenderer {
var (
r SPRenderer
smartAmpAngled = r.smartAmp(true, false)
smartAmpAngledNBSP = r.smartAmp(true, true)
smartAmpRegular = r.smartAmp(false, false)
smartAmpRegularNBSP = r.smartAmp(false, true)
addNBSP = flags&SmartypantsQuotesNBSP != 0
)
if flags&SmartypantsAngledQuotes == 0 {
r.callbacks['"'] = r.smartDoubleQuote
if !addNBSP {
r.callbacks['&'] = smartAmpRegular
} else {
r.callbacks['&'] = smartAmpRegularNBSP
}
} else {
r.callbacks['"'] = r.smartAngledDoubleQuote
if !addNBSP {
r.callbacks['&'] = smartAmpAngled
} else {
r.callbacks['&'] = smartAmpAngledNBSP
}
}
r.callbacks['\''] = r.smartSingleQuote
r.callbacks['('] = r.smartParens
if flags&SmartypantsDashes != 0 {
if flags&SmartypantsLatexDashes == 0 {
r.callbacks['-'] = r.smartDash
} else {
r.callbacks['-'] = r.smartDashLatex
}
}
r.callbacks['.'] = r.smartPeriod
if flags&SmartypantsFractions == 0 {
r.callbacks['1'] = r.smartNumber
r.callbacks['3'] = r.smartNumber
} else {
for ch := '1'; ch <= '9'; ch++ {
r.callbacks[ch] = r.smartNumberGeneric
}
}
r.callbacks['<'] = r.smartLeftAngle
r.callbacks['`'] = r.smartBacktick
return &r
}
// Process is the entry point of the Smartypants renderer.
func (r *SPRenderer) Process(w io.Writer, text []byte) {
mark := 0
for i := 0; i < len(text); i++ {
if action := r.callbacks[text[i]]; action != nil {
if i > mark {
w.Write(text[mark:i])
}
previousChar := byte(0)
if i > 0 {
previousChar = text[i-1]
}
var tmp bytes.Buffer
i += action(&tmp, previousChar, text[i:])
w.Write(tmp.Bytes())
mark = i + 1
}
}
if mark < len(text) {
w.Write(text[mark:])
}
}

@ -1,16 +0,0 @@
sudo: false
language: go
go:
- 1.x
- master
matrix:
allow_failures:
- go: master
fast_finish: true
install:
- # Do nothing. This is needed to prevent default install action "go get -t -v ./..." from happening here (we want it to happen inside script step).
script:
- go get -t -v ./...
- diff -u <(echo -n) <(gofmt -d -s .)
- go tool vet .
- go test -v -race ./...

@ -1,36 +0,0 @@
sanitized_anchor_name
=====================
[![Build Status](https://travis-ci.org/shurcooL/sanitized_anchor_name.svg?branch=master)](https://travis-ci.org/shurcooL/sanitized_anchor_name) [![GoDoc](https://godoc.org/github.com/shurcooL/sanitized_anchor_name?status.svg)](https://godoc.org/github.com/shurcooL/sanitized_anchor_name)
Package sanitized_anchor_name provides a func to create sanitized anchor names.
Its logic can be reused by multiple packages to create interoperable anchor names
and links to those anchors.
At this time, it does not try to ensure that generated anchor names
are unique, that responsibility falls on the caller.
Installation
------------
```bash
go get -u github.com/shurcooL/sanitized_anchor_name
```
Example
-------
```Go
anchorName := sanitized_anchor_name.Create("This is a header")
fmt.Println(anchorName)
// Output:
// this-is-a-header
```
License
-------
- [MIT License](LICENSE)

@ -1 +0,0 @@
module github.com/shurcooL/sanitized_anchor_name

@ -1,29 +0,0 @@
// Package sanitized_anchor_name provides a func to create sanitized anchor names.
//
// Its logic can be reused by multiple packages to create interoperable anchor names
// and links to those anchors.
//
// At this time, it does not try to ensure that generated anchor names
// are unique, that responsibility falls on the caller.
package sanitized_anchor_name // import "github.com/shurcooL/sanitized_anchor_name"
import "unicode"
// Create returns a sanitized anchor name for the given text.
func Create(text string) string {
var anchorName []rune
var futureDash = false
for _, r := range text {
switch {
case unicode.IsLetter(r) || unicode.IsNumber(r):
if futureDash && len(anchorName) > 0 {
anchorName = append(anchorName, '-')
}
futureDash = false
anchorName = append(anchorName, unicode.ToLower(r))
default:
futureDash = true
}
}
return string(anchorName)
}

@ -0,0 +1,19 @@
# Binaries for programs and plugins
*.exe
*.exe~
*.dll
*.so
*.dylib
# Test binary, build with `go test -c`
*.test
*.pprof
# Output of the go coverage tool, specifically when used with LiteIDE
*.out
.DS_Store
fuzz/corpus
fuzz/crashers
fuzz/suppressions
fuzz/fuzz-fuzz.zip

@ -1,6 +1,6 @@
MIT License
Copyright (c) 2015 Dmitri Shuralyov
Copyright (c) 2019 Yusuke Inuzuka
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal

@ -0,0 +1,16 @@
.PHONY: test fuzz
test:
go test -coverprofile=profile.out -coverpkg=github.com/yuin/goldmark,github.com/yuin/goldmark/ast,github.com/yuin/goldmark/extension,github.com/yuin/goldmark/extension/ast,github.com/yuin/goldmark/parser,github.com/yuin/goldmark/renderer,github.com/yuin/goldmark/renderer/html,github.com/yuin/goldmark/text,github.com/yuin/goldmark/util ./...
cov: test
go tool cover -html=profile.out
fuzz:
which go-fuzz > /dev/null 2>&1 || (GO111MODULE=off go get -u github.com/dvyukov/go-fuzz/go-fuzz github.com/dvyukov/go-fuzz/go-fuzz-build; GO111MODULE=off go get -d github.com/dvyukov/go-fuzz-corpus; true)
rm -rf ./fuzz/corpus
rm -rf ./fuzz/crashers
rm -rf ./fuzz/suppressions
rm -f ./fuzz/fuzz-fuzz.zip
cd ./fuzz && go-fuzz-build
cd ./fuzz && go-fuzz

@ -0,0 +1,343 @@
goldmark
==========================================
[![http://godoc.org/github.com/yuin/goldmark](https://godoc.org/github.com/yuin/goldmark?status.svg)](http://godoc.org/github.com/yuin/goldmark)
[![https://github.com/yuin/goldmark/actions?query=workflow:test](https://github.com/yuin/goldmark/workflows/test/badge.svg?branch=master&event=push)](https://github.com/yuin/goldmark/actions?query=workflow:test)
[![https://coveralls.io/github/yuin/goldmark](https://coveralls.io/repos/github/yuin/goldmark/badge.svg?branch=master)](https://coveralls.io/github/yuin/goldmark)
[![https://goreportcard.com/report/github.com/yuin/goldmark](https://goreportcard.com/badge/github.com/yuin/goldmark)](https://goreportcard.com/report/github.com/yuin/goldmark)
> A Markdown parser written in Go. Easy to extend, standard compliant, well structured.
goldmark is compliant with CommonMark 0.29.
Motivation
----------------------
I need a Markdown parser for Go that meets following conditions:
- Easy to extend.
- Markdown is poor in document expressions compared with other light markup languages like reStructuredText.
- We have extensions to the Markdown syntax, e.g. PHP Markdown Extra, GitHub Flavored Markdown.
- Standard compliant.
- Markdown has many dialects.
- GitHub Flavored Markdown is widely used and it is based on CommonMark aside from whether CommonMark is good specification or not.
- CommonMark is too complicated and hard to implement.
- Well structured.
- AST based, and preserves source position of nodes.
- Written in pure Go.
[golang-commonmark](https://gitlab.com/golang-commonmark/markdown) may be a good choice, but it seems to be a copy of [markdown-it](https://github.com/markdown-it).
[blackfriday.v2](https://github.com/russross/blackfriday/tree/v2) is a fast and widely used implementation, but it is not CommonMark compliant and cannot be extended from outside of the package since its AST uses structs instead of interfaces.
Furthermore, its behavior differs from other implementations in some cases, especially regarding lists: ([Deep nested lists don't output correctly #329](https://github.com/russross/blackfriday/issues/329), [List block cannot have a second line #244](https://github.com/russross/blackfriday/issues/244), etc).
This behavior sometimes causes problems. If you migrate your Markdown text to blackfriday-based wikis from GitHub, many lists will immediately be broken.
As mentioned above, CommonMark is too complicated and hard to implement, so Markdown parsers based on CommonMark barely exist.
Features
----------------------
- **Standard compliant.** goldmark gets full compliance with the latest CommonMark spec.
- **Extensible.** Do you want to add a `@username` mention syntax to Markdown?
You can easily do it in goldmark. You can add your AST nodes,
parsers for block level elements, parsers for inline level elements,
transformers for paragraphs, transformers for whole AST structure, and
renderers.
- **Performance.** goldmark performs pretty much equally to cmark,
the CommonMark reference implementation written in C.
- **Robust.** goldmark is tested with [go-fuzz](https://github.com/dvyukov/go-fuzz), a fuzz testing tool.
- **Builtin extensions.** goldmark ships with common extensions like tables, strikethrough,
task lists, and definition lists.
- **Depends only on standard libraries.**
Installation
----------------------
```bash
$ go get github.com/yuin/goldmark
```
Usage
----------------------
Import packages:
```
import (
"bytes"
"github.com/yuin/goldmark"
)
```
Convert Markdown documents with the CommonMark compliant mode:
```go
var buf bytes.Buffer
if err := goldmark.Convert(source, &buf); err != nil {
panic(err)
}
```
With options
------------------------------
```go
var buf bytes.Buffer
if err := goldmark.Convert(source, &buf, parser.WithContext(ctx)); err != nil {
panic(err)
}
```
| Functional option | Type | Description |
| ----------------- | ---- | ----------- |
| `parser.WithContext` | A `parser.Context` | Context for the parsing phase. |
Context options
----------------------
| Functional option | Type | Description |
| ----------------- | ---- | ----------- |
| `parser.WithIDs` | A `parser.IDs` | `IDs` allows you to change logics that are related to element id(ex: Auto heading id generation). |
Custom parser and renderer
--------------------------
```go
import (
"bytes"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/extension"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer/html"
)
md := goldmark.New(
goldmark.WithExtensions(extension.GFM),
goldmark.WithParserOptions(
parser.WithAutoHeadingID(),
),
goldmark.WithRendererOptions(
html.WithHardWraps(),
html.WithXHTML(),
),
)
var buf bytes.Buffer
if err := md.Convert(source, &buf); err != nil {
panic(err)
}
```
Parser and Renderer options
------------------------------
### Parser options
| Functional option | Type | Description |
| ----------------- | ---- | ----------- |
| `parser.WithBlockParsers` | A `util.PrioritizedSlice` whose elements are `parser.BlockParser` | Parsers for parsing block level elements. |
| `parser.WithInlineParsers` | A `util.PrioritizedSlice` whose elements are `parser.InlineParser` | Parsers for parsing inline level elements. |
| `parser.WithParagraphTransformers` | A `util.PrioritizedSlice` whose elements are `parser.ParagraphTransformer` | Transformers for transforming paragraph nodes. |
| `parser.WithASTTransformers` | A `util.PrioritizedSlice` whose elements are `parser.ASTTransformer` | Transformers for transforming an AST. |
| `parser.WithAutoHeadingID` | `-` | Enables auto heading ids. |
| `parser.WithAttribute` | `-` | Enables custom attributes. Currently only headings supports attributes. |
### HTML Renderer options
| Functional option | Type | Description |
| ----------------- | ---- | ----------- |
| `html.WithWriter` | `html.Writer` | `html.Writer` for writing contents to an `io.Writer`. |
| `html.WithHardWraps` | `-` | Render new lines as `<br>`.|
| `html.WithXHTML` | `-` | Render as XHTML. |
| `html.WithUnsafe` | `-` | By default, goldmark does not render raw HTML and potentially dangerous links. With this option, goldmark renders these contents as written. |
### Built-in extensions
- `extension.Table`
- [GitHub Flavored Markdown: Tables](https://github.github.com/gfm/#tables-extension-)
- `extension.Strikethrough`
- [GitHub Flavored Markdown: Strikethrough](https://github.github.com/gfm/#strikethrough-extension-)
- `extension.Linkify`
- [GitHub Flavored Markdown: Autolinks](https://github.github.com/gfm/#autolinks-extension-)
- `extension.TaskList`
- [GitHub Flavored Markdown: Task list items](https://github.github.com/gfm/#task-list-items-extension-)
- `extension.GFM`
- This extension enables Table, Strikethrough, Linkify and TaskList.
- This extension does not filter tags defined in [6.11: Disallowed Raw HTML (extension)](https://github.github.com/gfm/#disallowed-raw-html-extension-).
If you need to filter HTML tags, see [Security](#security)
- `extension.DefinitionList`
- [PHP Markdown Extra: Definition lists](https://michelf.ca/projects/php-markdown/extra/#def-list)
- `extension.Footnote`
- [PHP Markdown Extra: Footnotes](https://michelf.ca/projects/php-markdown/extra/#footnotes)
- `extension.Typographer`
- This extension substitutes punctuations with typographic entities like [smartypants](https://daringfireball.net/projects/smartypants/).
### Attributes
`parser.WithAttribute` option allows you to define attributes on some elements.
Currently only headings support attributes.
**Attributes are being discussed in the
[CommonMark forum](https://talk.commonmark.org/t/consistent-attribute-syntax/272).
This syntax may possibly change in the future.**
#### Headings
```
## heading ## {#id .className attrName=attrValue class="class1 class2"}
## heading {#id .className attrName=attrValue class="class1 class2"}
```
```
heading {#id .className attrName=attrValue}
============
```
### Typographer extension
Typographer extension translates plain ASCII punctuation characters into typographic punctuation HTML entities.
Default substitutions are:
| Punctuation | Default entity |
| ------------ | ---------- |
| `'` | `&lsquo;`, `&rsquo;` |
| `"` | `&ldquo;`, `&rdquo;` |
| `--` | `&ndash;` |
| `---` | `&mdash;` |
| `...` | `&hellip;` |
| `<<` | `&laquo;` |
| `>>` | `&raquo;` |
You can overwrite the substitutions by `extensions.WithTypographicSubstitutions`.
```go
markdown := goldmark.New(
goldmark.WithExtensions(
extension.NewTypographer(
extension.WithTypographicSubstitutions(extension.TypographicSubstitutions{
extension.LeftSingleQuote: []byte("&sbquo;"),
extension.RightSingleQuote: nil, // nil disables a substitution
}),
),
),
)
```
Security
--------------------
By default, goldmark does not render raw HTML and potentially dangerous URLs.
If you need to gain more control over untrusted contents, it is recommended to
use an HTML sanitizer such as [bluemonday](https://github.com/microcosm-cc/bluemonday).
Benchmark
--------------------
You can run this benchmark in the `_benchmark` directory.
### against other golang libraries
blackfriday v2 seems to be fastest, but it is not CommonMark compliant, so the performance of
blackfriday v2 cannot simply be compared with that of the other CommonMark compliant libraries.
Though goldmark builds clean extensible AST structure and get full compliance with
CommonMark, it is reasonably fast and has lower memory consumption.
```
goos: darwin
goarch: amd64
BenchmarkMarkdown/Blackfriday-v2-12 326 3465240 ns/op 3298861 B/op 20047 allocs/op
BenchmarkMarkdown/GoldMark-12 303 3927494 ns/op 2574809 B/op 13853 allocs/op
BenchmarkMarkdown/CommonMark-12 244 4900853 ns/op 2753851 B/op 20527 allocs/op
BenchmarkMarkdown/Lute-12 130 9195245 ns/op 9175030 B/op 123534 allocs/op
BenchmarkMarkdown/GoMarkdown-12 9 113541994 ns/op 2187472 B/op 22173 allocs/op
```
### against cmark (CommonMark reference implementation written in C)
```
----------- cmark -----------
file: _data.md
iteration: 50
average: 0.0037760639 sec
go run ./goldmark_benchmark.go
------- goldmark -------
file: _data.md
iteration: 50
average: 0.0040964230 sec
```
As you can see, goldmark performs pretty much equally to cmark.
Extensions
--------------------
- [goldmark-meta](https://github.com/yuin/goldmark-meta): A YAML metadata
extension for the goldmark Markdown parser.
- [goldmark-highlighting](https://github.com/yuin/goldmark-highlighting): A Syntax highlighting extension
for the goldmark markdown parser.
- [goldmark-mathjax](https://github.com/litao91/goldmark-mathjax): Mathjax support for goldmark markdown parser
goldmark internal(for extension developers)
----------------------------------------------
### Overview
goldmark's Markdown processing is outlined as a bellow diagram.
```
<Markdown in []byte, parser.Context>
|
V
+-------- parser.Parser ---------------------------
| 1. Parse block elements into AST
| 1. If a parsed block is a paragraph, apply
| ast.ParagraphTransformer
| 2. Traverse AST and parse blocks.
| 1. Process delimiters(emphasis) at the end of
| block parsing
| 3. Apply parser.ASTTransformers to AST
|
V
<ast.Node>
|
V
+------- renderer.Renderer ------------------------
| 1. Traverse AST and apply renderer.NodeRenderer
| corespond to the node type
|
V
<Output>
```
### Parsing
Markdown documents are read through `text.Reader` interface.
AST nodes do not have concrete text. AST nodes have segment information of the documents. It is represented by `text.Segment` .
`text.Segment` has 3 attributes: `Start`, `End`, `Padding` .
**TODO**
See `extension` directory for examples of extensions.
Summary:
1. Define AST Node as a struct in which `ast.BaseBlock` or `ast.BaseInline` is embedded.
2. Write a parser that implements `parser.BlockParser` or `parser.InlineParser`.
3. Write a renderer that implements `renderer.NodeRenderer`.
4. Define your goldmark extension that implements `goldmark.Extender`.
Donation
--------------------
BTC: 1NEDSyUmo4SMTDP83JJQSWi1MvQUGGNMZB
License
--------------------
MIT
Author
--------------------
Yusuke Inuzuka

@ -0,0 +1,485 @@
// Package ast defines AST nodes that represent markdown elements.
package ast
import (
"bytes"
"fmt"
"strings"
textm "github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// A NodeType indicates what type a node belongs to.
type NodeType int
const (
// TypeBlock indicates that a node is kind of block nodes.
TypeBlock NodeType = iota + 1
// TypeInline indicates that a node is kind of inline nodes.
TypeInline
// TypeDocument indicates that a node is kind of document nodes.
TypeDocument
)
// NodeKind indicates more specific type than NodeType.
type NodeKind int
func (k NodeKind) String() string {
return kindNames[k]
}
var kindMax NodeKind
var kindNames = []string{""}
// NewNodeKind returns a new Kind value.
func NewNodeKind(name string) NodeKind {
kindMax++
kindNames = append(kindNames, name)
return kindMax
}
// An Attribute is an attribute of the Node
type Attribute struct {
Name []byte
Value interface{}
}
var attrNameIDS = []byte("#")
var attrNameID = []byte("id")
var attrNameClassS = []byte(".")
var attrNameClass = []byte("class")
// A Node interface defines basic AST node functionalities.
type Node interface {
// Type returns a type of this node.
Type() NodeType
// Kind returns a kind of this node.
Kind() NodeKind
// NextSibling returns a next sibling node of this node.
NextSibling() Node
// PreviousSibling returns a previous sibling node of this node.
PreviousSibling() Node
// Parent returns a parent node of this node.
Parent() Node
// SetParent sets a parent node to this node.
SetParent(Node)
// SetPreviousSibling sets a previous sibling node to this node.
SetPreviousSibling(Node)
// SetNextSibling sets a next sibling node to this node.
SetNextSibling(Node)
// HasChildren returns true if this node has any children, otherwise false.
HasChildren() bool
// ChildCount returns a total number of children.
ChildCount() int
// FirstChild returns a first child of this node.
FirstChild() Node
// LastChild returns a last child of this node.
LastChild() Node
// AppendChild append a node child to the tail of the children.
AppendChild(self, child Node)
// RemoveChild removes a node child from this node.
// If a node child is not children of this node, RemoveChild nothing to do.
RemoveChild(self, child Node)
// RemoveChildren removes all children from this node.
RemoveChildren(self Node)
// SortChildren sorts childrens by comparator.
SortChildren(comparator func(n1, n2 Node) int)
// ReplaceChild replace a node v1 with a node insertee.
// If v1 is not children of this node, ReplaceChild append a insetee to the
// tail of the children.
ReplaceChild(self, v1, insertee Node)
// InsertBefore inserts a node insertee before a node v1.
// If v1 is not children of this node, InsertBefore append a insetee to the
// tail of the children.
InsertBefore(self, v1, insertee Node)
// InsertAfterinserts a node insertee after a node v1.
// If v1 is not children of this node, InsertBefore append a insetee to the
// tail of the children.
InsertAfter(self, v1, insertee Node)
// Dump dumps an AST tree structure to stdout.
// This function completely aimed for debugging.
// level is a indent level. Implementer should indent informations with
// 2 * level spaces.
Dump(source []byte, level int)
// Text returns text values of this node.
Text(source []byte) []byte
// HasBlankPreviousLines returns true if the row before this node is blank,
// otherwise false.
// This method is valid only for block nodes.
HasBlankPreviousLines() bool
// SetBlankPreviousLines sets whether the row before this node is blank.
// This method is valid only for block nodes.
SetBlankPreviousLines(v bool)
// Lines returns text segments that hold positions in a source.
// This method is valid only for block nodes.
Lines() *textm.Segments
// SetLines sets text segments that hold positions in a source.
// This method is valid only for block nodes.
SetLines(*textm.Segments)
// IsRaw returns true if contents should be rendered as 'raw' contents.
IsRaw() bool
// SetAttribute sets the given value to the attributes.
SetAttribute(name []byte, value interface{})
// SetAttributeString sets the given value to the attributes.
SetAttributeString(name string, value interface{})
// Attribute returns a (attribute value, true) if an attribute
// associated with the given name is found, otherwise
// (nil, false)
Attribute(name []byte) (interface{}, bool)
// AttributeString returns a (attribute value, true) if an attribute
// associated with the given name is found, otherwise
// (nil, false)
AttributeString(name string) (interface{}, bool)
// Attributes returns a list of attributes.
// This may be a nil if there are no attributes.
Attributes() []Attribute
// RemoveAttributes removes all attributes from this node.
RemoveAttributes()
}
// A BaseNode struct implements the Node interface.
type BaseNode struct {
firstChild Node
lastChild Node
parent Node
next Node
prev Node
childCount int
attributes []Attribute
}
func ensureIsolated(v Node) {
if p := v.Parent(); p != nil {
p.RemoveChild(p, v)
}
}
// HasChildren implements Node.HasChildren .
func (n *BaseNode) HasChildren() bool {
return n.firstChild != nil
}
// SetPreviousSibling implements Node.SetPreviousSibling .
func (n *BaseNode) SetPreviousSibling(v Node) {
n.prev = v
}
// SetNextSibling implements Node.SetNextSibling .
func (n *BaseNode) SetNextSibling(v Node) {
n.next = v
}
// PreviousSibling implements Node.PreviousSibling .
func (n *BaseNode) PreviousSibling() Node {
return n.prev
}
// NextSibling implements Node.NextSibling .
func (n *BaseNode) NextSibling() Node {
return n.next
}
// RemoveChild implements Node.RemoveChild .
func (n *BaseNode) RemoveChild(self, v Node) {
if v.Parent() != self {
return
}
n.childCount--
prev := v.PreviousSibling()
next := v.NextSibling()
if prev != nil {
prev.SetNextSibling(next)
} else {
n.firstChild = next
}
if next != nil {
next.SetPreviousSibling(prev)
} else {
n.lastChild = prev
}
v.SetParent(nil)
v.SetPreviousSibling(nil)
v.SetNextSibling(nil)
}
// RemoveChildren implements Node.RemoveChildren .
func (n *BaseNode) RemoveChildren(self Node) {
for c := n.firstChild; c != nil; c = c.NextSibling() {
c.SetParent(nil)
c.SetPreviousSibling(nil)
c.SetNextSibling(nil)
}
n.firstChild = nil
n.lastChild = nil
n.childCount = 0
}
// SortChildren implements Node.SortChildren
func (n *BaseNode) SortChildren(comparator func(n1, n2 Node) int) {
var sorted Node
current := n.firstChild
for current != nil {
next := current.NextSibling()
if sorted == nil || comparator(sorted, current) >= 0 {
current.SetNextSibling(sorted)
if sorted != nil {
sorted.SetPreviousSibling(current)
}
sorted = current
sorted.SetPreviousSibling(nil)
} else {
c := sorted
for c.NextSibling() != nil && comparator(c.NextSibling(), current) < 0 {
c = c.NextSibling()
}
current.SetNextSibling(c.NextSibling())
current.SetPreviousSibling(c)
if c.NextSibling() != nil {
c.NextSibling().SetPreviousSibling(current)
}
c.SetNextSibling(current)
}
current = next
}
n.firstChild = sorted
for c := n.firstChild; c != nil; c = c.NextSibling() {
n.lastChild = c
}
}
// FirstChild implements Node.FirstChild .
func (n *BaseNode) FirstChild() Node {
return n.firstChild
}
// LastChild implements Node.LastChild .
func (n *BaseNode) LastChild() Node {
return n.lastChild
}
// ChildCount implements Node.ChildCount .
func (n *BaseNode) ChildCount() int {
return n.childCount
}
// Parent implements Node.Parent .
func (n *BaseNode) Parent() Node {
return n.parent
}
// SetParent implements Node.SetParent .
func (n *BaseNode) SetParent(v Node) {
n.parent = v
}
// AppendChild implements Node.AppendChild .
func (n *BaseNode) AppendChild(self, v Node) {
ensureIsolated(v)
if n.firstChild == nil {
n.firstChild = v
v.SetNextSibling(nil)
v.SetPreviousSibling(nil)
} else {
last := n.lastChild
last.SetNextSibling(v)
v.SetPreviousSibling(last)
}
v.SetParent(self)
n.lastChild = v
n.childCount++
}
// ReplaceChild implements Node.ReplaceChild .
func (n *BaseNode) ReplaceChild(self, v1, insertee Node) {
n.InsertBefore(self, v1, insertee)
n.RemoveChild(self, v1)
}
// InsertAfter implements Node.InsertAfter .
func (n *BaseNode) InsertAfter(self, v1, insertee Node) {
n.InsertBefore(self, v1.NextSibling(), insertee)
}
// InsertBefore implements Node.InsertBefore .
func (n *BaseNode) InsertBefore(self, v1, insertee Node) {
n.childCount++
if v1 == nil {
n.AppendChild(self, insertee)
return
}
ensureIsolated(insertee)
if v1.Parent() == self {
c := v1
prev := c.PreviousSibling()
if prev != nil {
prev.SetNextSibling(insertee)
insertee.SetPreviousSibling(prev)
} else {
n.firstChild = insertee
insertee.SetPreviousSibling(nil)
}
insertee.SetNextSibling(c)
c.SetPreviousSibling(insertee)
insertee.SetParent(self)
}
}
// Text implements Node.Text .
func (n *BaseNode) Text(source []byte) []byte {
var buf bytes.Buffer
for c := n.firstChild; c != nil; c = c.NextSibling() {
buf.Write(c.Text(source))
}
return buf.Bytes()
}
// SetAttribute implements Node.SetAttribute.
func (n *BaseNode) SetAttribute(name []byte, value interface{}) {
if n.attributes == nil {
n.attributes = make([]Attribute, 0, 10)
} else {
for i, a := range n.attributes {
if bytes.Equal(a.Name, name) {
n.attributes[i].Name = name
n.attributes[i].Value = value
return
}
}
}
n.attributes = append(n.attributes, Attribute{name, value})
}
// SetAttributeString implements Node.SetAttributeString
func (n *BaseNode) SetAttributeString(name string, value interface{}) {
n.SetAttribute(util.StringToReadOnlyBytes(name), value)
}
// Attribute implements Node.Attribute.
func (n *BaseNode) Attribute(name []byte) (interface{}, bool) {
if n.attributes == nil {
return nil, false
}
for i, a := range n.attributes {
if bytes.Equal(a.Name, name) {
return n.attributes[i].Value, true
}
}
return nil, false
}
// AttributeString implements Node.AttributeString.
func (n *BaseNode) AttributeString(s string) (interface{}, bool) {
return n.Attribute(util.StringToReadOnlyBytes(s))
}
// Attributes implements Node.Attributes
func (n *BaseNode) Attributes() []Attribute {
return n.attributes
}
// RemoveAttributes implements Node.RemoveAttributes
func (n *BaseNode) RemoveAttributes() {
n.attributes = nil
}
// DumpHelper is a helper function to implement Node.Dump.
// kv is pairs of an attribute name and an attribute value.
// cb is a function called after wrote a name and attributes.
func DumpHelper(v Node, source []byte, level int, kv map[string]string, cb func(int)) {
name := v.Kind().String()
indent := strings.Repeat(" ", level)
fmt.Printf("%s%s {\n", indent, name)
indent2 := strings.Repeat(" ", level+1)
if v.Type() == TypeBlock {
fmt.Printf("%sRawText: \"", indent2)
for i := 0; i < v.Lines().Len(); i++ {
line := v.Lines().At(i)
fmt.Printf("%s", line.Value(source))
}
fmt.Printf("\"\n")
fmt.Printf("%sHasBlankPreviousLines: %v\n", indent2, v.HasBlankPreviousLines())
}
for name, value := range kv {
fmt.Printf("%s%s: %s\n", indent2, name, value)
}
if cb != nil {
cb(level + 1)
}
for c := v.FirstChild(); c != nil; c = c.NextSibling() {
c.Dump(source, level+1)
}
fmt.Printf("%s}\n", indent)
}
// WalkStatus represents a current status of the Walk function.
type WalkStatus int
const (
// WalkStop indicates no more walking needed.
WalkStop WalkStatus = iota + 1
// WalkSkipChildren indicates that Walk wont walk on children of current
// node.
WalkSkipChildren
// WalkContinue indicates that Walk can continue to walk.
WalkContinue
)
// Walker is a function that will be called when Walk find a
// new node.
// entering is set true before walks children, false after walked children.
// If Walker returns error, Walk function immediately stop walking.
type Walker func(n Node, entering bool) (WalkStatus, error)
// Walk walks a AST tree by the depth first search algorithm.
func Walk(n Node, walker Walker) error {
status, err := walker(n, true)
if err != nil || status == WalkStop {
return err
}
if status != WalkSkipChildren {
for c := n.FirstChild(); c != nil; c = c.NextSibling() {
if err = Walk(c, walker); err != nil {
return err
}
}
}
status, err = walker(n, false)
if err != nil || status == WalkStop {
return err
}
return nil
}

@ -0,0 +1,474 @@
package ast
import (
"fmt"
"strings"
textm "github.com/yuin/goldmark/text"
)
// A BaseBlock struct implements the Node interface.
type BaseBlock struct {
BaseNode
blankPreviousLines bool
lines *textm.Segments
}
// Type implements Node.Type
func (b *BaseBlock) Type() NodeType {
return TypeBlock
}
// IsRaw implements Node.IsRaw
func (b *BaseBlock) IsRaw() bool {
return false
}
// HasBlankPreviousLines implements Node.HasBlankPreviousLines.
func (b *BaseBlock) HasBlankPreviousLines() bool {
return b.blankPreviousLines
}
// SetBlankPreviousLines implements Node.SetBlankPreviousLines.
func (b *BaseBlock) SetBlankPreviousLines(v bool) {
b.blankPreviousLines = v
}
// Lines implements Node.Lines
func (b *BaseBlock) Lines() *textm.Segments {
if b.lines == nil {
b.lines = textm.NewSegments()
}
return b.lines
}
// SetLines implements Node.SetLines
func (b *BaseBlock) SetLines(v *textm.Segments) {
b.lines = v
}
// A Document struct is a root node of Markdown text.
type Document struct {
BaseBlock
}
// KindDocument is a NodeKind of the Document node.
var KindDocument = NewNodeKind("Document")
// Dump implements Node.Dump .
func (n *Document) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// Type implements Node.Type .
func (n *Document) Type() NodeType {
return TypeDocument
}
// Kind implements Node.Kind.
func (n *Document) Kind() NodeKind {
return KindDocument
}
// NewDocument returns a new Document node.
func NewDocument() *Document {
return &Document{
BaseBlock: BaseBlock{},
}
}
// A TextBlock struct is a node whose lines
// should be rendered without any containers.
type TextBlock struct {
BaseBlock
}
// Dump implements Node.Dump .
func (n *TextBlock) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindTextBlock is a NodeKind of the TextBlock node.
var KindTextBlock = NewNodeKind("TextBlock")
// Kind implements Node.Kind.
func (n *TextBlock) Kind() NodeKind {
return KindTextBlock
}
// NewTextBlock returns a new TextBlock node.
func NewTextBlock() *TextBlock {
return &TextBlock{
BaseBlock: BaseBlock{},
}
}
// A Paragraph struct represents a paragraph of Markdown text.
type Paragraph struct {
BaseBlock
}
// Dump implements Node.Dump .
func (n *Paragraph) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindParagraph is a NodeKind of the Paragraph node.
var KindParagraph = NewNodeKind("Paragraph")
// Kind implements Node.Kind.
func (n *Paragraph) Kind() NodeKind {
return KindParagraph
}
// NewParagraph returns a new Paragraph node.
func NewParagraph() *Paragraph {
return &Paragraph{
BaseBlock: BaseBlock{},
}
}
// IsParagraph returns true if the given node implements the Paragraph interface,
// otherwise false.
func IsParagraph(node Node) bool {
_, ok := node.(*Paragraph)
return ok
}
// A Heading struct represents headings like SetextHeading and ATXHeading.
type Heading struct {
BaseBlock
// Level returns a level of this heading.
// This value is between 1 and 6.
Level int
}
// Dump implements Node.Dump .
func (n *Heading) Dump(source []byte, level int) {
m := map[string]string{
"Level": fmt.Sprintf("%d", n.Level),
}
DumpHelper(n, source, level, m, nil)
}
// KindHeading is a NodeKind of the Heading node.
var KindHeading = NewNodeKind("Heading")
// Kind implements Node.Kind.
func (n *Heading) Kind() NodeKind {
return KindHeading
}
// NewHeading returns a new Heading node.
func NewHeading(level int) *Heading {
return &Heading{
BaseBlock: BaseBlock{},
Level: level,
}
}
// A ThematicBreak struct represents a thematic break of Markdown text.
type ThematicBreak struct {
BaseBlock
}
// Dump implements Node.Dump .
func (n *ThematicBreak) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindThematicBreak is a NodeKind of the ThematicBreak node.
var KindThematicBreak = NewNodeKind("ThematicBreak")
// Kind implements Node.Kind.
func (n *ThematicBreak) Kind() NodeKind {
return KindThematicBreak
}
// NewThematicBreak returns a new ThematicBreak node.
func NewThematicBreak() *ThematicBreak {
return &ThematicBreak{
BaseBlock: BaseBlock{},
}
}
// A CodeBlock interface represents an indented code block of Markdown text.
type CodeBlock struct {
BaseBlock
}
// IsRaw implements Node.IsRaw.
func (n *CodeBlock) IsRaw() bool {
return true
}
// Dump implements Node.Dump .
func (n *CodeBlock) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindCodeBlock is a NodeKind of the CodeBlock node.
var KindCodeBlock = NewNodeKind("CodeBlock")
// Kind implements Node.Kind.
func (n *CodeBlock) Kind() NodeKind {
return KindCodeBlock
}
// NewCodeBlock returns a new CodeBlock node.
func NewCodeBlock() *CodeBlock {
return &CodeBlock{
BaseBlock: BaseBlock{},
}
}
// A FencedCodeBlock struct represents a fenced code block of Markdown text.
type FencedCodeBlock struct {
BaseBlock
// Info returns a info text of this fenced code block.
Info *Text
language []byte
}
// Language returns an language in an info string.
// Language returns nil if this node does not have an info string.
func (n *FencedCodeBlock) Language(source []byte) []byte {
if n.language == nil && n.Info != nil {
segment := n.Info.Segment
info := segment.Value(source)
i := 0
for ; i < len(info); i++ {
if info[i] == ' ' {
break
}
}
n.language = info[:i]
}
return n.language
}
// IsRaw implements Node.IsRaw.
func (n *FencedCodeBlock) IsRaw() bool {
return true
}
// Dump implements Node.Dump .
func (n *FencedCodeBlock) Dump(source []byte, level int) {
m := map[string]string{}
if n.Info != nil {
m["Info"] = fmt.Sprintf("\"%s\"", n.Info.Text(source))
}
DumpHelper(n, source, level, m, nil)
}
// KindFencedCodeBlock is a NodeKind of the FencedCodeBlock node.
var KindFencedCodeBlock = NewNodeKind("FencedCodeBlock")
// Kind implements Node.Kind.
func (n *FencedCodeBlock) Kind() NodeKind {
return KindFencedCodeBlock
}
// NewFencedCodeBlock return a new FencedCodeBlock node.
func NewFencedCodeBlock(info *Text) *FencedCodeBlock {
return &FencedCodeBlock{
BaseBlock: BaseBlock{},
Info: info,
}
}
// A Blockquote struct represents an blockquote block of Markdown text.
type Blockquote struct {
BaseBlock
}
// Dump implements Node.Dump .
func (n *Blockquote) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindBlockquote is a NodeKind of the Blockquote node.
var KindBlockquote = NewNodeKind("Blockquote")
// Kind implements Node.Kind.
func (n *Blockquote) Kind() NodeKind {
return KindBlockquote
}
// NewBlockquote returns a new Blockquote node.
func NewBlockquote() *Blockquote {
return &Blockquote{
BaseBlock: BaseBlock{},
}
}
// A List structr represents a list of Markdown text.
type List struct {
BaseBlock
// Marker is a markar character like '-', '+', ')' and '.'.
Marker byte
// IsTight is a true if this list is a 'tight' list.
// See https://spec.commonmark.org/0.29/#loose for details.
IsTight bool
// Start is an initial number of this ordered list.
// If this list is not an ordered list, Start is 0.
Start int
}
// IsOrdered returns true if this list is an ordered list, otherwise false.
func (l *List) IsOrdered() bool {
return l.Marker == '.' || l.Marker == ')'
}
// CanContinue returns true if this list can continue with
// the given mark and a list type, otherwise false.
func (l *List) CanContinue(marker byte, isOrdered bool) bool {
return marker == l.Marker && isOrdered == l.IsOrdered()
}
// Dump implements Node.Dump.
func (l *List) Dump(source []byte, level int) {
m := map[string]string{
"Ordered": fmt.Sprintf("%v", l.IsOrdered()),
"Marker": fmt.Sprintf("%c", l.Marker),
"Tight": fmt.Sprintf("%v", l.IsTight),
}
if l.IsOrdered() {
m["Start"] = fmt.Sprintf("%d", l.Start)
}
DumpHelper(l, source, level, m, nil)
}
// KindList is a NodeKind of the List node.
var KindList = NewNodeKind("List")
// Kind implements Node.Kind.
func (l *List) Kind() NodeKind {
return KindList
}
// NewList returns a new List node.
func NewList(marker byte) *List {
return &List{
BaseBlock: BaseBlock{},
Marker: marker,
IsTight: true,
}
}
// A ListItem struct represents a list item of Markdown text.
type ListItem struct {
BaseBlock
// Offset is an offset potision of this item.
Offset int
}
// Dump implements Node.Dump.
func (n *ListItem) Dump(source []byte, level int) {
m := map[string]string{
"Offset": fmt.Sprintf("%d", n.Offset),
}
DumpHelper(n, source, level, m, nil)
}
// KindListItem is a NodeKind of the ListItem node.
var KindListItem = NewNodeKind("ListItem")
// Kind implements Node.Kind.
func (n *ListItem) Kind() NodeKind {
return KindListItem
}
// NewListItem returns a new ListItem node.
func NewListItem(offset int) *ListItem {
return &ListItem{
BaseBlock: BaseBlock{},
Offset: offset,
}
}
// HTMLBlockType represents kinds of an html blocks.
// See https://spec.commonmark.org/0.29/#html-blocks
type HTMLBlockType int
const (
// HTMLBlockType1 represents type 1 html blocks
HTMLBlockType1 HTMLBlockType = iota + 1
// HTMLBlockType2 represents type 2 html blocks
HTMLBlockType2
// HTMLBlockType3 represents type 3 html blocks
HTMLBlockType3
// HTMLBlockType4 represents type 4 html blocks
HTMLBlockType4
// HTMLBlockType5 represents type 5 html blocks
HTMLBlockType5
// HTMLBlockType6 represents type 6 html blocks
HTMLBlockType6
// HTMLBlockType7 represents type 7 html blocks
HTMLBlockType7
)
// An HTMLBlock struct represents an html block of Markdown text.
type HTMLBlock struct {
BaseBlock
// Type is a type of this html block.
HTMLBlockType HTMLBlockType
// ClosureLine is a line that closes this html block.
ClosureLine textm.Segment
}
// IsRaw implements Node.IsRaw.
func (n *HTMLBlock) IsRaw() bool {
return true
}
// HasClosure returns true if this html block has a closure line,
// otherwise false.
func (n *HTMLBlock) HasClosure() bool {
return n.ClosureLine.Start >= 0
}
// Dump implements Node.Dump.
func (n *HTMLBlock) Dump(source []byte, level int) {
indent := strings.Repeat(" ", level)
fmt.Printf("%s%s {\n", indent, "HTMLBlock")
indent2 := strings.Repeat(" ", level+1)
fmt.Printf("%sRawText: \"", indent2)
for i := 0; i < n.Lines().Len(); i++ {
s := n.Lines().At(i)
fmt.Print(string(source[s.Start:s.Stop]))
}
fmt.Printf("\"\n")
for c := n.FirstChild(); c != nil; c = c.NextSibling() {
c.Dump(source, level+1)
}
if n.HasClosure() {
cl := n.ClosureLine
fmt.Printf("%sClosure: \"%s\"\n", indent2, string(cl.Value(source)))
}
fmt.Printf("%s}\n", indent)
}
// KindHTMLBlock is a NodeKind of the HTMLBlock node.
var KindHTMLBlock = NewNodeKind("HTMLBlock")
// Kind implements Node.Kind.
func (n *HTMLBlock) Kind() NodeKind {
return KindHTMLBlock
}
// NewHTMLBlock returns a new HTMLBlock node.
func NewHTMLBlock(typ HTMLBlockType) *HTMLBlock {
return &HTMLBlock{
BaseBlock: BaseBlock{},
HTMLBlockType: typ,
ClosureLine: textm.NewSegment(-1, -1),
}
}

@ -0,0 +1,548 @@
package ast
import (
"fmt"
"strings"
textm "github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// A BaseInline struct implements the Node interface.
type BaseInline struct {
BaseNode
}
// Type implements Node.Type
func (b *BaseInline) Type() NodeType {
return TypeInline
}
// IsRaw implements Node.IsRaw
func (b *BaseInline) IsRaw() bool {
return false
}
// HasBlankPreviousLines implements Node.HasBlankPreviousLines.
func (b *BaseInline) HasBlankPreviousLines() bool {
panic("can not call with inline nodes.")
}
// SetBlankPreviousLines implements Node.SetBlankPreviousLines.
func (b *BaseInline) SetBlankPreviousLines(v bool) {
panic("can not call with inline nodes.")
}
// Lines implements Node.Lines
func (b *BaseInline) Lines() *textm.Segments {
panic("can not call with inline nodes.")
}
// SetLines implements Node.SetLines
func (b *BaseInline) SetLines(v *textm.Segments) {
panic("can not call with inline nodes.")
}
// A Text struct represents a textual content of the Markdown text.
type Text struct {
BaseInline
// Segment is a position in a source text.
Segment textm.Segment
flags uint8
}
const (
textSoftLineBreak = 1 << iota
textHardLineBreak
textRaw
textCode
)
func textFlagsString(flags uint8) string {
buf := []string{}
if flags&textSoftLineBreak != 0 {
buf = append(buf, "SoftLineBreak")
}
if flags&textHardLineBreak != 0 {
buf = append(buf, "HardLineBreak")
}
if flags&textRaw != 0 {
buf = append(buf, "Raw")
}
if flags&textCode != 0 {
buf = append(buf, "Code")
}
return strings.Join(buf, ", ")
}
// Inline implements Inline.Inline.
func (n *Text) Inline() {
}
// SoftLineBreak returns true if this node ends with a new line,
// otherwise false.
func (n *Text) SoftLineBreak() bool {
return n.flags&textSoftLineBreak != 0
}
// SetSoftLineBreak sets whether this node ends with a new line.
func (n *Text) SetSoftLineBreak(v bool) {
if v {
n.flags |= textSoftLineBreak
} else {
n.flags = n.flags &^ textHardLineBreak
}
}
// IsRaw returns true if this text should be rendered without unescaping
// back slash escapes and resolving references.
func (n *Text) IsRaw() bool {
return n.flags&textRaw != 0
}
// SetRaw sets whether this text should be rendered as raw contents.
func (n *Text) SetRaw(v bool) {
if v {
n.flags |= textRaw
} else {
n.flags = n.flags &^ textRaw
}
}
// HardLineBreak returns true if this node ends with a hard line break.
// See https://spec.commonmark.org/0.29/#hard-line-breaks for details.
func (n *Text) HardLineBreak() bool {
return n.flags&textHardLineBreak != 0
}
// SetHardLineBreak sets whether this node ends with a hard line break.
func (n *Text) SetHardLineBreak(v bool) {
if v {
n.flags |= textHardLineBreak
} else {
n.flags = n.flags &^ textHardLineBreak
}
}
// Merge merges a Node n into this node.
// Merge returns true if the given node has been merged, otherwise false.
func (n *Text) Merge(node Node, source []byte) bool {
t, ok := node.(*Text)
if !ok {
return false
}
if n.Segment.Stop != t.Segment.Start || t.Segment.Padding != 0 || source[n.Segment.Stop-1] == '\n' || t.IsRaw() != n.IsRaw() {
return false
}
n.Segment.Stop = t.Segment.Stop
n.SetSoftLineBreak(t.SoftLineBreak())
n.SetHardLineBreak(t.HardLineBreak())
return true
}
// Text implements Node.Text.
func (n *Text) Text(source []byte) []byte {
return n.Segment.Value(source)
}
// Dump implements Node.Dump.
func (n *Text) Dump(source []byte, level int) {
fs := textFlagsString(n.flags)
if len(fs) != 0 {
fs = "(" + fs + ")"
}
fmt.Printf("%sText%s: \"%s\"\n", strings.Repeat(" ", level), fs, strings.TrimRight(string(n.Text(source)), "\n"))
}
// KindText is a NodeKind of the Text node.
var KindText = NewNodeKind("Text")
// Kind implements Node.Kind.
func (n *Text) Kind() NodeKind {
return KindText
}
// NewText returns a new Text node.
func NewText() *Text {
return &Text{
BaseInline: BaseInline{},
}
}
// NewTextSegment returns a new Text node with the given source potision.
func NewTextSegment(v textm.Segment) *Text {
return &Text{
BaseInline: BaseInline{},
Segment: v,
}
}
// NewRawTextSegment returns a new Text node with the given source position.
// The new node should be rendered as raw contents.
func NewRawTextSegment(v textm.Segment) *Text {
t := &Text{
BaseInline: BaseInline{},
Segment: v,
}
t.SetRaw(true)
return t
}
// MergeOrAppendTextSegment merges a given s into the last child of the parent if
// it can be merged, otherwise creates a new Text node and appends it to after current
// last child.
func MergeOrAppendTextSegment(parent Node, s textm.Segment) {
last := parent.LastChild()
t, ok := last.(*Text)
if ok && t.Segment.Stop == s.Start && !t.SoftLineBreak() {
t.Segment = t.Segment.WithStop(s.Stop)
} else {
parent.AppendChild(parent, NewTextSegment(s))
}
}
// MergeOrReplaceTextSegment merges a given s into a previous sibling of the node n
// if a previous sibling of the node n is *Text, otherwise replaces Node n with s.
func MergeOrReplaceTextSegment(parent Node, n Node, s textm.Segment) {
prev := n.PreviousSibling()
if t, ok := prev.(*Text); ok && t.Segment.Stop == s.Start && !t.SoftLineBreak() {
t.Segment = t.Segment.WithStop(s.Stop)
parent.RemoveChild(parent, n)
} else {
parent.ReplaceChild(parent, n, NewTextSegment(s))
}
}
// A String struct is a textual content that has a concrete value
type String struct {
BaseInline
Value []byte
flags uint8
}
// Inline implements Inline.Inline.
func (n *String) Inline() {
}
// IsRaw returns true if this text should be rendered without unescaping
// back slash escapes and resolving references.
func (n *String) IsRaw() bool {
return n.flags&textRaw != 0
}
// SetRaw sets whether this text should be rendered as raw contents.
func (n *String) SetRaw(v bool) {
if v {
n.flags |= textRaw
} else {
n.flags = n.flags &^ textRaw
}
}
// IsCode returns true if this text should be rendered without any
// modifications.
func (n *String) IsCode() bool {
return n.flags&textCode != 0
}
// SetCode sets whether this text should be rendered without any modifications.
func (n *String) SetCode(v bool) {
if v {
n.flags |= textCode
} else {
n.flags = n.flags &^ textCode
}
}
// Text implements Node.Text.
func (n *String) Text(source []byte) []byte {
return n.Value
}
// Dump implements Node.Dump.
func (n *String) Dump(source []byte, level int) {
fs := textFlagsString(n.flags)
if len(fs) != 0 {
fs = "(" + fs + ")"
}
fmt.Printf("%sString%s: \"%s\"\n", strings.Repeat(" ", level), fs, strings.TrimRight(string(n.Value), "\n"))
}
// KindString is a NodeKind of the String node.
var KindString = NewNodeKind("String")
// Kind implements Node.Kind.
func (n *String) Kind() NodeKind {
return KindString
}
// NewString returns a new String node.
func NewString(v []byte) *String {
return &String{
Value: v,
}
}
// A CodeSpan struct represents a code span of Markdown text.
type CodeSpan struct {
BaseInline
}
// Inline implements Inline.Inline .
func (n *CodeSpan) Inline() {
}
// IsBlank returns true if this node consists of spaces, otherwise false.
func (n *CodeSpan) IsBlank(source []byte) bool {
for c := n.FirstChild(); c != nil; c = c.NextSibling() {
text := c.(*Text).Segment
if !util.IsBlank(text.Value(source)) {
return false
}
}
return true
}
// Dump implements Node.Dump
func (n *CodeSpan) Dump(source []byte, level int) {
DumpHelper(n, source, level, nil, nil)
}
// KindCodeSpan is a NodeKind of the CodeSpan node.
var KindCodeSpan = NewNodeKind("CodeSpan")
// Kind implements Node.Kind.
func (n *CodeSpan) Kind() NodeKind {
return KindCodeSpan
}
// NewCodeSpan returns a new CodeSpan node.
func NewCodeSpan() *CodeSpan {
return &CodeSpan{
BaseInline: BaseInline{},
}
}
// An Emphasis struct represents an emphasis of Markdown text.
type Emphasis struct {
BaseInline
// Level is a level of the emphasis.
Level int
}
// Dump implements Node.Dump.
func (n *Emphasis) Dump(source []byte, level int) {
m := map[string]string{
"Level": fmt.Sprintf("%v", n.Level),
}
DumpHelper(n, source, level, m, nil)
}
// KindEmphasis is a NodeKind of the Emphasis node.
var KindEmphasis = NewNodeKind("Emphasis")
// Kind implements Node.Kind.
func (n *Emphasis) Kind() NodeKind {
return KindEmphasis
}
// NewEmphasis returns a new Emphasis node with the given level.
func NewEmphasis(level int) *Emphasis {
return &Emphasis{
BaseInline: BaseInline{},
Level: level,
}
}
type baseLink struct {
BaseInline
// Destination is a destination(URL) of this link.
Destination []byte
// Title is a title of this link.
Title []byte
}
// Inline implements Inline.Inline.
func (n *baseLink) Inline() {
}
// A Link struct represents a link of the Markdown text.
type Link struct {
baseLink
}
// Dump implements Node.Dump.
func (n *Link) Dump(source []byte, level int) {
m := map[string]string{}
m["Destination"] = string(n.Destination)
m["Title"] = string(n.Title)
DumpHelper(n, source, level, m, nil)
}
// KindLink is a NodeKind of the Link node.
var KindLink = NewNodeKind("Link")
// Kind implements Node.Kind.
func (n *Link) Kind() NodeKind {
return KindLink
}
// NewLink returns a new Link node.
func NewLink() *Link {
c := &Link{
baseLink: baseLink{
BaseInline: BaseInline{},
},
}
return c
}
// An Image struct represents an image of the Markdown text.
type Image struct {
baseLink
}
// Dump implements Node.Dump.
func (n *Image) Dump(source []byte, level int) {
m := map[string]string{}
m["Destination"] = string(n.Destination)
m["Title"] = string(n.Title)
DumpHelper(n, source, level, m, nil)
}
// KindImage is a NodeKind of the Image node.
var KindImage = NewNodeKind("Image")
// Kind implements Node.Kind.
func (n *Image) Kind() NodeKind {
return KindImage
}
// NewImage returns a new Image node.
func NewImage(link *Link) *Image {
c := &Image{
baseLink: baseLink{
BaseInline: BaseInline{},
},
}
c.Destination = link.Destination
c.Title = link.Title
for n := link.FirstChild(); n != nil; {
next := n.NextSibling()
link.RemoveChild(link, n)
c.AppendChild(c, n)
n = next
}
return c
}
// AutoLinkType defines kind of auto links.
type AutoLinkType int
const (
// AutoLinkEmail indicates that an autolink is an email address.
AutoLinkEmail AutoLinkType = iota + 1
// AutoLinkURL indicates that an autolink is a generic URL.
AutoLinkURL
)
// An AutoLink struct represents an autolink of the Markdown text.
type AutoLink struct {
BaseInline
// Type is a type of this autolink.
AutoLinkType AutoLinkType
// Protocol specified a protocol of the link.
Protocol []byte
value *Text
}
// Inline implements Inline.Inline.
func (n *AutoLink) Inline() {}
// Dump implenets Node.Dump
func (n *AutoLink) Dump(source []byte, level int) {
segment := n.value.Segment
m := map[string]string{
"Value": string(segment.Value(source)),
}
DumpHelper(n, source, level, m, nil)
}
// KindAutoLink is a NodeKind of the AutoLink node.
var KindAutoLink = NewNodeKind("AutoLink")
// Kind implements Node.Kind.
func (n *AutoLink) Kind() NodeKind {
return KindAutoLink
}
// URL returns an url of this node.
func (n *AutoLink) URL(source []byte) []byte {
if n.Protocol != nil {
s := n.value.Segment
ret := make([]byte, 0, len(n.Protocol)+s.Len()+3)
ret = append(ret, n.Protocol...)
ret = append(ret, ':', '/', '/')
ret = append(ret, n.value.Text(source)...)
return ret
}
return n.value.Text(source)
}
// Label returns a label of this node.
func (n *AutoLink) Label(source []byte) []byte {
return n.value.Text(source)
}
// NewAutoLink returns a new AutoLink node.
func NewAutoLink(typ AutoLinkType, value *Text) *AutoLink {
return &AutoLink{
BaseInline: BaseInline{},
value: value,
AutoLinkType: typ,
}
}
// A RawHTML struct represents an inline raw HTML of the Markdown text.
type RawHTML struct {
BaseInline
Segments *textm.Segments
}
// Inline implements Inline.Inline.
func (n *RawHTML) Inline() {}
// Dump implements Node.Dump.
func (n *RawHTML) Dump(source []byte, level int) {
m := map[string]string{}
t := []string{}
for i := 0; i < n.Segments.Len(); i++ {
segment := n.Segments.At(i)
t = append(t, string(segment.Value(source)))
}
m["RawText"] = strings.Join(t, "")
DumpHelper(n, source, level, m, nil)
}
// KindRawHTML is a NodeKind of the RawHTML node.
var KindRawHTML = NewNodeKind("RawHTML")
// Kind implements Node.Kind.
func (n *RawHTML) Kind() NodeKind {
return KindRawHTML
}
// NewRawHTML returns a new RawHTML node.
func NewRawHTML() *RawHTML {
return &RawHTML{
Segments: textm.NewSegments(),
}
}

@ -0,0 +1,83 @@
package ast
import (
gast "github.com/yuin/goldmark/ast"
)
// A DefinitionList struct represents a definition list of Markdown
// (PHPMarkdownExtra) text.
type DefinitionList struct {
gast.BaseBlock
Offset int
TemporaryParagraph *gast.Paragraph
}
// Dump implements Node.Dump.
func (n *DefinitionList) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindDefinitionList is a NodeKind of the DefinitionList node.
var KindDefinitionList = gast.NewNodeKind("DefinitionList")
// Kind implements Node.Kind.
func (n *DefinitionList) Kind() gast.NodeKind {
return KindDefinitionList
}
// NewDefinitionList returns a new DefinitionList node.
func NewDefinitionList(offset int, para *gast.Paragraph) *DefinitionList {
return &DefinitionList{
Offset: offset,
TemporaryParagraph: para,
}
}
// A DefinitionTerm struct represents a definition list term of Markdown
// (PHPMarkdownExtra) text.
type DefinitionTerm struct {
gast.BaseBlock
}
// Dump implements Node.Dump.
func (n *DefinitionTerm) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindDefinitionTerm is a NodeKind of the DefinitionTerm node.
var KindDefinitionTerm = gast.NewNodeKind("DefinitionTerm")
// Kind implements Node.Kind.
func (n *DefinitionTerm) Kind() gast.NodeKind {
return KindDefinitionTerm
}
// NewDefinitionTerm returns a new DefinitionTerm node.
func NewDefinitionTerm() *DefinitionTerm {
return &DefinitionTerm{}
}
// A DefinitionDescription struct represents a definition list description of Markdown
// (PHPMarkdownExtra) text.
type DefinitionDescription struct {
gast.BaseBlock
IsTight bool
}
// Dump implements Node.Dump.
func (n *DefinitionDescription) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindDefinitionDescription is a NodeKind of the DefinitionDescription node.
var KindDefinitionDescription = gast.NewNodeKind("DefinitionDescription")
// Kind implements Node.Kind.
func (n *DefinitionDescription) Kind() gast.NodeKind {
return KindDefinitionDescription
}
// NewDefinitionDescription returns a new DefinitionDescription node.
func NewDefinitionDescription() *DefinitionDescription {
return &DefinitionDescription{}
}

@ -0,0 +1,125 @@
package ast
import (
"fmt"
gast "github.com/yuin/goldmark/ast"
)
// A FootnoteLink struct represents a link to a footnote of Markdown
// (PHP Markdown Extra) text.
type FootnoteLink struct {
gast.BaseInline
Index int
}
// Dump implements Node.Dump.
func (n *FootnoteLink) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
gast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteLink is a NodeKind of the FootnoteLink node.
var KindFootnoteLink = gast.NewNodeKind("FootnoteLink")
// Kind implements Node.Kind.
func (n *FootnoteLink) Kind() gast.NodeKind {
return KindFootnoteLink
}
// NewFootnoteLink returns a new FootnoteLink node.
func NewFootnoteLink(index int) *FootnoteLink {
return &FootnoteLink{
Index: index,
}
}
// A FootnoteBackLink struct represents a link to a footnote of Markdown
// (PHP Markdown Extra) text.
type FootnoteBackLink struct {
gast.BaseInline
Index int
}
// Dump implements Node.Dump.
func (n *FootnoteBackLink) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
gast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteBackLink is a NodeKind of the FootnoteBackLink node.
var KindFootnoteBackLink = gast.NewNodeKind("FootnoteBackLink")
// Kind implements Node.Kind.
func (n *FootnoteBackLink) Kind() gast.NodeKind {
return KindFootnoteBackLink
}
// NewFootnoteBackLink returns a new FootnoteBackLink node.
func NewFootnoteBackLink(index int) *FootnoteBackLink {
return &FootnoteBackLink{
Index: index,
}
}
// A Footnote struct represents a footnote of Markdown
// (PHP Markdown Extra) text.
type Footnote struct {
gast.BaseBlock
Ref []byte
Index int
}
// Dump implements Node.Dump.
func (n *Footnote) Dump(source []byte, level int) {
m := map[string]string{}
m["Index"] = fmt.Sprintf("%v", n.Index)
m["Ref"] = fmt.Sprintf("%s", n.Ref)
gast.DumpHelper(n, source, level, m, nil)
}
// KindFootnote is a NodeKind of the Footnote node.
var KindFootnote = gast.NewNodeKind("Footnote")
// Kind implements Node.Kind.
func (n *Footnote) Kind() gast.NodeKind {
return KindFootnote
}
// NewFootnote returns a new Footnote node.
func NewFootnote(ref []byte) *Footnote {
return &Footnote{
Ref: ref,
Index: -1,
}
}
// A FootnoteList struct represents footnotes of Markdown
// (PHP Markdown Extra) text.
type FootnoteList struct {
gast.BaseBlock
Count int
}
// Dump implements Node.Dump.
func (n *FootnoteList) Dump(source []byte, level int) {
m := map[string]string{}
m["Count"] = fmt.Sprintf("%v", n.Count)
gast.DumpHelper(n, source, level, m, nil)
}
// KindFootnoteList is a NodeKind of the FootnoteList node.
var KindFootnoteList = gast.NewNodeKind("FootnoteList")
// Kind implements Node.Kind.
func (n *FootnoteList) Kind() gast.NodeKind {
return KindFootnoteList
}
// NewFootnoteList returns a new FootnoteList node.
func NewFootnoteList() *FootnoteList {
return &FootnoteList{
Count: 0,
}
}

@ -0,0 +1,29 @@
// Package ast defines AST nodes that represents extension's elements
package ast
import (
gast "github.com/yuin/goldmark/ast"
)
// A Strikethrough struct represents a strikethrough of GFM text.
type Strikethrough struct {
gast.BaseInline
}
// Dump implements Node.Dump.
func (n *Strikethrough) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindStrikethrough is a NodeKind of the Strikethrough node.
var KindStrikethrough = gast.NewNodeKind("Strikethrough")
// Kind implements Node.Kind.
func (n *Strikethrough) Kind() gast.NodeKind {
return KindStrikethrough
}
// NewStrikethrough returns a new Strikethrough node.
func NewStrikethrough() *Strikethrough {
return &Strikethrough{}
}

@ -0,0 +1,157 @@
package ast
import (
"fmt"
gast "github.com/yuin/goldmark/ast"
"strings"
)
// Alignment is a text alignment of table cells.
type Alignment int
const (
// AlignLeft indicates text should be left justified.
AlignLeft Alignment = iota + 1
// AlignRight indicates text should be right justified.
AlignRight
// AlignCenter indicates text should be centered.
AlignCenter
// AlignNone indicates text should be aligned by default manner.
AlignNone
)
func (a Alignment) String() string {
switch a {
case AlignLeft:
return "left"
case AlignRight:
return "right"
case AlignCenter:
return "center"
case AlignNone:
return "none"
}
return ""
}
// A Table struct represents a table of Markdown(GFM) text.
type Table struct {
gast.BaseBlock
// Alignments returns alignments of the columns.
Alignments []Alignment
}
// Dump implements Node.Dump
func (n *Table) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, func(level int) {
indent := strings.Repeat(" ", level)
fmt.Printf("%sAlignments {\n", indent)
for i, alignment := range n.Alignments {
indent2 := strings.Repeat(" ", level+1)
fmt.Printf("%s%s", indent2, alignment.String())
if i != len(n.Alignments)-1 {
fmt.Println("")
}
}
fmt.Printf("\n%s}\n", indent)
})
}
// KindTable is a NodeKind of the Table node.
var KindTable = gast.NewNodeKind("Table")
// Kind implements Node.Kind.
func (n *Table) Kind() gast.NodeKind {
return KindTable
}
// NewTable returns a new Table node.
func NewTable() *Table {
return &Table{
Alignments: []Alignment{},
}
}
// A TableRow struct represents a table row of Markdown(GFM) text.
type TableRow struct {
gast.BaseBlock
Alignments []Alignment
}
// Dump implements Node.Dump.
func (n *TableRow) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindTableRow is a NodeKind of the TableRow node.
var KindTableRow = gast.NewNodeKind("TableRow")
// Kind implements Node.Kind.
func (n *TableRow) Kind() gast.NodeKind {
return KindTableRow
}
// NewTableRow returns a new TableRow node.
func NewTableRow(alignments []Alignment) *TableRow {
return &TableRow{}
}
// A TableHeader struct represents a table header of Markdown(GFM) text.
type TableHeader struct {
gast.BaseBlock
Alignments []Alignment
}
// KindTableHeader is a NodeKind of the TableHeader node.
var KindTableHeader = gast.NewNodeKind("TableHeader")
// Kind implements Node.Kind.
func (n *TableHeader) Kind() gast.NodeKind {
return KindTableHeader
}
// Dump implements Node.Dump.
func (n *TableHeader) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// NewTableHeader returns a new TableHeader node.
func NewTableHeader(row *TableRow) *TableHeader {
n := &TableHeader{}
for c := row.FirstChild(); c != nil; {
next := c.NextSibling()
n.AppendChild(n, c)
c = next
}
return n
}
// A TableCell struct represents a table cell of a Markdown(GFM) text.
type TableCell struct {
gast.BaseBlock
Alignment Alignment
}
// Dump implements Node.Dump.
func (n *TableCell) Dump(source []byte, level int) {
gast.DumpHelper(n, source, level, nil, nil)
}
// KindTableCell is a NodeKind of the TableCell node.
var KindTableCell = gast.NewNodeKind("TableCell")
// Kind implements Node.Kind.
func (n *TableCell) Kind() gast.NodeKind {
return KindTableCell
}
// NewTableCell returns a new TableCell node.
func NewTableCell() *TableCell {
return &TableCell{
Alignment: AlignNone,
}
}

@ -0,0 +1,35 @@
package ast
import (
"fmt"
gast "github.com/yuin/goldmark/ast"
)
// A TaskCheckBox struct represents a checkbox of a task list.
type TaskCheckBox struct {
gast.BaseInline
IsChecked bool
}
// Dump impelemtns Node.Dump.
func (n *TaskCheckBox) Dump(source []byte, level int) {
m := map[string]string{
"Checked": fmt.Sprintf("%v", n.IsChecked),
}
gast.DumpHelper(n, source, level, m, nil)
}
// KindTaskCheckBox is a NodeKind of the TaskCheckBox node.
var KindTaskCheckBox = gast.NewNodeKind("TaskCheckBox")
// Kind implements Node.Kind.
func (n *TaskCheckBox) Kind() gast.NodeKind {
return KindTaskCheckBox
}
// NewTaskCheckBox returns a new TaskCheckBox node.
func NewTaskCheckBox(checked bool) *TaskCheckBox {
return &TaskCheckBox{
IsChecked: checked,
}
}

@ -0,0 +1,270 @@
package extension
import (
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type definitionListParser struct {
}
var defaultDefinitionListParser = &definitionListParser{}
// NewDefinitionListParser return a new parser.BlockParser that
// can parse PHP Markdown Extra Definition lists.
func NewDefinitionListParser() parser.BlockParser {
return defaultDefinitionListParser
}
func (b *definitionListParser) Trigger() []byte {
return []byte{':'}
}
func (b *definitionListParser) Open(parent gast.Node, reader text.Reader, pc parser.Context) (gast.Node, parser.State) {
if _, ok := parent.(*ast.DefinitionList); ok {
return nil, parser.NoChildren
}
line, _ := reader.PeekLine()
pos := pc.BlockOffset()
indent := pc.BlockIndent()
if pos < 0 || line[pos] != ':' || indent != 0 {
return nil, parser.NoChildren
}
last := parent.LastChild()
// need 1 or more spaces after ':'
w, _ := util.IndentWidth(line[pos+1:], pos+1)
if w < 1 {
return nil, parser.NoChildren
}
if w >= 8 { // starts with indented code
w = 5
}
w += pos + 1 /* 1 = ':' */
para, lastIsParagraph := last.(*gast.Paragraph)
var list *ast.DefinitionList
status := parser.HasChildren
var ok bool
if lastIsParagraph {
list, ok = last.PreviousSibling().(*ast.DefinitionList)
if ok { // is not first item
list.Offset = w
list.TemporaryParagraph = para
} else { // is first item
list = ast.NewDefinitionList(w, para)
status |= parser.RequireParagraph
}
} else if list, ok = last.(*ast.DefinitionList); ok { // multiple description
list.Offset = w
list.TemporaryParagraph = nil
} else {
return nil, parser.NoChildren
}
return list, status
}
func (b *definitionListParser) Continue(node gast.Node, reader text.Reader, pc parser.Context) parser.State {
line, _ := reader.PeekLine()
if util.IsBlank(line) {
return parser.Continue | parser.HasChildren
}
list, _ := node.(*ast.DefinitionList)
w, _ := util.IndentWidth(line, reader.LineOffset())
if w < list.Offset {
return parser.Close
}
pos, padding := util.IndentPosition(line, reader.LineOffset(), list.Offset)
reader.AdvanceAndSetPadding(pos, padding)
return parser.Continue | parser.HasChildren
}
func (b *definitionListParser) Close(node gast.Node, reader text.Reader, pc parser.Context) {
// nothing to do
}
func (b *definitionListParser) CanInterruptParagraph() bool {
return true
}
func (b *definitionListParser) CanAcceptIndentedLine() bool {
return false
}
type definitionDescriptionParser struct {
}
var defaultDefinitionDescriptionParser = &definitionDescriptionParser{}
// NewDefinitionDescriptionParser return a new parser.BlockParser that
// can parse definition description starts with ':'.
func NewDefinitionDescriptionParser() parser.BlockParser {
return defaultDefinitionDescriptionParser
}
func (b *definitionDescriptionParser) Trigger() []byte {
return []byte{':'}
}
func (b *definitionDescriptionParser) Open(parent gast.Node, reader text.Reader, pc parser.Context) (gast.Node, parser.State) {
line, _ := reader.PeekLine()
pos := pc.BlockOffset()
indent := pc.BlockIndent()
if pos < 0 || line[pos] != ':' || indent != 0 {
return nil, parser.NoChildren
}
list, _ := parent.(*ast.DefinitionList)
if list == nil {
return nil, parser.NoChildren
}
para := list.TemporaryParagraph
list.TemporaryParagraph = nil
if para != nil {
lines := para.Lines()
l := lines.Len()
for i := 0; i < l; i++ {
term := ast.NewDefinitionTerm()
segment := lines.At(i)
term.Lines().Append(segment.TrimRightSpace(reader.Source()))
list.AppendChild(list, term)
}
para.Parent().RemoveChild(para.Parent(), para)
}
cpos, padding := util.IndentPosition(line[pos+1:], pos+1, list.Offset-pos-1)
reader.AdvanceAndSetPadding(cpos, padding)
return ast.NewDefinitionDescription(), parser.HasChildren
}
func (b *definitionDescriptionParser) Continue(node gast.Node, reader text.Reader, pc parser.Context) parser.State {
// definitionListParser detects end of the description.
// so this method will never be called.
return parser.Continue | parser.HasChildren
}
func (b *definitionDescriptionParser) Close(node gast.Node, reader text.Reader, pc parser.Context) {
desc := node.(*ast.DefinitionDescription)
desc.IsTight = !desc.HasBlankPreviousLines()
if desc.IsTight {
for gc := desc.FirstChild(); gc != nil; gc = gc.NextSibling() {
paragraph, ok := gc.(*gast.Paragraph)
if ok {
textBlock := gast.NewTextBlock()
textBlock.SetLines(paragraph.Lines())
desc.ReplaceChild(desc, paragraph, textBlock)
}
}
}
}
func (b *definitionDescriptionParser) CanInterruptParagraph() bool {
return true
}
func (b *definitionDescriptionParser) CanAcceptIndentedLine() bool {
return false
}
// DefinitionListHTMLRenderer is a renderer.NodeRenderer implementation that
// renders DefinitionList nodes.
type DefinitionListHTMLRenderer struct {
html.Config
}
// NewDefinitionListHTMLRenderer returns a new DefinitionListHTMLRenderer.
func NewDefinitionListHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &DefinitionListHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *DefinitionListHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(ast.KindDefinitionList, r.renderDefinitionList)
reg.Register(ast.KindDefinitionTerm, r.renderDefinitionTerm)
reg.Register(ast.KindDefinitionDescription, r.renderDefinitionDescription)
}
// DefinitionListAttributeFilter defines attribute names which dl elements can have.
var DefinitionListAttributeFilter = html.GlobalAttributeFilter
func (r *DefinitionListHTMLRenderer) renderDefinitionList(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<dl")
html.RenderAttributes(w, n, DefinitionListAttributeFilter)
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("<dl>\n")
}
} else {
_, _ = w.WriteString("</dl>\n")
}
return gast.WalkContinue, nil
}
// DefinitionTermAttributeFilter defines attribute names which dd elements can have.
var DefinitionTermAttributeFilter = html.GlobalAttributeFilter
func (r *DefinitionListHTMLRenderer) renderDefinitionTerm(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<dt")
html.RenderAttributes(w, n, DefinitionTermAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<dt>")
}
} else {
_, _ = w.WriteString("</dt>\n")
}
return gast.WalkContinue, nil
}
// DefinitionDescriptionAttributeFilter defines attribute names which dd elements can have.
var DefinitionDescriptionAttributeFilter = html.GlobalAttributeFilter
func (r *DefinitionListHTMLRenderer) renderDefinitionDescription(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
n := node.(*ast.DefinitionDescription)
_, _ = w.WriteString("<dd")
if n.Attributes() != nil {
html.RenderAttributes(w, n, DefinitionDescriptionAttributeFilter)
}
if n.IsTight {
_, _ = w.WriteString(">")
} else {
_, _ = w.WriteString(">\n")
}
} else {
_, _ = w.WriteString("</dd>\n")
}
return gast.WalkContinue, nil
}
type definitionList struct {
}
// DefinitionList is an extension that allow you to use PHP Markdown Extra Definition lists.
var DefinitionList = &definitionList{}
func (e *definitionList) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(parser.WithBlockParsers(
util.Prioritized(NewDefinitionListParser(), 101),
util.Prioritized(NewDefinitionDescriptionParser(), 102),
))
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewDefinitionListHTMLRenderer(), 500),
))
}

@ -0,0 +1,336 @@
package extension
import (
"bytes"
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"strconv"
)
var footnoteListKey = parser.NewContextKey()
type footnoteBlockParser struct {
}
var defaultFootnoteBlockParser = &footnoteBlockParser{}
// NewFootnoteBlockParser returns a new parser.BlockParser that can parse
// footnotes of the Markdown(PHP Markdown Extra) text.
func NewFootnoteBlockParser() parser.BlockParser {
return defaultFootnoteBlockParser
}
func (b *footnoteBlockParser) Trigger() []byte {
return []byte{'['}
}
func (b *footnoteBlockParser) Open(parent gast.Node, reader text.Reader, pc parser.Context) (gast.Node, parser.State) {
line, segment := reader.PeekLine()
pos := pc.BlockOffset()
if pos < 0 || line[pos] != '[' {
return nil, parser.NoChildren
}
pos++
if pos > len(line)-1 || line[pos] != '^' {
return nil, parser.NoChildren
}
open := pos + 1
closes := 0
closure := util.FindClosure(line[pos+1:], '[', ']', false, false)
closes = pos + 1 + closure
next := closes + 1
if closure > -1 {
if next >= len(line) || line[next] != ':' {
return nil, parser.NoChildren
}
} else {
return nil, parser.NoChildren
}
padding := segment.Padding
label := reader.Value(text.NewSegment(segment.Start+open-padding, segment.Start+closes-padding))
if util.IsBlank(label) {
return nil, parser.NoChildren
}
item := ast.NewFootnote(label)
pos = next + 1 - padding
if pos >= len(line) {
reader.Advance(pos)
return item, parser.NoChildren
}
reader.AdvanceAndSetPadding(pos, padding)
return item, parser.HasChildren
}
func (b *footnoteBlockParser) Continue(node gast.Node, reader text.Reader, pc parser.Context) parser.State {
line, _ := reader.PeekLine()
if util.IsBlank(line) {
return parser.Continue | parser.HasChildren
}
childpos, padding := util.IndentPosition(line, reader.LineOffset(), 4)
if childpos < 0 {
return parser.Close
}
reader.AdvanceAndSetPadding(childpos, padding)
return parser.Continue | parser.HasChildren
}
func (b *footnoteBlockParser) Close(node gast.Node, reader text.Reader, pc parser.Context) {
var list *ast.FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*ast.FootnoteList)
} else {
list = ast.NewFootnoteList()
pc.Set(footnoteListKey, list)
node.Parent().InsertBefore(node.Parent(), node, list)
}
node.Parent().RemoveChild(node.Parent(), node)
list.AppendChild(list, node)
}
func (b *footnoteBlockParser) CanInterruptParagraph() bool {
return true
}
func (b *footnoteBlockParser) CanAcceptIndentedLine() bool {
return false
}
type footnoteParser struct {
}
var defaultFootnoteParser = &footnoteParser{}
// NewFootnoteParser returns a new parser.InlineParser that can parse
// footnote links of the Markdown(PHP Markdown Extra) text.
func NewFootnoteParser() parser.InlineParser {
return defaultFootnoteParser
}
func (s *footnoteParser) Trigger() []byte {
// footnote syntax probably conflict with the image syntax.
// So we need trigger this parser with '!'.
return []byte{'!', '['}
}
func (s *footnoteParser) Parse(parent gast.Node, block text.Reader, pc parser.Context) gast.Node {
line, segment := block.PeekLine()
pos := 1
if len(line) > 0 && line[0] == '!' {
pos++
}
if pos >= len(line) || line[pos] != '^' {
return nil
}
pos++
if pos >= len(line) {
return nil
}
open := pos
closure := util.FindClosure(line[pos:], '[', ']', false, false)
if closure < 0 {
return nil
}
closes := pos + closure
value := block.Value(text.NewSegment(segment.Start+open, segment.Start+closes))
block.Advance(closes + 1)
var list *ast.FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*ast.FootnoteList)
}
if list == nil {
return nil
}
index := 0
for def := list.FirstChild(); def != nil; def = def.NextSibling() {
d := def.(*ast.Footnote)
if bytes.Equal(d.Ref, value) {
if d.Index < 0 {
list.Count += 1
d.Index = list.Count
}
index = d.Index
break
}
}
if index == 0 {
return nil
}
return ast.NewFootnoteLink(index)
}
type footnoteASTTransformer struct {
}
var defaultFootnoteASTTransformer = &footnoteASTTransformer{}
// NewFootnoteASTTransformer returns a new parser.ASTTransformer that
// insert a footnote list to the last of the document.
func NewFootnoteASTTransformer() parser.ASTTransformer {
return defaultFootnoteASTTransformer
}
func (a *footnoteASTTransformer) Transform(node *gast.Document, reader text.Reader, pc parser.Context) {
var list *ast.FootnoteList
if tlist := pc.Get(footnoteListKey); tlist != nil {
list = tlist.(*ast.FootnoteList)
} else {
return
}
pc.Set(footnoteListKey, nil)
for footnote := list.FirstChild(); footnote != nil; {
var container gast.Node = footnote
next := footnote.NextSibling()
if fc := container.LastChild(); fc != nil && gast.IsParagraph(fc) {
container = fc
}
index := footnote.(*ast.Footnote).Index
if index < 0 {
list.RemoveChild(list, footnote)
} else {
container.AppendChild(container, ast.NewFootnoteBackLink(index))
}
footnote = next
}
list.SortChildren(func(n1, n2 gast.Node) int {
if n1.(*ast.Footnote).Index < n2.(*ast.Footnote).Index {
return -1
}
return 1
})
if list.Count <= 0 {
list.Parent().RemoveChild(list.Parent(), list)
return
}
node.AppendChild(node, list)
}
// FootnoteHTMLRenderer is a renderer.NodeRenderer implementation that
// renders FootnoteLink nodes.
type FootnoteHTMLRenderer struct {
html.Config
}
// NewFootnoteHTMLRenderer returns a new FootnoteHTMLRenderer.
func NewFootnoteHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &FootnoteHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *FootnoteHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(ast.KindFootnoteLink, r.renderFootnoteLink)
reg.Register(ast.KindFootnoteBackLink, r.renderFootnoteBackLink)
reg.Register(ast.KindFootnote, r.renderFootnote)
reg.Register(ast.KindFootnoteList, r.renderFootnoteList)
}
func (r *FootnoteHTMLRenderer) renderFootnoteLink(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
n := node.(*ast.FootnoteLink)
is := strconv.Itoa(n.Index)
_, _ = w.WriteString(`<sup id="fnref:`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`"><a href="#fn:`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`" class="footnote-ref" role="doc-noteref">`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`</a></sup>`)
}
return gast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnoteBackLink(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
n := node.(*ast.FootnoteBackLink)
is := strconv.Itoa(n.Index)
_, _ = w.WriteString(` <a href="#fnref:`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`" class="footnote-backref" role="doc-backlink">`)
_, _ = w.WriteString("&#x21a9;&#xfe0e;")
_, _ = w.WriteString(`</a>`)
}
return gast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnote(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
n := node.(*ast.Footnote)
is := strconv.Itoa(n.Index)
if entering {
_, _ = w.WriteString(`<li id="fn:`)
_, _ = w.WriteString(is)
_, _ = w.WriteString(`" role="doc-endnote"`)
if node.Attributes() != nil {
html.RenderAttributes(w, node, html.ListItemAttributeFilter)
}
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("</li>\n")
}
return gast.WalkContinue, nil
}
func (r *FootnoteHTMLRenderer) renderFootnoteList(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
tag := "section"
if r.Config.XHTML {
tag = "div"
}
if entering {
_, _ = w.WriteString("<")
_, _ = w.WriteString(tag)
_, _ = w.WriteString(` class="footnotes" role="doc-endnotes"`)
if node.Attributes() != nil {
html.RenderAttributes(w, node, html.GlobalAttributeFilter)
}
_ = w.WriteByte('>')
if r.Config.XHTML {
_, _ = w.WriteString("\n<hr />\n")
} else {
_, _ = w.WriteString("\n<hr>\n")
}
_, _ = w.WriteString("<ol>\n")
} else {
_, _ = w.WriteString("</ol>\n")
_, _ = w.WriteString("</")
_, _ = w.WriteString(tag)
_, _ = w.WriteString(">\n")
}
return gast.WalkContinue, nil
}
type footnote struct {
}
// Footnote is an extension that allow you to use PHP Markdown Extra Footnotes.
var Footnote = &footnote{}
func (e *footnote) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(
parser.WithBlockParsers(
util.Prioritized(NewFootnoteBlockParser(), 999),
),
parser.WithInlineParsers(
util.Prioritized(NewFootnoteParser(), 101),
),
parser.WithASTTransformers(
util.Prioritized(NewFootnoteASTTransformer(), 999),
),
)
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewFootnoteHTMLRenderer(), 500),
))
}

@ -0,0 +1,18 @@
package extension
import (
"github.com/yuin/goldmark"
)
type gfm struct {
}
// GFM is an extension that provides Github Flavored markdown functionalities.
var GFM = &gfm{}
func (e *gfm) Extend(m goldmark.Markdown) {
Linkify.Extend(m)
Table.Extend(m)
Strikethrough.Extend(m)
TaskList.Extend(m)
}

@ -0,0 +1,150 @@
package extension
import (
"bytes"
"github.com/yuin/goldmark"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"regexp"
)
var wwwURLRegxp = regexp.MustCompile(`^www\.[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}((?:/|[#?])[-a-zA-Z0-9@:%_\+.~#!?&//=\(\);,'">\^{}\[\]` + "`" + `]*)?`)
var urlRegexp = regexp.MustCompile(`^(?:http|https|ftp):\/\/(?:www\.)?[-a-zA-Z0-9@:%._\+~#=]{2,256}\.[a-z]{2,6}((?:/|[#?])[-a-zA-Z0-9@:%_+.~#$!?&//=\(\);,'">\^{}\[\]` + "`" + `]*)?`)
type linkifyParser struct {
}
var defaultLinkifyParser = &linkifyParser{}
// NewLinkifyParser return a new InlineParser can parse
// text that seems like a URL.
func NewLinkifyParser() parser.InlineParser {
return defaultLinkifyParser
}
func (s *linkifyParser) Trigger() []byte {
// ' ' indicates any white spaces and a line head
return []byte{' ', '*', '_', '~', '('}
}
var protoHTTP = []byte("http:")
var protoHTTPS = []byte("https:")
var protoFTP = []byte("ftp:")
var domainWWW = []byte("www.")
func (s *linkifyParser) Parse(parent ast.Node, block text.Reader, pc parser.Context) ast.Node {
if pc.IsInLinkLabel() {
return nil
}
line, segment := block.PeekLine()
consumes := 0
start := segment.Start
c := line[0]
// advance if current position is not a line head.
if c == ' ' || c == '*' || c == '_' || c == '~' || c == '(' {
consumes++
start++
line = line[1:]
}
var m []int
var protocol []byte
var typ ast.AutoLinkType = ast.AutoLinkURL
if bytes.HasPrefix(line, protoHTTP) || bytes.HasPrefix(line, protoHTTPS) || bytes.HasPrefix(line, protoFTP) {
m = urlRegexp.FindSubmatchIndex(line)
}
if m == nil && bytes.HasPrefix(line, domainWWW) {
m = wwwURLRegxp.FindSubmatchIndex(line)
protocol = []byte("http")
}
if m != nil {
lastChar := line[m[1]-1]
if lastChar == '.' {
m[1]--
} else if lastChar == ')' {
closing := 0
for i := m[1] - 1; i >= m[0]; i-- {
if line[i] == ')' {
closing++
} else if line[i] == '(' {
closing--
}
}
if closing > 0 {
m[1] -= closing
}
} else if lastChar == ';' {
i := m[1] - 2
for ; i >= m[0]; i-- {
if util.IsAlphaNumeric(line[i]) {
continue
}
break
}
if i != m[1]-2 {
if line[i] == '&' {
m[1] -= m[1] - i
}
}
}
}
if m == nil {
if len(line) > 0 && util.IsPunct(line[0]) {
return nil
}
typ = ast.AutoLinkEmail
stop := util.FindEmailIndex(line)
if stop < 0 {
return nil
}
at := bytes.IndexByte(line, '@')
m = []int{0, stop, at, stop - 1}
if m == nil || bytes.IndexByte(line[m[2]:m[3]], '.') < 0 {
return nil
}
lastChar := line[m[1]-1]
if lastChar == '.' {
m[1]--
}
if m[1] < len(line) {
nextChar := line[m[1]]
if nextChar == '-' || nextChar == '_' {
return nil
}
}
}
if m == nil {
return nil
}
if consumes != 0 {
s := segment.WithStop(segment.Start + 1)
ast.MergeOrAppendTextSegment(parent, s)
}
consumes += m[1]
block.Advance(consumes)
n := ast.NewTextSegment(text.NewSegment(start, start+m[1]))
link := ast.NewAutoLink(typ, n)
link.Protocol = protocol
return link
}
func (s *linkifyParser) CloseBlock(parent ast.Node, pc parser.Context) {
// nothing to do
}
type linkify struct {
}
// Linkify is an extension that allow you to parse text that seems like a URL.
var Linkify = &linkify{}
func (e *linkify) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(
parser.WithInlineParsers(
util.Prioritized(NewLinkifyParser(), 999),
),
)
}

@ -0,0 +1,116 @@
package extension
import (
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type strikethroughDelimiterProcessor struct {
}
func (p *strikethroughDelimiterProcessor) IsDelimiter(b byte) bool {
return b == '~'
}
func (p *strikethroughDelimiterProcessor) CanOpenCloser(opener, closer *parser.Delimiter) bool {
return opener.Char == closer.Char
}
func (p *strikethroughDelimiterProcessor) OnMatch(consumes int) gast.Node {
return ast.NewStrikethrough()
}
var defaultStrikethroughDelimiterProcessor = &strikethroughDelimiterProcessor{}
type strikethroughParser struct {
}
var defaultStrikethroughParser = &strikethroughParser{}
// NewStrikethroughParser return a new InlineParser that parses
// strikethrough expressions.
func NewStrikethroughParser() parser.InlineParser {
return defaultStrikethroughParser
}
func (s *strikethroughParser) Trigger() []byte {
return []byte{'~'}
}
func (s *strikethroughParser) Parse(parent gast.Node, block text.Reader, pc parser.Context) gast.Node {
before := block.PrecendingCharacter()
line, segment := block.PeekLine()
node := parser.ScanDelimiter(line, before, 2, defaultStrikethroughDelimiterProcessor)
if node == nil {
return nil
}
node.Segment = segment.WithStop(segment.Start + node.OriginalLength)
block.Advance(node.OriginalLength)
pc.PushDelimiter(node)
return node
}
func (s *strikethroughParser) CloseBlock(parent gast.Node, pc parser.Context) {
// nothing to do
}
// StrikethroughHTMLRenderer is a renderer.NodeRenderer implementation that
// renders Strikethrough nodes.
type StrikethroughHTMLRenderer struct {
html.Config
}
// NewStrikethroughHTMLRenderer returns a new StrikethroughHTMLRenderer.
func NewStrikethroughHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &StrikethroughHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *StrikethroughHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(ast.KindStrikethrough, r.renderStrikethrough)
}
// StrikethroughAttributeFilter defines attribute names which dd elements can have.
var StrikethroughAttributeFilter = html.GlobalAttributeFilter
func (r *StrikethroughHTMLRenderer) renderStrikethrough(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<del")
html.RenderAttributes(w, n, StrikethroughAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<del>")
}
} else {
_, _ = w.WriteString("</del>")
}
return gast.WalkContinue, nil
}
type strikethrough struct {
}
// Strikethrough is an extension that allow you to use strikethrough expression like '~~text~~' .
var Strikethrough = &strikethrough{}
func (e *strikethrough) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(parser.WithInlineParsers(
util.Prioritized(NewStrikethroughParser(), 500),
))
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewStrikethroughHTMLRenderer(), 500),
))
}

@ -0,0 +1,319 @@
package extension
import (
"bytes"
"fmt"
"regexp"
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var tableDelimRegexp = regexp.MustCompile(`^[\s\-\|\:]+$`)
var tableDelimLeft = regexp.MustCompile(`^\s*\:\-+\s*$`)
var tableDelimRight = regexp.MustCompile(`^\s*\-+\:\s*$`)
var tableDelimCenter = regexp.MustCompile(`^\s*\:\-+\:\s*$`)
var tableDelimNone = regexp.MustCompile(`^\s*\-+\s*$`)
type tableParagraphTransformer struct {
}
var defaultTableParagraphTransformer = &tableParagraphTransformer{}
// NewTableParagraphTransformer returns a new ParagraphTransformer
// that can transform pargraphs into tables.
func NewTableParagraphTransformer() parser.ParagraphTransformer {
return defaultTableParagraphTransformer
}
func (b *tableParagraphTransformer) Transform(node *gast.Paragraph, reader text.Reader, pc parser.Context) {
lines := node.Lines()
if lines.Len() < 2 {
return
}
alignments := b.parseDelimiter(lines.At(1), reader)
if alignments == nil {
return
}
header := b.parseRow(lines.At(0), alignments, true, reader)
if header == nil || len(alignments) != header.ChildCount() {
return
}
table := ast.NewTable()
table.Alignments = alignments
table.AppendChild(table, ast.NewTableHeader(header))
for i := 2; i < lines.Len(); i++ {
table.AppendChild(table, b.parseRow(lines.At(i), alignments, false, reader))
}
node.Parent().InsertBefore(node.Parent(), node, table)
node.Parent().RemoveChild(node.Parent(), node)
}
func (b *tableParagraphTransformer) parseRow(segment text.Segment, alignments []ast.Alignment, isHeader bool, reader text.Reader) *ast.TableRow {
source := reader.Source()
line := segment.Value(source)
pos := 0
pos += util.TrimLeftSpaceLength(line)
limit := len(line)
limit -= util.TrimRightSpaceLength(line)
row := ast.NewTableRow(alignments)
if len(line) > 0 && line[pos] == '|' {
pos++
}
if len(line) > 0 && line[limit-1] == '|' {
limit--
}
i := 0
for ; pos < limit; i++ {
alignment := ast.AlignNone
if i >= len(alignments) {
if !isHeader {
return row
}
} else {
alignment = alignments[i]
}
closure := util.FindClosure(line[pos:], byte(0), '|', true, false)
if closure < 0 {
closure = len(line[pos:])
}
node := ast.NewTableCell()
seg := text.NewSegment(segment.Start+pos, segment.Start+pos+closure)
seg = seg.TrimLeftSpace(source)
seg = seg.TrimRightSpace(source)
node.Lines().Append(seg)
node.Alignment = alignment
row.AppendChild(row, node)
pos += closure + 1
}
for ; i < len(alignments); i++ {
row.AppendChild(row, ast.NewTableCell())
}
return row
}
func (b *tableParagraphTransformer) parseDelimiter(segment text.Segment, reader text.Reader) []ast.Alignment {
line := segment.Value(reader.Source())
if !tableDelimRegexp.Match(line) {
return nil
}
cols := bytes.Split(line, []byte{'|'})
if util.IsBlank(cols[0]) {
cols = cols[1:]
}
if len(cols) > 0 && util.IsBlank(cols[len(cols)-1]) {
cols = cols[:len(cols)-1]
}
var alignments []ast.Alignment
for _, col := range cols {
if tableDelimLeft.Match(col) {
alignments = append(alignments, ast.AlignLeft)
} else if tableDelimRight.Match(col) {
alignments = append(alignments, ast.AlignRight)
} else if tableDelimCenter.Match(col) {
alignments = append(alignments, ast.AlignCenter)
} else if tableDelimNone.Match(col) {
alignments = append(alignments, ast.AlignNone)
} else {
return nil
}
}
return alignments
}
// TableHTMLRenderer is a renderer.NodeRenderer implementation that
// renders Table nodes.
type TableHTMLRenderer struct {
html.Config
}
// NewTableHTMLRenderer returns a new TableHTMLRenderer.
func NewTableHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &TableHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *TableHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(ast.KindTable, r.renderTable)
reg.Register(ast.KindTableHeader, r.renderTableHeader)
reg.Register(ast.KindTableRow, r.renderTableRow)
reg.Register(ast.KindTableCell, r.renderTableCell)
}
// TableAttributeFilter defines attribute names which table elements can have.
var TableAttributeFilter = html.GlobalAttributeFilter.Extend(
[]byte("align"), // [Deprecated]
[]byte("bgcolor"), // [Deprecated]
[]byte("border"), // [Deprecated]
[]byte("cellpadding"), // [Deprecated]
[]byte("cellspacing"), // [Deprecated]
[]byte("frame"), // [Deprecated]
[]byte("rules"), // [Deprecated]
[]byte("summary"), // [Deprecated]
[]byte("width"), // [Deprecated]
)
func (r *TableHTMLRenderer) renderTable(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
_, _ = w.WriteString("<table")
if n.Attributes() != nil {
html.RenderAttributes(w, n, TableAttributeFilter)
}
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("</table>\n")
}
return gast.WalkContinue, nil
}
// TableHeaderAttributeFilter defines attribute names which <thead> elements can have.
var TableHeaderAttributeFilter = html.GlobalAttributeFilter.Extend(
[]byte("align"), // [Deprecated since HTML4] [Obsolete since HTML5]
[]byte("bgcolor"), // [Not Standardized]
[]byte("char"), // [Deprecated since HTML4] [Obsolete since HTML5]
[]byte("charoff"), // [Deprecated since HTML4] [Obsolete since HTML5]
[]byte("valign"), // [Deprecated since HTML4] [Obsolete since HTML5]
)
func (r *TableHTMLRenderer) renderTableHeader(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
_, _ = w.WriteString("<thead")
if n.Attributes() != nil {
html.RenderAttributes(w, n, TableHeaderAttributeFilter)
}
_, _ = w.WriteString(">\n")
_, _ = w.WriteString("<tr>\n") // Header <tr> has no separate handle
} else {
_, _ = w.WriteString("</tr>\n")
_, _ = w.WriteString("</thead>\n")
if n.NextSibling() != nil {
_, _ = w.WriteString("<tbody>\n")
}
}
return gast.WalkContinue, nil
}
// TableRowAttributeFilter defines attribute names which <tr> elements can have.
var TableRowAttributeFilter = html.GlobalAttributeFilter.Extend(
[]byte("align"), // [Obsolete since HTML5]
[]byte("bgcolor"), // [Obsolete since HTML5]
[]byte("char"), // [Obsolete since HTML5]
[]byte("charoff"), // [Obsolete since HTML5]
[]byte("valign"), // [Obsolete since HTML5]
)
func (r *TableHTMLRenderer) renderTableRow(w util.BufWriter, source []byte, n gast.Node, entering bool) (gast.WalkStatus, error) {
if entering {
_, _ = w.WriteString("<tr")
if n.Attributes() != nil {
html.RenderAttributes(w, n, TableRowAttributeFilter)
}
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("</tr>\n")
if n.Parent().LastChild() == n {
_, _ = w.WriteString("</tbody>\n")
}
}
return gast.WalkContinue, nil
}
// TableThCellAttributeFilter defines attribute names which table <th> cells can have.
var TableThCellAttributeFilter = html.GlobalAttributeFilter.Extend(
[]byte("abbr"), // [OK] Contains a short abbreviated description of the cell's content [NOT OK in <td>]
[]byte("align"), // [Obsolete since HTML5]
[]byte("axis"), // [Obsolete since HTML5]
[]byte("bgcolor"), // [Not Standardized]
[]byte("char"), // [Obsolete since HTML5]
[]byte("charoff"), // [Obsolete since HTML5]
[]byte("colspan"), // [OK] Number of columns that the cell is to span
[]byte("headers"), // [OK] This attribute contains a list of space-separated strings, each corresponding to the id attribute of the <th> elements that apply to this element
[]byte("height"), // [Deprecated since HTML4] [Obsolete since HTML5]
[]byte("rowspan"), // [OK] Number of rows that the cell is to span
[]byte("scope"), // [OK] This enumerated attribute defines the cells that the header (defined in the <th>) element relates to [NOT OK in <td>]
[]byte("valign"), // [Obsolete since HTML5]
[]byte("width"), // [Deprecated since HTML4] [Obsolete since HTML5]
)
// TableTdCellAttributeFilter defines attribute names which table <td> cells can have.
var TableTdCellAttributeFilter = html.GlobalAttributeFilter.Extend(
[]byte("abbr"), // [Obsolete since HTML5] [OK in <th>]
[]byte("align"), // [Obsolete since HTML5]
[]byte("axis"), // [Obsolete since HTML5]
[]byte("bgcolor"), // [Not Standardized]
[]byte("char"), // [Obsolete since HTML5]
[]byte("charoff"), // [Obsolete since HTML5]
[]byte("colspan"), // [OK] Number of columns that the cell is to span
[]byte("headers"), // [OK] This attribute contains a list of space-separated strings, each corresponding to the id attribute of the <th> elements that apply to this element
[]byte("height"), // [Deprecated since HTML4] [Obsolete since HTML5]
[]byte("rowspan"), // [OK] Number of rows that the cell is to span
[]byte("scope"), // [Obsolete since HTML5] [OK in <th>]
[]byte("valign"), // [Obsolete since HTML5]
[]byte("width"), // [Deprecated since HTML4] [Obsolete since HTML5]
)
func (r *TableHTMLRenderer) renderTableCell(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
n := node.(*ast.TableCell)
tag := "td"
if n.Parent().Kind() == ast.KindTableHeader {
tag = "th"
}
if entering {
align := ""
if n.Alignment != ast.AlignNone {
if _, ok := n.AttributeString("align"); !ok { // Skip align render if overridden
// TODO: "align" is deprecated. style="text-align:%s" instead?
align = fmt.Sprintf(` align="%s"`, n.Alignment.String())
}
}
fmt.Fprintf(w, "<%s", tag)
if n.Attributes() != nil {
if tag == "td" {
html.RenderAttributes(w, n, TableTdCellAttributeFilter) // <td>
} else {
html.RenderAttributes(w, n, TableThCellAttributeFilter) // <th>
}
}
fmt.Fprintf(w, "%s>", align)
} else {
fmt.Fprintf(w, "</%s>\n", tag)
}
return gast.WalkContinue, nil
}
type table struct {
}
// Table is an extension that allow you to use GFM tables .
var Table = &table{}
func (e *table) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(parser.WithParagraphTransformers(
util.Prioritized(NewTableParagraphTransformer(), 200),
))
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewTableHTMLRenderer(), 500),
))
}

@ -0,0 +1,115 @@
package extension
import (
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/extension/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"regexp"
)
var taskListRegexp = regexp.MustCompile(`^\[([\sxX])\]\s*`)
type taskCheckBoxParser struct {
}
var defaultTaskCheckBoxParser = &taskCheckBoxParser{}
// NewTaskCheckBoxParser returns a new InlineParser that can parse
// checkboxes in list items.
// This parser must take precedence over the parser.LinkParser.
func NewTaskCheckBoxParser() parser.InlineParser {
return defaultTaskCheckBoxParser
}
func (s *taskCheckBoxParser) Trigger() []byte {
return []byte{'['}
}
func (s *taskCheckBoxParser) Parse(parent gast.Node, block text.Reader, pc parser.Context) gast.Node {
// Given AST structure must be like
// - List
// - ListItem : parent.Parent
// - TextBlock : parent
// (current line)
if parent.Parent() == nil || parent.Parent().FirstChild() != parent {
return nil
}
if _, ok := parent.Parent().(*gast.ListItem); !ok {
return nil
}
line, _ := block.PeekLine()
m := taskListRegexp.FindSubmatchIndex(line)
if m == nil {
return nil
}
value := line[m[2]:m[3]][0]
block.Advance(m[1])
checked := value == 'x' || value == 'X'
return ast.NewTaskCheckBox(checked)
}
func (s *taskCheckBoxParser) CloseBlock(parent gast.Node, pc parser.Context) {
// nothing to do
}
// TaskCheckBoxHTMLRenderer is a renderer.NodeRenderer implementation that
// renders checkboxes in list items.
type TaskCheckBoxHTMLRenderer struct {
html.Config
}
// NewTaskCheckBoxHTMLRenderer returns a new TaskCheckBoxHTMLRenderer.
func NewTaskCheckBoxHTMLRenderer(opts ...html.Option) renderer.NodeRenderer {
r := &TaskCheckBoxHTMLRenderer{
Config: html.NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements renderer.NodeRenderer.RegisterFuncs.
func (r *TaskCheckBoxHTMLRenderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
reg.Register(ast.KindTaskCheckBox, r.renderTaskCheckBox)
}
func (r *TaskCheckBoxHTMLRenderer) renderTaskCheckBox(w util.BufWriter, source []byte, node gast.Node, entering bool) (gast.WalkStatus, error) {
if !entering {
return gast.WalkContinue, nil
}
n := node.(*ast.TaskCheckBox)
if n.IsChecked {
w.WriteString(`<input checked="" disabled="" type="checkbox"`)
} else {
w.WriteString(`<input disabled="" type="checkbox"`)
}
if r.XHTML {
w.WriteString(" />")
} else {
w.WriteString(">")
}
return gast.WalkContinue, nil
}
type taskList struct {
}
// TaskList is an extension that allow you to use GFM task lists.
var TaskList = &taskList{}
func (e *taskList) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(parser.WithInlineParsers(
util.Prioritized(NewTaskCheckBoxParser(), 0),
))
m.Renderer().AddOptions(renderer.WithNodeRenderers(
util.Prioritized(NewTaskCheckBoxHTMLRenderer(), 500),
))
}

@ -0,0 +1,245 @@
package extension
import (
"github.com/yuin/goldmark"
gast "github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// TypographicPunctuation is a key of the punctuations that can be replaced with
// typographic entities.
type TypographicPunctuation int
const (
// LeftSingleQuote is '
LeftSingleQuote TypographicPunctuation = iota + 1
// RightSingleQuote is '
RightSingleQuote
// LeftDoubleQuote is "
LeftDoubleQuote
// RightDoubleQuote is "
RightDoubleQuote
// EnDash is --
EnDash
// EmDash is ---
EmDash
// Ellipsis is ...
Ellipsis
// LeftAngleQuote is <<
LeftAngleQuote
// RightAngleQuote is >>
RightAngleQuote
typographicPunctuationMax
)
// An TypographerConfig struct is a data structure that holds configuration of the
// Typographer extension.
type TypographerConfig struct {
Substitutions [][]byte
}
func newDefaultSubstitutions() [][]byte {
replacements := make([][]byte, typographicPunctuationMax)
replacements[LeftSingleQuote] = []byte("&lsquo;")
replacements[RightSingleQuote] = []byte("&rsquo;")
replacements[LeftDoubleQuote] = []byte("&ldquo;")
replacements[RightDoubleQuote] = []byte("&rdquo;")
replacements[EnDash] = []byte("&ndash;")
replacements[EmDash] = []byte("&mdash;")
replacements[Ellipsis] = []byte("&hellip;")
replacements[LeftAngleQuote] = []byte("&laquo;")
replacements[RightAngleQuote] = []byte("&raquo;")
return replacements
}
// SetOption implements SetOptioner.
func (b *TypographerConfig) SetOption(name parser.OptionName, value interface{}) {
switch name {
case optTypographicSubstitutions:
b.Substitutions = value.([][]byte)
}
}
// A TypographerOption interface sets options for the TypographerParser.
type TypographerOption interface {
parser.Option
SetTypographerOption(*TypographerConfig)
}
const optTypographicSubstitutions parser.OptionName = "TypographicSubstitutions"
// TypographicSubstitutions is a list of the substitutions for the Typographer extension.
type TypographicSubstitutions map[TypographicPunctuation][]byte
type withTypographicSubstitutions struct {
value [][]byte
}
func (o *withTypographicSubstitutions) SetParserOption(c *parser.Config) {
c.Options[optTypographicSubstitutions] = o.value
}
func (o *withTypographicSubstitutions) SetTypographerOption(p *TypographerConfig) {
p.Substitutions = o.value
}
// WithTypographicSubstitutions is a functional otpion that specify replacement text
// for punctuations.
func WithTypographicSubstitutions(values map[TypographicPunctuation][]byte) TypographerOption {
replacements := newDefaultSubstitutions()
for k, v := range values {
replacements[k] = v
}
return &withTypographicSubstitutions{replacements}
}
type typographerDelimiterProcessor struct {
}
func (p *typographerDelimiterProcessor) IsDelimiter(b byte) bool {
return b == '\'' || b == '"'
}
func (p *typographerDelimiterProcessor) CanOpenCloser(opener, closer *parser.Delimiter) bool {
return opener.Char == closer.Char
}
func (p *typographerDelimiterProcessor) OnMatch(consumes int) gast.Node {
return nil
}
var defaultTypographerDelimiterProcessor = &typographerDelimiterProcessor{}
type typographerParser struct {
TypographerConfig
}
// NewTypographerParser return a new InlineParser that parses
// typographer expressions.
func NewTypographerParser(opts ...TypographerOption) parser.InlineParser {
p := &typographerParser{
TypographerConfig: TypographerConfig{
Substitutions: newDefaultSubstitutions(),
},
}
for _, o := range opts {
o.SetTypographerOption(&p.TypographerConfig)
}
return p
}
func (s *typographerParser) Trigger() []byte {
return []byte{'\'', '"', '-', '.', '<', '>'}
}
func (s *typographerParser) Parse(parent gast.Node, block text.Reader, pc parser.Context) gast.Node {
before := block.PrecendingCharacter()
line, _ := block.PeekLine()
c := line[0]
if len(line) > 2 {
if c == '-' {
if s.Substitutions[EmDash] != nil && line[1] == '-' && line[2] == '-' { // ---
node := gast.NewString(s.Substitutions[EmDash])
node.SetCode(true)
block.Advance(3)
return node
}
} else if c == '.' {
if s.Substitutions[Ellipsis] != nil && line[1] == '.' && line[2] == '.' { // ...
node := gast.NewString(s.Substitutions[Ellipsis])
node.SetCode(true)
block.Advance(3)
return node
}
return nil
}
}
if len(line) > 1 {
if c == '<' {
if s.Substitutions[LeftAngleQuote] != nil && line[1] == '<' { // <<
node := gast.NewString(s.Substitutions[LeftAngleQuote])
node.SetCode(true)
block.Advance(2)
return node
}
return nil
} else if c == '>' {
if s.Substitutions[RightAngleQuote] != nil && line[1] == '>' { // >>
node := gast.NewString(s.Substitutions[RightAngleQuote])
node.SetCode(true)
block.Advance(2)
return node
}
return nil
} else if s.Substitutions[EnDash] != nil && c == '-' && line[1] == '-' { // --
node := gast.NewString(s.Substitutions[EnDash])
node.SetCode(true)
block.Advance(2)
return node
}
}
if c == '\'' || c == '"' {
d := parser.ScanDelimiter(line, before, 1, defaultTypographerDelimiterProcessor)
if d == nil {
return nil
}
if c == '\'' {
if s.Substitutions[LeftSingleQuote] != nil && d.CanOpen && !d.CanClose {
node := gast.NewString(s.Substitutions[LeftSingleQuote])
node.SetCode(true)
block.Advance(1)
return node
}
if s.Substitutions[RightSingleQuote] != nil && d.CanClose && !d.CanOpen {
node := gast.NewString(s.Substitutions[RightSingleQuote])
node.SetCode(true)
block.Advance(1)
return node
}
}
if c == '"' {
if s.Substitutions[LeftDoubleQuote] != nil && d.CanOpen && !d.CanClose {
node := gast.NewString(s.Substitutions[LeftDoubleQuote])
node.SetCode(true)
block.Advance(1)
return node
}
if s.Substitutions[RightDoubleQuote] != nil && d.CanClose && !d.CanOpen {
node := gast.NewString(s.Substitutions[RightDoubleQuote])
node.SetCode(true)
block.Advance(1)
return node
}
}
}
return nil
}
func (s *typographerParser) CloseBlock(parent gast.Node, pc parser.Context) {
// nothing to do
}
type typographer struct {
options []TypographerOption
}
// Typographer is an extension that repalace punctuations with typographic entities.
var Typographer = &typographer{}
// NewTypographer returns a new Entender that repalace punctuations with typographic entities.
func NewTypographer(opts ...TypographerOption) goldmark.Extender {
return &typographer{
options: opts,
}
}
func (e *typographer) Extend(m goldmark.Markdown) {
m.Parser().AddOptions(parser.WithInlineParsers(
util.Prioritized(NewTypographerParser(e.options...), 9999),
))
}

@ -0,0 +1,3 @@
module github.com/yuin/goldmark
go 1.13

@ -0,0 +1,140 @@
// Package goldmark implements functions to convert markdown text to a desired format.
package goldmark
import (
"github.com/yuin/goldmark/parser"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/renderer/html"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"io"
)
// DefaultParser returns a new Parser that is configured by default values.
func DefaultParser() parser.Parser {
return parser.NewParser(parser.WithBlockParsers(parser.DefaultBlockParsers()...),
parser.WithInlineParsers(parser.DefaultInlineParsers()...),
parser.WithParagraphTransformers(parser.DefaultParagraphTransformers()...),
)
}
// DefaultRenderer returns a new Renderer that is configured by default values.
func DefaultRenderer() renderer.Renderer {
return renderer.NewRenderer(renderer.WithNodeRenderers(util.Prioritized(html.NewRenderer(), 1000)))
}
var defaultMarkdown = New()
// Convert interprets a UTF-8 bytes source in Markdown and
// write rendered contents to a writer w.
func Convert(source []byte, w io.Writer, opts ...parser.ParseOption) error {
return defaultMarkdown.Convert(source, w, opts...)
}
// A Markdown interface offers functions to convert Markdown text to
// a desired format.
type Markdown interface {
// Convert interprets a UTF-8 bytes source in Markdown and write rendered
// contents to a writer w.
Convert(source []byte, writer io.Writer, opts ...parser.ParseOption) error
// Parser returns a Parser that will be used for conversion.
Parser() parser.Parser
// SetParser sets a Parser to this object.
SetParser(parser.Parser)
// Parser returns a Renderer that will be used for conversion.
Renderer() renderer.Renderer
// SetRenderer sets a Renderer to this object.
SetRenderer(renderer.Renderer)
}
// Option is a functional option type for Markdown objects.
type Option func(*markdown)
// WithExtensions adds extensions.
func WithExtensions(ext ...Extender) Option {
return func(m *markdown) {
m.extensions = append(m.extensions, ext...)
}
}
// WithParser allows you to override the default parser.
func WithParser(p parser.Parser) Option {
return func(m *markdown) {
m.parser = p
}
}
// WithParserOptions applies options for the parser.
func WithParserOptions(opts ...parser.Option) Option {
return func(m *markdown) {
m.parser.AddOptions(opts...)
}
}
// WithRenderer allows you to override the default renderer.
func WithRenderer(r renderer.Renderer) Option {
return func(m *markdown) {
m.renderer = r
}
}
// WithRendererOptions applies options for the renderer.
func WithRendererOptions(opts ...renderer.Option) Option {
return func(m *markdown) {
m.renderer.AddOptions(opts...)
}
}
type markdown struct {
parser parser.Parser
renderer renderer.Renderer
extensions []Extender
}
// New returns a new Markdown with given options.
func New(options ...Option) Markdown {
md := &markdown{
parser: DefaultParser(),
renderer: DefaultRenderer(),
extensions: []Extender{},
}
for _, opt := range options {
opt(md)
}
for _, e := range md.extensions {
e.Extend(md)
}
return md
}
func (m *markdown) Convert(source []byte, writer io.Writer, opts ...parser.ParseOption) error {
reader := text.NewReader(source)
doc := m.parser.Parse(reader, opts...)
return m.renderer.Render(writer, source, doc)
}
func (m *markdown) Parser() parser.Parser {
return m.parser
}
func (m *markdown) SetParser(v parser.Parser) {
m.parser = v
}
func (m *markdown) Renderer() renderer.Renderer {
return m.renderer
}
func (m *markdown) SetRenderer(v renderer.Renderer) {
m.renderer = v
}
// An Extender interface is used for extending Markdown.
type Extender interface {
// Extend extends the Markdown.
Extend(Markdown)
}

@ -0,0 +1,319 @@
package parser
import (
"bytes"
"io"
"strconv"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var attrNameID = []byte("id")
var attrNameClass = []byte("class")
// An Attribute is an attribute of the markdown elements
type Attribute struct {
Name []byte
Value interface{}
}
// An Attributes is a collection of attributes.
type Attributes []Attribute
// Find returns a (value, true) if an attribute correspond with given name is found, otherwise (nil, false).
func (as Attributes) Find(name []byte) (interface{}, bool) {
for _, a := range as {
if bytes.Equal(a.Name, name) {
return a.Value, true
}
}
return nil, false
}
func (as Attributes) findUpdate(name []byte, cb func(v interface{}) interface{}) bool {
for i, a := range as {
if bytes.Equal(a.Name, name) {
as[i].Value = cb(a.Value)
return true
}
}
return false
}
// ParseAttributes parses attributes into a map.
// ParseAttributes returns a parsed attributes and true if could parse
// attributes, otherwise nil and false.
func ParseAttributes(reader text.Reader) (Attributes, bool) {
savedLine, savedPosition := reader.Position()
reader.SkipSpaces()
if reader.Peek() != '{' {
reader.SetPosition(savedLine, savedPosition)
return nil, false
}
reader.Advance(1)
attrs := Attributes{}
for {
if reader.Peek() == '}' {
reader.Advance(1)
return attrs, true
}
attr, ok := parseAttribute(reader)
if !ok {
reader.SetPosition(savedLine, savedPosition)
return nil, false
}
if bytes.Equal(attr.Name, attrNameClass) {
if !attrs.findUpdate(attrNameClass, func(v interface{}) interface{} {
ret := make([]byte, 0, len(v.([]byte))+1+len(attr.Value.([]byte)))
ret = append(ret, v.([]byte)...)
return append(append(ret, ' '), attr.Value.([]byte)...)
}) {
attrs = append(attrs, attr)
}
} else {
attrs = append(attrs, attr)
}
reader.SkipSpaces()
if reader.Peek() == ',' {
reader.Advance(1)
reader.SkipSpaces()
}
}
}
func parseAttribute(reader text.Reader) (Attribute, bool) {
reader.SkipSpaces()
c := reader.Peek()
if c == '#' || c == '.' {
reader.Advance(1)
line, _ := reader.PeekLine()
i := 0
for ; i < len(line) && !util.IsSpace(line[i]) && (!util.IsPunct(line[i]) || line[i] == '_' || line[i] == '-'); i++ {
}
name := attrNameClass
if c == '#' {
name = attrNameID
}
reader.Advance(i)
return Attribute{Name: name, Value: line[0:i]}, true
}
line, _ := reader.PeekLine()
if len(line) == 0 {
return Attribute{}, false
}
c = line[0]
if !((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
c == '_' || c == ':') {
return Attribute{}, false
}
i := 0
for ; i < len(line); i++ {
c = line[i]
if !((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
(c >= '0' && c <= '9') ||
c == '_' || c == ':' || c == '.' || c == '-') {
break
}
}
name := line[:i]
reader.Advance(i)
reader.SkipSpaces()
c = reader.Peek()
if c != '=' {
return Attribute{}, false
}
reader.Advance(1)
reader.SkipSpaces()
value, ok := parseAttributeValue(reader)
if !ok {
return Attribute{}, false
}
return Attribute{Name: name, Value: value}, true
}
func parseAttributeValue(reader text.Reader) (interface{}, bool) {
reader.SkipSpaces()
c := reader.Peek()
var value interface{}
ok := false
switch c {
case text.EOF:
return Attribute{}, false
case '{':
value, ok = ParseAttributes(reader)
case '[':
value, ok = parseAttributeArray(reader)
case '"':
value, ok = parseAttributeString(reader)
default:
if c == '-' || c == '+' || util.IsNumeric(c) {
value, ok = parseAttributeNumber(reader)
} else {
value, ok = parseAttributeOthers(reader)
}
}
if !ok {
return nil, false
}
return value, true
}
func parseAttributeArray(reader text.Reader) ([]interface{}, bool) {
reader.Advance(1) // skip [
ret := []interface{}{}
for i := 0; ; i++ {
c := reader.Peek()
comma := false
if i != 0 && c == ',' {
reader.Advance(1)
comma = true
}
if c == ']' {
if !comma {
reader.Advance(1)
return ret, true
}
return nil, false
}
reader.SkipSpaces()
value, ok := parseAttributeValue(reader)
if !ok {
return nil, false
}
ret = append(ret, value)
reader.SkipSpaces()
}
}
func parseAttributeString(reader text.Reader) ([]byte, bool) {
reader.Advance(1) // skip "
line, _ := reader.PeekLine()
i := 0
l := len(line)
var buf bytes.Buffer
for i < l {
c := line[i]
if c == '\\' && i != l-1 {
n := line[i+1]
switch n {
case '"', '/', '\\':
buf.WriteByte(n)
i += 2
case 'b':
buf.WriteString("\b")
i += 2
case 'f':
buf.WriteString("\f")
i += 2
case 'n':
buf.WriteString("\n")
i += 2
case 'r':
buf.WriteString("\r")
i += 2
case 't':
buf.WriteString("\t")
i += 2
default:
buf.WriteByte('\\')
i++
}
continue
}
if c == '"' {
reader.Advance(i + 1)
return buf.Bytes(), true
}
buf.WriteByte(c)
i++
}
return nil, false
}
func scanAttributeDecimal(reader text.Reader, w io.ByteWriter) {
for {
c := reader.Peek()
if util.IsNumeric(c) {
w.WriteByte(c)
} else {
return
}
reader.Advance(1)
}
}
func parseAttributeNumber(reader text.Reader) (float64, bool) {
sign := 1
c := reader.Peek()
if c == '-' {
sign = -1
reader.Advance(1)
} else if c == '+' {
reader.Advance(1)
}
var buf bytes.Buffer
if !util.IsNumeric(reader.Peek()) {
return 0, false
}
scanAttributeDecimal(reader, &buf)
if buf.Len() == 0 {
return 0, false
}
c = reader.Peek()
if c == '.' {
buf.WriteByte(c)
reader.Advance(1)
scanAttributeDecimal(reader, &buf)
}
c = reader.Peek()
if c == 'e' || c == 'E' {
buf.WriteByte(c)
reader.Advance(1)
c = reader.Peek()
if c == '-' || c == '+' {
buf.WriteByte(c)
reader.Advance(1)
}
scanAttributeDecimal(reader, &buf)
}
f, err := strconv.ParseFloat(buf.String(), 10)
if err != nil {
return 0, false
}
return float64(sign) * f, true
}
var bytesTrue = []byte("true")
var bytesFalse = []byte("false")
var bytesNull = []byte("null")
func parseAttributeOthers(reader text.Reader) (interface{}, bool) {
line, _ := reader.PeekLine()
c := line[0]
if !((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
c == '_' || c == ':') {
return nil, false
}
i := 0
for ; i < len(line); i++ {
c := line[i]
if !((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z') ||
(c >= '0' && c <= '9') ||
c == '_' || c == ':' || c == '.' || c == '-') {
break
}
}
value := line[:i]
reader.Advance(i)
if bytes.Equal(value, bytesTrue) {
return true, true
}
if bytes.Equal(value, bytesFalse) {
return false, true
}
if bytes.Equal(value, bytesNull) {
return nil, true
}
return value, true
}

@ -0,0 +1,242 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// A HeadingConfig struct is a data structure that holds configuration of the renderers related to headings.
type HeadingConfig struct {
AutoHeadingID bool
Attribute bool
}
// SetOption implements SetOptioner.
func (b *HeadingConfig) SetOption(name OptionName, value interface{}) {
switch name {
case optAutoHeadingID:
b.AutoHeadingID = true
case optAttribute:
b.Attribute = true
}
}
// A HeadingOption interface sets options for heading parsers.
type HeadingOption interface {
Option
SetHeadingOption(*HeadingConfig)
}
// AutoHeadingID is an option name that enables auto IDs for headings.
const optAutoHeadingID OptionName = "AutoHeadingID"
type withAutoHeadingID struct {
}
func (o *withAutoHeadingID) SetParserOption(c *Config) {
c.Options[optAutoHeadingID] = true
}
func (o *withAutoHeadingID) SetHeadingOption(p *HeadingConfig) {
p.AutoHeadingID = true
}
// WithAutoHeadingID is a functional option that enables custom heading ids and
// auto generated heading ids.
func WithAutoHeadingID() HeadingOption {
return &withAutoHeadingID{}
}
type withHeadingAttribute struct {
Option
}
func (o *withHeadingAttribute) SetHeadingOption(p *HeadingConfig) {
p.Attribute = true
}
// WithHeadingAttribute is a functional option that enables custom heading attributes.
func WithHeadingAttribute() HeadingOption {
return &withHeadingAttribute{WithAttribute()}
}
type atxHeadingParser struct {
HeadingConfig
}
// NewATXHeadingParser return a new BlockParser that can parse ATX headings.
func NewATXHeadingParser(opts ...HeadingOption) BlockParser {
p := &atxHeadingParser{}
for _, o := range opts {
o.SetHeadingOption(&p.HeadingConfig)
}
return p
}
func (b *atxHeadingParser) Trigger() []byte {
return []byte{'#'}
}
func (b *atxHeadingParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
line, segment := reader.PeekLine()
pos := pc.BlockOffset()
if pos < 0 {
return nil, NoChildren
}
i := pos
for ; i < len(line) && line[i] == '#'; i++ {
}
level := i - pos
if i == pos || level > 6 {
return nil, NoChildren
}
l := util.TrimLeftSpaceLength(line[i:])
if l == 0 {
return nil, NoChildren
}
start := i + l
if start >= len(line) {
start = len(line) - 1
}
origstart := start
stop := len(line) - util.TrimRightSpaceLength(line)
node := ast.NewHeading(level)
parsed := false
if b.Attribute { // handles special case like ### heading ### {#id}
start--
closureClose := -1
closureOpen := -1
for j := start; j < stop; {
c := line[j]
if util.IsEscapedPunctuation(line, j) {
j += 2
} else if util.IsSpace(c) && j < stop-1 && line[j+1] == '#' {
closureOpen = j + 1
k := j + 1
for ; k < stop && line[k] == '#'; k++ {
}
closureClose = k
break
} else {
j++
}
}
if closureClose > 0 {
reader.Advance(closureClose)
attrs, ok := ParseAttributes(reader)
parsed = ok
if parsed {
for _, attr := range attrs {
node.SetAttribute(attr.Name, attr.Value)
}
node.Lines().Append(text.NewSegment(segment.Start+start+1-segment.Padding, segment.Start+closureOpen-segment.Padding))
}
}
}
if !parsed {
start = origstart
stop := len(line) - util.TrimRightSpaceLength(line)
if stop <= start { // empty headings like '##[space]'
stop = start
} else {
i = stop - 1
for ; line[i] == '#' && i >= start; i-- {
}
if i != stop-1 && !util.IsSpace(line[i]) {
i = stop - 1
}
i++
stop = i
}
if len(util.TrimRight(line[start:stop], []byte{'#'})) != 0 { // empty heading like '### ###'
node.Lines().Append(text.NewSegment(segment.Start+start-segment.Padding, segment.Start+stop-segment.Padding))
}
}
return node, NoChildren
}
func (b *atxHeadingParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
return Close
}
func (b *atxHeadingParser) Close(node ast.Node, reader text.Reader, pc Context) {
if b.Attribute {
_, ok := node.AttributeString("id")
if !ok {
parseLastLineAttributes(node, reader, pc)
}
}
if b.AutoHeadingID {
id, ok := node.AttributeString("id")
if !ok {
generateAutoHeadingID(node.(*ast.Heading), reader, pc)
} else {
pc.IDs().Put(id.([]byte))
}
}
}
func (b *atxHeadingParser) CanInterruptParagraph() bool {
return true
}
func (b *atxHeadingParser) CanAcceptIndentedLine() bool {
return false
}
func generateAutoHeadingID(node *ast.Heading, reader text.Reader, pc Context) {
var line []byte
lastIndex := node.Lines().Len() - 1
if lastIndex > -1 {
lastLine := node.Lines().At(lastIndex)
line = lastLine.Value(reader.Source())
}
headingID := pc.IDs().Generate(line, ast.KindHeading)
node.SetAttribute(attrNameID, headingID)
}
func parseLastLineAttributes(node ast.Node, reader text.Reader, pc Context) {
lastIndex := node.Lines().Len() - 1
if lastIndex < 0 { // empty headings
return
}
lastLine := node.Lines().At(lastIndex)
line := lastLine.Value(reader.Source())
lr := text.NewReader(line)
var attrs Attributes
var ok bool
var start text.Segment
var sl int
var end text.Segment
for {
c := lr.Peek()
if c == text.EOF {
break
}
if c == '\\' {
lr.Advance(1)
if lr.Peek() == '{' {
lr.Advance(1)
}
continue
}
if c == '{' {
sl, start = lr.Position()
attrs, ok = ParseAttributes(lr)
_, end = lr.Position()
lr.SetPosition(sl, start)
}
lr.Advance(1)
}
if ok && util.IsBlank(line[end.Stop:]) {
for _, attr := range attrs {
node.SetAttribute(attr.Name, attr.Value)
}
lastLine.Stop = lastLine.Start + start.Start
node.Lines().Set(lastIndex, lastLine)
}
}

@ -0,0 +1,42 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type autoLinkParser struct {
}
var defaultAutoLinkParser = &autoLinkParser{}
// NewAutoLinkParser returns a new InlineParser that parses autolinks
// surrounded by '<' and '>' .
func NewAutoLinkParser() InlineParser {
return defaultAutoLinkParser
}
func (s *autoLinkParser) Trigger() []byte {
return []byte{'<'}
}
func (s *autoLinkParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
line, segment := block.PeekLine()
stop := util.FindEmailIndex(line[1:])
typ := ast.AutoLinkType(ast.AutoLinkEmail)
if stop < 0 {
stop = util.FindURLIndex(line[1:])
typ = ast.AutoLinkURL
}
if stop < 0 {
return nil
}
stop++
if stop >= len(line) || line[stop] != '>' {
return nil
}
value := ast.NewTextSegment(text.NewSegment(segment.Start+1, segment.Start+stop))
block.Advance(stop + 1)
return ast.NewAutoLink(typ, value)
}

@ -0,0 +1,69 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type blockquoteParser struct {
}
var defaultBlockquoteParser = &blockquoteParser{}
// NewBlockquoteParser returns a new BlockParser that
// parses blockquotes.
func NewBlockquoteParser() BlockParser {
return defaultBlockquoteParser
}
func (b *blockquoteParser) process(reader text.Reader) bool {
line, _ := reader.PeekLine()
w, pos := util.IndentWidth(line, reader.LineOffset())
if w > 3 || pos >= len(line) || line[pos] != '>' {
return false
}
pos++
if pos >= len(line) || line[pos] == '\n' {
reader.Advance(pos)
return true
}
if line[pos] == ' ' || line[pos] == '\t' {
pos++
}
reader.Advance(pos)
if line[pos-1] == '\t' {
reader.SetPadding(2)
}
return true
}
func (b *blockquoteParser) Trigger() []byte {
return []byte{'>'}
}
func (b *blockquoteParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
if b.process(reader) {
return ast.NewBlockquote(), HasChildren
}
return nil, NoChildren
}
func (b *blockquoteParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
if b.process(reader) {
return Continue | HasChildren
}
return Close
}
func (b *blockquoteParser) Close(node ast.Node, reader text.Reader, pc Context) {
// nothing to do
}
func (b *blockquoteParser) CanInterruptParagraph() bool {
return true
}
func (b *blockquoteParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,79 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type codeBlockParser struct {
}
// CodeBlockParser is a BlockParser implementation that parses indented code blocks.
var defaultCodeBlockParser = &codeBlockParser{}
// NewCodeBlockParser returns a new BlockParser that
// parses code blocks.
func NewCodeBlockParser() BlockParser {
return defaultCodeBlockParser
}
func (b *codeBlockParser) Trigger() []byte {
return nil
}
func (b *codeBlockParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
line, segment := reader.PeekLine()
pos, padding := util.IndentPosition(line, reader.LineOffset(), 4)
if pos < 0 || util.IsBlank(line) {
return nil, NoChildren
}
node := ast.NewCodeBlock()
reader.AdvanceAndSetPadding(pos, padding)
_, segment = reader.PeekLine()
node.Lines().Append(segment)
reader.Advance(segment.Len() - 1)
return node, NoChildren
}
func (b *codeBlockParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
line, segment := reader.PeekLine()
if util.IsBlank(line) {
node.Lines().Append(segment.TrimLeftSpaceWidth(4, reader.Source()))
return Continue | NoChildren
}
pos, padding := util.IndentPosition(line, reader.LineOffset(), 4)
if pos < 0 {
return Close
}
reader.AdvanceAndSetPadding(pos, padding)
_, segment = reader.PeekLine()
node.Lines().Append(segment)
reader.Advance(segment.Len() - 1)
return Continue | NoChildren
}
func (b *codeBlockParser) Close(node ast.Node, reader text.Reader, pc Context) {
// trim trailing blank lines
lines := node.Lines()
length := lines.Len() - 1
source := reader.Source()
for length >= 0 {
line := lines.At(length)
if util.IsBlank(line.Value(source)) {
length--
} else {
break
}
}
lines.SetSliced(0, length+1)
}
func (b *codeBlockParser) CanInterruptParagraph() bool {
return false
}
func (b *codeBlockParser) CanAcceptIndentedLine() bool {
return true
}

@ -0,0 +1,83 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type codeSpanParser struct {
}
var defaultCodeSpanParser = &codeSpanParser{}
// NewCodeSpanParser return a new InlineParser that parses inline codes
// surrounded by '`' .
func NewCodeSpanParser() InlineParser {
return defaultCodeSpanParser
}
func (s *codeSpanParser) Trigger() []byte {
return []byte{'`'}
}
func (s *codeSpanParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
line, startSegment := block.PeekLine()
opener := 0
for ; opener < len(line) && line[opener] == '`'; opener++ {
}
block.Advance(opener)
l, pos := block.Position()
node := ast.NewCodeSpan()
for {
line, segment := block.PeekLine()
if line == nil {
block.SetPosition(l, pos)
return ast.NewTextSegment(startSegment.WithStop(startSegment.Start + opener))
}
for i := 0; i < len(line); i++ {
c := line[i]
if c == '`' {
oldi := i
for ; i < len(line) && line[i] == '`'; i++ {
}
closure := i - oldi
if closure == opener && (i >= len(line) || line[i] != '`') {
segment = segment.WithStop(segment.Start + i - closure)
if !segment.IsEmpty() {
node.AppendChild(node, ast.NewRawTextSegment(segment))
}
block.Advance(i)
goto end
}
}
}
if !util.IsBlank(line) {
node.AppendChild(node, ast.NewRawTextSegment(segment))
}
block.AdvanceLine()
}
end:
if !node.IsBlank(block.Source()) {
// trim first halfspace and last halfspace
segment := node.FirstChild().(*ast.Text).Segment
shouldTrimmed := true
if !(!segment.IsEmpty() && block.Source()[segment.Start] == ' ') {
shouldTrimmed = false
}
segment = node.LastChild().(*ast.Text).Segment
if !(!segment.IsEmpty() && block.Source()[segment.Stop-1] == ' ') {
shouldTrimmed = false
}
if shouldTrimmed {
t := node.FirstChild().(*ast.Text)
segment := t.Segment
t.Segment = segment.WithStart(segment.Start + 1)
t = node.LastChild().(*ast.Text)
segment = node.LastChild().(*ast.Text).Segment
t.Segment = segment.WithStop(segment.Stop - 1)
}
}
return node
}

@ -0,0 +1,242 @@
package parser
import (
"fmt"
"strings"
"unicode"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
// A DelimiterProcessor interface provides a set of functions about
// Deliiter nodes.
type DelimiterProcessor interface {
// IsDelimiter returns true if given character is a delimiter, otherwise false.
IsDelimiter(byte) bool
// CanOpenCloser returns true if given opener can close given closer, otherwise false.
CanOpenCloser(opener, closer *Delimiter) bool
// OnMatch will be called when new matched delimiter found.
// OnMatch should return a new Node correspond to the matched delimiter.
OnMatch(consumes int) ast.Node
}
// A Delimiter struct represents a delimiter like '*' of the Markdown text.
type Delimiter struct {
ast.BaseInline
Segment text.Segment
// CanOpen is set true if this delimiter can open a span for a new node.
// See https://spec.commonmark.org/0.29/#can-open-emphasis for details.
CanOpen bool
// CanClose is set true if this delimiter can close a span for a new node.
// See https://spec.commonmark.org/0.29/#can-open-emphasis for details.
CanClose bool
// Length is a remaining length of this delmiter.
Length int
// OriginalLength is a original length of this delimiter.
OriginalLength int
// Char is a character of this delimiter.
Char byte
// PreviousDelimiter is a previous sibling delimiter node of this delimiter.
PreviousDelimiter *Delimiter
// NextDelimiter is a next sibling delimiter node of this delimiter.
NextDelimiter *Delimiter
// Processor is a DelimiterProcessor associated with this delimiter.
Processor DelimiterProcessor
}
// Inline implements Inline.Inline.
func (d *Delimiter) Inline() {}
// Dump implements Node.Dump.
func (d *Delimiter) Dump(source []byte, level int) {
fmt.Printf("%sDelimiter: \"%s\"\n", strings.Repeat(" ", level), string(d.Text(source)))
}
var kindDelimiter = ast.NewNodeKind("Delimiter")
// Kind implements Node.Kind
func (d *Delimiter) Kind() ast.NodeKind {
return kindDelimiter
}
// Text implements Node.Text
func (d *Delimiter) Text(source []byte) []byte {
return d.Segment.Value(source)
}
// ConsumeCharacters consumes delimiters.
func (d *Delimiter) ConsumeCharacters(n int) {
d.Length -= n
d.Segment = d.Segment.WithStop(d.Segment.Start + d.Length)
}
// CalcComsumption calculates how many characters should be used for opening
// a new span correspond to given closer.
func (d *Delimiter) CalcComsumption(closer *Delimiter) int {
if (d.CanClose || closer.CanOpen) && (d.OriginalLength+closer.OriginalLength)%3 == 0 && closer.OriginalLength%3 != 0 {
return 0
}
if d.Length >= 2 && closer.Length >= 2 {
return 2
}
return 1
}
// NewDelimiter returns a new Delimiter node.
func NewDelimiter(canOpen, canClose bool, length int, char byte, processor DelimiterProcessor) *Delimiter {
c := &Delimiter{
BaseInline: ast.BaseInline{},
CanOpen: canOpen,
CanClose: canClose,
Length: length,
OriginalLength: length,
Char: char,
PreviousDelimiter: nil,
NextDelimiter: nil,
Processor: processor,
}
return c
}
// ScanDelimiter scans a delimiter by given DelimiterProcessor.
func ScanDelimiter(line []byte, before rune, min int, processor DelimiterProcessor) *Delimiter {
i := 0
c := line[i]
j := i
if !processor.IsDelimiter(c) {
return nil
}
for ; j < len(line) && c == line[j]; j++ {
}
if (j - i) >= min {
after := rune(' ')
if j != len(line) {
after = util.ToRune(line, j)
}
canOpen, canClose := false, false
beforeIsPunctuation := unicode.IsPunct(before)
beforeIsWhitespace := unicode.IsSpace(before)
afterIsPunctuation := unicode.IsPunct(after)
afterIsWhitespace := unicode.IsSpace(after)
isLeft := !afterIsWhitespace &&
(!afterIsPunctuation || beforeIsWhitespace || beforeIsPunctuation)
isRight := !beforeIsWhitespace &&
(!beforeIsPunctuation || afterIsWhitespace || afterIsPunctuation)
if line[i] == '_' {
canOpen = isLeft && (!isRight || beforeIsPunctuation)
canClose = isRight && (!isLeft || afterIsPunctuation)
} else {
canOpen = isLeft
canClose = isRight
}
return NewDelimiter(canOpen, canClose, j-i, c, processor)
}
return nil
}
// ProcessDelimiters processes the delimiter list in the context.
// Processing will be stop when reaching the bottom.
//
// If you implement an inline parser that can have other inline nodes as
// children, you should call this function when nesting span has closed.
func ProcessDelimiters(bottom ast.Node, pc Context) {
lastDelimiter := pc.LastDelimiter()
if lastDelimiter == nil {
return
}
var closer *Delimiter
if bottom != nil {
if bottom != lastDelimiter {
for c := lastDelimiter.PreviousSibling(); c != nil; {
if d, ok := c.(*Delimiter); ok {
closer = d
}
prev := c.PreviousSibling()
if prev == bottom {
break
}
c = prev
}
}
} else {
closer = pc.FirstDelimiter()
}
if closer == nil {
pc.ClearDelimiters(bottom)
return
}
for closer != nil {
if !closer.CanClose {
closer = closer.NextDelimiter
continue
}
consume := 0
found := false
maybeOpener := false
var opener *Delimiter
for opener = closer.PreviousDelimiter; opener != nil; opener = opener.PreviousDelimiter {
if opener.CanOpen && opener.Processor.CanOpenCloser(opener, closer) {
maybeOpener = true
consume = opener.CalcComsumption(closer)
if consume > 0 {
found = true
break
}
}
}
if !found {
if !maybeOpener && !closer.CanOpen {
pc.RemoveDelimiter(closer)
}
closer = closer.NextDelimiter
continue
}
opener.ConsumeCharacters(consume)
closer.ConsumeCharacters(consume)
node := opener.Processor.OnMatch(consume)
parent := opener.Parent()
child := opener.NextSibling()
for child != nil && child != closer {
next := child.NextSibling()
node.AppendChild(node, child)
child = next
}
parent.InsertAfter(parent, opener, node)
for c := opener.NextDelimiter; c != nil && c != closer; {
next := c.NextDelimiter
pc.RemoveDelimiter(c)
c = next
}
if opener.Length == 0 {
pc.RemoveDelimiter(opener)
}
if closer.Length == 0 {
next := closer.NextDelimiter
pc.RemoveDelimiter(closer)
closer = next
}
}
pc.ClearDelimiters(bottom)
}

@ -0,0 +1,50 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
)
type emphasisDelimiterProcessor struct {
}
func (p *emphasisDelimiterProcessor) IsDelimiter(b byte) bool {
return b == '*' || b == '_'
}
func (p *emphasisDelimiterProcessor) CanOpenCloser(opener, closer *Delimiter) bool {
return opener.Char == closer.Char
}
func (p *emphasisDelimiterProcessor) OnMatch(consumes int) ast.Node {
return ast.NewEmphasis(consumes)
}
var defaultEmphasisDelimiterProcessor = &emphasisDelimiterProcessor{}
type emphasisParser struct {
}
var defaultEmphasisParser = &emphasisParser{}
// NewEmphasisParser return a new InlineParser that parses emphasises.
func NewEmphasisParser() InlineParser {
return defaultEmphasisParser
}
func (s *emphasisParser) Trigger() []byte {
return []byte{'*', '_'}
}
func (s *emphasisParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
before := block.PrecendingCharacter()
line, segment := block.PeekLine()
node := ScanDelimiter(line, before, 1, defaultEmphasisDelimiterProcessor)
if node == nil {
return nil
}
node.Segment = segment.WithStop(segment.Start + node.OriginalLength)
block.Advance(node.OriginalLength)
pc.PushDelimiter(node)
return node
}

@ -0,0 +1,110 @@
package parser
import (
"bytes"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type fencedCodeBlockParser struct {
}
var defaultFencedCodeBlockParser = &fencedCodeBlockParser{}
// NewFencedCodeBlockParser returns a new BlockParser that
// parses fenced code blocks.
func NewFencedCodeBlockParser() BlockParser {
return defaultFencedCodeBlockParser
}
type fenceData struct {
char byte
indent int
length int
node ast.Node
}
var fencedCodeBlockInfoKey = NewContextKey()
func (b *fencedCodeBlockParser) Trigger() []byte {
return []byte{'~', '`'}
}
func (b *fencedCodeBlockParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
line, segment := reader.PeekLine()
pos := pc.BlockOffset()
if pos < 0 || (line[pos] != '`' && line[pos] != '~') {
return nil, NoChildren
}
findent := pos
fenceChar := line[pos]
i := pos
for ; i < len(line) && line[i] == fenceChar; i++ {
}
oFenceLength := i - pos
if oFenceLength < 3 {
return nil, NoChildren
}
var info *ast.Text
if i < len(line)-1 {
rest := line[i:]
left := util.TrimLeftSpaceLength(rest)
right := util.TrimRightSpaceLength(rest)
if left < len(rest)-right {
infoStart, infoStop := segment.Start-segment.Padding+i+left, segment.Stop-right
value := rest[left : len(rest)-right]
if fenceChar == '`' && bytes.IndexByte(value, '`') > -1 {
return nil, NoChildren
} else if infoStart != infoStop {
info = ast.NewTextSegment(text.NewSegment(infoStart, infoStop))
}
}
}
node := ast.NewFencedCodeBlock(info)
pc.Set(fencedCodeBlockInfoKey, &fenceData{fenceChar, findent, oFenceLength, node})
return node, NoChildren
}
func (b *fencedCodeBlockParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
line, segment := reader.PeekLine()
fdata := pc.Get(fencedCodeBlockInfoKey).(*fenceData)
w, pos := util.IndentWidth(line, reader.LineOffset())
if w < 4 {
i := pos
for ; i < len(line) && line[i] == fdata.char; i++ {
}
length := i - pos
if length >= fdata.length && util.IsBlank(line[i:]) {
newline := 1
if line[len(line)-1] != '\n' {
newline = 0
}
reader.Advance(segment.Stop - segment.Start - newline - segment.Padding)
return Close
}
}
pos, padding := util.DedentPositionPadding(line, reader.LineOffset(), segment.Padding, fdata.indent)
seg := text.NewSegmentPadding(segment.Start+pos, segment.Stop, padding)
node.Lines().Append(seg)
reader.AdvanceAndSetPadding(segment.Stop-segment.Start-pos-1, padding)
return Continue | NoChildren
}
func (b *fencedCodeBlockParser) Close(node ast.Node, reader text.Reader, pc Context) {
fdata := pc.Get(fencedCodeBlockInfoKey).(*fenceData)
if fdata.node == node {
pc.Set(fencedCodeBlockInfoKey, nil)
}
}
func (b *fencedCodeBlockParser) CanInterruptParagraph() bool {
return true
}
func (b *fencedCodeBlockParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,228 @@
package parser
import (
"bytes"
"regexp"
"strings"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var allowedBlockTags = map[string]bool{
"address": true,
"article": true,
"aside": true,
"base": true,
"basefont": true,
"blockquote": true,
"body": true,
"caption": true,
"center": true,
"col": true,
"colgroup": true,
"dd": true,
"details": true,
"dialog": true,
"dir": true,
"div": true,
"dl": true,
"dt": true,
"fieldset": true,
"figcaption": true,
"figure": true,
"footer": true,
"form": true,
"frame": true,
"frameset": true,
"h1": true,
"h2": true,
"h3": true,
"h4": true,
"h5": true,
"h6": true,
"head": true,
"header": true,
"hr": true,
"html": true,
"iframe": true,
"legend": true,
"li": true,
"link": true,
"main": true,
"menu": true,
"menuitem": true,
"meta": true,
"nav": true,
"noframes": true,
"ol": true,
"optgroup": true,
"option": true,
"p": true,
"param": true,
"section": true,
"source": true,
"summary": true,
"table": true,
"tbody": true,
"td": true,
"tfoot": true,
"th": true,
"thead": true,
"title": true,
"tr": true,
"track": true,
"ul": true,
}
var htmlBlockType1OpenRegexp = regexp.MustCompile(`(?i)^[ ]{0,3}<(script|pre|style)(?:\s.*|>.*|/>.*|)\n?$`)
var htmlBlockType1CloseRegexp = regexp.MustCompile(`(?i)^.*</(?:script|pre|style)>.*`)
var htmlBlockType2OpenRegexp = regexp.MustCompile(`^[ ]{0,3}<!\-\-`)
var htmlBlockType2Close = []byte{'-', '-', '>'}
var htmlBlockType3OpenRegexp = regexp.MustCompile(`^[ ]{0,3}<\?`)
var htmlBlockType3Close = []byte{'?', '>'}
var htmlBlockType4OpenRegexp = regexp.MustCompile(`^[ ]{0,3}<![A-Z]+.*\n?$`)
var htmlBlockType4Close = []byte{'>'}
var htmlBlockType5OpenRegexp = regexp.MustCompile(`^[ ]{0,3}<\!\[CDATA\[`)
var htmlBlockType5Close = []byte{']', ']', '>'}
var htmlBlockType6Regexp = regexp.MustCompile(`^[ ]{0,3}</?([a-zA-Z0-9]+)(?:\s.*|>.*|/>.*|)\n?$`)
var htmlBlockType7Regexp = regexp.MustCompile(`^[ ]{0,3}<(/)?([a-zA-Z0-9]+)(` + attributePattern + `*)(:?>|/>)\s*\n?$`)
type htmlBlockParser struct {
}
var defaultHTMLBlockParser = &htmlBlockParser{}
// NewHTMLBlockParser return a new BlockParser that can parse html
// blocks.
func NewHTMLBlockParser() BlockParser {
return defaultHTMLBlockParser
}
func (b *htmlBlockParser) Trigger() []byte {
return []byte{'<'}
}
func (b *htmlBlockParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
var node *ast.HTMLBlock
line, segment := reader.PeekLine()
last := pc.LastOpenedBlock().Node
if pos := pc.BlockOffset(); pos < 0 || line[pos] != '<' {
return nil, NoChildren
}
if m := htmlBlockType1OpenRegexp.FindSubmatchIndex(line); m != nil {
node = ast.NewHTMLBlock(ast.HTMLBlockType1)
} else if htmlBlockType2OpenRegexp.Match(line) {
node = ast.NewHTMLBlock(ast.HTMLBlockType2)
} else if htmlBlockType3OpenRegexp.Match(line) {
node = ast.NewHTMLBlock(ast.HTMLBlockType3)
} else if htmlBlockType4OpenRegexp.Match(line) {
node = ast.NewHTMLBlock(ast.HTMLBlockType4)
} else if htmlBlockType5OpenRegexp.Match(line) {
node = ast.NewHTMLBlock(ast.HTMLBlockType5)
} else if match := htmlBlockType7Regexp.FindSubmatchIndex(line); match != nil {
isCloseTag := match[2] > -1 && bytes.Equal(line[match[2]:match[3]], []byte("/"))
hasAttr := match[6] != match[7]
tagName := strings.ToLower(string(line[match[4]:match[5]]))
_, ok := allowedBlockTags[tagName]
if ok {
node = ast.NewHTMLBlock(ast.HTMLBlockType6)
} else if tagName != "script" && tagName != "style" && tagName != "pre" && !ast.IsParagraph(last) && !(isCloseTag && hasAttr) { // type 7 can not interrupt paragraph
node = ast.NewHTMLBlock(ast.HTMLBlockType7)
}
}
if node == nil {
if match := htmlBlockType6Regexp.FindSubmatchIndex(line); match != nil {
tagName := string(line[match[2]:match[3]])
_, ok := allowedBlockTags[strings.ToLower(tagName)]
if ok {
node = ast.NewHTMLBlock(ast.HTMLBlockType6)
}
}
}
if node != nil {
reader.Advance(segment.Len() - 1)
node.Lines().Append(segment)
return node, NoChildren
}
return nil, NoChildren
}
func (b *htmlBlockParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
htmlBlock := node.(*ast.HTMLBlock)
lines := htmlBlock.Lines()
line, segment := reader.PeekLine()
var closurePattern []byte
switch htmlBlock.HTMLBlockType {
case ast.HTMLBlockType1:
if lines.Len() == 1 {
firstLine := lines.At(0)
if htmlBlockType1CloseRegexp.Match(firstLine.Value(reader.Source())) {
return Close
}
}
if htmlBlockType1CloseRegexp.Match(line) {
htmlBlock.ClosureLine = segment
reader.Advance(segment.Len() - 1)
return Close
}
case ast.HTMLBlockType2:
closurePattern = htmlBlockType2Close
fallthrough
case ast.HTMLBlockType3:
if closurePattern == nil {
closurePattern = htmlBlockType3Close
}
fallthrough
case ast.HTMLBlockType4:
if closurePattern == nil {
closurePattern = htmlBlockType4Close
}
fallthrough
case ast.HTMLBlockType5:
if closurePattern == nil {
closurePattern = htmlBlockType5Close
}
if lines.Len() == 1 {
firstLine := lines.At(0)
if bytes.Contains(firstLine.Value(reader.Source()), closurePattern) {
return Close
}
}
if bytes.Contains(line, closurePattern) {
htmlBlock.ClosureLine = segment
reader.Advance(segment.Len() - 1)
return Close
}
case ast.HTMLBlockType6, ast.HTMLBlockType7:
if util.IsBlank(line) {
return Close
}
}
node.Lines().Append(segment)
reader.Advance(segment.Len() - 1)
return Continue | NoChildren
}
func (b *htmlBlockParser) Close(node ast.Node, reader text.Reader, pc Context) {
// nothing to do
}
func (b *htmlBlockParser) CanInterruptParagraph() bool {
return true
}
func (b *htmlBlockParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,375 @@
package parser
import (
"fmt"
"regexp"
"strings"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var linkLabelStateKey = NewContextKey()
type linkLabelState struct {
ast.BaseInline
Segment text.Segment
IsImage bool
Prev *linkLabelState
Next *linkLabelState
First *linkLabelState
Last *linkLabelState
}
func newLinkLabelState(segment text.Segment, isImage bool) *linkLabelState {
return &linkLabelState{
Segment: segment,
IsImage: isImage,
}
}
func (s *linkLabelState) Text(source []byte) []byte {
return s.Segment.Value(source)
}
func (s *linkLabelState) Dump(source []byte, level int) {
fmt.Printf("%slinkLabelState: \"%s\"\n", strings.Repeat(" ", level), s.Text(source))
}
var kindLinkLabelState = ast.NewNodeKind("LinkLabelState")
func (s *linkLabelState) Kind() ast.NodeKind {
return kindLinkLabelState
}
func pushLinkLabelState(pc Context, v *linkLabelState) {
tlist := pc.Get(linkLabelStateKey)
var list *linkLabelState
if tlist == nil {
list = v
v.First = v
v.Last = v
pc.Set(linkLabelStateKey, list)
} else {
list = tlist.(*linkLabelState)
l := list.Last
list.Last = v
l.Next = v
v.Prev = l
}
}
func removeLinkLabelState(pc Context, d *linkLabelState) {
tlist := pc.Get(linkLabelStateKey)
var list *linkLabelState
if tlist == nil {
return
}
list = tlist.(*linkLabelState)
if d.Prev == nil {
list = d.Next
if list != nil {
list.First = d
list.Last = d.Last
list.Prev = nil
pc.Set(linkLabelStateKey, list)
} else {
pc.Set(linkLabelStateKey, nil)
}
} else {
d.Prev.Next = d.Next
if d.Next != nil {
d.Next.Prev = d.Prev
}
}
if list != nil && d.Next == nil {
list.Last = d.Prev
}
d.Next = nil
d.Prev = nil
d.First = nil
d.Last = nil
}
type linkParser struct {
}
var defaultLinkParser = &linkParser{}
// NewLinkParser return a new InlineParser that parses links.
func NewLinkParser() InlineParser {
return defaultLinkParser
}
func (s *linkParser) Trigger() []byte {
return []byte{'!', '[', ']'}
}
var linkDestinationRegexp = regexp.MustCompile(`\s*([^\s].+)`)
var linkTitleRegexp = regexp.MustCompile(`\s+(\)|["'\(].+)`)
var linkBottom = NewContextKey()
func (s *linkParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
line, segment := block.PeekLine()
if line[0] == '!' {
if len(line) > 1 && line[1] == '[' {
block.Advance(1)
pc.Set(linkBottom, pc.LastDelimiter())
return processLinkLabelOpen(block, segment.Start+1, true, pc)
}
return nil
}
if line[0] == '[' {
pc.Set(linkBottom, pc.LastDelimiter())
return processLinkLabelOpen(block, segment.Start, false, pc)
}
// line[0] == ']'
tlist := pc.Get(linkLabelStateKey)
if tlist == nil {
return nil
}
last := tlist.(*linkLabelState).Last
if last == nil {
return nil
}
block.Advance(1)
removeLinkLabelState(pc, last)
if s.containsLink(last) { // a link in a link text is not allowed
ast.MergeOrReplaceTextSegment(last.Parent(), last, last.Segment)
return nil
}
labelValue := block.Value(text.NewSegment(last.Segment.Start+1, segment.Start))
if util.IsBlank(labelValue) && !last.IsImage {
ast.MergeOrReplaceTextSegment(last.Parent(), last, last.Segment)
return nil
}
c := block.Peek()
l, pos := block.Position()
var link *ast.Link
var hasValue bool
if c == '(' { // normal link
link = s.parseLink(parent, last, block, pc)
} else if c == '[' { // reference link
link, hasValue = s.parseReferenceLink(parent, last, block, pc)
if link == nil && hasValue {
ast.MergeOrReplaceTextSegment(last.Parent(), last, last.Segment)
return nil
}
}
if link == nil {
// maybe shortcut reference link
block.SetPosition(l, pos)
ssegment := text.NewSegment(last.Segment.Stop, segment.Start)
maybeReference := block.Value(ssegment)
ref, ok := pc.Reference(util.ToLinkReference(maybeReference))
if !ok {
ast.MergeOrReplaceTextSegment(last.Parent(), last, last.Segment)
return nil
}
link = ast.NewLink()
s.processLinkLabel(parent, link, last, pc)
link.Title = ref.Title()
link.Destination = ref.Destination()
}
if last.IsImage {
last.Parent().RemoveChild(last.Parent(), last)
return ast.NewImage(link)
}
last.Parent().RemoveChild(last.Parent(), last)
return link
}
func (s *linkParser) containsLink(last *linkLabelState) bool {
if last.IsImage {
return false
}
var c ast.Node
for c = last; c != nil; c = c.NextSibling() {
if _, ok := c.(*ast.Link); ok {
return true
}
}
return false
}
func processLinkLabelOpen(block text.Reader, pos int, isImage bool, pc Context) *linkLabelState {
start := pos
if isImage {
start--
}
state := newLinkLabelState(text.NewSegment(start, pos+1), isImage)
pushLinkLabelState(pc, state)
block.Advance(1)
return state
}
func (s *linkParser) processLinkLabel(parent ast.Node, link *ast.Link, last *linkLabelState, pc Context) {
var bottom ast.Node
if v := pc.Get(linkBottom); v != nil {
bottom = v.(ast.Node)
}
pc.Set(linkBottom, nil)
ProcessDelimiters(bottom, pc)
for c := last.NextSibling(); c != nil; {
next := c.NextSibling()
parent.RemoveChild(parent, c)
link.AppendChild(link, c)
c = next
}
}
func (s *linkParser) parseReferenceLink(parent ast.Node, last *linkLabelState, block text.Reader, pc Context) (*ast.Link, bool) {
_, orgpos := block.Position()
block.Advance(1) // skip '['
line, segment := block.PeekLine()
endIndex := util.FindClosure(line, '[', ']', false, true)
if endIndex < 0 {
return nil, false
}
block.Advance(endIndex + 1)
ssegment := segment.WithStop(segment.Start + endIndex)
maybeReference := block.Value(ssegment)
if util.IsBlank(maybeReference) { // collapsed reference link
ssegment = text.NewSegment(last.Segment.Stop, orgpos.Start-1)
maybeReference = block.Value(ssegment)
}
ref, ok := pc.Reference(util.ToLinkReference(maybeReference))
if !ok {
return nil, true
}
link := ast.NewLink()
s.processLinkLabel(parent, link, last, pc)
link.Title = ref.Title()
link.Destination = ref.Destination()
return link, true
}
func (s *linkParser) parseLink(parent ast.Node, last *linkLabelState, block text.Reader, pc Context) *ast.Link {
block.Advance(1) // skip '('
block.SkipSpaces()
var title []byte
var destination []byte
var ok bool
if block.Peek() == ')' { // empty link like '[link]()'
block.Advance(1)
} else {
destination, ok = parseLinkDestination(block)
if !ok {
return nil
}
block.SkipSpaces()
if block.Peek() == ')' {
block.Advance(1)
} else {
title, ok = parseLinkTitle(block)
if !ok {
return nil
}
block.SkipSpaces()
if block.Peek() == ')' {
block.Advance(1)
} else {
return nil
}
}
}
link := ast.NewLink()
s.processLinkLabel(parent, link, last, pc)
link.Destination = destination
link.Title = title
return link
}
func parseLinkDestination(block text.Reader) ([]byte, bool) {
block.SkipSpaces()
line, _ := block.PeekLine()
buf := []byte{}
if block.Peek() == '<' {
i := 1
for i < len(line) {
c := line[i]
if c == '\\' && i < len(line)-1 && util.IsPunct(line[i+1]) {
buf = append(buf, '\\', line[i+1])
i += 2
continue
} else if c == '>' {
block.Advance(i + 1)
return line[1:i], true
}
buf = append(buf, c)
i++
}
return nil, false
}
opened := 0
i := 0
for i < len(line) {
c := line[i]
if c == '\\' && i < len(line)-1 && util.IsPunct(line[i+1]) {
buf = append(buf, '\\', line[i+1])
i += 2
continue
} else if c == '(' {
opened++
} else if c == ')' {
opened--
if opened < 0 {
break
}
} else if util.IsSpace(c) {
break
}
buf = append(buf, c)
i++
}
block.Advance(i)
return line[:i], len(line[:i]) != 0
}
func parseLinkTitle(block text.Reader) ([]byte, bool) {
block.SkipSpaces()
opener := block.Peek()
if opener != '"' && opener != '\'' && opener != '(' {
return nil, false
}
closer := opener
if opener == '(' {
closer = ')'
}
line, _ := block.PeekLine()
pos := util.FindClosure(line[1:], opener, closer, false, true)
if pos < 0 {
return nil, false
}
pos += 2 // opener + closer
block.Advance(pos)
return line[1 : pos-1], true
}
func (s *linkParser) CloseBlock(parent ast.Node, block text.Reader, pc Context) {
tlist := pc.Get(linkLabelStateKey)
if tlist == nil {
return
}
for s := tlist.(*linkLabelState); s != nil; {
next := s.Next
removeLinkLabelState(pc, s)
s.Parent().ReplaceChild(s.Parent(), s, ast.NewTextSegment(s.Segment))
s = next
}
}

@ -0,0 +1,163 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type linkReferenceParagraphTransformer struct {
}
// LinkReferenceParagraphTransformer is a ParagraphTransformer implementation
// that parses and extracts link reference from paragraphs.
var LinkReferenceParagraphTransformer = &linkReferenceParagraphTransformer{}
func (p *linkReferenceParagraphTransformer) Transform(node *ast.Paragraph, reader text.Reader, pc Context) {
lines := node.Lines()
block := text.NewBlockReader(reader.Source(), lines)
removes := [][2]int{}
for {
start, end := parseLinkReferenceDefinition(block, pc)
if start > -1 {
if start == end {
end++
}
removes = append(removes, [2]int{start, end})
continue
}
break
}
offset := 0
for _, remove := range removes {
if lines.Len() == 0 {
break
}
s := lines.Sliced(remove[1]-offset, lines.Len())
lines.SetSliced(0, remove[0]-offset)
lines.AppendAll(s)
offset = remove[1]
}
if lines.Len() == 0 {
t := ast.NewTextBlock()
t.SetBlankPreviousLines(node.HasBlankPreviousLines())
node.Parent().ReplaceChild(node.Parent(), node, t)
return
}
node.SetLines(lines)
}
func parseLinkReferenceDefinition(block text.Reader, pc Context) (int, int) {
block.SkipSpaces()
line, segment := block.PeekLine()
if line == nil {
return -1, -1
}
startLine, _ := block.Position()
width, pos := util.IndentWidth(line, 0)
if width > 3 {
return -1, -1
}
if width != 0 {
pos++
}
if line[pos] != '[' {
return -1, -1
}
open := segment.Start + pos + 1
closes := -1
block.Advance(pos + 1)
for {
line, segment = block.PeekLine()
if line == nil {
return -1, -1
}
closure := util.FindClosure(line, '[', ']', false, false)
if closure > -1 {
closes = segment.Start + closure
next := closure + 1
if next >= len(line) || line[next] != ':' {
return -1, -1
}
block.Advance(next + 1)
break
}
block.AdvanceLine()
}
if closes < 0 {
return -1, -1
}
label := block.Value(text.NewSegment(open, closes))
if util.IsBlank(label) {
return -1, -1
}
block.SkipSpaces()
destination, ok := parseLinkDestination(block)
if !ok {
return -1, -1
}
line, segment = block.PeekLine()
isNewLine := line == nil || util.IsBlank(line)
endLine, _ := block.Position()
_, spaces, _ := block.SkipSpaces()
opener := block.Peek()
if opener != '"' && opener != '\'' && opener != '(' {
if !isNewLine {
return -1, -1
}
ref := NewReference(label, destination, nil)
pc.AddReference(ref)
return startLine, endLine + 1
}
if spaces == 0 {
return -1, -1
}
block.Advance(1)
open = -1
closes = -1
closer := opener
if opener == '(' {
closer = ')'
}
for {
line, segment = block.PeekLine()
if line == nil {
return -1, -1
}
if open < 0 {
open = segment.Start
}
closure := util.FindClosure(line, opener, closer, false, true)
if closure > -1 {
closes = segment.Start + closure
block.Advance(closure + 1)
break
}
block.AdvanceLine()
}
if closes < 0 {
return -1, -1
}
line, segment = block.PeekLine()
if line != nil && !util.IsBlank(line) {
if !isNewLine {
return -1, -1
}
title := block.Value(text.NewSegment(open, closes))
ref := NewReference(label, destination, title)
pc.AddReference(ref)
return startLine, endLine
}
title := block.Value(text.NewSegment(open, closes))
endLine, _ = block.Position()
ref := NewReference(label, destination, title)
pc.AddReference(ref)
return startLine, endLine + 1
}

@ -0,0 +1,250 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"strconv"
)
type listItemType int
const (
notList listItemType = iota
bulletList
orderedList
)
// Same as
// `^(([ ]*)([\-\*\+]))(\s+.*)?\n?$`.FindSubmatchIndex or
// `^(([ ]*)(\d{1,9}[\.\)]))(\s+.*)?\n?$`.FindSubmatchIndex
func parseListItem(line []byte) ([6]int, listItemType) {
i := 0
l := len(line)
ret := [6]int{}
for ; i < l && line[i] == ' '; i++ {
c := line[i]
if c == '\t' {
return ret, notList
}
}
if i > 3 {
return ret, notList
}
ret[0] = 0
ret[1] = i
ret[2] = i
var typ listItemType
if i < l && (line[i] == '-' || line[i] == '*' || line[i] == '+') {
i++
ret[3] = i
typ = bulletList
} else if i < l {
for ; i < l && util.IsNumeric(line[i]); i++ {
}
ret[3] = i
if ret[3] == ret[2] || ret[3]-ret[2] > 9 {
return ret, notList
}
if i < l && (line[i] == '.' || line[i] == ')') {
i++
ret[3] = i
} else {
return ret, notList
}
typ = orderedList
} else {
return ret, notList
}
if i < l && line[i] != '\n' {
w, _ := util.IndentWidth(line[i:], 0)
if w == 0 {
return ret, notList
}
}
if i >= l {
ret[4] = -1
ret[5] = -1
return ret, typ
}
ret[4] = i
ret[5] = len(line)
if line[ret[5]-1] == '\n' && line[i] != '\n' {
ret[5]--
}
return ret, typ
}
func matchesListItem(source []byte, strict bool) ([6]int, listItemType) {
m, typ := parseListItem(source)
if typ != notList && (!strict || strict && m[1] < 4) {
return m, typ
}
return m, notList
}
func calcListOffset(source []byte, match [6]int) int {
offset := 0
if match[4] < 0 || util.IsBlank(source[match[4]:]) { // list item starts with a blank line
offset = 1
} else {
offset, _ = util.IndentWidth(source[match[4]:], match[4])
if offset > 4 { // offseted codeblock
offset = 1
}
}
return offset
}
func lastOffset(node ast.Node) int {
lastChild := node.LastChild()
if lastChild != nil {
return lastChild.(*ast.ListItem).Offset
}
return 0
}
type listParser struct {
}
var defaultListParser = &listParser{}
// NewListParser returns a new BlockParser that
// parses lists.
// This parser must take precedence over the ListItemParser.
func NewListParser() BlockParser {
return defaultListParser
}
func (b *listParser) Trigger() []byte {
return []byte{'-', '+', '*', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9'}
}
func (b *listParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
last := pc.LastOpenedBlock().Node
if _, lok := last.(*ast.List); lok || pc.Get(skipListParser) != nil {
pc.Set(skipListParser, nil)
return nil, NoChildren
}
line, _ := reader.PeekLine()
match, typ := matchesListItem(line, true)
if typ == notList {
return nil, NoChildren
}
start := -1
if typ == orderedList {
number := line[match[2] : match[3]-1]
start, _ = strconv.Atoi(string(number))
}
if ast.IsParagraph(last) && last.Parent() == parent {
// we allow only lists starting with 1 to interrupt paragraphs.
if typ == orderedList && start != 1 {
return nil, NoChildren
}
//an empty list item cannot interrupt a paragraph:
if match[5]-match[4] == 1 {
return nil, NoChildren
}
}
marker := line[match[3]-1]
node := ast.NewList(marker)
if start > -1 {
node.Start = start
}
return node, HasChildren
}
func (b *listParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
list := node.(*ast.List)
line, _ := reader.PeekLine()
if util.IsBlank(line) {
// A list item can begin with at most one blank line
if node.ChildCount() == 1 && node.LastChild().ChildCount() == 0 {
return Close
}
return Continue | HasChildren
}
// Thematic Breaks take precedence over lists
if isThematicBreak(line, reader.LineOffset()) {
isHeading := false
last := pc.LastOpenedBlock().Node
if ast.IsParagraph(last) {
c, ok := matchesSetextHeadingBar(line)
if ok && c == '-' {
isHeading = true
}
}
if !isHeading {
return Close
}
}
// "offset" means a width that bar indicates.
// - aaaaaaaa
// |----|
//
// If the indent is less than the last offset like
// - a
// - b <--- current line
// it maybe a new child of the list.
offset := lastOffset(node)
indent, _ := util.IndentWidth(line, reader.LineOffset())
if indent < offset {
if indent < 4 {
match, typ := matchesListItem(line, false) // may have a leading spaces more than 3
if typ != notList && match[1]-offset < 4 {
marker := line[match[3]-1]
if !list.CanContinue(marker, typ == orderedList) {
return Close
}
return Continue | HasChildren
}
}
return Close
}
return Continue | HasChildren
}
func (b *listParser) Close(node ast.Node, reader text.Reader, pc Context) {
list := node.(*ast.List)
for c := node.FirstChild(); c != nil && list.IsTight; c = c.NextSibling() {
if c.FirstChild() != nil && c.FirstChild() != c.LastChild() {
for c1 := c.FirstChild().NextSibling(); c1 != nil; c1 = c1.NextSibling() {
if bl, ok := c1.(ast.Node); ok && bl.HasBlankPreviousLines() {
list.IsTight = false
break
}
}
}
if c != node.FirstChild() {
if bl, ok := c.(ast.Node); ok && bl.HasBlankPreviousLines() {
list.IsTight = false
}
}
}
if list.IsTight {
for child := node.FirstChild(); child != nil; child = child.NextSibling() {
for gc := child.FirstChild(); gc != nil; gc = gc.NextSibling() {
paragraph, ok := gc.(*ast.Paragraph)
if ok {
textBlock := ast.NewTextBlock()
textBlock.SetLines(paragraph.Lines())
child.ReplaceChild(child, paragraph, textBlock)
}
}
}
}
}
func (b *listParser) CanInterruptParagraph() bool {
return true
}
func (b *listParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,85 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type listItemParser struct {
}
var defaultListItemParser = &listItemParser{}
// NewListItemParser returns a new BlockParser that
// parses list items.
func NewListItemParser() BlockParser {
return defaultListItemParser
}
var skipListParser = NewContextKey()
var skipListParserValue interface{} = true
func (b *listItemParser) Trigger() []byte {
return []byte{'-', '+', '*', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9'}
}
func (b *listItemParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
list, lok := parent.(*ast.List)
if !lok { // list item must be a child of a list
return nil, NoChildren
}
offset := lastOffset(list)
line, _ := reader.PeekLine()
match, typ := matchesListItem(line, false)
if typ == notList {
return nil, NoChildren
}
if match[1]-offset > 3 {
return nil, NoChildren
}
itemOffset := calcListOffset(line, match)
node := ast.NewListItem(match[3] + itemOffset)
if match[4] < 0 || match[5]-match[4] == 1 {
return node, NoChildren
}
pos, padding := util.IndentPosition(line[match[4]:], match[4], itemOffset)
child := match[3] + pos
reader.AdvanceAndSetPadding(child, padding)
return node, HasChildren
}
func (b *listItemParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
line, _ := reader.PeekLine()
if util.IsBlank(line) {
return Continue | HasChildren
}
indent, _ := util.IndentWidth(line, reader.LineOffset())
offset := lastOffset(node.Parent())
if indent < offset && indent < 4 {
_, typ := matchesListItem(line, true)
// new list item found
if typ != notList {
pc.Set(skipListParser, skipListParserValue)
}
return Close
}
pos, padding := util.IndentPosition(line, reader.LineOffset(), offset)
reader.AdvanceAndSetPadding(pos, padding)
return Continue | HasChildren
}
func (b *listItemParser) Close(node ast.Node, reader text.Reader, pc Context) {
// nothing to do
}
func (b *listItemParser) CanInterruptParagraph() bool {
return true
}
func (b *listItemParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,71 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
)
type paragraphParser struct {
}
var defaultParagraphParser = &paragraphParser{}
// NewParagraphParser returns a new BlockParser that
// parses paragraphs.
func NewParagraphParser() BlockParser {
return defaultParagraphParser
}
func (b *paragraphParser) Trigger() []byte {
return nil
}
func (b *paragraphParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
_, segment := reader.PeekLine()
segment = segment.TrimLeftSpace(reader.Source())
if segment.IsEmpty() {
return nil, NoChildren
}
node := ast.NewParagraph()
node.Lines().Append(segment)
reader.Advance(segment.Len() - 1)
return node, NoChildren
}
func (b *paragraphParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
_, segment := reader.PeekLine()
segment = segment.TrimLeftSpace(reader.Source())
if segment.IsEmpty() {
return Close
}
node.Lines().Append(segment)
reader.Advance(segment.Len() - 1)
return Continue | NoChildren
}
func (b *paragraphParser) Close(node ast.Node, reader text.Reader, pc Context) {
parent := node.Parent()
if parent == nil {
// paragraph has been transformed
return
}
lines := node.Lines()
if lines.Len() != 0 {
// trim trailing spaces
length := lines.Len()
lastLine := node.Lines().At(length - 1)
node.Lines().Set(length-1, lastLine.TrimRightSpace(reader.Source()))
}
if lines.Len() == 0 {
node.Parent().RemoveChild(node.Parent(), node)
return
}
}
func (b *paragraphParser) CanInterruptParagraph() bool {
return false
}
func (b *paragraphParser) CanAcceptIndentedLine() bool {
return false
}

File diff suppressed because it is too large Load Diff

@ -0,0 +1,108 @@
package parser
import (
"bytes"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
"regexp"
)
type rawHTMLParser struct {
}
var defaultRawHTMLParser = &rawHTMLParser{}
// NewRawHTMLParser return a new InlineParser that can parse
// inline htmls
func NewRawHTMLParser() InlineParser {
return defaultRawHTMLParser
}
func (s *rawHTMLParser) Trigger() []byte {
return []byte{'<'}
}
func (s *rawHTMLParser) Parse(parent ast.Node, block text.Reader, pc Context) ast.Node {
line, _ := block.PeekLine()
if len(line) > 1 && util.IsAlphaNumeric(line[1]) {
return s.parseMultiLineRegexp(openTagRegexp, block, pc)
}
if len(line) > 2 && line[1] == '/' && util.IsAlphaNumeric(line[2]) {
return s.parseMultiLineRegexp(closeTagRegexp, block, pc)
}
if bytes.HasPrefix(line, []byte("<!--")) {
return s.parseMultiLineRegexp(commentRegexp, block, pc)
}
if bytes.HasPrefix(line, []byte("<?")) {
return s.parseSingleLineRegexp(processingInstructionRegexp, block, pc)
}
if len(line) > 2 && line[1] == '!' && line[2] >= 'A' && line[2] <= 'Z' {
return s.parseSingleLineRegexp(declRegexp, block, pc)
}
if bytes.HasPrefix(line, []byte("<![CDATA[")) {
return s.parseMultiLineRegexp(cdataRegexp, block, pc)
}
return nil
}
var tagnamePattern = `([A-Za-z][A-Za-z0-9-]*)`
var attributePattern = `(?:\s+[a-zA-Z_:][a-zA-Z0-9:._-]*(?:\s*=\s*(?:[^\"'=<>` + "`" + `\x00-\x20]+|'[^']*'|"[^"]*"))?)`
var openTagRegexp = regexp.MustCompile("^<" + tagnamePattern + attributePattern + `*\s*/?>`)
var closeTagRegexp = regexp.MustCompile("^</" + tagnamePattern + `\s*>`)
var commentRegexp = regexp.MustCompile(`^<!---->|<!--(?:-?[^>-])(?:-?[^-])*-->`)
var processingInstructionRegexp = regexp.MustCompile(`^(?:<\?).*?(?:\?>)`)
var declRegexp = regexp.MustCompile(`^<![A-Z]+\s+[^>]*>`)
var cdataRegexp = regexp.MustCompile(`<!\[CDATA\[[\s\S]*?\]\]>`)
func (s *rawHTMLParser) parseSingleLineRegexp(reg *regexp.Regexp, block text.Reader, pc Context) ast.Node {
line, segment := block.PeekLine()
match := reg.FindSubmatchIndex(line)
if match == nil {
return nil
}
node := ast.NewRawHTML()
node.Segments.Append(segment.WithStop(segment.Start + match[1]))
block.Advance(match[1])
return node
}
var dummyMatch = [][]byte{}
func (s *rawHTMLParser) parseMultiLineRegexp(reg *regexp.Regexp, block text.Reader, pc Context) ast.Node {
sline, ssegment := block.Position()
if block.Match(reg) {
node := ast.NewRawHTML()
eline, esegment := block.Position()
block.SetPosition(sline, ssegment)
for {
line, segment := block.PeekLine()
if line == nil {
break
}
l, _ := block.Position()
start := segment.Start
if l == sline {
start = ssegment.Start
}
end := segment.Stop
if l == eline {
end = esegment.Start
}
node.Segments.Append(text.NewSegment(start, end))
if l == eline {
block.Advance(end - start)
break
} else {
block.AdvanceLine()
}
}
return node
}
return nil
}
func (s *rawHTMLParser) CloseBlock(parent ast.Node, pc Context) {
// nothing to do
}

@ -0,0 +1,126 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
var temporaryParagraphKey = NewContextKey()
type setextHeadingParser struct {
HeadingConfig
}
func matchesSetextHeadingBar(line []byte) (byte, bool) {
start := 0
end := len(line)
space := util.TrimLeftLength(line, []byte{' '})
if space > 3 {
return 0, false
}
start += space
level1 := util.TrimLeftLength(line[start:end], []byte{'='})
c := byte('=')
var level2 int
if level1 == 0 {
level2 = util.TrimLeftLength(line[start:end], []byte{'-'})
c = '-'
}
if util.IsSpace(line[end-1]) {
end -= util.TrimRightSpaceLength(line[start:end])
}
if !((level1 > 0 && start+level1 == end) || (level2 > 0 && start+level2 == end)) {
return 0, false
}
return c, true
}
// NewSetextHeadingParser return a new BlockParser that can parse Setext headings.
func NewSetextHeadingParser(opts ...HeadingOption) BlockParser {
p := &setextHeadingParser{}
for _, o := range opts {
o.SetHeadingOption(&p.HeadingConfig)
}
return p
}
func (b *setextHeadingParser) Trigger() []byte {
return []byte{'-', '='}
}
func (b *setextHeadingParser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
last := pc.LastOpenedBlock().Node
if last == nil {
return nil, NoChildren
}
paragraph, ok := last.(*ast.Paragraph)
if !ok || paragraph.Parent() != parent {
return nil, NoChildren
}
line, segment := reader.PeekLine()
c, ok := matchesSetextHeadingBar(line)
if !ok {
return nil, NoChildren
}
level := 1
if c == '-' {
level = 2
}
node := ast.NewHeading(level)
node.Lines().Append(segment)
pc.Set(temporaryParagraphKey, last)
return node, NoChildren | RequireParagraph
}
func (b *setextHeadingParser) Continue(node ast.Node, reader text.Reader, pc Context) State {
return Close
}
func (b *setextHeadingParser) Close(node ast.Node, reader text.Reader, pc Context) {
heading := node.(*ast.Heading)
segment := node.Lines().At(0)
heading.Lines().Clear()
tmp := pc.Get(temporaryParagraphKey).(*ast.Paragraph)
pc.Set(temporaryParagraphKey, nil)
if tmp.Lines().Len() == 0 {
next := heading.NextSibling()
segment = segment.TrimLeftSpace(reader.Source())
if next == nil || !ast.IsParagraph(next) {
para := ast.NewParagraph()
para.Lines().Append(segment)
heading.Parent().InsertAfter(heading.Parent(), heading, para)
} else {
next.(ast.Node).Lines().Unshift(segment)
}
heading.Parent().RemoveChild(heading.Parent(), heading)
} else {
heading.SetLines(tmp.Lines())
heading.SetBlankPreviousLines(tmp.HasBlankPreviousLines())
tp := tmp.Parent()
if tp != nil {
tp.RemoveChild(tp, tmp)
}
}
if b.Attribute {
parseLastLineAttributes(node, reader, pc)
}
if b.AutoHeadingID {
id, ok := node.AttributeString("id")
if !ok {
generateAutoHeadingID(heading, reader, pc)
} else {
pc.IDs().Put(id.([]byte))
}
}
}
func (b *setextHeadingParser) CanInterruptParagraph() bool {
return true
}
func (b *setextHeadingParser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,75 @@
package parser
import (
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/text"
"github.com/yuin/goldmark/util"
)
type thematicBreakPraser struct {
}
var defaultThematicBreakPraser = &thematicBreakPraser{}
// NewThematicBreakParser returns a new BlockParser that
// parses thematic breaks.
func NewThematicBreakParser() BlockParser {
return defaultThematicBreakPraser
}
func isThematicBreak(line []byte, offset int) bool {
w, pos := util.IndentWidth(line, offset)
if w > 3 {
return false
}
mark := byte(0)
count := 0
for i := pos; i < len(line); i++ {
c := line[i]
if util.IsSpace(c) {
continue
}
if mark == 0 {
mark = c
count = 1
if mark == '*' || mark == '-' || mark == '_' {
continue
}
return false
}
if c != mark {
return false
}
count++
}
return count > 2
}
func (b *thematicBreakPraser) Trigger() []byte {
return []byte{'-', '*', '_'}
}
func (b *thematicBreakPraser) Open(parent ast.Node, reader text.Reader, pc Context) (ast.Node, State) {
line, segment := reader.PeekLine()
if isThematicBreak(line, reader.LineOffset()) {
reader.Advance(segment.Len() - 1)
return ast.NewThematicBreak(), NoChildren
}
return nil, NoChildren
}
func (b *thematicBreakPraser) Continue(node ast.Node, reader text.Reader, pc Context) State {
return Close
}
func (b *thematicBreakPraser) Close(node ast.Node, reader text.Reader, pc Context) {
// nothing to do
}
func (b *thematicBreakPraser) CanInterruptParagraph() bool {
return true
}
func (b *thematicBreakPraser) CanAcceptIndentedLine() bool {
return false
}

@ -0,0 +1,804 @@
package html
import (
"bytes"
"fmt"
"strconv"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/renderer"
"github.com/yuin/goldmark/util"
)
// A Config struct has configurations for the HTML based renderers.
type Config struct {
Writer Writer
HardWraps bool
XHTML bool
Unsafe bool
}
// NewConfig returns a new Config with defaults.
func NewConfig() Config {
return Config{
Writer: DefaultWriter,
HardWraps: false,
XHTML: false,
Unsafe: false,
}
}
// SetOption implements renderer.NodeRenderer.SetOption.
func (c *Config) SetOption(name renderer.OptionName, value interface{}) {
switch name {
case optHardWraps:
c.HardWraps = value.(bool)
case optXHTML:
c.XHTML = value.(bool)
case optUnsafe:
c.Unsafe = value.(bool)
case optTextWriter:
c.Writer = value.(Writer)
}
}
// An Option interface sets options for HTML based renderers.
type Option interface {
SetHTMLOption(*Config)
}
// TextWriter is an option name used in WithWriter.
const optTextWriter renderer.OptionName = "Writer"
type withWriter struct {
value Writer
}
func (o *withWriter) SetConfig(c *renderer.Config) {
c.Options[optTextWriter] = o.value
}
func (o *withWriter) SetHTMLOption(c *Config) {
c.Writer = o.value
}
// WithWriter is a functional option that allow you to set the given writer to
// the renderer.
func WithWriter(writer Writer) interface {
renderer.Option
Option
} {
return &withWriter{writer}
}
// HardWraps is an option name used in WithHardWraps.
const optHardWraps renderer.OptionName = "HardWraps"
type withHardWraps struct {
}
func (o *withHardWraps) SetConfig(c *renderer.Config) {
c.Options[optHardWraps] = true
}
func (o *withHardWraps) SetHTMLOption(c *Config) {
c.HardWraps = true
}
// WithHardWraps is a functional option that indicates whether softline breaks
// should be rendered as '<br>'.
func WithHardWraps() interface {
renderer.Option
Option
} {
return &withHardWraps{}
}
// XHTML is an option name used in WithXHTML.
const optXHTML renderer.OptionName = "XHTML"
type withXHTML struct {
}
func (o *withXHTML) SetConfig(c *renderer.Config) {
c.Options[optXHTML] = true
}
func (o *withXHTML) SetHTMLOption(c *Config) {
c.XHTML = true
}
// WithXHTML is a functional option indicates that nodes should be rendered in
// xhtml instead of HTML5.
func WithXHTML() interface {
Option
renderer.Option
} {
return &withXHTML{}
}
// Unsafe is an option name used in WithUnsafe.
const optUnsafe renderer.OptionName = "Unsafe"
type withUnsafe struct {
}
func (o *withUnsafe) SetConfig(c *renderer.Config) {
c.Options[optUnsafe] = true
}
func (o *withUnsafe) SetHTMLOption(c *Config) {
c.Unsafe = true
}
// WithUnsafe is a functional option that renders dangerous contents
// (raw htmls and potentially dangerous links) as it is.
func WithUnsafe() interface {
renderer.Option
Option
} {
return &withUnsafe{}
}
// A Renderer struct is an implementation of renderer.NodeRenderer that renders
// nodes as (X)HTML.
type Renderer struct {
Config
}
// NewRenderer returns a new Renderer with given options.
func NewRenderer(opts ...Option) renderer.NodeRenderer {
r := &Renderer{
Config: NewConfig(),
}
for _, opt := range opts {
opt.SetHTMLOption(&r.Config)
}
return r
}
// RegisterFuncs implements NodeRenderer.RegisterFuncs .
func (r *Renderer) RegisterFuncs(reg renderer.NodeRendererFuncRegisterer) {
// blocks
reg.Register(ast.KindDocument, r.renderDocument)
reg.Register(ast.KindHeading, r.renderHeading)
reg.Register(ast.KindBlockquote, r.renderBlockquote)
reg.Register(ast.KindCodeBlock, r.renderCodeBlock)
reg.Register(ast.KindFencedCodeBlock, r.renderFencedCodeBlock)
reg.Register(ast.KindHTMLBlock, r.renderHTMLBlock)
reg.Register(ast.KindList, r.renderList)
reg.Register(ast.KindListItem, r.renderListItem)
reg.Register(ast.KindParagraph, r.renderParagraph)
reg.Register(ast.KindTextBlock, r.renderTextBlock)
reg.Register(ast.KindThematicBreak, r.renderThematicBreak)
// inlines
reg.Register(ast.KindAutoLink, r.renderAutoLink)
reg.Register(ast.KindCodeSpan, r.renderCodeSpan)
reg.Register(ast.KindEmphasis, r.renderEmphasis)
reg.Register(ast.KindImage, r.renderImage)
reg.Register(ast.KindLink, r.renderLink)
reg.Register(ast.KindRawHTML, r.renderRawHTML)
reg.Register(ast.KindText, r.renderText)
reg.Register(ast.KindString, r.renderString)
}
func (r *Renderer) writeLines(w util.BufWriter, source []byte, n ast.Node) {
l := n.Lines().Len()
for i := 0; i < l; i++ {
line := n.Lines().At(i)
r.Writer.RawWrite(w, line.Value(source))
}
}
// GlobalAttributeFilter defines attribute names which any elements can have.
var GlobalAttributeFilter = util.NewBytesFilter(
[]byte("accesskey"),
[]byte("autocapitalize"),
[]byte("class"),
[]byte("contenteditable"),
[]byte("contextmenu"),
[]byte("dir"),
[]byte("draggable"),
[]byte("dropzone"),
[]byte("hidden"),
[]byte("id"),
[]byte("itemprop"),
[]byte("lang"),
[]byte("slot"),
[]byte("spellcheck"),
[]byte("style"),
[]byte("tabindex"),
[]byte("title"),
[]byte("translate"),
)
func (r *Renderer) renderDocument(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
// nothing to do
return ast.WalkContinue, nil
}
// HeadingAttributeFilter defines attribute names which heading elements can have
var HeadingAttributeFilter = GlobalAttributeFilter
func (r *Renderer) renderHeading(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.Heading)
if entering {
_, _ = w.WriteString("<h")
_ = w.WriteByte("0123456"[n.Level])
if n.Attributes() != nil {
RenderAttributes(w, node, HeadingAttributeFilter)
}
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("</h")
_ = w.WriteByte("0123456"[n.Level])
_, _ = w.WriteString(">\n")
}
return ast.WalkContinue, nil
}
// BlockquoteAttributeFilter defines attribute names which blockquote elements can have
var BlockquoteAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("cite"),
)
func (r *Renderer) renderBlockquote(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<blockquote")
RenderAttributes(w, n, BlockquoteAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<blockquote>\n")
}
} else {
_, _ = w.WriteString("</blockquote>\n")
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderCodeBlock(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
_, _ = w.WriteString("<pre><code>")
r.writeLines(w, source, n)
} else {
_, _ = w.WriteString("</code></pre>\n")
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderFencedCodeBlock(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.FencedCodeBlock)
if entering {
_, _ = w.WriteString("<pre><code")
language := n.Language(source)
if language != nil {
_, _ = w.WriteString(" class=\"language-")
r.Writer.Write(w, language)
_, _ = w.WriteString("\"")
}
_ = w.WriteByte('>')
r.writeLines(w, source, n)
} else {
_, _ = w.WriteString("</code></pre>\n")
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderHTMLBlock(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.HTMLBlock)
if entering {
if r.Unsafe {
l := n.Lines().Len()
for i := 0; i < l; i++ {
line := n.Lines().At(i)
_, _ = w.Write(line.Value(source))
}
} else {
_, _ = w.WriteString("<!-- raw HTML omitted -->\n")
}
} else {
if n.HasClosure() {
if r.Unsafe {
closure := n.ClosureLine
_, _ = w.Write(closure.Value(source))
} else {
_, _ = w.WriteString("<!-- raw HTML omitted -->\n")
}
}
}
return ast.WalkContinue, nil
}
// ListAttributeFilter defines attribute names which list elements can have.
var ListAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("start"),
[]byte("reversed"),
)
func (r *Renderer) renderList(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.List)
tag := "ul"
if n.IsOrdered() {
tag = "ol"
}
if entering {
_ = w.WriteByte('<')
_, _ = w.WriteString(tag)
if n.IsOrdered() && n.Start != 1 {
fmt.Fprintf(w, " start=\"%d\"", n.Start)
}
if n.Attributes() != nil {
RenderAttributes(w, n, ListAttributeFilter)
}
_, _ = w.WriteString(">\n")
} else {
_, _ = w.WriteString("</")
_, _ = w.WriteString(tag)
_, _ = w.WriteString(">\n")
}
return ast.WalkContinue, nil
}
// ListItemAttributeFilter defines attribute names which list item elements can have.
var ListItemAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("value"),
)
func (r *Renderer) renderListItem(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<li")
RenderAttributes(w, n, ListItemAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<li>")
}
fc := n.FirstChild()
if fc != nil {
if _, ok := fc.(*ast.TextBlock); !ok {
_ = w.WriteByte('\n')
}
}
} else {
_, _ = w.WriteString("</li>\n")
}
return ast.WalkContinue, nil
}
// ParagraphAttributeFilter defines attribute names which paragraph elements can have.
var ParagraphAttributeFilter = GlobalAttributeFilter
func (r *Renderer) renderParagraph(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<p")
RenderAttributes(w, n, ParagraphAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<p>")
}
} else {
_, _ = w.WriteString("</p>\n")
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderTextBlock(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
if _, ok := n.NextSibling().(ast.Node); ok && n.FirstChild() != nil {
_ = w.WriteByte('\n')
}
}
return ast.WalkContinue, nil
}
// ThematicAttributeFilter defines attribute names which hr elements can have.
var ThematicAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("align"), // [Deprecated]
[]byte("color"), // [Not Standardized]
[]byte("noshade"), // [Deprecated]
[]byte("size"), // [Deprecated]
[]byte("width"), // [Deprecated]
)
func (r *Renderer) renderThematicBreak(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
_, _ = w.WriteString("<hr")
if n.Attributes() != nil {
RenderAttributes(w, n, ThematicAttributeFilter)
}
if r.XHTML {
_, _ = w.WriteString(" />\n")
} else {
_, _ = w.WriteString(">\n")
}
return ast.WalkContinue, nil
}
// LinkAttributeFilter defines attribute names which link elements can have.
var LinkAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("download"),
// []byte("href"),
[]byte("hreflang"),
[]byte("media"),
[]byte("ping"),
[]byte("referrerpolicy"),
[]byte("rel"),
[]byte("shape"),
[]byte("target"),
)
func (r *Renderer) renderAutoLink(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.AutoLink)
if !entering {
return ast.WalkContinue, nil
}
_, _ = w.WriteString(`<a href="`)
url := n.URL(source)
label := n.Label(source)
if n.AutoLinkType == ast.AutoLinkEmail && !bytes.HasPrefix(bytes.ToLower(url), []byte("mailto:")) {
_, _ = w.WriteString("mailto:")
}
_, _ = w.Write(util.EscapeHTML(util.URLEscape(url, false)))
if n.Attributes() != nil {
_ = w.WriteByte('"')
RenderAttributes(w, n, LinkAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString(`">`)
}
_, _ = w.Write(util.EscapeHTML(label))
_, _ = w.WriteString(`</a>`)
return ast.WalkContinue, nil
}
// CodeAttributeFilter defines attribute names which code elements can have.
var CodeAttributeFilter = GlobalAttributeFilter
func (r *Renderer) renderCodeSpan(w util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error) {
if entering {
if n.Attributes() != nil {
_, _ = w.WriteString("<code")
RenderAttributes(w, n, CodeAttributeFilter)
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("<code>")
}
for c := n.FirstChild(); c != nil; c = c.NextSibling() {
segment := c.(*ast.Text).Segment
value := segment.Value(source)
if bytes.HasSuffix(value, []byte("\n")) {
r.Writer.RawWrite(w, value[:len(value)-1])
if c != n.LastChild() {
r.Writer.RawWrite(w, []byte(" "))
}
} else {
r.Writer.RawWrite(w, value)
}
}
return ast.WalkSkipChildren, nil
}
_, _ = w.WriteString("</code>")
return ast.WalkContinue, nil
}
// EmphasisAttributeFilter defines attribute names which emphasis elements can have.
var EmphasisAttributeFilter = GlobalAttributeFilter
func (r *Renderer) renderEmphasis(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.Emphasis)
tag := "em"
if n.Level == 2 {
tag = "strong"
}
if entering {
_ = w.WriteByte('<')
_, _ = w.WriteString(tag)
if n.Attributes() != nil {
RenderAttributes(w, n, EmphasisAttributeFilter)
}
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("</")
_, _ = w.WriteString(tag)
_ = w.WriteByte('>')
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderLink(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
n := node.(*ast.Link)
if entering {
_, _ = w.WriteString("<a href=\"")
if r.Unsafe || !IsDangerousURL(n.Destination) {
_, _ = w.Write(util.EscapeHTML(util.URLEscape(n.Destination, true)))
}
_ = w.WriteByte('"')
if n.Title != nil {
_, _ = w.WriteString(` title="`)
r.Writer.Write(w, n.Title)
_ = w.WriteByte('"')
}
if n.Attributes() != nil {
RenderAttributes(w, n, LinkAttributeFilter)
}
_ = w.WriteByte('>')
} else {
_, _ = w.WriteString("</a>")
}
return ast.WalkContinue, nil
}
// ImageAttributeFilter defines attribute names which image elements can have.
var ImageAttributeFilter = GlobalAttributeFilter.Extend(
[]byte("align"),
[]byte("border"),
[]byte("crossorigin"),
[]byte("decoding"),
[]byte("height"),
[]byte("importance"),
[]byte("intrinsicsize"),
[]byte("ismap"),
[]byte("loading"),
[]byte("referrerpolicy"),
[]byte("sizes"),
[]byte("srcset"),
[]byte("usemap"),
[]byte("width"),
)
func (r *Renderer) renderImage(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
n := node.(*ast.Image)
_, _ = w.WriteString("<img src=\"")
if r.Unsafe || !IsDangerousURL(n.Destination) {
_, _ = w.Write(util.EscapeHTML(util.URLEscape(n.Destination, true)))
}
_, _ = w.WriteString(`" alt="`)
_, _ = w.Write(n.Text(source))
_ = w.WriteByte('"')
if n.Title != nil {
_, _ = w.WriteString(` title="`)
r.Writer.Write(w, n.Title)
_ = w.WriteByte('"')
}
if n.Attributes() != nil {
RenderAttributes(w, n, ImageAttributeFilter)
}
if r.XHTML {
_, _ = w.WriteString(" />")
} else {
_, _ = w.WriteString(">")
}
return ast.WalkSkipChildren, nil
}
func (r *Renderer) renderRawHTML(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkSkipChildren, nil
}
if r.Unsafe {
n := node.(*ast.RawHTML)
l := n.Segments.Len()
for i := 0; i < l; i++ {
segment := n.Segments.At(i)
_, _ = w.Write(segment.Value(source))
}
return ast.WalkSkipChildren, nil
}
_, _ = w.WriteString("<!-- raw HTML omitted -->")
return ast.WalkSkipChildren, nil
}
func (r *Renderer) renderText(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
n := node.(*ast.Text)
segment := n.Segment
if n.IsRaw() {
r.Writer.RawWrite(w, segment.Value(source))
} else {
r.Writer.Write(w, segment.Value(source))
if n.HardLineBreak() || (n.SoftLineBreak() && r.HardWraps) {
if r.XHTML {
_, _ = w.WriteString("<br />\n")
} else {
_, _ = w.WriteString("<br>\n")
}
} else if n.SoftLineBreak() {
_ = w.WriteByte('\n')
}
}
return ast.WalkContinue, nil
}
func (r *Renderer) renderString(w util.BufWriter, source []byte, node ast.Node, entering bool) (ast.WalkStatus, error) {
if !entering {
return ast.WalkContinue, nil
}
n := node.(*ast.String)
if n.IsCode() {
_, _ = w.Write(n.Value)
} else {
if n.IsRaw() {
r.Writer.RawWrite(w, n.Value)
} else {
r.Writer.Write(w, n.Value)
}
}
return ast.WalkContinue, nil
}
var dataPrefix = []byte("data-")
// RenderAttributes renders given node's attributes.
// You can specify attribute names to render by the filter.
// If filter is nil, RenderAttributes renders all attributes.
func RenderAttributes(w util.BufWriter, node ast.Node, filter util.BytesFilter) {
for _, attr := range node.Attributes() {
if filter != nil && !filter.Contains(attr.Name) {
if !bytes.HasPrefix(attr.Name, dataPrefix) {
continue
}
}
_, _ = w.WriteString(" ")
_, _ = w.Write(attr.Name)
_, _ = w.WriteString(`="`)
// TODO: convert numeric values to strings
_, _ = w.Write(util.EscapeHTML(attr.Value.([]byte)))
_ = w.WriteByte('"')
}
}
// A Writer interface wirtes textual contents to a writer.
type Writer interface {
// Write writes the given source to writer with resolving references and unescaping
// backslash escaped characters.
Write(writer util.BufWriter, source []byte)
// RawWrite wirtes the given source to writer without resolving references and
// unescaping backslash escaped characters.
RawWrite(writer util.BufWriter, source []byte)
}
type defaultWriter struct {
}
func escapeRune(writer util.BufWriter, r rune) {
if r < 256 {
v := util.EscapeHTMLByte(byte(r))
if v != nil {
_, _ = writer.Write(v)
return
}
}
_, _ = writer.WriteRune(util.ToValidRune(r))
}
func (d *defaultWriter) RawWrite(writer util.BufWriter, source []byte) {
n := 0
l := len(source)
for i := 0; i < l; i++ {
v := util.EscapeHTMLByte(source[i])
if v != nil {
_, _ = writer.Write(source[i-n : i])
n = 0
_, _ = writer.Write(v)
continue
}
n++
}
if n != 0 {
_, _ = writer.Write(source[l-n:])
}
}
func (d *defaultWriter) Write(writer util.BufWriter, source []byte) {
escaped := false
var ok bool
limit := len(source)
n := 0
for i := 0; i < limit; i++ {
c := source[i]
if escaped {
if util.IsPunct(c) {
d.RawWrite(writer, source[n:i-1])
n = i
escaped = false
continue
}
}
if c == '&' {
pos := i
next := i + 1
if next < limit && source[next] == '#' {
nnext := next + 1
if nnext < limit {
nc := source[nnext]
// code point like #x22;
if nnext < limit && nc == 'x' || nc == 'X' {
start := nnext + 1
i, ok = util.ReadWhile(source, [2]int{start, limit}, util.IsHexDecimal)
if ok && i < limit && source[i] == ';' {
v, _ := strconv.ParseUint(util.BytesToReadOnlyString(source[start:i]), 16, 32)
d.RawWrite(writer, source[n:pos])
n = i + 1
escapeRune(writer, rune(v))
continue
}
// code point like #1234;
} else if nc >= '0' && nc <= '9' {
start := nnext
i, ok = util.ReadWhile(source, [2]int{start, limit}, util.IsNumeric)
if ok && i < limit && i-start < 8 && source[i] == ';' {
v, _ := strconv.ParseUint(util.BytesToReadOnlyString(source[start:i]), 0, 32)
d.RawWrite(writer, source[n:pos])
n = i + 1
escapeRune(writer, rune(v))
continue
}
}
}
} else {
start := next
i, ok = util.ReadWhile(source, [2]int{start, limit}, util.IsAlphaNumeric)
// entity reference
if ok && i < limit && source[i] == ';' {
name := util.BytesToReadOnlyString(source[start:i])
entity, ok := util.LookUpHTML5EntityByName(name)
if ok {
d.RawWrite(writer, source[n:pos])
n = i + 1
d.RawWrite(writer, entity.Characters)
continue
}
}
}
i = next - 1
}
if c == '\\' {
escaped = true
continue
}
escaped = false
}
d.RawWrite(writer, source[n:])
}
// DefaultWriter is a default implementation of the Writer.
var DefaultWriter = &defaultWriter{}
var bDataImage = []byte("data:image/")
var bPng = []byte("png;")
var bGif = []byte("gif;")
var bJpeg = []byte("jpeg;")
var bWebp = []byte("webp;")
var bJs = []byte("javascript:")
var bVb = []byte("vbscript:")
var bFile = []byte("file:")
var bData = []byte("data:")
// IsDangerousURL returns true if the given url seems a potentially dangerous url,
// otherwise false.
func IsDangerousURL(url []byte) bool {
if bytes.HasPrefix(url, bDataImage) && len(url) >= 11 {
v := url[11:]
if bytes.HasPrefix(v, bPng) || bytes.HasPrefix(v, bGif) ||
bytes.HasPrefix(v, bJpeg) || bytes.HasPrefix(v, bWebp) {
return false
}
return true
}
return bytes.HasPrefix(url, bJs) || bytes.HasPrefix(url, bVb) ||
bytes.HasPrefix(url, bFile) || bytes.HasPrefix(url, bData)
}

@ -0,0 +1,174 @@
// Package renderer renders the given AST to certain formats.
package renderer
import (
"bufio"
"io"
"sync"
"github.com/yuin/goldmark/ast"
"github.com/yuin/goldmark/util"
)
// A Config struct is a data structure that holds configuration of the Renderer.
type Config struct {
Options map[OptionName]interface{}
NodeRenderers util.PrioritizedSlice
}
// NewConfig returns a new Config
func NewConfig() *Config {
return &Config{
Options: map[OptionName]interface{}{},
NodeRenderers: util.PrioritizedSlice{},
}
}
// An OptionName is a name of the option.
type OptionName string
// An Option interface is a functional option type for the Renderer.
type Option interface {
SetConfig(*Config)
}
type withNodeRenderers struct {
value []util.PrioritizedValue
}
func (o *withNodeRenderers) SetConfig(c *Config) {
c.NodeRenderers = append(c.NodeRenderers, o.value...)
}
// WithNodeRenderers is a functional option that allow you to add
// NodeRenderers to the renderer.
func WithNodeRenderers(ps ...util.PrioritizedValue) Option {
return &withNodeRenderers{ps}
}
type withOption struct {
name OptionName
value interface{}
}
func (o *withOption) SetConfig(c *Config) {
c.Options[o.name] = o.value
}
// WithOption is a functional option that allow you to set
// an arbitrary option to the parser.
func WithOption(name OptionName, value interface{}) Option {
return &withOption{name, value}
}
// A SetOptioner interface sets given option to the object.
type SetOptioner interface {
// SetOption sets given option to the object.
// Unacceptable options may be passed.
// Thus implementations must ignore unacceptable options.
SetOption(name OptionName, value interface{})
}
// NodeRendererFunc is a function that renders a given node.
type NodeRendererFunc func(writer util.BufWriter, source []byte, n ast.Node, entering bool) (ast.WalkStatus, error)
// A NodeRenderer interface offers NodeRendererFuncs.
type NodeRenderer interface {
// RendererFuncs registers NodeRendererFuncs to given NodeRendererFuncRegisterer.
RegisterFuncs(NodeRendererFuncRegisterer)
}
// A NodeRendererFuncRegisterer registers
type NodeRendererFuncRegisterer interface {
// Register registers given NodeRendererFunc to this object.
Register(ast.NodeKind, NodeRendererFunc)
}
// A Renderer interface renders given AST node to given
// writer with given Renderer.
type Renderer interface {
Render(w io.Writer, source []byte, n ast.Node) error
// AddOptions adds given option to this renderer.
AddOptions(...Option)
}
type renderer struct {
config *Config
options map[OptionName]interface{}
nodeRendererFuncsTmp map[ast.NodeKind]NodeRendererFunc
maxKind int
nodeRendererFuncs []NodeRendererFunc
initSync sync.Once
}
// NewRenderer returns a new Renderer with given options.
func NewRenderer(options ...Option) Renderer {
config := NewConfig()
for _, opt := range options {
opt.SetConfig(config)
}
r := &renderer{
options: map[OptionName]interface{}{},
config: config,
nodeRendererFuncsTmp: map[ast.NodeKind]NodeRendererFunc{},
}
return r
}
func (r *renderer) AddOptions(opts ...Option) {
for _, opt := range opts {
opt.SetConfig(r.config)
}
}
func (r *renderer) Register(kind ast.NodeKind, v NodeRendererFunc) {
r.nodeRendererFuncsTmp[kind] = v
if int(kind) > r.maxKind {
r.maxKind = int(kind)
}
}
// Render renders the given AST node to the given writer with the given Renderer.
func (r *renderer) Render(w io.Writer, source []byte, n ast.Node) error {
r.initSync.Do(func() {
r.options = r.config.Options
r.config.NodeRenderers.Sort()
l := len(r.config.NodeRenderers)
for i := l - 1; i >= 0; i-- {
v := r.config.NodeRenderers[i]
nr, _ := v.Value.(NodeRenderer)
if se, ok := v.Value.(SetOptioner); ok {
for oname, ovalue := range r.options {
se.SetOption(oname, ovalue)
}
}
nr.RegisterFuncs(r)
}
r.nodeRendererFuncs = make([]NodeRendererFunc, r.maxKind+1)
for kind, nr := range r.nodeRendererFuncsTmp {
r.nodeRendererFuncs[kind] = nr
}
r.config = nil
r.nodeRendererFuncsTmp = nil
})
writer, ok := w.(util.BufWriter)
if !ok {
writer = bufio.NewWriter(w)
}
err := ast.Walk(n, func(n ast.Node, entering bool) (ast.WalkStatus, error) {
s := ast.WalkStatus(ast.WalkContinue)
var err error
f := r.nodeRendererFuncs[n.Kind()]
if f != nil {
s, err = f(writer, source, n, entering)
}
return s, err
})
if err != nil {
return err
}
return writer.Flush()
}

@ -0,0 +1,539 @@
package text
import (
"io"
"regexp"
"unicode/utf8"
"github.com/yuin/goldmark/util"
)
const invalidValue = -1
// EOF indicates the end of file.
const EOF = byte(0xff)
// A Reader interface provides abstracted method for reading text.
type Reader interface {
io.RuneReader
// Source returns a source of the reader.
Source() []byte
// ResetPosition resets positions.
ResetPosition()
// Peek returns a byte at current position without advancing the internal pointer.
Peek() byte
// PeekLine returns the current line without advancing the internal pointer.
PeekLine() ([]byte, Segment)
// PrecendingCharacter returns a character just before current internal pointer.
PrecendingCharacter() rune
// Value returns a value of the given segment.
Value(Segment) []byte
// LineOffset returns a distance from the line head to current position.
LineOffset() int
// Position returns current line number and position.
Position() (int, Segment)
// SetPosition sets current line number and position.
SetPosition(int, Segment)
// SetPadding sets padding to the reader.
SetPadding(int)
// Advance advances the internal pointer.
Advance(int)
// AdvanceAndSetPadding advances the internal pointer and add padding to the
// reader.
AdvanceAndSetPadding(int, int)
// AdvanceLine advances the internal pointer to the next line head.
AdvanceLine()
// SkipSpaces skips space characters and returns a non-blank line.
// If it reaches EOF, returns false.
SkipSpaces() (Segment, int, bool)
// SkipSpaces skips blank lines and returns a non-blank line.
// If it reaches EOF, returns false.
SkipBlankLines() (Segment, int, bool)
// Match performs regular expression matching to current line.
Match(reg *regexp.Regexp) bool
// Match performs regular expression searching to current line.
FindSubMatch(reg *regexp.Regexp) [][]byte
}
type reader struct {
source []byte
sourceLength int
line int
peekedLine []byte
pos Segment
head int
lineOffset int
}
// NewReader return a new Reader that can read UTF-8 bytes .
func NewReader(source []byte) Reader {
r := &reader{
source: source,
sourceLength: len(source),
}
r.ResetPosition()
return r
}
func (r *reader) ResetPosition() {
r.line = -1
r.head = 0
r.lineOffset = -1
r.AdvanceLine()
}
func (r *reader) Source() []byte {
return r.source
}
func (r *reader) Value(seg Segment) []byte {
return seg.Value(r.source)
}
func (r *reader) Peek() byte {
if r.pos.Start >= 0 && r.pos.Start < r.sourceLength {
if r.pos.Padding != 0 {
return space[0]
}
return r.source[r.pos.Start]
}
return EOF
}
func (r *reader) PeekLine() ([]byte, Segment) {
if r.pos.Start >= 0 && r.pos.Start < r.sourceLength {
if r.peekedLine == nil {
r.peekedLine = r.pos.Value(r.Source())
}
return r.peekedLine, r.pos
}
return nil, r.pos
}
// io.RuneReader interface
func (r *reader) ReadRune() (rune, int, error) {
return readRuneReader(r)
}
func (r *reader) LineOffset() int {
if r.lineOffset < 0 {
v := 0
for i := r.head; i < r.pos.Start; i++ {
if r.source[i] == '\t' {
v += util.TabWidth(v)
} else {
v++
}
}
r.lineOffset = v - r.pos.Padding
}
return r.lineOffset
}
func (r *reader) PrecendingCharacter() rune {
if r.pos.Start <= 0 {
if r.pos.Padding != 0 {
return rune(' ')
}
return rune('\n')
}
i := r.pos.Start - 1
for ; i >= 0; i-- {
if utf8.RuneStart(r.source[i]) {
break
}
}
rn, _ := utf8.DecodeRune(r.source[i:])
return rn
}
func (r *reader) Advance(n int) {
r.lineOffset = -1
if n < len(r.peekedLine) && r.pos.Padding == 0 {
r.pos.Start += n
r.peekedLine = nil
return
}
r.peekedLine = nil
l := r.sourceLength
for ; n > 0 && r.pos.Start < l; n-- {
if r.pos.Padding != 0 {
r.pos.Padding--
continue
}
if r.source[r.pos.Start] == '\n' {
r.AdvanceLine()
continue
}
r.pos.Start++
}
}
func (r *reader) AdvanceAndSetPadding(n, padding int) {
r.Advance(n)
if padding > r.pos.Padding {
r.SetPadding(padding)
}
}
func (r *reader) AdvanceLine() {
r.lineOffset = -1
r.peekedLine = nil
r.pos.Start = r.pos.Stop
r.head = r.pos.Start
if r.pos.Start < 0 {
return
}
r.pos.Stop = r.sourceLength
for i := r.pos.Start; i < r.sourceLength; i++ {
c := r.source[i]
if c == '\n' {
r.pos.Stop = i + 1
break
}
}
r.line++
r.pos.Padding = 0
}
func (r *reader) Position() (int, Segment) {
return r.line, r.pos
}
func (r *reader) SetPosition(line int, pos Segment) {
r.lineOffset = -1
r.line = line
r.pos = pos
}
func (r *reader) SetPadding(v int) {
r.pos.Padding = v
}
func (r *reader) SkipSpaces() (Segment, int, bool) {
return skipSpacesReader(r)
}
func (r *reader) SkipBlankLines() (Segment, int, bool) {
return skipBlankLinesReader(r)
}
func (r *reader) Match(reg *regexp.Regexp) bool {
return matchReader(r, reg)
}
func (r *reader) FindSubMatch(reg *regexp.Regexp) [][]byte {
return findSubMatchReader(r, reg)
}
// A BlockReader interface is a reader that is optimized for Blocks.
type BlockReader interface {
Reader
// Reset resets current state and sets new segments to the reader.
Reset(segment *Segments)
}
type blockReader struct {
source []byte
segments *Segments
segmentsLength int
line int
pos Segment
head int
last int
lineOffset int
}
// NewBlockReader returns a new BlockReader.
func NewBlockReader(source []byte, segments *Segments) BlockReader {
r := &blockReader{
source: source,
}
if segments != nil {
r.Reset(segments)
}
return r
}
func (r *blockReader) ResetPosition() {
r.line = -1
r.head = 0
r.last = 0
r.lineOffset = -1
r.pos.Start = -1
r.pos.Stop = -1
r.pos.Padding = 0
if r.segmentsLength > 0 {
last := r.segments.At(r.segmentsLength - 1)
r.last = last.Stop
}
r.AdvanceLine()
}
func (r *blockReader) Reset(segments *Segments) {
r.segments = segments
r.segmentsLength = segments.Len()
r.ResetPosition()
}
func (r *blockReader) Source() []byte {
return r.source
}
func (r *blockReader) Value(seg Segment) []byte {
line := r.segmentsLength - 1
ret := make([]byte, 0, seg.Stop-seg.Start+1)
for ; line >= 0; line-- {
if seg.Start >= r.segments.At(line).Start {
break
}
}
i := seg.Start
for ; line < r.segmentsLength; line++ {
s := r.segments.At(line)
if i < 0 {
i = s.Start
}
ret = s.ConcatPadding(ret)
for ; i < seg.Stop && i < s.Stop; i++ {
ret = append(ret, r.source[i])
}
i = -1
if s.Stop > seg.Stop {
break
}
}
return ret
}
// io.RuneReader interface
func (r *blockReader) ReadRune() (rune, int, error) {
return readRuneReader(r)
}
func (r *blockReader) PrecendingCharacter() rune {
if r.pos.Padding != 0 {
return rune(' ')
}
if r.pos.Start <= 0 {
return rune('\n')
}
l := len(r.source)
i := r.pos.Start - 1
for ; i < l && i >= 0; i-- {
if utf8.RuneStart(r.source[i]) {
break
}
}
if i < 0 || i >= l {
return rune('\n')
}
rn, _ := utf8.DecodeRune(r.source[i:])
return rn
}
func (r *blockReader) LineOffset() int {
if r.lineOffset < 0 {
v := 0
for i := r.head; i < r.pos.Start; i++ {
if r.source[i] == '\t' {
v += util.TabWidth(v)
} else {
v++
}
}
r.lineOffset = v - r.pos.Padding
}
return r.lineOffset
}
func (r *blockReader) Peek() byte {
if r.line < r.segmentsLength && r.pos.Start >= 0 && r.pos.Start < r.last {
if r.pos.Padding != 0 {
return space[0]
}
return r.source[r.pos.Start]
}
return EOF
}
func (r *blockReader) PeekLine() ([]byte, Segment) {
if r.line < r.segmentsLength && r.pos.Start >= 0 && r.pos.Start < r.last {
return r.pos.Value(r.source), r.pos
}
return nil, r.pos
}
func (r *blockReader) Advance(n int) {
r.lineOffset = -1
if n < r.pos.Stop-r.pos.Start && r.pos.Padding == 0 {
r.pos.Start += n
return
}
for ; n > 0; n-- {
if r.pos.Padding != 0 {
r.pos.Padding--
continue
}
if r.pos.Start >= r.pos.Stop-1 && r.pos.Stop < r.last {
r.AdvanceLine()
continue
}
r.pos.Start++
}
}
func (r *blockReader) AdvanceAndSetPadding(n, padding int) {
r.Advance(n)
if padding > r.pos.Padding {
r.SetPadding(padding)
}
}
func (r *blockReader) AdvanceLine() {
r.SetPosition(r.line+1, NewSegment(invalidValue, invalidValue))
r.head = r.pos.Start
}
func (r *blockReader) Position() (int, Segment) {
return r.line, r.pos
}
func (r *blockReader) SetPosition(line int, pos Segment) {
r.lineOffset = -1
r.line = line
if pos.Start == invalidValue {
if r.line < r.segmentsLength {
s := r.segments.At(line)
r.head = s.Start
r.pos = s
}
} else {
r.pos = pos
if r.line < r.segmentsLength {
s := r.segments.At(line)
r.head = s.Start
}
}
}
func (r *blockReader) SetPadding(v int) {
r.lineOffset = -1
r.pos.Padding = v
}
func (r *blockReader) SkipSpaces() (Segment, int, bool) {
return skipSpacesReader(r)
}
func (r *blockReader) SkipBlankLines() (Segment, int, bool) {
return skipBlankLinesReader(r)
}
func (r *blockReader) Match(reg *regexp.Regexp) bool {
return matchReader(r, reg)
}
func (r *blockReader) FindSubMatch(reg *regexp.Regexp) [][]byte {
return findSubMatchReader(r, reg)
}
func skipBlankLinesReader(r Reader) (Segment, int, bool) {
lines := 0
for {
line, seg := r.PeekLine()
if line == nil {
return seg, lines, false
}
if util.IsBlank(line) {
lines++
r.AdvanceLine()
} else {
return seg, lines, true
}
}
}
func skipSpacesReader(r Reader) (Segment, int, bool) {
chars := 0
for {
line, segment := r.PeekLine()
if line == nil {
return segment, chars, false
}
for i, c := range line {
if util.IsSpace(c) {
chars++
r.Advance(1)
continue
}
return segment.WithStart(segment.Start + i + 1), chars, true
}
}
}
func matchReader(r Reader, reg *regexp.Regexp) bool {
oldline, oldseg := r.Position()
match := reg.FindReaderSubmatchIndex(r)
r.SetPosition(oldline, oldseg)
if match == nil {
return false
}
r.Advance(match[1] - match[0])
return true
}
func findSubMatchReader(r Reader, reg *regexp.Regexp) [][]byte {
oldline, oldseg := r.Position()
match := reg.FindReaderSubmatchIndex(r)
r.SetPosition(oldline, oldseg)
if match == nil {
return nil
}
runes := make([]rune, 0, match[1]-match[0])
for i := 0; i < match[1]; {
r, size, _ := readRuneReader(r)
i += size
runes = append(runes, r)
}
result := [][]byte{}
for i := 0; i < len(match); i += 2 {
result = append(result, []byte(string(runes[match[i]:match[i+1]])))
}
r.SetPosition(oldline, oldseg)
r.Advance(match[1] - match[0])
return result
}
func readRuneReader(r Reader) (rune, int, error) {
line, _ := r.PeekLine()
if line == nil {
return 0, 0, io.EOF
}
rn, size := utf8.DecodeRune(line)
if rn == utf8.RuneError {
return 0, 0, io.EOF
}
r.Advance(size)
return rn, size, nil
}

@ -0,0 +1,209 @@
package text
import (
"bytes"
"github.com/yuin/goldmark/util"
)
var space = []byte(" ")
// A Segment struct holds information about source potisions.
type Segment struct {
// Start is a start position of the segment.
Start int
// Stop is a stop position of the segment.
// This value should be excluded.
Stop int
// Padding is a padding length of the segment.
Padding int
}
// NewSegment return a new Segment.
func NewSegment(start, stop int) Segment {
return Segment{
Start: start,
Stop: stop,
Padding: 0,
}
}
// NewSegmentPadding returns a new Segment with the given padding.
func NewSegmentPadding(start, stop, n int) Segment {
return Segment{
Start: start,
Stop: stop,
Padding: n,
}
}
// Value returns a value of the segment.
func (t *Segment) Value(buffer []byte) []byte {
if t.Padding == 0 {
return buffer[t.Start:t.Stop]
}
result := make([]byte, 0, t.Padding+t.Stop-t.Start+1)
result = append(result, bytes.Repeat(space, t.Padding)...)
return append(result, buffer[t.Start:t.Stop]...)
}
// Len returns a length of the segment.
func (t *Segment) Len() int {
return t.Stop - t.Start + t.Padding
}
// Between returns a segment between this segment and the given segment.
func (t *Segment) Between(other Segment) Segment {
if t.Stop != other.Stop {
panic("invalid state")
}
return NewSegmentPadding(
t.Start,
other.Start,
t.Padding-other.Padding,
)
}
// IsEmpty returns true if this segment is empty, otherwise false.
func (t *Segment) IsEmpty() bool {
return t.Start >= t.Stop && t.Padding == 0
}
// TrimRightSpace returns a new segment by slicing off all trailing
// space characters.
func (t *Segment) TrimRightSpace(buffer []byte) Segment {
v := buffer[t.Start:t.Stop]
l := util.TrimRightSpaceLength(v)
if l == len(v) {
return NewSegment(t.Start, t.Start)
}
return NewSegmentPadding(t.Start, t.Stop-l, t.Padding)
}
// TrimLeftSpace returns a new segment by slicing off all leading
// space characters including padding.
func (t *Segment) TrimLeftSpace(buffer []byte) Segment {
v := buffer[t.Start:t.Stop]
l := util.TrimLeftSpaceLength(v)
return NewSegment(t.Start+l, t.Stop)
}
// TrimLeftSpaceWidth returns a new segment by slicing off leading space
// characters until the given width.
func (t *Segment) TrimLeftSpaceWidth(width int, buffer []byte) Segment {
padding := t.Padding
for ; width > 0; width-- {
if padding == 0 {
break
}
padding--
}
if width == 0 {
return NewSegmentPadding(t.Start, t.Stop, padding)
}
text := buffer[t.Start:t.Stop]
start := t.Start
for _, c := range text {
if start >= t.Stop-1 || width <= 0 {
break
}
if c == ' ' {
width--
} else if c == '\t' {
width -= 4
} else {
break
}
start++
}
if width < 0 {
padding = width * -1
}
return NewSegmentPadding(start, t.Stop, padding)
}
// WithStart returns a new Segment with same value except Start.
func (t *Segment) WithStart(v int) Segment {
return NewSegmentPadding(v, t.Stop, t.Padding)
}
// WithStop returns a new Segment with same value except Stop.
func (t *Segment) WithStop(v int) Segment {
return NewSegmentPadding(t.Start, v, t.Padding)
}
// ConcatPadding concats the padding to the given slice.
func (t *Segment) ConcatPadding(v []byte) []byte {
if t.Padding > 0 {
return append(v, bytes.Repeat(space, t.Padding)...)
}
return v
}
// Segments is a collection of the Segment.
type Segments struct {
values []Segment
}
// NewSegments return a new Segments.
func NewSegments() *Segments {
return &Segments{
values: nil,
}
}
// Append appends the given segment after the tail of the collection.
func (s *Segments) Append(t Segment) {
if s.values == nil {
s.values = make([]Segment, 0, 20)
}
s.values = append(s.values, t)
}
// AppendAll appends all elements of given segments after the tail of the collection.
func (s *Segments) AppendAll(t []Segment) {
if s.values == nil {
s.values = make([]Segment, 0, 20)
}
s.values = append(s.values, t...)
}
// Len returns the length of the collection.
func (s *Segments) Len() int {
if s.values == nil {
return 0
}
return len(s.values)
}
// At returns a segment at the given index.
func (s *Segments) At(i int) Segment {
return s.values[i]
}
// Set sets the given Segment.
func (s *Segments) Set(i int, v Segment) {
s.values[i] = v
}
// SetSliced replace the collection with a subsliced value.
func (s *Segments) SetSliced(lo, hi int) {
s.values = s.values[lo:hi]
}
// Sliced returns a subslice of the collection.
func (s *Segments) Sliced(lo, hi int) []Segment {
return s.values[lo:hi]
}
// Clear delete all element of the collction.
func (s *Segments) Clear() {
s.values = nil
}
// Unshift insert the given Segment to head of the collection.
func (s *Segments) Unshift(v Segment) {
s.values = append(s.values[0:1], s.values[0:]...)
s.values[0] = v
}

File diff suppressed because it is too large Load Diff

@ -0,0 +1,897 @@
// Package util provides utility functions for the goldmark.
package util
import (
"bytes"
"io"
"net/url"
"regexp"
"sort"
"strconv"
"strings"
"unicode/utf8"
)
// A CopyOnWriteBuffer is a byte buffer that copies buffer when
// it need to be changed.
type CopyOnWriteBuffer struct {
buffer []byte
copied bool
}
// NewCopyOnWriteBuffer returns a new CopyOnWriteBuffer.
func NewCopyOnWriteBuffer(buffer []byte) CopyOnWriteBuffer {
return CopyOnWriteBuffer{
buffer: buffer,
copied: false,
}
}
// Write writes given bytes to the buffer.
func (b *CopyOnWriteBuffer) Write(value []byte) {
if !b.copied {
b.buffer = make([]byte, 0, len(b.buffer)+20)
b.copied = true
}
b.buffer = append(b.buffer, value...)
}
// WriteByte writes the given byte to the buffer.
func (b *CopyOnWriteBuffer) WriteByte(c byte) {
if !b.copied {
b.buffer = make([]byte, 0, len(b.buffer)+20)
b.copied = true
}
b.buffer = append(b.buffer, c)
}
// Bytes returns bytes of this buffer.
func (b *CopyOnWriteBuffer) Bytes() []byte {
return b.buffer
}
// IsCopied returns true if buffer has been copied, otherwise false.
func (b *CopyOnWriteBuffer) IsCopied() bool {
return b.copied
}
// IsEscapedPunctuation returns true if caracter at a given index i
// is an escaped punctuation, otherwise false.
func IsEscapedPunctuation(source []byte, i int) bool {
return source[i] == '\\' && i < len(source)-1 && IsPunct(source[i+1])
}
// ReadWhile read the given source while pred is true.
func ReadWhile(source []byte, index [2]int, pred func(byte) bool) (int, bool) {
j := index[0]
ok := false
for ; j < index[1]; j++ {
c1 := source[j]
if pred(c1) {
ok = true
continue
}
break
}
return j, ok
}
// IsBlank returns true if the given string is all space characters.
func IsBlank(bs []byte) bool {
for _, b := range bs {
if !IsSpace(b) {
return false
}
}
return true
}
// VisualizeSpaces visualize invisible space characters.
func VisualizeSpaces(bs []byte) []byte {
bs = bytes.Replace(bs, []byte(" "), []byte("[SPACE]"), -1)
bs = bytes.Replace(bs, []byte("\t"), []byte("[TAB]"), -1)
bs = bytes.Replace(bs, []byte("\n"), []byte("[NEWLINE]\n"), -1)
bs = bytes.Replace(bs, []byte("\r"), []byte("[CR]"), -1)
return bs
}
// TabWidth calculates actual width of a tab at the given position.
func TabWidth(currentPos int) int {
return 4 - currentPos%4
}
// IndentPosition searches an indent position with the given width for the given line.
// If the line contains tab characters, paddings may be not zero.
// currentPos==0 and width==2:
//
// position: 0 1
// [TAB]aaaa
// width: 1234 5678
//
// width=2 is in the tab character. In this case, IndentPosition returns
// (pos=1, padding=2)
func IndentPosition(bs []byte, currentPos, width int) (pos, padding int) {
if width == 0 {
return 0, 0
}
w := 0
l := len(bs)
i := 0
hasTab := false
for ; i < l; i++ {
if bs[i] == '\t' {
w += TabWidth(currentPos + w)
hasTab = true
} else if bs[i] == ' ' {
w++
} else {
break
}
}
if w >= width {
if !hasTab {
return width, 0
}
return i, w - width
}
return -1, -1
}
// IndentPositionPadding searches an indent position with the given width for the given line.
// This function is mostly same as IndentPosition except this function
// takes account into additional paddings.
func IndentPositionPadding(bs []byte, currentPos, paddingv, width int) (pos, padding int) {
if width == 0 {
return 0, paddingv
}
w := 0
i := 0
l := len(bs)
for ; i < l; i++ {
if bs[i] == '\t' {
w += TabWidth(currentPos + w)
} else if bs[i] == ' ' {
w++
} else {
break
}
}
if w >= width {
return i - paddingv, w - width
}
return -1, -1
}
// DedentPosition dedents lines by the given width.
func DedentPosition(bs []byte, currentPos, width int) (pos, padding int) {
if width == 0 {
return 0, 0
}
w := 0
l := len(bs)
i := 0
for ; i < l; i++ {
if bs[i] == '\t' {
w += TabWidth(currentPos + w)
} else if bs[i] == ' ' {
w++
} else {
break
}
}
if w >= width {
return i, w - width
}
return i, 0
}
// DedentPositionPadding dedents lines by the given width.
// This function is mostly same as DedentPosition except this function
// takes account into additional paddings.
func DedentPositionPadding(bs []byte, currentPos, paddingv, width int) (pos, padding int) {
if width == 0 {
return 0, paddingv
}
w := 0
i := 0
l := len(bs)
for ; i < l; i++ {
if bs[i] == '\t' {
w += TabWidth(currentPos + w)
} else if bs[i] == ' ' {
w++
} else {
break
}
}
if w >= width {
return i - paddingv, w - width
}
return i - paddingv, 0
}
// IndentWidth calculate an indent width for the given line.
func IndentWidth(bs []byte, currentPos int) (width, pos int) {
l := len(bs)
for i := 0; i < l; i++ {
b := bs[i]
if b == ' ' {
width++
pos++
} else if b == '\t' {
width += TabWidth(currentPos + width)
pos++
} else {
break
}
}
return
}
// FirstNonSpacePosition returns a potisoin line that is a first nonspace
// character.
func FirstNonSpacePosition(bs []byte) int {
i := 0
for ; i < len(bs); i++ {
c := bs[i]
if c == ' ' || c == '\t' {
continue
}
if c == '\n' {
return -1
}
return i
}
return -1
}
// FindClosure returns a position that closes the given opener.
// If codeSpan is set true, it ignores characters in code spans.
// If allowNesting is set true, closures correspond to nested opener will be
// ignored.
func FindClosure(bs []byte, opener, closure byte, codeSpan, allowNesting bool) int {
i := 0
opened := 1
codeSpanOpener := 0
for i < len(bs) {
c := bs[i]
if codeSpan && codeSpanOpener != 0 && c == '`' {
codeSpanCloser := 0
for ; i < len(bs); i++ {
if bs[i] == '`' {
codeSpanCloser++
} else {
i--
break
}
}
if codeSpanCloser == codeSpanOpener {
codeSpanOpener = 0
}
} else if c == '\\' && i < len(bs)-1 && IsPunct(bs[i+1]) {
i += 2
continue
} else if codeSpan && codeSpanOpener == 0 && c == '`' {
for ; i < len(bs); i++ {
if bs[i] == '`' {
codeSpanOpener++
} else {
i--
break
}
}
} else if (codeSpan && codeSpanOpener == 0) || !codeSpan {
if c == closure {
opened--
if opened == 0 {
return i
}
} else if c == opener {
if !allowNesting {
return -1
}
opened++
}
}
i++
}
return -1
}
// TrimLeft trims characters in the given s from head of the source.
// bytes.TrimLeft offers same functionalities, but bytes.TrimLeft
// allocates new buffer for the result.
func TrimLeft(source, b []byte) []byte {
i := 0
for ; i < len(source); i++ {
c := source[i]
found := false
for j := 0; j < len(b); j++ {
if c == b[j] {
found = true
break
}
}
if !found {
break
}
}
return source[i:]
}
// TrimRight trims characters in the given s from tail of the source.
func TrimRight(source, b []byte) []byte {
i := len(source) - 1
for ; i >= 0; i-- {
c := source[i]
found := false
for j := 0; j < len(b); j++ {
if c == b[j] {
found = true
break
}
}
if !found {
break
}
}
return source[:i+1]
}
// TrimLeftLength returns a length of leading specified characters.
func TrimLeftLength(source, s []byte) int {
return len(source) - len(TrimLeft(source, s))
}
// TrimRightLength returns a length of trailing specified characters.
func TrimRightLength(source, s []byte) int {
return len(source) - len(TrimRight(source, s))
}
// TrimLeftSpaceLength returns a length of leading space characters.
func TrimLeftSpaceLength(source []byte) int {
i := 0
for ; i < len(source); i++ {
if !IsSpace(source[i]) {
break
}
}
return i
}
// TrimRightSpaceLength returns a length of trailing space characters.
func TrimRightSpaceLength(source []byte) int {
l := len(source)
i := l - 1
for ; i >= 0; i-- {
if !IsSpace(source[i]) {
break
}
}
if i < 0 {
return l
}
return l - 1 - i
}
// TrimLeftSpace returns a subslice of the given string by slicing off all leading
// space characters.
func TrimLeftSpace(source []byte) []byte {
return TrimLeft(source, spaces)
}
// TrimRightSpace returns a subslice of the given string by slicing off all trailing
// space characters.
func TrimRightSpace(source []byte) []byte {
return TrimRight(source, spaces)
}
// ReplaceSpaces replaces sequence of spaces with the given repl.
func ReplaceSpaces(source []byte, repl byte) []byte {
var ret []byte
start := -1
for i, c := range source {
iss := IsSpace(c)
if start < 0 && iss {
start = i
continue
} else if start >= 0 && iss {
continue
} else if start >= 0 {
if ret == nil {
ret = make([]byte, 0, len(source))
ret = append(ret, source[:start]...)
}
ret = append(ret, repl)
start = -1
}
if ret != nil {
ret = append(ret, c)
}
}
if start >= 0 && ret != nil {
ret = append(ret, repl)
}
if ret == nil {
return source
}
return ret
}
// ToRune decode given bytes start at pos and returns a rune.
func ToRune(source []byte, pos int) rune {
i := pos
for ; i >= 0; i-- {
if utf8.RuneStart(source[i]) {
break
}
}
r, _ := utf8.DecodeRune(source[i:])
return r
}
// ToValidRune returns 0xFFFD if the given rune is invalid, otherwise v.
func ToValidRune(v rune) rune {
if v == 0 || !utf8.ValidRune(v) {
return rune(0xFFFD)
}
return v
}
// ToLinkReference convert given bytes into a valid link reference string.
// ToLinkReference trims leading and trailing spaces and convert into lower
// case and replace spaces with a single space character.
func ToLinkReference(v []byte) string {
v = TrimLeftSpace(v)
v = TrimRightSpace(v)
return strings.ToLower(string(ReplaceSpaces(v, ' ')))
}
var htmlEscapeTable = [256][]byte{nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, []byte("&quot;"), nil, nil, nil, []byte("&amp;"), nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, []byte("&lt;"), nil, []byte("&gt;"), nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil, nil}
// EscapeHTMLByte returns HTML escaped bytes if the given byte should be escaped,
// otherwise nil.
func EscapeHTMLByte(b byte) []byte {
return htmlEscapeTable[b]
}
// EscapeHTML escapes characters that should be escaped in HTML text.
func EscapeHTML(v []byte) []byte {
cob := NewCopyOnWriteBuffer(v)
n := 0
for i := 0; i < len(v); i++ {
c := v[i]
escaped := htmlEscapeTable[c]
if escaped != nil {
cob.Write(v[n:i])
cob.Write(escaped)
n = i + 1
}
}
if cob.IsCopied() {
cob.Write(v[n:])
}
return cob.Bytes()
}
// UnescapePunctuations unescapes blackslash escaped punctuations.
func UnescapePunctuations(source []byte) []byte {
cob := NewCopyOnWriteBuffer(source)
limit := len(source)
n := 0
for i := 0; i < limit; {
c := source[i]
if i < limit-1 && c == '\\' && IsPunct(source[i+1]) {
cob.Write(source[n:i])
cob.WriteByte(source[i+1])
i += 2
n = i
continue
}
i++
}
if cob.IsCopied() {
cob.Write(source[n:])
}
return cob.Bytes()
}
// ResolveNumericReferences resolve numeric references like '&#1234;" .
func ResolveNumericReferences(source []byte) []byte {
cob := NewCopyOnWriteBuffer(source)
buf := make([]byte, 6, 6)
limit := len(source)
ok := false
n := 0
for i := 0; i < limit; i++ {
if source[i] == '&' {
pos := i
next := i + 1
if next < limit && source[next] == '#' {
nnext := next + 1
if nnext < limit {
nc := source[nnext]
// code point like #x22;
if nnext < limit && nc == 'x' || nc == 'X' {
start := nnext + 1
i, ok = ReadWhile(source, [2]int{start, limit}, IsHexDecimal)
if ok && i < limit && source[i] == ';' {
v, _ := strconv.ParseUint(BytesToReadOnlyString(source[start:i]), 16, 32)
cob.Write(source[n:pos])
n = i + 1
runeSize := utf8.EncodeRune(buf, ToValidRune(rune(v)))
cob.Write(buf[:runeSize])
continue
}
// code point like #1234;
} else if nc >= '0' && nc <= '9' {
start := nnext
i, ok = ReadWhile(source, [2]int{start, limit}, IsNumeric)
if ok && i < limit && i-start < 8 && source[i] == ';' {
v, _ := strconv.ParseUint(BytesToReadOnlyString(source[start:i]), 0, 32)
cob.Write(source[n:pos])
n = i + 1
runeSize := utf8.EncodeRune(buf, ToValidRune(rune(v)))
cob.Write(buf[:runeSize])
continue
}
}
}
}
i = next - 1
}
}
if cob.IsCopied() {
cob.Write(source[n:])
}
return cob.Bytes()
}
// ResolveEntityNames resolve entity references like '&ouml;" .
func ResolveEntityNames(source []byte) []byte {
cob := NewCopyOnWriteBuffer(source)
limit := len(source)
ok := false
n := 0
for i := 0; i < limit; i++ {
if source[i] == '&' {
pos := i
next := i + 1
if !(next < limit && source[next] == '#') {
start := next
i, ok = ReadWhile(source, [2]int{start, limit}, IsAlphaNumeric)
if ok && i < limit && source[i] == ';' {
name := BytesToReadOnlyString(source[start:i])
entity, ok := LookUpHTML5EntityByName(name)
if ok {
cob.Write(source[n:pos])
n = i + 1
cob.Write(entity.Characters)
continue
}
}
}
i = next - 1
}
}
if cob.IsCopied() {
cob.Write(source[n:])
}
return cob.Bytes()
}
var htmlSpace = []byte("%20")
// URLEscape escape the given URL.
// If resolveReference is set true:
// 1. unescape punctuations
// 2. resolve numeric references
// 3. resolve entity references
//
// URL encoded values (%xx) are keeped as is.
func URLEscape(v []byte, resolveReference bool) []byte {
if resolveReference {
v = UnescapePunctuations(v)
v = ResolveNumericReferences(v)
v = ResolveEntityNames(v)
}
cob := NewCopyOnWriteBuffer(v)
limit := len(v)
n := 0
for i := 0; i < limit; {
c := v[i]
if urlEscapeTable[c] == 1 {
i++
continue
}
if c == '%' && i+2 < limit && IsHexDecimal(v[i+1]) && IsHexDecimal(v[i+1]) {
i += 3
continue
}
u8len := utf8lenTable[c]
if u8len == 99 { // invalid utf8 leading byte, skip it
i++
continue
}
if c == ' ' {
cob.Write(v[n:i])
cob.Write(htmlSpace)
i++
n = i
continue
}
if int(u8len) >= len(v) {
u8len = int8(len(v) - 1)
}
if u8len == 0 {
i++
n = i
continue
}
cob.Write(v[n:i])
stop := i + int(u8len)
if stop > len(v) {
i++
n = i
continue
}
cob.Write(StringToReadOnlyBytes(url.QueryEscape(string(v[i:stop]))))
i += int(u8len)
n = i
}
if cob.IsCopied() && n < limit {
cob.Write(v[n:])
}
return cob.Bytes()
}
// FindURLIndex returns a stop index value if the given bytes seem an URL.
// This function is equivalent to [A-Za-z][A-Za-z0-9.+-]{1,31}:[^<>\x00-\x20]* .
func FindURLIndex(b []byte) int {
i := 0
if !(len(b) > 0 && urlTable[b[i]]&7 == 7) {
return -1
}
i++
for ; i < len(b); i++ {
c := b[i]
if urlTable[c]&4 != 4 {
break
}
}
if i == 1 || i > 33 || i >= len(b) {
return -1
}
if b[i] != ':' {
return -1
}
i++
for ; i < len(b); i++ {
c := b[i]
if urlTable[c]&1 != 1 {
break
}
}
return i
}
var emailDomainRegexp = regexp.MustCompile(`^[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*`)
// FindEmailIndex returns a stop index value if the given bytes seem an email address.
func FindEmailIndex(b []byte) int {
// TODO: eliminate regexps
i := 0
for ; i < len(b); i++ {
c := b[i]
if emailTable[c]&1 != 1 {
break
}
}
if i == 0 {
return -1
}
if i >= len(b) || b[i] != '@' {
return -1
}
i++
if i >= len(b) {
return -1
}
match := emailDomainRegexp.FindSubmatchIndex(b[i:])
if match == nil {
return -1
}
return i + match[1]
}
var spaces = []byte(" \t\n\x0b\x0c\x0d")
var spaceTable = [256]int8{0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
var punctTable = [256]int8{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
// a-zA-Z0-9, ;/?:@&=+$,-_.!~*'()#
var urlEscapeTable = [256]int8{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
var utf8lenTable = [256]int8{1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 3, 4, 4, 4, 4, 4, 4, 4, 4, 99, 99, 99, 99, 99, 99, 99, 99}
var urlTable = [256]uint8{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 5, 1, 5, 5, 1, 5, 5, 5, 5, 5, 5, 5, 5, 5, 5, 1, 1, 0, 1, 0, 1, 1, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 1, 1, 1, 1, 1, 1, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 7, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1}
var emailTable = [256]uint8{0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 1, 1, 1, 1, 0, 0, 1, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0}
// UTF8Len returns a byte length of the utf-8 character.
func UTF8Len(b byte) int8 {
return utf8lenTable[b]
}
// IsPunct returns true if the given character is a punctuation, otherwise false.
func IsPunct(c byte) bool {
return punctTable[c] == 1
}
// IsSpace returns true if the given character is a space, otherwise false.
func IsSpace(c byte) bool {
return spaceTable[c] == 1
}
// IsNumeric returns true if the given character is a numeric, otherwise false.
func IsNumeric(c byte) bool {
return c >= '0' && c <= '9'
}
// IsHexDecimal returns true if the given character is a hexdecimal, otherwise false.
func IsHexDecimal(c byte) bool {
return c >= '0' && c <= '9' || c >= 'a' && c <= 'f' || c >= 'A' && c <= 'F'
}
// IsAlphaNumeric returns true if the given character is a alphabet or a numeric, otherwise false.
func IsAlphaNumeric(c byte) bool {
return c >= 'a' && c <= 'z' || c >= 'A' && c <= 'Z' || c >= '0' && c <= '9'
}
// A BufWriter is a subset of the bufio.Writer .
type BufWriter interface {
io.Writer
Available() int
Buffered() int
Flush() error
WriteByte(c byte) error
WriteRune(r rune) (size int, err error)
WriteString(s string) (int, error)
}
// A PrioritizedValue struct holds pair of an arbitrary value and a priority.
type PrioritizedValue struct {
// Value is an arbitrary value that you want to prioritize.
Value interface{}
// Priority is a priority of the value.
Priority int
}
// PrioritizedSlice is a slice of the PrioritizedValues
type PrioritizedSlice []PrioritizedValue
// Sort sorts the PrioritizedSlice in ascending order.
func (s PrioritizedSlice) Sort() {
sort.Slice(s, func(i, j int) bool {
return s[i].Priority < s[j].Priority
})
}
// Remove removes the given value from this slice.
func (s PrioritizedSlice) Remove(v interface{}) PrioritizedSlice {
i := 0
found := false
for ; i < len(s); i++ {
if s[i].Value == v {
found = true
break
}
}
if !found {
return s
}
return append(s[:i], s[i+1:]...)
}
// Prioritized returns a new PrioritizedValue.
func Prioritized(v interface{}, priority int) PrioritizedValue {
return PrioritizedValue{v, priority}
}
func bytesHash(b []byte) uint64 {
var hash uint64 = 5381
for _, c := range b {
hash = ((hash << 5) + hash) + uint64(c)
}
return hash
}
// BytesFilter is a efficient data structure for checking whether bytes exist or not.
// BytesFilter is thread-safe.
type BytesFilter interface {
// Add adds given bytes to this set.
Add([]byte)
// Contains return true if this set contains given bytes, otherwise false.
Contains([]byte) bool
// Extend copies this filter and adds given bytes to new filter.
Extend(...[]byte) BytesFilter
}
type bytesFilter struct {
chars [256]uint8
threshold int
slots [][][]byte
}
// NewBytesFilter returns a new BytesFilter.
func NewBytesFilter(elements ...[]byte) BytesFilter {
s := &bytesFilter{
threshold: 3,
slots: make([][][]byte, 64),
}
for _, element := range elements {
s.Add(element)
}
return s
}
func (s *bytesFilter) Add(b []byte) {
l := len(b)
m := s.threshold
if l < s.threshold {
m = l
}
for i := 0; i < m; i++ {
s.chars[b[i]] |= 1 << uint8(i)
}
h := bytesHash(b) % uint64(len(s.slots))
slot := s.slots[h]
if slot == nil {
slot = [][]byte{}
}
s.slots[h] = append(slot, b)
}
func (s *bytesFilter) Extend(bs ...[]byte) BytesFilter {
newFilter := NewBytesFilter().(*bytesFilter)
newFilter.chars = s.chars
newFilter.threshold = s.threshold
for k, v := range s.slots {
newSlot := make([][]byte, len(v))
copy(newSlot, v)
newFilter.slots[k] = v
}
for _, b := range bs {
newFilter.Add(b)
}
return newFilter
}
func (s *bytesFilter) Contains(b []byte) bool {
l := len(b)
m := s.threshold
if l < s.threshold {
m = l
}
for i := 0; i < m; i++ {
if (s.chars[b[i]] & (1 << uint8(i))) == 0 {
return false
}
}
h := bytesHash(b) % uint64(len(s.slots))
slot := s.slots[h]
if slot == nil || len(slot) == 0 {
return false
}
for _, element := range slot {
if bytes.Equal(element, b) {
return true
}
}
return false
}

@ -0,0 +1,13 @@
// +build appengine,js
package util
// BytesToReadOnlyString returns a string converted from given bytes.
func BytesToReadOnlyString(b []byte) string {
return string(b)
}
// StringToReadOnlyBytes returns bytes converted from given string.
func StringToReadOnlyBytes(s string) []byte {
return []byte(s)
}

@ -0,0 +1,20 @@
// +build !appengine,!js
package util
import (
"reflect"
"unsafe"
)
// BytesToReadOnlyString returns a string converted from given bytes.
func BytesToReadOnlyString(b []byte) string {
return *(*string)(unsafe.Pointer(&b))
}
// StringToReadOnlyBytes returns bytes converted from given string.
func StringToReadOnlyBytes(s string) []byte {
sh := (*reflect.StringHeader)(unsafe.Pointer(&s))
bh := reflect.SliceHeader{Data: sh.Data, Len: sh.Len, Cap: sh.Len}
return *(*[]byte)(unsafe.Pointer(&bh))
}

14
vendor/modules.txt vendored

@ -361,16 +361,12 @@ github.com/prometheus/procfs/internal/util
# github.com/quasoft/websspi v1.0.0
github.com/quasoft/websspi
github.com/quasoft/websspi/secctx
# github.com/russross/blackfriday/v2 v2.0.1
github.com/russross/blackfriday/v2
# github.com/satori/go.uuid v1.2.0
github.com/satori/go.uuid
# github.com/sergi/go-diff v1.0.0
github.com/sergi/go-diff/diffmatchpatch
# github.com/shurcooL/httpfs v0.0.0-20190527155220-6a4d4a70508b
github.com/shurcooL/httpfs/vfsutil
# github.com/shurcooL/sanitized_anchor_name v1.0.0
github.com/shurcooL/sanitized_anchor_name
# github.com/shurcooL/vfsgen v0.0.0-20181202132449-6a9ea43bcacd
github.com/shurcooL/vfsgen
# github.com/siddontang/go-snappy v0.0.0-20140704025258-d8f7bb82a96d
@ -432,6 +428,16 @@ github.com/willf/bitset
github.com/xanzy/ssh-agent
# github.com/yohcop/openid-go v0.0.0-20160914080427-2c050d2dae53
github.com/yohcop/openid-go
# github.com/yuin/goldmark v1.1.19
github.com/yuin/goldmark
github.com/yuin/goldmark/ast
github.com/yuin/goldmark/extension
github.com/yuin/goldmark/extension/ast
github.com/yuin/goldmark/parser
github.com/yuin/goldmark/renderer
github.com/yuin/goldmark/renderer/html
github.com/yuin/goldmark/text
github.com/yuin/goldmark/util
# go.mongodb.org/mongo-driver v1.1.1
go.mongodb.org/mongo-driver/bson
go.mongodb.org/mongo-driver/bson/bsoncodec

Loading…
Cancel
Save