1
0
Fork 0
forked from forgejo/forgejo

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>
This commit is contained in:
zeripath 2019-12-31 01:53:28 +00:00 committed by Lauris BH
parent 0c07f1de5b
commit 27757714d0
83 changed files with 13838 additions and 6297 deletions

View file

@ -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),
))
}

View file

@ -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?://")
)

View file

@ -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),
),
)
}

View file

@ -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
}

View file

@ -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>`)
}

View file

@ -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
}

View file

@ -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"
giteautil "code.gitea.io/gitea/modules/util"
"github.com/russross/blackfriday/v2"
"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 urlPrefixKey = parser.NewContextKey()
var isWikiKey = parser.NewContextKey()
// 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
}
var byteMailto = []byte("mailto:")
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])
}
}
// 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())
}
})
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)
}
exts := blackfridayExtensions
if setting.Markdown.EnableHardLineBreak {
exts |= blackfriday.HardLineBreak
}
// 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 {

View file

@ -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)

View file

@ -6,33 +6,86 @@ package mdstripper
import (
"bytes"
"sync"
"io"
"github.com/russross/blackfriday/v2"
"code.gitea.io/gitea/modules/log"
"code.gitea.io/gitea/modules/markup/common"
"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"
)
// MarkdownStripper extends blackfriday.Renderer
type MarkdownStripper struct {
links []string
coallesce bool
empty bool
type stripRenderer struct {
links []string
empty bool
}
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
)
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
})
}
func (r *stripRenderer) doubleSpace(w io.Writer) {
if !r.empty {
_, _ = w.Write([]byte{'\n'})
}
}
func (r *stripRenderer) processString(w io.Writer, text []byte, coalesce bool) {
// Always break-up words
if !coalesce {
r.doubleSpace(w)
}
_, _ = w.Write(text)
r.empty = false
}
func (r *stripRenderer) processLink(w io.Writer, link []byte) {
// Links are processed out of band
r.links = append(r.links, string(link))
}
// GetLinks returns the list of link data collected while parsing
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
@ -41,78 +94,40 @@ func StripMarkdown(rawBytes []byte) (string, []string) {
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) {
stripper := &MarkdownStripper{
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,
}
parser := blackfriday.New(blackfriday.WithRenderer(stripper), blackfriday.WithExtensions(blackfridayExtensions))
ast := parser.Parse(rawBytes)
reader := text.NewReader(rawBytes)
doc := stripParser.Parse(reader)
var buf bytes.Buffer
stripper.RenderHeader(&buf, ast)
ast.Walk(func(node *blackfriday.Node, entering bool) blackfriday.WalkStatus {
return stripper.RenderNode(&buf, node, entering)
})
stripper.RenderFooter(&buf, ast)
if err := stripper.Render(&buf, rawBytes, doc); err != nil {
log.Error("Unable to strip: %v", err)
}
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) {
if !r.empty {
_, _ = w.Write([]byte{'\n'})
}
}
func (r *MarkdownStripper) processString(w io.Writer, text []byte, coallesce bool) {
// Always break-up words
if !coallesce || !r.coallesce {
r.doubleSpace(w)
}
_, _ = w.Write(text)
r.coallesce = coallesce
r.empty = false
}
func (r *MarkdownStripper) 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 {
return r.links
}

View file

@ -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 {

View file

@ -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 {

View file

@ -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{