StarMirror/blackfriday
1
0
Fork 0
mirror of https://github.com/russross/blackfriday.git synced 2024-03-22 13:40:34 +08:00
Code Issues Projects Releases Wiki Activity

Merge pull request #365 from Ambrevar/v2

Browse Source 
Merge Parser into Processor
This commit is contained in:
Vytautas Šaltenis 2017-06-09 19:18:05 +03:00 committed by GitHub
commit 70c446a327
5 changed files with 133 additions and 146 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

90
block.go
View File

@ -34,7 +34,7 @@ var (
// Parse block-level data. // Parse block-level data.
// Note: this function and many that it calls assume that // Note: this function and many that it calls assume that
// the input buffer ends with a newline. // the input buffer ends with a newline.
func (p *Parser) block(data []byte) { func (p *Markdown) block(data []byte) {
// this is called recursively: enforce a maximum depth // this is called recursively: enforce a maximum depth
if p.nesting >= p.maxNesting { if p.nesting >= p.maxNesting {
return return
@ -71,7 +71,7 @@ func (p *Parser) block(data []byte) {
// % stuff // % stuff
// % more stuff // % more stuff
// % even more stuff // % even more stuff
if p.flags&Titleblock != 0 { if p.extensions&Titleblock != 0 {
if data[0] == '%' { if data[0] == '%' {
if i := p.titleBlock(data, true); i > 0 { if i := p.titleBlock(data, true); i > 0 {
data = data[i:] data = data[i:]
@ -109,7 +109,7 @@ func (p *Parser) block(data []byte) {
// return n * fact(n-1) // return n * fact(n-1)
// } // }
// ``` // ```
if p.flags&FencedCode != 0 { if p.extensions&FencedCode != 0 {
if i := p.fencedCodeBlock(data, true); i > 0 { if i := p.fencedCodeBlock(data, true); i > 0 {
data = data[i:] data = data[i:]
continue continue
@ -147,7 +147,7 @@ func (p *Parser) block(data []byte) {
// ------|-----|--------- // ------|-----|---------
// Bob | 31 | 555-1234 // Bob | 31 | 555-1234
// Alice | 27 | 555-4321 // Alice | 27 | 555-4321
if p.flags&Tables != 0 { if p.extensions&Tables != 0 {
if i := p.table(data); i > 0 { if i := p.table(data); i > 0 {
data = data[i:] data = data[i:]
continue continue
@ -182,7 +182,7 @@ func (p *Parser) block(data []byte) {
// //
// Term 2 // Term 2
// : Definition c // : Definition c
if p.flags&DefinitionLists != 0 { if p.extensions&DefinitionLists != 0 {
if p.dliPrefix(data) > 0 { if p.dliPrefix(data) > 0 {
data = data[p.list(data, ListTypeDefinition):] data = data[p.list(data, ListTypeDefinition):]
continue continue
@ -197,19 +197,19 @@ func (p *Parser) block(data []byte) {
p.nesting-- p.nesting--
} }
func (p *Parser) addBlock(typ NodeType, content []byte) *Node { func (p *Markdown) addBlock(typ NodeType, content []byte) *Node {
p.closeUnmatchedBlocks() p.closeUnmatchedBlocks()
container := p.addChild(typ, 0) container := p.addChild(typ, 0)
container.content = content container.content = content
return container return container
} }
func (p *Parser) isPrefixHeading(data []byte) bool { func (p *Markdown) isPrefixHeading(data []byte) bool {
if data[0] != '#' { if data[0] != '#' {
return false return false
} }
if p.flags&SpaceHeadings != 0 { if p.extensions&SpaceHeadings != 0 {
level := 0 level := 0
for level < 6 && level < len(data) && data[level] == '#' { for level < 6 && level < len(data) && data[level] == '#' {
level++ level++
@ -221,7 +221,7 @@ func (p *Parser) isPrefixHeading(data []byte) bool {
return true return true
} }
func (p *Parser) prefixHeading(data []byte) int { func (p *Markdown) prefixHeading(data []byte) int {
level := 0 level := 0
for level < 6 && level < len(data) && data[level] == '#' { for level < 6 && level < len(data) && data[level] == '#' {
level++ level++
@ -230,7 +230,7 @@ func (p *Parser) prefixHeading(data []byte) int {
end := skipUntilChar(data, i, '\n') end := skipUntilChar(data, i, '\n')
skip := end skip := end
id := "" id := ""
if p.flags&HeadingIDs != 0 { if p.extensions&HeadingIDs != 0 {
j, k := 0, 0 j, k := 0, 0
// find start/end of heading id // find start/end of heading id
for j = i; j < end-1 && (data[j] != '{' || data[j+1] != '#'); j++ { for j = i; j < end-1 && (data[j] != '{' || data[j+1] != '#'); j++ {
@ -257,7 +257,7 @@ func (p *Parser) prefixHeading(data []byte) int {
end-- end--
} }
if end > i { if end > i {
if id == "" && p.flags&AutoHeadingIDs != 0 { if id == "" && p.extensions&AutoHeadingIDs != 0 {
id = sanitized_anchor_name.Create(string(data[i:end])) id = sanitized_anchor_name.Create(string(data[i:end]))
} }
block := p.addBlock(Heading, data[i:end]) block := p.addBlock(Heading, data[i:end])
@ -267,7 +267,7 @@ func (p *Parser) prefixHeading(data []byte) int {
return skip return skip
} }
func (p *Parser) isUnderlinedHeading(data []byte) int { func (p *Markdown) isUnderlinedHeading(data []byte) int {
// test of level 1 heading // test of level 1 heading
if data[0] == '=' { if data[0] == '=' {
i := skipChar(data, 1, '=') i := skipChar(data, 1, '=')
@ -291,7 +291,7 @@ func (p *Parser) isUnderlinedHeading(data []byte) int {
return 0 return 0
} }
func (p *Parser) titleBlock(data []byte, doRender bool) int { func (p *Markdown) titleBlock(data []byte, doRender bool) int {
if data[0] != '%' { if data[0] != '%' {
return 0 return 0
} }
@ -315,7 +315,7 @@ func (p *Parser) titleBlock(data []byte, doRender bool) int {
return consumed return consumed
} }
func (p *Parser) html(data []byte, doRender bool) int { func (p *Markdown) html(data []byte, doRender bool) int {
var i, j int var i, j int
// identify the opening tag // identify the opening tag
@ -419,7 +419,7 @@ func finalizeHTMLBlock(block *Node) {
} }
// HTML comment, lax form // HTML comment, lax form
func (p *Parser) htmlComment(data []byte, doRender bool) int { func (p *Markdown) htmlComment(data []byte, doRender bool) int {
i := p.inlineHTMLComment(data) i := p.inlineHTMLComment(data)
// needs to end with a blank line // needs to end with a blank line
if j := p.isEmpty(data[i:]); j > 0 { if j := p.isEmpty(data[i:]); j > 0 {
@ -439,7 +439,7 @@ func (p *Parser) htmlComment(data []byte, doRender bool) int {
} }
// HR, which is the only self-closing block tag considered // HR, which is the only self-closing block tag considered
func (p *Parser) htmlHr(data []byte, doRender bool) int { func (p *Markdown) htmlHr(data []byte, doRender bool) int {
if len(data) < 4 { if len(data) < 4 {
return 0 return 0
} }
@ -472,7 +472,7 @@ func (p *Parser) htmlHr(data []byte, doRender bool) int {
return 0 return 0
} }
func (p *Parser) htmlFindTag(data []byte) (string, bool) { func (p *Markdown) htmlFindTag(data []byte) (string, bool) {
i := 0 i := 0
for i < len(data) && isalnum(data[i]) { for i < len(data) && isalnum(data[i]) {
i++ i++
@ -484,7 +484,7 @@ func (p *Parser) htmlFindTag(data []byte) (string, bool) {
return "", false return "", false
} }
func (p *Parser) htmlFindEnd(tag string, data []byte) int { func (p *Markdown) htmlFindEnd(tag string, data []byte) int {
// assume data[0] == '<' && data[1] == '/' already tested // assume data[0] == '<' && data[1] == '/' already tested
if tag == "hr" { if tag == "hr" {
return 2 return 2
@ -508,7 +508,7 @@ func (p *Parser) htmlFindEnd(tag string, data []byte) int {
return i return i
} }
if p.flags&LaxHTMLBlocks != 0 { if p.extensions&LaxHTMLBlocks != 0 {
return i return i
} }
if skip = p.isEmpty(data[i:]); skip == 0 { if skip = p.isEmpty(data[i:]); skip == 0 {
@ -519,7 +519,7 @@ func (p *Parser) htmlFindEnd(tag string, data []byte) int {
return i + skip return i + skip
} }
func (*Parser) isEmpty(data []byte) int { func (*Markdown) isEmpty(data []byte) int {
// it is okay to call isEmpty on an empty buffer // it is okay to call isEmpty on an empty buffer
if len(data) == 0 { if len(data) == 0 {
return 0 return 0
@ -537,7 +537,7 @@ func (*Parser) isEmpty(data []byte) int {
return i return i
} }
func (*Parser) isHRule(data []byte) bool { func (*Markdown) isHRule(data []byte) bool {
i := 0 i := 0
// skip up to three spaces // skip up to three spaces
@ -667,7 +667,7 @@ func isFenceLine(data []byte, syntax *string, oldmarker string) (end int, marker
// fencedCodeBlock returns the end index if data contains a fenced code block at the beginning, // fencedCodeBlock returns the end index if data contains a fenced code block at the beginning,
// or 0 otherwise. It writes to out if doRender is true, otherwise it has no side effects. // or 0 otherwise. It writes to out if doRender is true, otherwise it has no side effects.
// If doRender is true, a final newline is mandatory to recognize the fenced code block. // If doRender is true, a final newline is mandatory to recognize the fenced code block.
func (p *Parser) fencedCodeBlock(data []byte, doRender bool) int { func (p *Markdown) fencedCodeBlock(data []byte, doRender bool) int {
var syntax string var syntax string
beg, marker := isFenceLine(data, &syntax, "") beg, marker := isFenceLine(data, &syntax, "")
if beg == 0 || beg >= len(data) { if beg == 0 || beg >= len(data) {
@ -739,7 +739,7 @@ func finalizeCodeBlock(block *Node) {
block.content = nil block.content = nil
} }
func (p *Parser) table(data []byte) int { func (p *Markdown) table(data []byte) int {
table := p.addBlock(Table, nil) table := p.addBlock(Table, nil)
i, columns := p.tableHeader(data) i, columns := p.tableHeader(data)
if i == 0 { if i == 0 {
@ -782,7 +782,7 @@ func isBackslashEscaped(data []byte, i int) bool {
return backslashes&1 == 1 return backslashes&1 == 1
} }
func (p *Parser) tableHeader(data []byte) (size int, columns []CellAlignFlags) { func (p *Markdown) tableHeader(data []byte) (size int, columns []CellAlignFlags) {
i := 0 i := 0
colCount := 1 colCount := 1
for i = 0; i < len(data) && data[i] != '\n'; i++ { for i = 0; i < len(data) && data[i] != '\n'; i++ {
@ -895,7 +895,7 @@ func (p *Parser) tableHeader(data []byte) (size int, columns []CellAlignFlags) {
return return
} }
func (p *Parser) tableRow(data []byte, columns []CellAlignFlags, header bool) { func (p *Markdown) tableRow(data []byte, columns []CellAlignFlags, header bool) {
p.addBlock(TableRow, nil) p.addBlock(TableRow, nil)
i, col := 0, 0 i, col := 0, 0
@ -939,7 +939,7 @@ func (p *Parser) tableRow(data []byte, columns []CellAlignFlags, header bool) {
} }
// returns blockquote prefix length // returns blockquote prefix length
func (p *Parser) quotePrefix(data []byte) int { func (p *Markdown) quotePrefix(data []byte) int {
i := 0 i := 0
for i < 3 && i < len(data) && data[i] == ' ' { for i < 3 && i < len(data) && data[i] == ' ' {
i++ i++
@ -955,7 +955,7 @@ func (p *Parser) quotePrefix(data []byte) int {
// blockquote ends with at least one blank line // blockquote ends with at least one blank line
// followed by something without a blockquote prefix // followed by something without a blockquote prefix
func (p *Parser) terminateBlockquote(data []byte, beg, end int) bool { func (p *Markdown) terminateBlockquote(data []byte, beg, end int) bool {
if p.isEmpty(data[beg:]) <= 0 { if p.isEmpty(data[beg:]) <= 0 {
return false return false
} }
@ -966,7 +966,7 @@ func (p *Parser) terminateBlockquote(data []byte, beg, end int) bool {
} }
// parse a blockquote fragment // parse a blockquote fragment
func (p *Parser) quote(data []byte) int { func (p *Markdown) quote(data []byte) int {
block := p.addBlock(BlockQuote, nil) block := p.addBlock(BlockQuote, nil)
var raw bytes.Buffer var raw bytes.Buffer
beg, end := 0, 0 beg, end := 0, 0
@ -976,7 +976,7 @@ func (p *Parser) quote(data []byte) int {
// fenced code and if one's found, incorporate it altogether, // fenced code and if one's found, incorporate it altogether,
// irregardless of any contents inside it // irregardless of any contents inside it
for end < len(data) && data[end] != '\n' { for end < len(data) && data[end] != '\n' {
if p.flags&FencedCode != 0 { if p.extensions&FencedCode != 0 {
if i := p.fencedCodeBlock(data[end:], false); i > 0 { if i := p.fencedCodeBlock(data[end:], false); i > 0 {
// -1 to compensate for the extra end++ after the loop: // -1 to compensate for the extra end++ after the loop:
end += i - 1 end += i - 1
@ -1004,7 +1004,7 @@ func (p *Parser) quote(data []byte) int {
} }
// returns prefix length for block code // returns prefix length for block code
func (p *Parser) codePrefix(data []byte) int { func (p *Markdown) codePrefix(data []byte) int {
if len(data) >= 1 && data[0] == '\t' { if len(data) >= 1 && data[0] == '\t' {
return 1 return 1
} }
@ -1014,7 +1014,7 @@ func (p *Parser) codePrefix(data []byte) int {
return 0 return 0
} }
func (p *Parser) code(data []byte) int { func (p *Markdown) code(data []byte) int {
var work bytes.Buffer var work bytes.Buffer
i := 0 i := 0
@ -1064,7 +1064,7 @@ func (p *Parser) code(data []byte) int {
} }
// returns unordered list item prefix // returns unordered list item prefix
func (p *Parser) uliPrefix(data []byte) int { func (p *Markdown) uliPrefix(data []byte) int {
i := 0 i := 0
// start with up to 3 spaces // start with up to 3 spaces
for i < len(data) && i < 3 && data[i] == ' ' { for i < len(data) && i < 3 && data[i] == ' ' {
@ -1082,7 +1082,7 @@ func (p *Parser) uliPrefix(data []byte) int {
} }
// returns ordered list item prefix // returns ordered list item prefix
func (p *Parser) oliPrefix(data []byte) int { func (p *Markdown) oliPrefix(data []byte) int {
i := 0 i := 0
// start with up to 3 spaces // start with up to 3 spaces
@ -1107,7 +1107,7 @@ func (p *Parser) oliPrefix(data []byte) int {
} }
// returns definition list item prefix // returns definition list item prefix
func (p *Parser) dliPrefix(data []byte) int { func (p *Markdown) dliPrefix(data []byte) int {
if len(data) < 2 { if len(data) < 2 {
return 0 return 0
} }
@ -1123,7 +1123,7 @@ func (p *Parser) dliPrefix(data []byte) int {
} }
// parse ordered or unordered list block // parse ordered or unordered list block
func (p *Parser) list(data []byte, flags ListType) int { func (p *Markdown) list(data []byte, flags ListType) int {
i := 0 i := 0
flags |= ListItemBeginningOfList flags |= ListItemBeginningOfList
block := p.addBlock(List, nil) block := p.addBlock(List, nil)
@ -1191,7 +1191,7 @@ func finalizeList(block *Node) {
// Parse a single list item. // Parse a single list item.
// Assumes initial prefix is already removed if this is a sublist. // Assumes initial prefix is already removed if this is a sublist.
func (p *Parser) listItem(data []byte, flags *ListType) int { func (p *Markdown) listItem(data []byte, flags *ListType) int {
// keep track of the indentation of the first line // keep track of the indentation of the first line
itemIndent := 0 itemIndent := 0
if data[0] == '\t' { if data[0] == '\t' {
@ -1383,7 +1383,7 @@ gatherlines:
} }
// render a single paragraph that has already been parsed out // render a single paragraph that has already been parsed out
func (p *Parser) renderParagraph(data []byte) { func (p *Markdown) renderParagraph(data []byte) {
if len(data) == 0 { if len(data) == 0 {
return return
} }
@ -1408,13 +1408,13 @@ func (p *Parser) renderParagraph(data []byte) {
p.addBlock(Paragraph, data[beg:end]) p.addBlock(Paragraph, data[beg:end])
} }
func (p *Parser) paragraph(data []byte) int { func (p *Markdown) paragraph(data []byte) int {
// prev: index of 1st char of previous line // prev: index of 1st char of previous line
// line: index of 1st char of current line // line: index of 1st char of current line
// i: index of cursor/end of current line // i: index of cursor/end of current line
var prev, line, i int var prev, line, i int
tabSize := TabSizeDefault tabSize := TabSizeDefault
if p.flags&TabSizeEight != 0 { if p.extensions&TabSizeEight != 0 {
tabSize = TabSizeDouble tabSize = TabSizeDouble
} }
// keep going until we find something to mark the end of the paragraph // keep going until we find something to mark the end of the paragraph
@ -1435,7 +1435,7 @@ func (p *Parser) paragraph(data []byte) int {
// did we find a blank line marking the end of the paragraph? // did we find a blank line marking the end of the paragraph?
if n := p.isEmpty(current); n > 0 { if n := p.isEmpty(current); n > 0 {
// did this blank line followed by a definition list item? // did this blank line followed by a definition list item?
if p.flags&DefinitionLists != 0 { if p.extensions&DefinitionLists != 0 {
if i < len(data)-1 && data[i+1] == ':' { if i < len(data)-1 && data[i+1] == ':' {
return p.list(data[prev:], ListTypeDefinition) return p.list(data[prev:], ListTypeDefinition)
} }
@ -1461,7 +1461,7 @@ func (p *Parser) paragraph(data []byte) int {
} }
id := "" id := ""
if p.flags&AutoHeadingIDs != 0 { if p.extensions&AutoHeadingIDs != 0 {
id = sanitized_anchor_name.Create(string(data[prev:eol])) id = sanitized_anchor_name.Create(string(data[prev:eol]))
} }
@ -1478,7 +1478,7 @@ func (p *Parser) paragraph(data []byte) int {
} }
// if the next line starts a block of HTML, then the paragraph ends here // if the next line starts a block of HTML, then the paragraph ends here
if p.flags&LaxHTMLBlocks != 0 { if p.extensions&LaxHTMLBlocks != 0 {
if data[i] == '<' && p.html(current, false) > 0 { if data[i] == '<' && p.html(current, false) > 0 {
// rewind to before the HTML block // rewind to before the HTML block
p.renderParagraph(data[:i]) p.renderParagraph(data[:i])
@ -1493,7 +1493,7 @@ func (p *Parser) paragraph(data []byte) int {
} }
// if there's a fenced code block, paragraph is over // if there's a fenced code block, paragraph is over
if p.flags&FencedCode != 0 { if p.extensions&FencedCode != 0 {
if p.fencedCodeBlock(current, false) > 0 { if p.fencedCodeBlock(current, false) > 0 {
p.renderParagraph(data[:i]) p.renderParagraph(data[:i])
return i return i
@ -1501,7 +1501,7 @@ func (p *Parser) paragraph(data []byte) int {
} }
// if there's a definition list item, prev line is a definition term // if there's a definition list item, prev line is a definition term
if p.flags&DefinitionLists != 0 { if p.extensions&DefinitionLists != 0 {
if p.dliPrefix(current) != 0 { if p.dliPrefix(current) != 0 {
ret := p.list(data[prev:], ListTypeDefinition) ret := p.list(data[prev:], ListTypeDefinition)
return ret return ret
@ -1509,7 +1509,7 @@ func (p *Parser) paragraph(data []byte) int {
} }
// if there's a list after this, paragraph is over // if there's a list after this, paragraph is over
if p.flags&NoEmptyLineBeforeBlock != 0 { if p.extensions&NoEmptyLineBeforeBlock != 0 {
if p.uliPrefix(current) != 0 || if p.uliPrefix(current) != 0 ||
p.oliPrefix(current) != 0 || p.oliPrefix(current) != 0 ||
p.quotePrefix(current) != 0 || p.quotePrefix(current) != 0 ||

14
doc.go
View File

@ -4,14 +4,14 @@
// then be further processed to HTML (provided by Blackfriday itself) or other // then be further processed to HTML (provided by Blackfriday itself) or other
// formats (provided by the community). // formats (provided by the community).
// //
// The simplest way to invoke Blackfriday is to call one of Markdown* // The simplest way to invoke Blackfriday is to call the Markdown function. It
// functions. It will take a text input and produce a text output in HTML (or // will take a text input and produce a text output in HTML (or other format).
// other format).
// //
// A slightly more sophisticated way to use Blackfriday is to call Parse, which // A slightly more sophisticated way to use Blackfriday is to create a Markdown
// returns a syntax tree for the input document. You can use that to write your // processor and to call Parse, which returns a syntax tree for the input
// own renderer or, for example, to leverage Blackfriday's parsing for content // document. You can leverage Blackfriday's parsing for content extraction from
// extraction from markdown documents. // 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 // If you're interested in calling Blackfriday from command line, see
// https://github.com/russross/blackfriday-tool. // https://github.com/russross/blackfriday-tool.

2
helpers_test.go
View File

@ -46,7 +46,7 @@ func execRecoverableTestSuite(t *testing.T, tests []string, params TestParams, s
func runMarkdown(input string, params TestParams) string { func runMarkdown(input string, params TestParams) string {
params.HTMLRendererParameters.Flags = params.HTMLFlags params.HTMLRendererParameters.Flags = params.HTMLFlags
renderer := NewHTMLRenderer(params.HTMLRendererParameters) renderer := NewHTMLRenderer(params.HTMLRendererParameters)
return string(Markdown([]byte(input), WithRenderer(renderer), return string(Run([]byte(input), WithRenderer(renderer),
WithExtensions(params.extensions), WithExtensions(params.extensions),
WithRefOverride(params.referenceOverride))) WithRefOverride(params.referenceOverride)))
} }

44
inline.go
View File

@ -32,7 +32,7 @@ var (
// data is the complete block being rendered // data is the complete block being rendered
// offset is the number of valid chars before the current cursor // offset is the number of valid chars before the current cursor
func (p *Parser) inline(currBlock *Node, data []byte) { func (p *Markdown) inline(currBlock *Node, data []byte) {
// handlers might call us recursively: enforce a maximum depth // handlers might call us recursively: enforce a maximum depth
if p.nesting >= p.maxNesting || len(data) == 0 { if p.nesting >= p.maxNesting || len(data) == 0 {
return return
@ -69,7 +69,7 @@ func (p *Parser) inline(currBlock *Node, data []byte) {
} }
// single and double emphasis parsing // single and double emphasis parsing
func emphasis(p *Parser, data []byte, offset int) (int, *Node) { func emphasis(p *Markdown, data []byte, offset int) (int, *Node) {
data = data[offset:] data = data[offset:]
c := data[0] c := data[0]
@ -114,7 +114,7 @@ func emphasis(p *Parser, data []byte, offset int) (int, *Node) {
return 0, nil return 0, nil
} }
func codeSpan(p *Parser, data []byte, offset int) (int, *Node) { func codeSpan(p *Markdown, data []byte, offset int) (int, *Node) {
data = data[offset:] data = data[offset:]
nb := 0 nb := 0
@ -161,7 +161,7 @@ func codeSpan(p *Parser, data []byte, offset int) (int, *Node) {
} }
// newline preceded by two spaces becomes <br> // newline preceded by two spaces becomes <br>
func maybeLineBreak(p *Parser, data []byte, offset int) (int, *Node) { func maybeLineBreak(p *Markdown, data []byte, offset int) (int, *Node) {
origOffset := offset origOffset := offset
for offset < len(data) && data[offset] == ' ' { for offset < len(data) && data[offset] == ' ' {
offset++ offset++
@ -177,8 +177,8 @@ func maybeLineBreak(p *Parser, data []byte, offset int) (int, *Node) {
} }
// newline without two spaces works when HardLineBreak is enabled // newline without two spaces works when HardLineBreak is enabled
func lineBreak(p *Parser, data []byte, offset int) (int, *Node) { func lineBreak(p *Markdown, data []byte, offset int) (int, *Node) {
if p.flags&HardLineBreak != 0 { if p.extensions&HardLineBreak != 0 {
return 1, NewNode(Hardbreak) return 1, NewNode(Hardbreak)
} }
return 0, nil return 0, nil
@ -200,14 +200,14 @@ func isReferenceStyleLink(data []byte, pos int, t linkType) bool {
return pos < len(data)-1 && data[pos] == '[' && data[pos+1] != '^' return pos < len(data)-1 && data[pos] == '[' && data[pos+1] != '^'
} }
func maybeImage(p *Parser, data []byte, offset int) (int, *Node) { func maybeImage(p *Markdown, data []byte, offset int) (int, *Node) {
if offset < len(data)-1 && data[offset+1] == '[' { if offset < len(data)-1 && data[offset+1] == '[' {
return link(p, data, offset) return link(p, data, offset)
} }
return 0, nil return 0, nil
} }
func maybeInlineFootnote(p *Parser, data []byte, offset int) (int, *Node) { func maybeInlineFootnote(p *Markdown, data []byte, offset int) (int, *Node) {
if offset < len(data)-1 && data[offset+1] == '[' { if offset < len(data)-1 && data[offset+1] == '[' {
return link(p, data, offset) return link(p, data, offset)
} }
@ -215,7 +215,7 @@ func maybeInlineFootnote(p *Parser, data []byte, offset int) (int, *Node) {
} }
// '[': parse a link or an image or a footnote // '[': parse a link or an image or a footnote
func link(p *Parser, data []byte, offset int) (int, *Node) { func link(p *Markdown, data []byte, offset int) (int, *Node) {
// no links allowed inside regular links, footnote, and deferred footnotes // no links allowed inside regular links, footnote, and deferred footnotes
if p.insideLink && (offset > 0 && data[offset-1] == '[' || len(data)-1 > offset && data[offset+1] == '^') { if p.insideLink && (offset > 0 && data[offset-1] == '[' || len(data)-1 > offset && data[offset+1] == '^') {
return 0, nil return 0, nil
@ -225,7 +225,7 @@ func link(p *Parser, data []byte, offset int) (int, *Node) {
switch { switch {
// special case: ![^text] == deferred footnote (that follows something with // special case: ![^text] == deferred footnote (that follows something with
// an exclamation point) // an exclamation point)
case p.flags&Footnotes != 0 && len(data)-1 > offset && data[offset+1] == '^': case p.extensions&Footnotes != 0 && len(data)-1 > offset && data[offset+1] == '^':
t = linkDeferredFootnote t = linkDeferredFootnote
// ![alt] == image // ![alt] == image
case offset >= 0 && data[offset] == '!': case offset >= 0 && data[offset] == '!':
@ -233,7 +233,7 @@ func link(p *Parser, data []byte, offset int) (int, *Node) {
offset++ offset++
// ^[text] == inline footnote // ^[text] == inline footnote
// [^refId] == deferred footnote // [^refId] == deferred footnote
case p.flags&Footnotes != 0: case p.extensions&Footnotes != 0:
if offset >= 0 && data[offset] == '^' { if offset >= 0 && data[offset] == '^' {
t = linkInlineFootnote t = linkInlineFootnote
offset++ offset++
@ -573,7 +573,7 @@ func link(p *Parser, data []byte, offset int) (int, *Node) {
return i, linkNode return i, linkNode
} }
func (p *Parser) inlineHTMLComment(data []byte) int { func (p *Markdown) inlineHTMLComment(data []byte) int {
if len(data) < 5 { if len(data) < 5 {
return 0 return 0
} }
@ -613,7 +613,7 @@ const (
) )
// '<' when tags or autolinks are allowed // '<' when tags or autolinks are allowed
func leftAngle(p *Parser, data []byte, offset int) (int, *Node) { func leftAngle(p *Markdown, data []byte, offset int) (int, *Node) {
data = data[offset:] data = data[offset:]
altype, end := tagLength(data) altype, end := tagLength(data)
if size := p.inlineHTMLComment(data); size > 0 { if size := p.inlineHTMLComment(data); size > 0 {
@ -646,11 +646,11 @@ func leftAngle(p *Parser, data []byte, offset int) (int, *Node) {
// '\\' backslash escape // '\\' backslash escape
var escapeChars = []byte("\\`*_{}[]()#+-.!:|&<>~") var escapeChars = []byte("\\`*_{}[]()#+-.!:|&<>~")
func escape(p *Parser, data []byte, offset int) (int, *Node) { func escape(p *Markdown, data []byte, offset int) (int, *Node) {
data = data[offset:] data = data[offset:]
if len(data) > 1 { if len(data) > 1 {
if p.flags&BackslashLineBreak != 0 && data[1] == '\n' { if p.extensions&BackslashLineBreak != 0 && data[1] == '\n' {
return 2, NewNode(Hardbreak) return 2, NewNode(Hardbreak)
} }
if bytes.IndexByte(escapeChars, data[1]) < 0 { if bytes.IndexByte(escapeChars, data[1]) < 0 {
@ -686,7 +686,7 @@ func unescapeText(ob *bytes.Buffer, src []byte) {
// '&' escaped when it doesn't belong to an entity // '&' escaped when it doesn't belong to an entity
// valid entities are assumed to be anything matching &#?[A-Za-z0-9]+; // valid entities are assumed to be anything matching &#?[A-Za-z0-9]+;
func entity(p *Parser, data []byte, offset int) (int, *Node) { func entity(p *Markdown, data []byte, offset int) (int, *Node) {
data = data[offset:] data = data[offset:]
end := 1 end := 1
@ -748,7 +748,7 @@ var protocolPrefixes = [][]byte{
const shortestPrefix = 6 // len("ftp://"), the shortest of the above const shortestPrefix = 6 // len("ftp://"), the shortest of the above
func maybeAutoLink(p *Parser, data []byte, offset int) (int, *Node) { func maybeAutoLink(p *Markdown, data []byte, offset int) (int, *Node) {
// quick check to rule out most false hits // quick check to rule out most false hits
if p.insideLink || len(data) < offset+shortestPrefix { if p.insideLink || len(data) < offset+shortestPrefix {
return 0, nil return 0, nil
@ -765,7 +765,7 @@ func maybeAutoLink(p *Parser, data []byte, offset int) (int, *Node) {
return 0, nil return 0, nil
} }
func autoLink(p *Parser, data []byte, offset int) (int, *Node) { func autoLink(p *Markdown, data []byte, offset int) (int, *Node) {
// Now a more expensive check to see if we're not inside an anchor element // Now a more expensive check to see if we're not inside an anchor element
anchorStart := offset anchorStart := offset
offsetFromAnchor := 0 offsetFromAnchor := 0
@ -1095,7 +1095,7 @@ func helperFindEmphChar(data []byte, c byte) int {
return 0 return 0
} }
func helperEmphasis(p *Parser, data []byte, c byte) (int, *Node) { func helperEmphasis(p *Markdown, data []byte, c byte) (int, *Node) {
i := 0 i := 0
// skip one symbol if coming from emph3 // skip one symbol if coming from emph3
@ -1120,7 +1120,7 @@ func helperEmphasis(p *Parser, data []byte, c byte) (int, *Node) {
if data[i] == c && !isspace(data[i-1]) { if data[i] == c && !isspace(data[i-1]) {
if p.flags&NoIntraEmphasis != 0 { if p.extensions&NoIntraEmphasis != 0 {
if !(i+1 == len(data) || isspace(data[i+1]) || ispunct(data[i+1])) { if !(i+1 == len(data) || isspace(data[i+1]) || ispunct(data[i+1])) {
continue continue
} }
@ -1135,7 +1135,7 @@ func helperEmphasis(p *Parser, data []byte, c byte) (int, *Node) {
return 0, nil return 0, nil
} }
func helperDoubleEmphasis(p *Parser, data []byte, c byte) (int, *Node) { func helperDoubleEmphasis(p *Markdown, data []byte, c byte) (int, *Node) {
i := 0 i := 0
for i < len(data) { for i < len(data) {
@ -1159,7 +1159,7 @@ func helperDoubleEmphasis(p *Parser, data []byte, c byte) (int, *Node) {
return 0, nil return 0, nil
} }
func helperTripleEmphasis(p *Parser, data []byte, offset int, c byte) (int, *Node) { func helperTripleEmphasis(p *Markdown, data []byte, offset int, c byte) (int, *Node) {
i := 0 i := 0
origData := data origData := data
data = data[offset:] data = data[offset:]

129
markdown.go
View File

@ -154,18 +154,20 @@ type Renderer interface {
// Callback functions for inline parsing. One such function is defined // Callback functions for inline parsing. One such function is defined
// for each character that triggers a response when parsing inline data. // for each character that triggers a response when parsing inline data.
type inlineParser func(p *Parser, data []byte, offset int) (int, *Node) type inlineParser func(p *Markdown, data []byte, offset int) (int, *Node)
// Parser holds runtime state used by the parser. // Markdown is a type that holds:
// This is constructed by the Markdown function. // - extensions and the runtime state used by Parse,
type Parser struct { // - the renderer.
refOverride ReferenceOverrideFunc type Markdown struct {
refs map[string]*reference renderer Renderer
inlineCallback [256]inlineParser referenceOverride ReferenceOverrideFunc
flags Extensions refs map[string]*reference
nesting int inlineCallback [256]inlineParser
maxNesting int extensions Extensions
insideLink bool nesting int
maxNesting int
insideLink bool
// Footnotes need to be ordered as well as available to quickly check for // 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 // presence. If a ref is also a footnote, it's stored both in refs and here
@ -179,9 +181,9 @@ type Parser struct {
allClosed bool allClosed bool
} }
func (p *Parser) getRef(refid string) (ref *reference, found bool) { func (p *Markdown) getRef(refid string) (ref *reference, found bool) {
if p.refOverride != nil { if p.referenceOverride != nil {
r, overridden := p.refOverride(refid) r, overridden := p.referenceOverride(refid)
if overridden { if overridden {
if r == nil { if r == nil {
return nil, false return nil, false
@ -199,17 +201,17 @@ func (p *Parser) getRef(refid string) (ref *reference, found bool) {
return ref, found return ref, found
} }
func (p *Parser) finalize(block *Node) { func (p *Markdown) finalize(block *Node) {
above := block.Parent above := block.Parent
block.open = false block.open = false
p.tip = above p.tip = above
} }
func (p *Parser) addChild(node NodeType, offset uint32) *Node { func (p *Markdown) addChild(node NodeType, offset uint32) *Node {
return p.addExistingChild(NewNode(node), offset) return p.addExistingChild(NewNode(node), offset)
} }
func (p *Parser) addExistingChild(node *Node, offset uint32) *Node { func (p *Markdown) addExistingChild(node *Node, offset uint32) *Node {
for !p.tip.canContain(node.Type) { for !p.tip.canContain(node.Type) {
p.finalize(p.tip) p.finalize(p.tip)
} }
@ -218,7 +220,7 @@ func (p *Parser) addExistingChild(node *Node, offset uint32) *Node {
return node return node
} }
func (p *Parser) closeUnmatchedBlocks() { func (p *Markdown) closeUnmatchedBlocks() {
if !p.allClosed { if !p.allClosed {
for p.oldTip != p.lastMatchedContainer { for p.oldTip != p.lastMatchedContainer {
parent := p.oldTip.Parent parent := p.oldTip.Parent
@ -253,32 +255,13 @@ type Reference struct {
// See the documentation in Options for more details on use-case. // See the documentation in Options for more details on use-case.
type ReferenceOverrideFunc func(reference string) (ref *Reference, overridden bool) type ReferenceOverrideFunc func(reference string) (ref *Reference, overridden bool)
// Processor contains all the state necessary for Blackfriday to operate. // New constructs a Markdown processor. You can use the same With* functions as
type Processor struct { // for Run() to customize parser's behavior and the renderer.
r Renderer func New(opts ...Option) *Markdown {
extensions Extensions var p Markdown
referenceOverride ReferenceOverrideFunc
}
// DefaultProcessor creates the processor tuned to the most common behavior.
func DefaultProcessor() *Processor {
return &Processor{
r: NewHTMLRenderer(HTMLRendererParameters{
Flags: CommonHTMLFlags,
}),
extensions: CommonExtensions,
}
}
// NewParser constructs a Parser. You can use the same With* functions as for
// Markdown() to customize parser's behavior.
func (proc *Processor) NewParser(opts ...Option) *Parser {
for _, opt := range opts { for _, opt := range opts {
opt(proc) opt(&p)
} }
var p Parser
p.flags = proc.extensions
p.refOverride = proc.referenceOverride
p.refs = make(map[string]*reference) p.refs = make(map[string]*reference)
p.maxNesting = 16 p.maxNesting = 16
p.insideLink = false p.insideLink = false
@ -292,7 +275,7 @@ func (proc *Processor) NewParser(opts ...Option) *Parser {
p.inlineCallback[' '] = maybeLineBreak p.inlineCallback[' '] = maybeLineBreak
p.inlineCallback['*'] = emphasis p.inlineCallback['*'] = emphasis
p.inlineCallback['_'] = emphasis p.inlineCallback['_'] = emphasis
if proc.extensions&Strikethrough != 0 { if p.extensions&Strikethrough != 0 {
p.inlineCallback['~'] = emphasis p.inlineCallback['~'] = emphasis
} }
p.inlineCallback['`'] = codeSpan p.inlineCallback['`'] = codeSpan
@ -303,7 +286,7 @@ func (proc *Processor) NewParser(opts ...Option) *Parser {
p.inlineCallback['&'] = entity p.inlineCallback['&'] = entity
p.inlineCallback['!'] = maybeImage p.inlineCallback['!'] = maybeImage
p.inlineCallback['^'] = maybeInlineFootnote p.inlineCallback['^'] = maybeInlineFootnote
if proc.extensions&Autolink != 0 { if p.extensions&Autolink != 0 {
p.inlineCallback['h'] = maybeAutoLink p.inlineCallback['h'] = maybeAutoLink
p.inlineCallback['m'] = maybeAutoLink p.inlineCallback['m'] = maybeAutoLink
p.inlineCallback['f'] = maybeAutoLink p.inlineCallback['f'] = maybeAutoLink
@ -311,35 +294,35 @@ func (proc *Processor) NewParser(opts ...Option) *Parser {
p.inlineCallback['M'] = maybeAutoLink p.inlineCallback['M'] = maybeAutoLink
p.inlineCallback['F'] = maybeAutoLink p.inlineCallback['F'] = maybeAutoLink
} }
if proc.extensions&Footnotes != 0 { if p.extensions&Footnotes != 0 {
p.notes = make([]*reference, 0) p.notes = make([]*reference, 0)
} }
return &p return &p
} }
// Option customizes Processor's default behavior. // Option customizes the Markdown processor's default behavior.
type Option func(*Processor) type Option func(*Markdown)
// WithRenderer allows you to override the default renderer. // WithRenderer allows you to override the default renderer.
func WithRenderer(r Renderer) Option { func WithRenderer(r Renderer) Option {
return func(p *Processor) { return func(p *Markdown) {
p.r = r p.renderer = r
} }
} }
// WithExtensions allows you to pick some of the many extensions provided by // WithExtensions allows you to pick some of the many extensions provided by
// Blackfriday. You can bitwise OR them. // Blackfriday. You can bitwise OR them.
func WithExtensions(e Extensions) Option { func WithExtensions(e Extensions) Option {
return func(p *Processor) { return func(p *Markdown) {
p.extensions = e p.extensions = e
} }
} }
// WithNoExtensions turns off all extensions and custom behavior. // WithNoExtensions turns off all extensions and custom behavior.
func WithNoExtensions() Option { func WithNoExtensions() Option {
return func(p *Processor) { return func(p *Markdown) {
p.extensions = NoExtensions p.extensions = NoExtensions
p.r = NewHTMLRenderer(HTMLRendererParameters{ p.renderer = NewHTMLRenderer(HTMLRendererParameters{
Flags: HTMLFlagsNone, Flags: HTMLFlagsNone,
}) })
} }
@ -360,41 +343,45 @@ func WithNoExtensions() Option {
// the override function indicates an override did not occur, the refids at // the override function indicates an override did not occur, the refids at
// the bottom will be used to fill in the link details. // the bottom will be used to fill in the link details.
func WithRefOverride(o ReferenceOverrideFunc) Option { func WithRefOverride(o ReferenceOverrideFunc) Option {
return func(p *Processor) { return func(p *Markdown) {
p.referenceOverride = o p.referenceOverride = o
} }
} }
// Markdown is the main entry point to Blackfriday. It parses and renders a // Run is the main entry point to Blackfriday. It parses and renders a
// block of markdown-encoded text. // block of markdown-encoded text.
// //
// The simplest invocation of Markdown takes one argument, input: // The simplest invocation of Run takes one argument, input:
// output := Markdown(input) // output := Run(input)
// This will parse the input with CommonExtensions enabled and render it with // This will parse the input with CommonExtensions enabled and render it with
// the default HTMLRenderer (with CommonHTMLFlags). // the default HTMLRenderer (with CommonHTMLFlags).
// //
// Variadic arguments opts can customize the default behavior. Since Processor // Variadic arguments opts can customize the default behavior. Since Markdown
// type does not contain exported fields, you can not use it directly. Instead, // 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 // use the With* functions. For example, this will call the most basic
// functionality, with no extensions: // functionality, with no extensions:
// output := Markdown(input, WithNoExtensions()) // output := Run(input, WithNoExtensions())
// //
// You can use any number of With* arguments, even contradicting ones. They // 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 // will be applied in order of appearance and the latter will override the
// former: // former:
// output := Markdown(input, WithNoExtensions(), WithExtensions(exts), // output := Run(input, WithNoExtensions(), WithExtensions(exts),
// WithRenderer(yourRenderer)) // WithRenderer(yourRenderer))
func Markdown(input []byte, opts ...Option) []byte { func Run(input []byte, opts ...Option) []byte {
p := DefaultProcessor() r := NewHTMLRenderer(HTMLRendererParameters{
parser := p.NewParser(opts...) Flags: CommonHTMLFlags,
return p.r.Render(parser.Parse(input)) })
optList := []Option{WithRenderer(r), WithExtensions(CommonExtensions)}
optList = append(optList, opts...)
parser := New(optList...)
return parser.renderer.Render(parser.Parse(input))
} }
// Parse is an entry point to the parsing part of Blackfriday. It takes an // 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 // input markdown document and produces a syntax tree for its contents. This
// tree can then be rendered with a default or custom renderer, or // tree can then be rendered with a default or custom renderer, or
// analyzed/transformed by the caller to whatever non-standard needs they have. // analyzed/transformed by the caller to whatever non-standard needs they have.
func (p *Parser) Parse(input []byte) *Node { func (p *Markdown) Parse(input []byte) *Node {
p.block(input) p.block(input)
// Walk the tree and finish up some of unfinished blocks // Walk the tree and finish up some of unfinished blocks
for p.tip != nil { for p.tip != nil {
@ -412,8 +399,8 @@ func (p *Parser) Parse(input []byte) *Node {
return p.doc return p.doc
} }
func (p *Parser) parseRefsToAST() { func (p *Markdown) parseRefsToAST() {
if p.flags&Footnotes == 0 || len(p.notes) == 0 { if p.extensions&Footnotes == 0 || len(p.notes) == 0 {
return return
} }
p.tip = p.doc p.tip = p.doc
@ -537,7 +524,7 @@ func (r *reference) String() string {
// (in the render struct). // (in the render struct).
// Returns the number of bytes to skip to move past it, // Returns the number of bytes to skip to move past it,
// or zero if the first line is not a reference. // or zero if the first line is not a reference.
func isReference(p *Parser, data []byte, tabSize int) int { func isReference(p *Markdown, data []byte, tabSize int) int {
// up to 3 optional leading spaces // up to 3 optional leading spaces
if len(data) < 4 { if len(data) < 4 {
return 0 return 0
@ -554,7 +541,7 @@ func isReference(p *Parser, data []byte, tabSize int) int {
return 0 return 0
} }
i++ i++
if p.flags&Footnotes != 0 { if p.extensions&Footnotes != 0 {
if i < len(data) && data[i] == '^' { if i < len(data) && data[i] == '^' {
// we can set it to anything here because the proper noteIds will // 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 // be assigned later during the second pass. It just has to be != 0
@ -605,7 +592,7 @@ func isReference(p *Parser, data []byte, tabSize int) int {
hasBlock bool hasBlock bool
) )
if p.flags&Footnotes != 0 && noteID != 0 { if p.extensions&Footnotes != 0 && noteID != 0 {
linkOffset, linkEnd, raw, hasBlock = scanFootnote(p, data, i, tabSize) linkOffset, linkEnd, raw, hasBlock = scanFootnote(p, data, i, tabSize)
lineEnd = linkEnd lineEnd = linkEnd
} else { } else {
@ -640,7 +627,7 @@ func isReference(p *Parser, data []byte, tabSize int) int {
return lineEnd return lineEnd
} }
func scanLinkRef(p *Parser, data []byte, i int) (linkOffset, linkEnd, titleOffset, titleEnd, lineEnd int) { func scanLinkRef(p *Markdown, data []byte, i int) (linkOffset, linkEnd, titleOffset, titleEnd, lineEnd int) {
// link: whitespace-free sequence, optionally between angle brackets // link: whitespace-free sequence, optionally between angle brackets
if data[i] == '<' { if data[i] == '<' {
i++ i++
@ -714,7 +701,7 @@ func scanLinkRef(p *Parser, data []byte, i int) (linkOffset, linkEnd, titleOffse
// blockEnd is the end of the section in the input buffer, and contents is the // 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 // extracted text that was shifted over one tab. It will need to be rendered at
// the end of the document. // the end of the document.
func scanFootnote(p *Parser, data []byte, i, indentSize int) (blockStart, blockEnd int, contents []byte, hasBlock bool) { func scanFootnote(p *Markdown, data []byte, i, indentSize int) (blockStart, blockEnd int, contents []byte, hasBlock bool) {
if i == 0 || len(data) == 0 { if i == 0 || len(data) == 0 {
return return
} }