doc improvements, commenting

This commit is contained in:
Russ Ross 2011-07-07 11:56:45 -06:00
parent 256ff94689
commit bb8ee591d1
7 changed files with 263 additions and 221 deletions

View File

@ -164,8 +164,6 @@ Todo
----
* More unit testing
* Code cleanup
* Better code documentation
* Markdown pretty-printer output engine
* Improve unicode support. It does not understand all unicode
rules (about what constitutes a letter, a punctuation symbol,

244
block.go
View File

@ -20,16 +20,16 @@ import (
// Parse block-level data.
// Note: this function and many that it calls assume that
// the input buffer ends with a newline.
func (parser *Parser) block(out *bytes.Buffer, data []byte) {
func (p *parser) block(out *bytes.Buffer, data []byte) {
if len(data) == 0 || data[len(data)-1] != '\n' {
panic("block input is missing terminating newline")
}
// this is called recursively: enforce a maximum depth
if parser.nesting >= parser.maxNesting {
if p.nesting >= p.maxNesting {
return
}
parser.nesting++
p.nesting++
// parse out one block-level construct at a time
for len(data) > 0 {
@ -39,8 +39,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// ## Header 2
// ...
// ###### Header 6
if parser.isPrefixHeader(data) {
data = data[parser.prefixHeader(out, data):]
if p.isPrefixHeader(data) {
data = data[p.prefixHeader(out, data):]
continue
}
@ -50,14 +50,14 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// ...
// </div>
if data[0] == '<' {
if i := parser.html(out, data, true); i > 0 {
if i := p.html(out, data, true); i > 0 {
data = data[i:]
continue
}
}
// blank lines. note: returns the # of bytes to skip
if i := parser.isEmpty(data); i > 0 {
if i := p.isEmpty(data); i > 0 {
data = data[i:]
continue
}
@ -70,8 +70,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// }
// return b
// }
if parser.codePrefix(data) > 0 {
data = data[parser.code(out, data):]
if p.codePrefix(data) > 0 {
data = data[p.code(out, data):]
continue
}
@ -85,8 +85,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// return n * fact(n-1)
// }
// ```
if parser.flags&EXTENSION_FENCED_CODE != 0 {
if i := parser.fencedCode(out, data); i > 0 {
if p.flags&EXTENSION_FENCED_CODE != 0 {
if i := p.fencedCode(out, data); i > 0 {
data = data[i:]
continue
}
@ -99,8 +99,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// ******
// or
// ______
if parser.isHRule(data) {
parser.r.HRule(out)
if p.isHRule(data) {
p.r.HRule(out)
var i int
for i = 0; data[i] != '\n'; i++ {
}
@ -112,8 +112,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
//
// > A big quote I found somewhere
// > on the web
if parser.quotePrefix(data) > 0 {
data = data[parser.quote(out, data):]
if p.quotePrefix(data) > 0 {
data = data[p.quote(out, data):]
continue
}
@ -123,8 +123,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// ------|-----|---------
// Bob | 31 | 555-1234
// Alice | 27 | 555-4321
if parser.flags&EXTENSION_TABLES != 0 {
if i := parser.table(out, data); i > 0 {
if p.flags&EXTENSION_TABLES != 0 {
if i := p.table(out, data); i > 0 {
data = data[i:]
continue
}
@ -136,8 +136,8 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
// * Item 2
//
// also works with + or -
if parser.uliPrefix(data) > 0 {
data = data[parser.list(out, data, 0):]
if p.uliPrefix(data) > 0 {
data = data[p.list(out, data, 0):]
continue
}
@ -145,25 +145,25 @@ func (parser *Parser) block(out *bytes.Buffer, data []byte) {
//
// 1. Item 1
// 2. Item 2
if parser.oliPrefix(data) > 0 {
data = data[parser.list(out, data, LIST_TYPE_ORDERED):]
if p.oliPrefix(data) > 0 {
data = data[p.list(out, data, LIST_TYPE_ORDERED):]
continue
}
// anything else must look like a normal paragraph
// note: this finds underlined headers, too
data = data[parser.paragraph(out, data):]
data = data[p.paragraph(out, data):]
}
parser.nesting--
p.nesting--
}
func (parser *Parser) isPrefixHeader(data []byte) bool {
func (p *parser) isPrefixHeader(data []byte) bool {
if data[0] != '#' {
return false
}
if parser.flags&EXTENSION_SPACE_HEADERS != 0 {
if p.flags&EXTENSION_SPACE_HEADERS != 0 {
level := 0
for level < 6 && data[level] == '#' {
level++
@ -175,7 +175,7 @@ func (parser *Parser) isPrefixHeader(data []byte) bool {
return true
}
func (parser *Parser) prefixHeader(out *bytes.Buffer, data []byte) int {
func (p *parser) prefixHeader(out *bytes.Buffer, data []byte) int {
level := 0
for level < 6 && data[level] == '#' {
level++
@ -194,15 +194,15 @@ func (parser *Parser) prefixHeader(out *bytes.Buffer, data []byte) int {
}
if end > i {
work := func() bool {
parser.inline(out, data[i:end])
p.inline(out, data[i:end])
return true
}
parser.r.Header(out, work, level)
p.r.Header(out, work, level)
}
return skip
}
func (parser *Parser) isUnderlinedHeader(data []byte) int {
func (p *parser) isUnderlinedHeader(data []byte) int {
// test of level 1 header
if data[0] == '=' {
i := 1
@ -238,24 +238,24 @@ func (parser *Parser) isUnderlinedHeader(data []byte) int {
return 0
}
func (parser *Parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
func (p *parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
var i, j int
// identify the opening tag
if data[0] != '<' {
return 0
}
curtag, tagfound := parser.htmlFindTag(data[1:])
curtag, tagfound := p.htmlFindTag(data[1:])
// handle special cases
if !tagfound {
// check for an HTML comment
if size := parser.htmlComment(out, data, doRender); size > 0 {
if size := p.htmlComment(out, data, doRender); size > 0 {
return size
}
// check for an <hr> tag
if size := parser.htmlHr(out, data, doRender); size > 0 {
if size := p.htmlHr(out, data, doRender); size > 0 {
return size
}
@ -278,14 +278,14 @@ func (parser *Parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
}
// see if it is the only thing on the line
if skip := parser.isEmpty(data[j:]); skip > 0 {
if skip := p.isEmpty(data[j:]); skip > 0 {
// see if it is followed by a blank line/eof
j += skip
if j >= len(data) {
found = true
i = j
} else {
if skip := parser.isEmpty(data[j:]); skip > 0 {
if skip := p.isEmpty(data[j:]); skip > 0 {
j += skip
found = true
i = j
@ -309,7 +309,7 @@ func (parser *Parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
break
}
j = parser.htmlFindEnd(curtag, data[i-1:])
j = p.htmlFindEnd(curtag, data[i-1:])
if j > 0 {
i += j - 1
@ -330,14 +330,14 @@ func (parser *Parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
for end > 0 && data[end-1] == '\n' {
end--
}
parser.r.BlockHtml(out, data[:end])
p.r.BlockHtml(out, data[:end])
}
return i
}
// HTML comment, lax form
func (parser *Parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool) int {
func (p *parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool) int {
if data[0] != '<' || data[1] != '!' || data[2] != '-' || data[3] != '-' {
return 0
}
@ -356,7 +356,7 @@ func (parser *Parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool)
}
// needs to end with a blank line
if j := parser.isEmpty(data[i:]); j > 0 {
if j := p.isEmpty(data[i:]); j > 0 {
size := i + j
if doRender {
// trim trailing newlines
@ -364,7 +364,7 @@ func (parser *Parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool)
for end > 0 && data[end-1] == '\n' {
end--
}
parser.r.BlockHtml(out, data[:end])
p.r.BlockHtml(out, data[:end])
}
return size
}
@ -373,7 +373,7 @@ func (parser *Parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool)
}
// HR, which is the only self-closing block tag considered
func (parser *Parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
func (p *parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
if data[0] != '<' || (data[1] != 'h' && data[1] != 'H') || (data[2] != 'r' && data[2] != 'R') {
return 0
}
@ -389,7 +389,7 @@ func (parser *Parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int
if data[i] == '>' {
i++
if j := parser.isEmpty(data[i:]); j > 0 {
if j := p.isEmpty(data[i:]); j > 0 {
size := i + j
if doRender {
// trim newlines
@ -397,7 +397,7 @@ func (parser *Parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int
for end > 0 && data[end-1] == '\n' {
end--
}
parser.r.BlockHtml(out, data[:end])
p.r.BlockHtml(out, data[:end])
}
return size
}
@ -406,7 +406,7 @@ func (parser *Parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int
return 0
}
func (parser *Parser) htmlFindTag(data []byte) (string, bool) {
func (p *parser) htmlFindTag(data []byte) (string, bool) {
i := 0
for isalnum(data[i]) {
i++
@ -418,7 +418,7 @@ func (parser *Parser) htmlFindTag(data []byte) (string, bool) {
return "", false
}
func (parser *Parser) htmlFindEnd(tag string, data []byte) int {
func (p *parser) htmlFindEnd(tag string, data []byte) int {
// assume data[0] == '<' && data[1] == '/' already tested
// check if tag is a match
@ -430,7 +430,7 @@ func (parser *Parser) htmlFindEnd(tag string, data []byte) int {
// check that the rest of the line is blank
skip := 0
if skip = parser.isEmpty(data[i:]); skip == 0 {
if skip = p.isEmpty(data[i:]); skip == 0 {
return 0
}
i += skip
@ -440,10 +440,10 @@ func (parser *Parser) htmlFindEnd(tag string, data []byte) int {
return i
}
if parser.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
return i
}
if skip = parser.isEmpty(data[i:]); skip == 0 {
if skip = p.isEmpty(data[i:]); skip == 0 {
// following line must be blank
return 0
}
@ -451,7 +451,7 @@ func (parser *Parser) htmlFindEnd(tag string, data []byte) int {
return i + skip
}
func (parser *Parser) isEmpty(data []byte) int {
func (p *parser) isEmpty(data []byte) int {
// it is okay to call isEmpty on an empty buffer
if len(data) == 0 {
return 0
@ -466,7 +466,7 @@ func (parser *Parser) isEmpty(data []byte) int {
return i + 1
}
func (parser *Parser) isHRule(data []byte) bool {
func (p *parser) isHRule(data []byte) bool {
i := 0
// skip up to three spaces
@ -495,7 +495,7 @@ func (parser *Parser) isHRule(data []byte) bool {
return n >= 3
}
func (parser *Parser) isFencedCode(data []byte, syntax **string, oldmarker string) (skip int, marker string) {
func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (skip int, marker string) {
i, size := 0, 0
skip = 0
@ -584,9 +584,9 @@ func (parser *Parser) isFencedCode(data []byte, syntax **string, oldmarker strin
return
}
func (parser *Parser) fencedCode(out *bytes.Buffer, data []byte) int {
func (p *parser) fencedCode(out *bytes.Buffer, data []byte) int {
var lang *string
beg, marker := parser.isFencedCode(data, &lang, "")
beg, marker := p.isFencedCode(data, &lang, "")
if beg == 0 || beg >= len(data) {
return 0
}
@ -597,7 +597,7 @@ func (parser *Parser) fencedCode(out *bytes.Buffer, data []byte) int {
// safe to assume beg < len(data)
// check for the end of the code block
fenceEnd, _ := parser.isFencedCode(data[beg:], nil, marker)
fenceEnd, _ := p.isFencedCode(data[beg:], nil, marker)
if fenceEnd != 0 {
beg += fenceEnd
break
@ -625,14 +625,14 @@ func (parser *Parser) fencedCode(out *bytes.Buffer, data []byte) int {
syntax = *lang
}
parser.r.BlockCode(out, work.Bytes(), syntax)
p.r.BlockCode(out, work.Bytes(), syntax)
return beg
}
func (parser *Parser) table(out *bytes.Buffer, data []byte) int {
func (p *parser) table(out *bytes.Buffer, data []byte) int {
var header bytes.Buffer
i, columns := parser.tableHeader(&header, data)
i, columns := p.tableHeader(&header, data)
if i == 0 {
return 0
}
@ -654,15 +654,15 @@ func (parser *Parser) table(out *bytes.Buffer, data []byte) int {
// include the newline in data sent to tableRow
i++
parser.tableRow(&body, data[rowStart:i], columns)
p.tableRow(&body, data[rowStart:i], columns)
}
parser.r.Table(out, header.Bytes(), body.Bytes(), columns)
p.r.Table(out, header.Bytes(), body.Bytes(), columns)
return i
}
func (parser *Parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns []int) {
func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns []int) {
i := 0
colCount := 1
for i = 0; data[i] != '\n'; i++ {
@ -757,12 +757,12 @@ func (parser *Parser) tableHeader(out *bytes.Buffer, data []byte) (size int, col
return
}
parser.tableRow(out, header, columns)
p.tableRow(out, header, columns)
size = i + 1
return
}
func (parser *Parser) tableRow(out *bytes.Buffer, data []byte, columns []int) {
func (p *parser) tableRow(out *bytes.Buffer, data []byte, columns []int) {
i, col := 0, 0
var rowWork bytes.Buffer
@ -789,22 +789,22 @@ func (parser *Parser) tableRow(out *bytes.Buffer, data []byte, columns []int) {
}
var cellWork bytes.Buffer
parser.inline(&cellWork, data[cellStart:cellEnd])
parser.r.TableCell(&rowWork, cellWork.Bytes(), columns[col])
p.inline(&cellWork, data[cellStart:cellEnd])
p.r.TableCell(&rowWork, cellWork.Bytes(), columns[col])
}
// pad it out with empty columns to get the right number
for ; col < len(columns); col++ {
parser.r.TableCell(&rowWork, nil, columns[col])
p.r.TableCell(&rowWork, nil, columns[col])
}
// silently ignore rows with too many cells
parser.r.TableRow(out, rowWork.Bytes())
p.r.TableRow(out, rowWork.Bytes())
}
// returns blockquote prefix length
func (parser *Parser) quotePrefix(data []byte) int {
func (p *parser) quotePrefix(data []byte) int {
i := 0
for i < 3 && data[i] == ' ' {
i++
@ -819,7 +819,7 @@ func (parser *Parser) quotePrefix(data []byte) int {
}
// parse a blockquote fragment
func (parser *Parser) quote(out *bytes.Buffer, data []byte) int {
func (p *parser) quote(out *bytes.Buffer, data []byte) int {
var raw bytes.Buffer
beg, end := 0, 0
for beg < len(data) {
@ -829,12 +829,12 @@ func (parser *Parser) quote(out *bytes.Buffer, data []byte) int {
}
end++
if pre := parser.quotePrefix(data[beg:]); pre > 0 {
// string the prefix
if pre := p.quotePrefix(data[beg:]); pre > 0 {
// skip the prefix
beg += pre
} else if parser.isEmpty(data[beg:]) > 0 &&
} else if p.isEmpty(data[beg:]) > 0 &&
(end >= len(data) ||
(parser.quotePrefix(data[end:]) == 0 && parser.isEmpty(data[end:]) == 0)) {
(p.quotePrefix(data[end:]) == 0 && p.isEmpty(data[end:]) == 0)) {
// blockquote ends with at least one blank line
// followed by something without a blockquote prefix
break
@ -846,20 +846,20 @@ func (parser *Parser) quote(out *bytes.Buffer, data []byte) int {
}
var cooked bytes.Buffer
parser.block(&cooked, raw.Bytes())
parser.r.BlockQuote(out, cooked.Bytes())
p.block(&cooked, raw.Bytes())
p.r.BlockQuote(out, cooked.Bytes())
return end
}
// returns prefix length for block code
func (parser *Parser) codePrefix(data []byte) int {
func (p *parser) codePrefix(data []byte) int {
if data[0] == ' ' && data[1] == ' ' && data[2] == ' ' && data[3] == ' ' {
return 4
}
return 0
}
func (parser *Parser) code(out *bytes.Buffer, data []byte) int {
func (p *parser) code(out *bytes.Buffer, data []byte) int {
var work bytes.Buffer
i := 0
@ -870,8 +870,8 @@ func (parser *Parser) code(out *bytes.Buffer, data []byte) int {
}
i++
blankline := parser.isEmpty(data[beg:i]) > 0
if pre := parser.codePrefix(data[beg:i]); pre > 0 {
blankline := p.isEmpty(data[beg:i]) > 0
if pre := p.codePrefix(data[beg:i]); pre > 0 {
beg += pre
} else if !blankline {
// non-empty, non-prefixed line breaks the pre
@ -899,13 +899,13 @@ func (parser *Parser) code(out *bytes.Buffer, data []byte) int {
work.WriteByte('\n')
parser.r.BlockCode(out, work.Bytes(), "")
p.r.BlockCode(out, work.Bytes(), "")
return i
}
// returns unordered list item prefix
func (parser *Parser) uliPrefix(data []byte) int {
func (p *parser) uliPrefix(data []byte) int {
i := 0
// start with up to 3 spaces
@ -913,7 +913,7 @@ func (parser *Parser) uliPrefix(data []byte) int {
i++
}
// need a *, +, or - followed by a space/tab
// need a *, +, or - followed by a space
if (data[i] != '*' && data[i] != '+' && data[i] != '-') ||
data[i+1] != ' ' {
return 0
@ -922,7 +922,7 @@ func (parser *Parser) uliPrefix(data []byte) int {
}
// returns ordered list item prefix
func (parser *Parser) oliPrefix(data []byte) int {
func (p *parser) oliPrefix(data []byte) int {
i := 0
// start with up to 3 spaces
@ -936,7 +936,7 @@ func (parser *Parser) oliPrefix(data []byte) int {
i++
}
// we need >= 1 digits followed by a dot and a space/tab
// we need >= 1 digits followed by a dot and a space
if start == i || data[i] != '.' || data[i+1] != ' ' {
return 0
}
@ -944,12 +944,12 @@ func (parser *Parser) oliPrefix(data []byte) int {
}
// parse ordered or unordered list block
func (parser *Parser) list(out *bytes.Buffer, data []byte, flags int) int {
func (p *parser) list(out *bytes.Buffer, data []byte, flags int) int {
i := 0
flags |= LIST_ITEM_BEGINNING_OF_LIST
work := func() bool {
for i < len(data) {
skip := parser.listItem(out, data[i:], &flags)
skip := p.listItem(out, data[i:], &flags)
i += skip
if skip == 0 || flags&LIST_ITEM_END_OF_LIST != 0 {
@ -960,22 +960,22 @@ func (parser *Parser) list(out *bytes.Buffer, data []byte, flags int) int {
return true
}
parser.r.List(out, work, flags)
p.r.List(out, work, flags)
return i
}
// Parse a single list item.
// Assumes initial prefix is already removed if this is a sublist.
func (parser *Parser) listItem(out *bytes.Buffer, data []byte, flags *int) int {
func (p *parser) listItem(out *bytes.Buffer, data []byte, flags *int) int {
// keep track of the indentation of the first line
itemIndent := 0
for itemIndent < 3 && data[itemIndent] == ' ' {
itemIndent++
}
i := parser.uliPrefix(data)
i := p.uliPrefix(data)
if i == 0 {
i = parser.oliPrefix(data)
i = p.oliPrefix(data)
}
if i == 0 {
return 0
@ -1003,7 +1003,7 @@ func (parser *Parser) listItem(out *bytes.Buffer, data []byte, flags *int) int {
containsBlankLine := false
sublist := 0
loop:
gatherlines:
for line < len(data) {
i++
@ -1014,7 +1014,7 @@ loop:
// if it is an empty line, guess that it is part of this item
// and move on to the next line
if parser.isEmpty(data[line:i]) > 0 {
if p.isEmpty(data[line:i]) > 0 {
containsBlankLine = true
line = i
continue
@ -1031,8 +1031,8 @@ loop:
// evaluate how this line fits in
switch {
// is this a nested list item?
case (parser.uliPrefix(chunk) > 0 && !parser.isHRule(chunk)) ||
parser.oliPrefix(chunk) > 0:
case (p.uliPrefix(chunk) > 0 && !p.isHRule(chunk)) ||
p.oliPrefix(chunk) > 0:
if containsBlankLine {
*flags |= LIST_ITEM_CONTAINS_BLOCK
@ -1041,7 +1041,7 @@ loop:
// to be a nested list, it must be indented more
// if not, it is the next item in the same list
if indent <= itemIndent {
break loop
break gatherlines
}
// is this the first item in the the nested list?
@ -1050,12 +1050,12 @@ loop:
}
// is this a nested prefix header?
case parser.isPrefixHeader(chunk):
case p.isPrefixHeader(chunk):
// if the header is not indented, it is not nested in the list
// and thus ends the list
if containsBlankLine && indent < 4 {
*flags |= LIST_ITEM_END_OF_LIST
break loop
break gatherlines
}
*flags |= LIST_ITEM_CONTAINS_BLOCK
@ -1064,7 +1064,7 @@ loop:
// (regardless of the indentation of the beginning of the item)
case containsBlankLine && indent < 4:
*flags |= LIST_ITEM_END_OF_LIST
break loop
break gatherlines
// a blank line means this should be parsed as a block
case containsBlankLine:
@ -1085,18 +1085,18 @@ loop:
if *flags&LIST_ITEM_CONTAINS_BLOCK != 0 {
// intermediate render of block li
if sublist > 0 {
parser.block(&cooked, rawBytes[:sublist])
parser.block(&cooked, rawBytes[sublist:])
p.block(&cooked, rawBytes[:sublist])
p.block(&cooked, rawBytes[sublist:])
} else {
parser.block(&cooked, rawBytes)
p.block(&cooked, rawBytes)
}
} else {
// intermediate render of inline li
if sublist > 0 {
parser.inline(&cooked, rawBytes[:sublist])
parser.block(&cooked, rawBytes[sublist:])
p.inline(&cooked, rawBytes[:sublist])
p.block(&cooked, rawBytes[sublist:])
} else {
parser.inline(&cooked, rawBytes)
p.inline(&cooked, rawBytes)
}
}
@ -1108,13 +1108,13 @@ loop:
for parsedEnd > 0 && cookedBytes[parsedEnd-1] == '\n' {
parsedEnd--
}
parser.r.ListItem(out, cookedBytes[:parsedEnd], *flags)
p.r.ListItem(out, cookedBytes[:parsedEnd], *flags)
return line
}
// render a single paragraph that has already been parsed out
func (parser *Parser) renderParagraph(out *bytes.Buffer, data []byte) {
func (p *parser) renderParagraph(out *bytes.Buffer, data []byte) {
if len(data) == 0 {
return
}
@ -1134,13 +1134,13 @@ func (parser *Parser) renderParagraph(out *bytes.Buffer, data []byte) {
}
work := func() bool {
parser.inline(out, data[beg:end])
p.inline(out, data[beg:end])
return true
}
parser.r.Paragraph(out, work)
p.r.Paragraph(out, work)
}
func (parser *Parser) paragraph(out *bytes.Buffer, data []byte) int {
func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
// prev: index of 1st char of previous line
// line: index of 1st char of current line
// i: index of cursor/end of current line
@ -1154,16 +1154,16 @@ func (parser *Parser) paragraph(out *bytes.Buffer, data []byte) int {
line = i
// did we find a blank line marking the end of the paragraph?
if n := parser.isEmpty(current); n > 0 {
parser.renderParagraph(out, data[:i])
if n := p.isEmpty(current); n > 0 {
p.renderParagraph(out, data[:i])
return i + n
}
// an underline under some text marks a header, so our paragraph ended on prev line
if i > 0 {
if level := parser.isUnderlinedHeader(current); level > 0 {
if level := p.isUnderlinedHeader(current); level > 0 {
// render the paragraph
parser.renderParagraph(out, data[:prev])
p.renderParagraph(out, data[:prev])
// ignore leading and trailing whitespace
eol := i - 1
@ -1176,13 +1176,13 @@ func (parser *Parser) paragraph(out *bytes.Buffer, data []byte) int {
// render the header
// this ugly double closure avoids forcing variables onto the heap
work := func(o *bytes.Buffer, p *Parser, d []byte) func() bool {
work := func(o *bytes.Buffer, pp *parser, d []byte) func() bool {
return func() bool {
p.inline(o, d)
pp.inline(o, d)
return true
}
}(out, parser, data[prev:eol])
parser.r.Header(out, work, level)
}(out, p, data[prev:eol])
p.r.Header(out, work, level)
// find the end of the underline
for data[i] != '\n' {
@ -1193,17 +1193,17 @@ func (parser *Parser) paragraph(out *bytes.Buffer, data []byte) int {
}
// if the next line starts a block of HTML, then the paragraph ends here
if parser.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
if data[i] == '<' && parser.html(out, current, false) > 0 {
if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
if data[i] == '<' && p.html(out, current, false) > 0 {
// rewind to before the HTML block
parser.renderParagraph(out, data[:i])
p.renderParagraph(out, data[:i])
return i
}
}
// if there's a prefixed header or a horizontal rule after this, paragraph is over
if parser.isPrefixHeader(current) || parser.isHRule(current) {
parser.renderParagraph(out, data[:i])
if p.isPrefixHeader(current) || p.isHRule(current) {
p.renderParagraph(out, data[:i])
return i
}
@ -1214,6 +1214,6 @@ func (parser *Parser) paragraph(out *bytes.Buffer, data []byte) int {
i++
}
parser.renderParagraph(out, data[:i])
p.renderParagraph(out, data[:i])
return i
}

14
html.go
View File

@ -38,6 +38,9 @@ const (
HTML_SMARTYPANTS_LATEX_DASHES
)
// Html is a type that implements the Renderer interface for HTML output.
//
// Do not create this directly, instead use the HtmlRenderer function.
type Html struct {
flags int // HTML_* options
closeTag string // how to end singleton tags: either " />\n" or ">\n"
@ -50,7 +53,7 @@ type Html struct {
currentLevel int
toc *bytes.Buffer
smartypants *SmartypantsRenderer
smartypants *smartypantsRenderer
}
const (
@ -58,6 +61,13 @@ const (
htmlClose = ">\n"
)
// HtmlRenderer creates and configures an Html object, which
// satisfies the Renderer interface.
//
// flags is a set of HTML_* options ORed together.
// title is the title of the document, and css is a URL for the document's
// stylesheet.
// title and css are only used when HTML_COMPLETE_PAGE is selected.
func HtmlRenderer(flags int, title string, css string) Renderer {
// configure the rendering engine
closeTag := htmlClose
@ -75,7 +85,7 @@ func HtmlRenderer(flags int, title string, css string) Renderer {
currentLevel: 0,
toc: new(bytes.Buffer),
smartypants: Smartypants(flags),
smartypants: smartypants(flags),
}
}

100
inline.go
View File

@ -22,21 +22,21 @@ import (
// data is the complete block being rendered
// offset is the number of valid chars before the current cursor
func (parser *Parser) inline(out *bytes.Buffer, data []byte) {
func (p *parser) inline(out *bytes.Buffer, data []byte) {
// this is called recursively: enforce a maximum depth
if parser.nesting >= parser.maxNesting {
if p.nesting >= p.maxNesting {
return
}
parser.nesting++
p.nesting++
i, end := 0, 0
for i < len(data) {
// copy inactive chars into the output
for end < len(data) && parser.inlineCallback[data[end]] == nil {
for end < len(data) && p.inlineCallback[data[end]] == nil {
end++
}
parser.r.NormalText(out, data[i:end])
p.r.NormalText(out, data[i:end])
if end >= len(data) {
break
@ -44,8 +44,8 @@ func (parser *Parser) inline(out *bytes.Buffer, data []byte) {
i = end
// call the trigger
handler := parser.inlineCallback[data[end]]
if consumed := handler(parser, out, data, i); consumed == 0 {
handler := p.inlineCallback[data[end]]
if consumed := handler(p, out, data, i); consumed == 0 {
// no action from the callback; buffer the byte for later
end = i + 1
} else {
@ -55,11 +55,11 @@ func (parser *Parser) inline(out *bytes.Buffer, data []byte) {
}
}
parser.nesting--
p.nesting--
}
// single and double emphasis parsing
func emphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func emphasis(p *parser, out *bytes.Buffer, data []byte, offset int) int {
data = data[offset:]
c := data[0]
ret := 0
@ -70,7 +70,7 @@ func emphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
if c == '~' || isspace(data[1]) {
return 0
}
if ret = helperEmphasis(parser, out, data[1:], c); ret == 0 {
if ret = helperEmphasis(p, out, data[1:], c); ret == 0 {
return 0
}
@ -81,7 +81,7 @@ func emphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
if isspace(data[2]) {
return 0
}
if ret = helperDoubleEmphasis(parser, out, data[2:], c); ret == 0 {
if ret = helperDoubleEmphasis(p, out, data[2:], c); ret == 0 {
return 0
}
@ -92,7 +92,7 @@ func emphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
if c == '~' || isspace(data[3]) {
return 0
}
if ret = helperTripleEmphasis(parser, out, data, 3, c); ret == 0 {
if ret = helperTripleEmphasis(p, out, data, 3, c); ret == 0 {
return 0
}
@ -102,7 +102,7 @@ func emphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
return 0
}
func codeSpan(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func codeSpan(p *parser, out *bytes.Buffer, data []byte, offset int) int {
data = data[offset:]
nb := 0
@ -140,7 +140,7 @@ func codeSpan(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
// render the code span
if fBegin != fEnd {
parser.r.CodeSpan(out, data[fBegin:fEnd])
p.r.CodeSpan(out, data[fBegin:fEnd])
}
return end
@ -149,7 +149,7 @@ func codeSpan(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
// newline preceded by two spaces becomes <br>
// newline without two spaces works when EXTENSION_HARD_LINE_BREAK is enabled
func lineBreak(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func lineBreak(p *parser, out *bytes.Buffer, data []byte, offset int) int {
// remove trailing spaces from out
outBytes := out.Bytes()
end := len(outBytes)
@ -160,18 +160,18 @@ func lineBreak(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
out.Truncate(eol)
// should there be a hard line break here?
if parser.flags&EXTENSION_HARD_LINE_BREAK == 0 && end-eol < 2 {
if p.flags&EXTENSION_HARD_LINE_BREAK == 0 && end-eol < 2 {
return 0
}
parser.r.LineBreak(out)
p.r.LineBreak(out)
return 1
}
// '[': parse a link or an image
func link(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func link(p *parser, out *bytes.Buffer, data []byte, offset int) int {
// no links allowed inside other links
if parser.insideLink {
if p.insideLink {
return 0
}
@ -348,7 +348,7 @@ func link(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
// find the reference with matching id (ids are case-insensitive)
key := string(bytes.ToLower(id))
lr, ok := parser.refs[key]
lr, ok := p.refs[key]
if !ok {
return 0
}
@ -382,7 +382,7 @@ func link(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
// find the reference with matching id
key := string(bytes.ToLower(id))
lr, ok := parser.refs[key]
lr, ok := p.refs[key]
if !ok {
return 0
}
@ -402,10 +402,10 @@ func link(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
content.Write(data[1:txtE])
} else {
// links cannot contain other links, so turn off link parsing temporarily
insideLink := parser.insideLink
parser.insideLink = true
parser.inline(&content, data[1:txtE])
parser.insideLink = insideLink
insideLink := p.insideLink
p.insideLink = true
p.inline(&content, data[1:txtE])
p.insideLink = insideLink
}
}
@ -429,16 +429,16 @@ func link(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
out.Truncate(outSize - 1)
}
parser.r.Image(out, uLink, title, content.Bytes())
p.r.Image(out, uLink, title, content.Bytes())
} else {
parser.r.Link(out, uLink, title, content.Bytes())
p.r.Link(out, uLink, title, content.Bytes())
}
return i
}
// '<' when tags or autolinks are allowed
func leftAngle(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func leftAngle(p *parser, out *bytes.Buffer, data []byte, offset int) int {
data = data[offset:]
altype := LINK_TYPE_NOT_AUTOLINK
end := tagLength(data, &altype)
@ -448,10 +448,10 @@ func leftAngle(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
var uLink bytes.Buffer
unescapeText(&uLink, data[1:end+1-2])
if uLink.Len() > 0 {
parser.r.AutoLink(out, uLink.Bytes(), altype)
p.r.AutoLink(out, uLink.Bytes(), altype)
}
} else {
parser.r.RawHtmlTag(out, data[:end])
p.r.RawHtmlTag(out, data[:end])
}
}
@ -461,7 +461,7 @@ func leftAngle(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
// '\\' backslash escape
var escapeChars = []byte("\\`*_{}[]()#+-.!:|&<>")
func escape(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func escape(p *parser, out *bytes.Buffer, data []byte, offset int) int {
data = data[offset:]
if len(data) > 1 {
@ -469,7 +469,7 @@ func escape(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
return 0
}
parser.r.NormalText(out, data[1:2])
p.r.NormalText(out, data[1:2])
}
return 2
@ -498,7 +498,7 @@ func unescapeText(ob *bytes.Buffer, src []byte) {
// '&' escaped when it doesn't belong to an entity
// valid entities are assumed to be anything matching &#?[A-Za-z0-9]+;
func entity(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func entity(p *parser, out *bytes.Buffer, data []byte, offset int) int {
data = data[offset:]
end := 1
@ -517,14 +517,14 @@ func entity(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
return 0 // lone '&'
}
parser.r.Entity(out, data[:end])
p.r.Entity(out, data[:end])
return end
}
func autoLink(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
func autoLink(p *parser, out *bytes.Buffer, data []byte, offset int) int {
// quick check to rule out most false hits on ':'
if parser.insideLink || len(data) < offset+3 || data[offset+1] != '/' || data[offset+2] != '/' {
if p.insideLink || len(data) < offset+3 || data[offset+1] != '/' || data[offset+2] != '/' {
return 0
}
@ -622,7 +622,7 @@ func autoLink(parser *Parser, out *bytes.Buffer, data []byte, offset int) int {
unescapeText(&uLink, data[:linkEnd])
if uLink.Len() > 0 {
parser.r.AutoLink(out, uLink.Bytes(), LINK_TYPE_NORMAL)
p.r.AutoLink(out, uLink.Bytes(), LINK_TYPE_NORMAL)
}
return linkEnd - rewind
@ -832,7 +832,7 @@ func helperFindEmphChar(data []byte, c byte) int {
return 0
}
func helperEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte) int {
func helperEmphasis(p *parser, out *bytes.Buffer, data []byte, c byte) int {
i := 0
// skip one symbol if coming from emph3
@ -857,15 +857,15 @@ func helperEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte) int
if data[i] == c && !isspace(data[i-1]) {
if parser.flags&EXTENSION_NO_INTRA_EMPHASIS != 0 {
if p.flags&EXTENSION_NO_INTRA_EMPHASIS != 0 {
if !(i+1 == len(data) || isspace(data[i+1]) || ispunct(data[i+1])) {
continue
}
}
var work bytes.Buffer
parser.inline(&work, data[:i])
parser.r.Emphasis(out, work.Bytes())
p.inline(&work, data[:i])
p.r.Emphasis(out, work.Bytes())
return i + 1
}
}
@ -873,7 +873,7 @@ func helperEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte) int
return 0
}
func helperDoubleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte) int {
func helperDoubleEmphasis(p *parser, out *bytes.Buffer, data []byte, c byte) int {
i := 0
for i < len(data) {
@ -885,14 +885,14 @@ func helperDoubleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte
if i+1 < len(data) && data[i] == c && data[i+1] == c && i > 0 && !isspace(data[i-1]) {
var work bytes.Buffer
parser.inline(&work, data[:i])
p.inline(&work, data[:i])
if work.Len() > 0 {
// pick the right renderer
if c == '~' {
parser.r.StrikeThrough(out, work.Bytes())
p.r.StrikeThrough(out, work.Bytes())
} else {
parser.r.DoubleEmphasis(out, work.Bytes())
p.r.DoubleEmphasis(out, work.Bytes())
}
}
return i + 2
@ -902,7 +902,7 @@ func helperDoubleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, c byte
return 0
}
func helperTripleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, offset int, c byte) int {
func helperTripleEmphasis(p *parser, out *bytes.Buffer, data []byte, offset int, c byte) int {
i := 0
origData := data
data = data[offset:]
@ -924,14 +924,14 @@ func helperTripleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, offset
// triple symbol found
var work bytes.Buffer
parser.inline(&work, data[:i])
p.inline(&work, data[:i])
if work.Len() > 0 {
parser.r.TripleEmphasis(out, work.Bytes())
p.r.TripleEmphasis(out, work.Bytes())
}
return i + 3
case (i+1 < len(data) && data[i+1] == c):
// double symbol found, hand over to emph1
length = helperEmphasis(parser, out, origData[offset-2:], c)
length = helperEmphasis(p, out, origData[offset-2:], c)
if length == 0 {
return 0
} else {
@ -939,7 +939,7 @@ func helperTripleEmphasis(parser *Parser, out *bytes.Buffer, data []byte, offset
}
default:
// single symbol found, hand over to emph2
length = helperDoubleEmphasis(parser, out, origData[offset-1:], c)
length = helperDoubleEmphasis(p, out, origData[offset-1:], c)
if length == 0 {
return 0
} else {

View File

@ -19,10 +19,18 @@ import (
"bytes"
)
// Latex is a type that implements the Renderer interface for LaTeX output.
//
// Do not create this directly, instead use the LatexRenderer function.
type Latex struct {
}
// LatexRenderer creates and configures a Latex object, which
// satisfies the Renderer interface.
//
// flags is a set of LATEX_* options ORed together (currently no such options
// are defined).
func LatexRenderer(flags int) Renderer {
return &Latex{}
}

View File

@ -45,7 +45,7 @@ const (
LINK_TYPE_EMAIL
)
// These are the possible flag values for the listitem renderer.
// 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 (
@ -97,8 +97,17 @@ var blockTags = map[string]bool{
"blockquote": true,
}
// This interface defines the rendering interface.
// Renderer is the rendering interface.
// This is mostly of interest if you are implementing a new rendering format.
//
// When a byte slice is provided, it contains the (rendered) contents of the
// element.
//
// When a callback is provided instead, it will write the contents of the
// respective element directly to the output buffer and return true on success.
// If the callback returns false, the rendering function should reset the
// output buffer as though it had never been called.
//
// Currently Html and Latex implementations are provided
type Renderer interface {
// block-level callbacks
@ -137,12 +146,11 @@ type Renderer interface {
// Callback functions for inline parsing. One such function is defined
// for each character that triggers a response when parsing inline data.
type inlineParser func(parser *Parser, out *bytes.Buffer, data []byte, offset int) int
type inlineParser func(p *parser, out *bytes.Buffer, data []byte, offset int) int
// The main parser object.
// This is constructed by the Markdown function and
// contains state used during the parsing process.
type Parser struct {
// Parser holds runtime state used by the parser.
// This is constructed by the Markdown function.
type parser struct {
r Renderer
refs map[string]*reference
inlineCallback [256]inlineParser
@ -159,7 +167,8 @@ type Parser struct {
//
//
// Call Markdown with no extensions
// MarkdownBasic is a convenience function for simple rendering.
// It processes markdown input with no extensions enabled.
func MarkdownBasic(input []byte) []byte {
// set up the HTML renderer
htmlFlags := HTML_USE_XHTML
@ -172,6 +181,22 @@ func MarkdownBasic(input []byte) []byte {
}
// Call Markdown with most useful extensions enabled
// MarkdownCommon is a convenience function for simple rendering.
// It processes markdown input with common extensions enabled, including:
//
// * Smartypants processing with smart fractions and LaTeX dashes
//
// * Intra-word emphasis supression
//
// * Tables
//
// * Fenced code blocks
//
// * Autolinking
//
// * Strikethrough support
//
// * Strict header parsing
func MarkdownCommon(input []byte) []byte {
// set up the HTML renderer
htmlFlags := 0
@ -193,9 +218,13 @@ func MarkdownCommon(input []byte) []byte {
return Markdown(input, renderer, extensions)
}
// Parse and render a block of markdown-encoded text.
// The renderer is used to format the output, and extensions dictates which
// non-standard extensions are enabled.
// Markdown is the main rendering function.
// It parses and renders a block of markdown-encoded text.
// The supplied Renderer is used to format the output, and extensions dictates
// which non-standard extensions are enabled.
//
// To use the supplied Html or LaTeX renderers, see HtmlRenderer and
// LatexRenderer, respectively.
func Markdown(input []byte, renderer Renderer, extensions int) []byte {
// no point in parsing if we can't render
if renderer == nil {
@ -203,32 +232,32 @@ func Markdown(input []byte, renderer Renderer, extensions int) []byte {
}
// fill in the render structure
parser := new(Parser)
parser.r = renderer
parser.flags = extensions
parser.refs = make(map[string]*reference)
parser.maxNesting = 16
parser.insideLink = false
p := new(parser)
p.r = renderer
p.flags = extensions
p.refs = make(map[string]*reference)
p.maxNesting = 16
p.insideLink = false
// register inline parsers
parser.inlineCallback['*'] = emphasis
parser.inlineCallback['_'] = emphasis
p.inlineCallback['*'] = emphasis
p.inlineCallback['_'] = emphasis
if extensions&EXTENSION_STRIKETHROUGH != 0 {
parser.inlineCallback['~'] = emphasis
p.inlineCallback['~'] = emphasis
}
parser.inlineCallback['`'] = codeSpan
parser.inlineCallback['\n'] = lineBreak
parser.inlineCallback['['] = link
parser.inlineCallback['<'] = leftAngle
parser.inlineCallback['\\'] = escape
parser.inlineCallback['&'] = entity
p.inlineCallback['`'] = codeSpan
p.inlineCallback['\n'] = lineBreak
p.inlineCallback['['] = link
p.inlineCallback['<'] = leftAngle
p.inlineCallback['\\'] = escape
p.inlineCallback['&'] = entity
if extensions&EXTENSION_AUTOLINK != 0 {
parser.inlineCallback[':'] = autoLink
p.inlineCallback[':'] = autoLink
}
first := firstPass(parser, input)
second := secondPass(parser, first)
first := firstPass(p, input)
second := secondPass(p, first)
return second
}
@ -238,15 +267,15 @@ func Markdown(input []byte, renderer Renderer, extensions int) []byte {
// - expand tabs
// - normalize newlines
// - copy everything else
func firstPass(parser *Parser, input []byte) []byte {
func firstPass(p *parser, input []byte) []byte {
var out bytes.Buffer
tabSize := TAB_SIZE_DEFAULT
if parser.flags&EXTENSION_TAB_SIZE_EIGHT != 0 {
if p.flags&EXTENSION_TAB_SIZE_EIGHT != 0 {
tabSize = TAB_SIZE_EIGHT
}
beg, end := 0, 0
for beg < len(input) { // iterate over lines
if end = isReference(parser, input[beg:]); end > 0 {
if end = isReference(p, input[beg:]); end > 0 {
beg += end
} else { // skip to the next line
end = beg
@ -280,14 +309,14 @@ func firstPass(parser *Parser, input []byte) []byte {
}
// second pass: actual rendering
func secondPass(parser *Parser, input []byte) []byte {
func secondPass(p *parser, input []byte) []byte {
var output bytes.Buffer
parser.r.DocumentHeader(&output)
parser.block(&output, input)
parser.r.DocumentFooter(&output)
p.r.DocumentHeader(&output)
p.block(&output, input)
p.r.DocumentFooter(&output)
if parser.nesting != 0 {
if p.nesting != 0 {
panic("Nesting level did not end at zero")
}
@ -321,7 +350,7 @@ type reference struct {
// (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(parser *Parser, data []byte) int {
func isReference(p *parser, data []byte) int {
// up to 3 optional leading spaces
if len(data) < 4 {
return 0
@ -437,13 +466,10 @@ func isReference(parser *Parser, data []byte) int {
}
// a valid ref has been found
if parser == nil {
return lineEnd
}
// id matches are case-insensitive
id := string(bytes.ToLower(data[idOffset:idEnd]))
parser.refs[id] = &reference{
p.refs[id] = &reference{
link: data[linkOffset:linkEnd],
title: data[titleOffset:titleEnd],
}

View File

@ -343,10 +343,10 @@ func smartLeftAngle(out *bytes.Buffer, smrt *smartypantsData, previousChar byte,
type smartCallback func(out *bytes.Buffer, smrt *smartypantsData, previousChar byte, text []byte) int
type SmartypantsRenderer [256]smartCallback
type smartypantsRenderer [256]smartCallback
func Smartypants(flags int) *SmartypantsRenderer {
r := new(SmartypantsRenderer)
func smartypants(flags int) *smartypantsRenderer {
r := new(smartypantsRenderer)
r['"'] = smartDoubleQuote
r['&'] = smartAmp
r['\''] = smartSingleQuote