2011-05-30 01:00:31 +02:00
|
|
|
//
|
2011-06-28 04:11:32 +02:00
|
|
|
// Blackfriday Markdown Processor
|
|
|
|
// Available at http://github.com/russross/blackfriday
|
|
|
|
//
|
|
|
|
// Copyright © 2011 Russ Ross <russ@russross.com>.
|
2011-06-28 19:30:10 +02:00
|
|
|
// Distributed under the Simplified BSD License.
|
2011-06-28 04:11:32 +02:00
|
|
|
// See README.md for details.
|
2011-05-30 01:00:31 +02:00
|
|
|
//
|
|
|
|
|
|
|
|
//
|
|
|
|
// Functions to parse block-level elements.
|
|
|
|
//
|
|
|
|
|
|
|
|
package blackfriday
|
|
|
|
|
|
|
|
import (
|
|
|
|
"bytes"
|
|
|
|
)
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// Parse block-level data.
|
|
|
|
// Note: this function and many that it calls assume that
|
|
|
|
// the input buffer ends with a newline.
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) block(out *bytes.Buffer, data []byte) {
|
2011-06-30 04:15:58 +02:00
|
|
|
if len(data) == 0 || data[len(data)-1] != '\n' {
|
2011-07-05 22:22:21 +02:00
|
|
|
panic("block input is missing terminating newline")
|
2011-06-30 04:15:58 +02:00
|
|
|
}
|
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// this is called recursively: enforce a maximum depth
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.nesting >= p.maxNesting {
|
2011-05-30 01:00:31 +02:00
|
|
|
return
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.nesting++
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// parse out one block-level construct at a time
|
2011-05-30 01:00:31 +02:00
|
|
|
for len(data) > 0 {
|
2011-06-01 00:07:15 +02:00
|
|
|
// prefixed header:
|
|
|
|
//
|
|
|
|
// # Header 1
|
|
|
|
// ## Header 2
|
|
|
|
// ...
|
|
|
|
// ###### Header 6
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.isPrefixHeader(data) {
|
|
|
|
data = data[p.prefixHeader(out, data):]
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// block of preformatted HTML:
|
|
|
|
//
|
|
|
|
// <div>
|
|
|
|
// ...
|
|
|
|
// </div>
|
2011-06-29 19:13:17 +02:00
|
|
|
if data[0] == '<' {
|
2011-07-07 19:56:45 +02:00
|
|
|
if i := p.html(out, data, true); i > 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
data = data[i:]
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// blank lines. note: returns the # of bytes to skip
|
2011-07-07 19:56:45 +02:00
|
|
|
if i := p.isEmpty(data); i > 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
data = data[i:]
|
|
|
|
continue
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// indented code block:
|
2011-06-01 00:07:15 +02:00
|
|
|
//
|
2011-07-01 22:13:26 +02:00
|
|
|
// func max(a, b int) int {
|
|
|
|
// if a > b {
|
|
|
|
// return a
|
|
|
|
// }
|
|
|
|
// return b
|
|
|
|
// }
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.codePrefix(data) > 0 {
|
|
|
|
data = data[p.code(out, data):]
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// fenced code block:
|
|
|
|
//
|
|
|
|
// ``` go
|
|
|
|
// func fact(n int) int {
|
|
|
|
// if n <= 1 {
|
|
|
|
// return n
|
|
|
|
// }
|
|
|
|
// return n * fact(n-1)
|
|
|
|
// }
|
|
|
|
// ```
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.flags&EXTENSION_FENCED_CODE != 0 {
|
|
|
|
if i := p.fencedCode(out, data); i > 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
data = data[i:]
|
|
|
|
continue
|
|
|
|
}
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// horizontal rule:
|
2011-06-01 00:07:15 +02:00
|
|
|
//
|
2011-07-01 22:13:26 +02:00
|
|
|
// ------
|
|
|
|
// or
|
|
|
|
// ******
|
|
|
|
// or
|
|
|
|
// ______
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.isHRule(data) {
|
|
|
|
p.r.HRule(out)
|
2011-07-01 22:13:26 +02:00
|
|
|
var i int
|
|
|
|
for i = 0; data[i] != '\n'; i++ {
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
data = data[i:]
|
|
|
|
continue
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// block quote:
|
|
|
|
//
|
|
|
|
// > A big quote I found somewhere
|
|
|
|
// > on the web
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.quotePrefix(data) > 0 {
|
|
|
|
data = data[p.quote(out, data):]
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// table:
|
2011-06-01 00:07:15 +02:00
|
|
|
//
|
2011-07-01 22:13:26 +02:00
|
|
|
// Name | Age | Phone
|
|
|
|
// ------|-----|---------
|
|
|
|
// Bob | 31 | 555-1234
|
|
|
|
// Alice | 27 | 555-4321
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.flags&EXTENSION_TABLES != 0 {
|
|
|
|
if i := p.table(out, data); i > 0 {
|
2011-07-01 22:13:26 +02:00
|
|
|
data = data[i:]
|
|
|
|
continue
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// an itemized/unordered list:
|
|
|
|
//
|
|
|
|
// * Item 1
|
|
|
|
// * Item 2
|
|
|
|
//
|
|
|
|
// also works with + or -
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.uliPrefix(data) > 0 {
|
|
|
|
data = data[p.list(out, data, 0):]
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// a numbered/ordered list:
|
|
|
|
//
|
|
|
|
// 1. Item 1
|
|
|
|
// 2. Item 2
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.oliPrefix(data) > 0 {
|
|
|
|
data = data[p.list(out, data, LIST_TYPE_ORDERED):]
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// anything else must look like a normal paragraph
|
|
|
|
// note: this finds underlined headers, too
|
2011-07-07 19:56:45 +02:00
|
|
|
data = data[p.paragraph(out, data):]
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.nesting--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) isPrefixHeader(data []byte) bool {
|
2011-05-30 01:00:31 +02:00
|
|
|
if data[0] != '#' {
|
|
|
|
return false
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.flags&EXTENSION_SPACE_HEADERS != 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
level := 0
|
2011-06-30 04:15:58 +02:00
|
|
|
for level < 6 && data[level] == '#' {
|
2011-05-30 01:00:31 +02:00
|
|
|
level++
|
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
if data[level] != ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return false
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return true
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) prefixHeader(out *bytes.Buffer, data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
level := 0
|
2011-06-30 04:15:58 +02:00
|
|
|
for level < 6 && data[level] == '#' {
|
2011-05-30 01:00:31 +02:00
|
|
|
level++
|
|
|
|
}
|
|
|
|
i, end := 0, 0
|
2011-07-01 18:03:03 +02:00
|
|
|
for i = level; data[i] == ' '; i++ {
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
for end = i; data[end] != '\n'; end++ {
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
skip := end
|
|
|
|
for end > 0 && data[end-1] == '#' {
|
|
|
|
end--
|
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
for end > 0 && data[end-1] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
end--
|
|
|
|
}
|
|
|
|
if end > i {
|
2011-06-29 19:13:17 +02:00
|
|
|
work := func() bool {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.inline(out, data[i:end])
|
2011-06-29 19:13:17 +02:00
|
|
|
return true
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.Header(out, work, level)
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
return skip
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) isUnderlinedHeader(data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
// test of level 1 header
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[0] == '=' {
|
|
|
|
i := 1
|
|
|
|
for data[i] == '=' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
for data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] == '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 1
|
|
|
|
} else {
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// test of level 2 header
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[0] == '-' {
|
|
|
|
i := 1
|
|
|
|
for data[i] == '-' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
for data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] == '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 2
|
|
|
|
} else {
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
var i, j int
|
|
|
|
|
|
|
|
// identify the opening tag
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[0] != '<' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
curtag, tagfound := p.htmlFindTag(data[1:])
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
// handle special cases
|
|
|
|
if !tagfound {
|
2011-06-30 04:15:58 +02:00
|
|
|
// check for an HTML comment
|
2011-07-07 19:56:45 +02:00
|
|
|
if size := p.htmlComment(out, data, doRender); size > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
return size
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// check for an <hr> tag
|
2011-07-07 19:56:45 +02:00
|
|
|
if size := p.htmlHr(out, data, doRender); size > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
return size
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// no special case recognized
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// look for an unindented matching closing tag
|
2011-06-30 04:15:58 +02:00
|
|
|
// followed by a blank line
|
2011-05-30 01:00:31 +02:00
|
|
|
found := false
|
2011-06-30 04:15:58 +02:00
|
|
|
/*
|
|
|
|
closetag := []byte("\n</" + curtag + ">")
|
|
|
|
j = len(curtag) + 1
|
|
|
|
for !found {
|
|
|
|
// scan for a closing tag at the beginning of a line
|
|
|
|
if skip := bytes.Index(data[j:], closetag); skip >= 0 {
|
|
|
|
j += skip + len(closetag)
|
|
|
|
} else {
|
|
|
|
break
|
|
|
|
}
|
|
|
|
|
|
|
|
// see if it is the only thing on the line
|
2011-07-07 19:56:45 +02:00
|
|
|
if skip := p.isEmpty(data[j:]); skip > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
// see if it is followed by a blank line/eof
|
|
|
|
j += skip
|
|
|
|
if j >= len(data) {
|
|
|
|
found = true
|
|
|
|
i = j
|
|
|
|
} else {
|
2011-07-07 19:56:45 +02:00
|
|
|
if skip := p.isEmpty(data[j:]); skip > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
j += skip
|
|
|
|
found = true
|
|
|
|
i = j
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
*/
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
// if not found, try a second pass looking for indented match
|
|
|
|
// but not if tag is "ins" or "del" (following original Markdown.pl)
|
2011-06-30 04:15:58 +02:00
|
|
|
if !found && curtag != "ins" && curtag != "del" {
|
2011-05-30 01:00:31 +02:00
|
|
|
i = 1
|
|
|
|
for i < len(data) {
|
|
|
|
i++
|
|
|
|
for i < len(data) && !(data[i-1] == '<' && data[i] == '/') {
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
|
|
|
if i+2+len(curtag) >= len(data) {
|
|
|
|
break
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
j = p.htmlFindEnd(curtag, data[i-1:])
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
if j > 0 {
|
|
|
|
i += j - 1
|
|
|
|
found = true
|
|
|
|
break
|
|
|
|
}
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
if !found {
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// the end of the block has been found
|
2011-06-29 19:13:17 +02:00
|
|
|
if doRender {
|
2011-06-29 23:38:35 +02:00
|
|
|
// trim newlines
|
|
|
|
end := i
|
|
|
|
for end > 0 && data[end-1] == '\n' {
|
|
|
|
end--
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.BlockHtml(out, data[:end])
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
return i
|
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// HTML comment, lax form
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool) int {
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[0] != '<' || data[1] != '!' || data[2] != '-' || data[3] != '-' {
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
i := 5
|
|
|
|
|
|
|
|
// scan for an end-of-comment marker, across lines if necessary
|
|
|
|
for i < len(data) && !(data[i-2] == '-' && data[i-1] == '-' && data[i] == '>') {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
i++
|
|
|
|
|
|
|
|
// no end-of-comment marker
|
2011-05-30 01:00:31 +02:00
|
|
|
if i >= len(data) {
|
2011-06-30 04:15:58 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// needs to end with a blank line
|
2011-07-07 19:56:45 +02:00
|
|
|
if j := p.isEmpty(data[i:]); j > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
size := i + j
|
|
|
|
if doRender {
|
|
|
|
// trim trailing newlines
|
|
|
|
end := size
|
|
|
|
for end > 0 && data[end-1] == '\n' {
|
|
|
|
end--
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.BlockHtml(out, data[:end])
|
2011-06-30 04:15:58 +02:00
|
|
|
}
|
|
|
|
return size
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// HR, which is the only self-closing block tag considered
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[0] != '<' || (data[1] != 'h' && data[1] != 'H') || (data[2] != 'r' && data[2] != 'R') {
|
|
|
|
return 0
|
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
if data[3] != ' ' && data[3] != '/' && data[3] != '>' {
|
2011-06-30 04:15:58 +02:00
|
|
|
// not an <hr> tag after all; at least not a valid one
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
i := 3
|
|
|
|
for data[i] != '>' && data[i] != '\n' {
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
|
|
|
if data[i] == '>' {
|
|
|
|
i++
|
2011-07-07 19:56:45 +02:00
|
|
|
if j := p.isEmpty(data[i:]); j > 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
size := i + j
|
|
|
|
if doRender {
|
|
|
|
// trim newlines
|
|
|
|
end := size
|
|
|
|
for end > 0 && data[end-1] == '\n' {
|
|
|
|
end--
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.BlockHtml(out, data[:end])
|
2011-06-30 04:15:58 +02:00
|
|
|
}
|
|
|
|
return size
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) htmlFindTag(data []byte) (string, bool) {
|
2011-06-30 04:15:58 +02:00
|
|
|
i := 0
|
|
|
|
for isalnum(data[i]) {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
key := string(data[:i])
|
2011-06-29 00:02:12 +02:00
|
|
|
if blockTags[key] {
|
2011-05-30 01:00:31 +02:00
|
|
|
return key, true
|
|
|
|
}
|
|
|
|
return "", false
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) htmlFindEnd(tag string, data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
// assume data[0] == '<' && data[1] == '/' already tested
|
|
|
|
|
|
|
|
// check if tag is a match
|
2011-06-30 04:15:58 +02:00
|
|
|
closetag := []byte("</" + tag + ">")
|
|
|
|
if !bytes.HasPrefix(data, closetag) {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
i := len(closetag)
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// check that the rest of the line is blank
|
|
|
|
skip := 0
|
2011-07-07 19:56:45 +02:00
|
|
|
if skip = p.isEmpty(data[i:]); skip == 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
i += skip
|
|
|
|
skip = 0
|
|
|
|
|
|
|
|
if i >= len(data) {
|
|
|
|
return i
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
return i
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
if skip = p.isEmpty(data[i:]); skip == 0 {
|
2011-06-30 04:15:58 +02:00
|
|
|
// following line must be blank
|
|
|
|
return 0
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
return i + skip
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) isEmpty(data []byte) int {
|
2011-06-30 04:15:58 +02:00
|
|
|
// it is okay to call isEmpty on an empty buffer
|
|
|
|
if len(data) == 0 {
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2011-05-30 01:00:31 +02:00
|
|
|
var i int
|
2011-06-30 04:15:58 +02:00
|
|
|
for i = 0; data[i] != '\n'; i++ {
|
2011-07-01 18:03:03 +02:00
|
|
|
if data[i] != ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
}
|
|
|
|
return i + 1
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) isHRule(data []byte) bool {
|
2011-05-30 01:00:31 +02:00
|
|
|
i := 0
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// skip up to three spaces
|
|
|
|
for i < 3 && data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
|
|
|
// look at the hrule char
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] != '*' && data[i] != '-' && data[i] != '_' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return false
|
|
|
|
}
|
|
|
|
c := data[i]
|
|
|
|
|
|
|
|
// the whole line must be the char or whitespace
|
|
|
|
n := 0
|
2011-06-30 04:15:58 +02:00
|
|
|
for data[i] != '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
switch {
|
|
|
|
case data[i] == c:
|
|
|
|
n++
|
2011-07-01 18:03:03 +02:00
|
|
|
case data[i] != ' ':
|
2011-05-30 01:00:31 +02:00
|
|
|
return false
|
|
|
|
}
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
|
|
|
return n >= 3
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (skip int, marker string) {
|
2011-06-28 18:30:25 +02:00
|
|
|
i, size := 0, 0
|
|
|
|
skip = 0
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// skip up to three spaces
|
|
|
|
for i < 3 && data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-06-28 18:30:25 +02:00
|
|
|
// check for the marker characters: ~ or `
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] != '~' && data[i] != '`' {
|
2011-06-28 18:30:25 +02:00
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
c := data[i]
|
|
|
|
|
2011-06-28 18:30:25 +02:00
|
|
|
// the whole line must be the same char or whitespace
|
2011-06-30 04:15:58 +02:00
|
|
|
for data[i] == c {
|
2011-06-28 18:30:25 +02:00
|
|
|
size++
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-06-28 18:30:25 +02:00
|
|
|
// the marker char must occur at least 3 times
|
|
|
|
if size < 3 {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
marker = string(data[i-size : i])
|
|
|
|
|
|
|
|
// if this is the end marker, it must match the beginning marker
|
|
|
|
if oldmarker != "" && marker != oldmarker {
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
if syntax != nil {
|
|
|
|
syn := 0
|
|
|
|
|
2011-07-01 18:03:03 +02:00
|
|
|
for data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
syntaxStart := i
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] == '{' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
2011-06-29 00:02:12 +02:00
|
|
|
syntaxStart++
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
for data[i] != '}' && data[i] != '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
syn++
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] != '}' {
|
2011-06-28 18:30:25 +02:00
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-28 18:30:25 +02:00
|
|
|
// strip all whitespace at the beginning and the end
|
2011-05-30 01:00:31 +02:00
|
|
|
// of the {} block
|
2011-06-29 00:02:12 +02:00
|
|
|
for syn > 0 && isspace(data[syntaxStart]) {
|
|
|
|
syntaxStart++
|
2011-05-30 01:00:31 +02:00
|
|
|
syn--
|
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
for syn > 0 && isspace(data[syntaxStart+syn-1]) {
|
2011-05-30 01:00:31 +02:00
|
|
|
syn--
|
|
|
|
}
|
|
|
|
|
|
|
|
i++
|
|
|
|
} else {
|
2011-06-30 04:15:58 +02:00
|
|
|
for !isspace(data[i]) {
|
2011-05-30 01:00:31 +02:00
|
|
|
syn++
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
language := string(data[syntaxStart : syntaxStart+syn])
|
2011-05-30 01:00:31 +02:00
|
|
|
*syntax = &language
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] == ' ' {
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
if data[i] != '\n' {
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-28 18:30:25 +02:00
|
|
|
skip = i + 1
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) fencedCode(out *bytes.Buffer, data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
var lang *string
|
2011-07-07 19:56:45 +02:00
|
|
|
beg, marker := p.isFencedCode(data, &lang, "")
|
2011-07-01 22:13:26 +02:00
|
|
|
if beg == 0 || beg >= len(data) {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2011-05-31 19:11:04 +02:00
|
|
|
var work bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
for {
|
|
|
|
// safe to assume beg < len(data)
|
|
|
|
|
|
|
|
// check for the end of the code block
|
2011-07-07 19:56:45 +02:00
|
|
|
fenceEnd, _ := p.isFencedCode(data[beg:], nil, marker)
|
2011-06-29 00:02:12 +02:00
|
|
|
if fenceEnd != 0 {
|
|
|
|
beg += fenceEnd
|
2011-05-30 01:00:31 +02:00
|
|
|
break
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// copy the current line
|
|
|
|
end := beg
|
|
|
|
for data[end] != '\n' {
|
|
|
|
end++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
end++
|
2011-06-28 18:30:25 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// did we reach the end of the buffer without a closing marker?
|
|
|
|
if end >= len(data) {
|
2011-06-28 18:30:25 +02:00
|
|
|
return 0
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// verbatim copy to the working buffer
|
|
|
|
work.Write(data[beg:end])
|
|
|
|
beg = end
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-29 19:13:17 +02:00
|
|
|
syntax := ""
|
|
|
|
if lang != nil {
|
|
|
|
syntax = *lang
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.BlockCode(out, work.Bytes(), syntax)
|
2011-06-29 19:13:17 +02:00
|
|
|
|
2011-05-30 01:00:31 +02:00
|
|
|
return beg
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) table(out *bytes.Buffer, data []byte) int {
|
2011-07-01 22:13:26 +02:00
|
|
|
var header bytes.Buffer
|
2011-07-07 19:56:45 +02:00
|
|
|
i, columns := p.tableHeader(&header, data)
|
2011-06-01 00:28:07 +02:00
|
|
|
if i == 0 {
|
|
|
|
return 0
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
var body bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-01 00:28:07 +02:00
|
|
|
for i < len(data) {
|
2011-06-29 00:02:12 +02:00
|
|
|
pipes, rowStart := 0, i
|
2011-07-01 22:13:26 +02:00
|
|
|
for ; data[i] != '\n'; i++ {
|
2011-06-01 00:28:07 +02:00
|
|
|
if data[i] == '|' {
|
|
|
|
pipes++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
if pipes == 0 {
|
2011-06-29 00:02:12 +02:00
|
|
|
i = rowStart
|
2011-06-01 00:28:07 +02:00
|
|
|
break
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-01 00:28:07 +02:00
|
|
|
|
2011-07-05 22:22:21 +02:00
|
|
|
// include the newline in data sent to tableRow
|
2011-06-01 00:28:07 +02:00
|
|
|
i++
|
2011-07-07 19:56:45 +02:00
|
|
|
p.tableRow(&body, data[rowStart:i], columns)
|
2011-06-01 00:28:07 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.Table(out, header.Bytes(), body.Bytes(), columns)
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
return i
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns []int) {
|
2011-07-01 22:13:26 +02:00
|
|
|
i := 0
|
|
|
|
colCount := 1
|
|
|
|
for i = 0; data[i] != '\n'; i++ {
|
2011-05-30 01:00:31 +02:00
|
|
|
if data[i] == '|' {
|
2011-07-01 22:13:26 +02:00
|
|
|
colCount++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// doesn't look like a table header
|
|
|
|
if colCount == 1 {
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 22:22:21 +02:00
|
|
|
// include the newline in the data sent to tableRow
|
2011-07-01 22:13:26 +02:00
|
|
|
header := data[:i+1]
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// column count ignores pipes at beginning or end of line
|
2011-05-30 01:00:31 +02:00
|
|
|
if data[0] == '|' {
|
2011-07-01 22:13:26 +02:00
|
|
|
colCount--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
if i > 2 && data[i-1] == '|' {
|
2011-07-01 22:13:26 +02:00
|
|
|
colCount--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
columns = make([]int, colCount)
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// move on to the header underline
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
2011-07-01 22:13:26 +02:00
|
|
|
if i >= len(data) {
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
if data[i] == '|' {
|
|
|
|
i++
|
|
|
|
}
|
|
|
|
for data[i] == ' ' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// each column header is of form: / *:?-+:? *|/ with # dashes + # colons >= 3
|
|
|
|
// and trailing | optional on last column
|
2011-05-30 01:00:31 +02:00
|
|
|
col := 0
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] != '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
dashes := 0
|
|
|
|
|
|
|
|
if data[i] == ':' {
|
|
|
|
i++
|
2011-07-01 22:13:26 +02:00
|
|
|
columns[col] |= TABLE_ALIGNMENT_LEFT
|
2011-05-30 01:00:31 +02:00
|
|
|
dashes++
|
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] == '-' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
dashes++
|
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
if data[i] == ':' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
2011-07-01 22:13:26 +02:00
|
|
|
columns[col] |= TABLE_ALIGNMENT_RIGHT
|
2011-05-30 01:00:31 +02:00
|
|
|
dashes++
|
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// end of column test is messy
|
|
|
|
switch {
|
|
|
|
case dashes < 3:
|
|
|
|
// not a valid column
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
case data[i] == '|':
|
|
|
|
// marker found, now skip past trailing whitespace
|
|
|
|
col++
|
|
|
|
i++
|
|
|
|
for data[i] == ' ' {
|
|
|
|
i++
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-25 19:39:02 +02:00
|
|
|
// trailing junk found after last column
|
|
|
|
if col >= colCount && data[i] != '\n' {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
case data[i] != '|' && col+1 < colCount:
|
|
|
|
// something else found where marker was required
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
case data[i] == '\n':
|
|
|
|
// marker is optional for the last column
|
|
|
|
col++
|
|
|
|
|
|
|
|
default:
|
|
|
|
// trailing junk found after last column
|
|
|
|
return
|
|
|
|
}
|
|
|
|
}
|
|
|
|
if col != colCount {
|
|
|
|
return
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.tableRow(out, header, columns)
|
2011-07-01 22:13:26 +02:00
|
|
|
size = i + 1
|
2011-05-30 01:00:31 +02:00
|
|
|
return
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) tableRow(out *bytes.Buffer, data []byte, columns []int) {
|
2011-05-30 01:00:31 +02:00
|
|
|
i, col := 0, 0
|
2011-06-29 00:02:12 +02:00
|
|
|
var rowWork bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
if data[i] == '|' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-07-25 18:23:31 +02:00
|
|
|
for col = 0; col < len(columns) && i < len(data); col++ {
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
cellStart := i
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
for data[i] != '|' && data[i] != '\n' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
cellEnd := i
|
2011-07-25 18:23:31 +02:00
|
|
|
|
|
|
|
// skip the end-of-cell marker, possibly taking us past end of buffer
|
2011-07-01 22:13:26 +02:00
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
for cellEnd > cellStart && data[cellEnd-1] == ' ' {
|
2011-06-29 00:02:12 +02:00
|
|
|
cellEnd--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
var cellWork bytes.Buffer
|
2011-07-07 19:56:45 +02:00
|
|
|
p.inline(&cellWork, data[cellStart:cellEnd])
|
|
|
|
p.r.TableCell(&rowWork, cellWork.Bytes(), columns[col])
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// pad it out with empty columns to get the right number
|
|
|
|
for ; col < len(columns); col++ {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.TableCell(&rowWork, nil, columns[col])
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-01 22:13:26 +02:00
|
|
|
// silently ignore rows with too many cells
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.TableRow(out, rowWork.Bytes())
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// returns blockquote prefix length
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) quotePrefix(data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
i := 0
|
2011-06-30 04:15:58 +02:00
|
|
|
for i < 3 && data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-30 04:15:58 +02:00
|
|
|
if data[i] == '>' {
|
2011-07-01 18:03:03 +02:00
|
|
|
if data[i+1] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return i + 2
|
|
|
|
}
|
|
|
|
return i + 1
|
|
|
|
}
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// parse a blockquote fragment
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) quote(out *bytes.Buffer, data []byte) int {
|
2011-06-30 04:15:58 +02:00
|
|
|
var raw bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
beg, end := 0, 0
|
|
|
|
for beg < len(data) {
|
2011-07-01 22:13:26 +02:00
|
|
|
end = beg
|
|
|
|
for data[end] != '\n' {
|
|
|
|
end++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-01 22:13:26 +02:00
|
|
|
end++
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
if pre := p.quotePrefix(data[beg:]); pre > 0 {
|
|
|
|
// skip the prefix
|
2011-06-30 04:15:58 +02:00
|
|
|
beg += pre
|
2011-07-07 19:56:45 +02:00
|
|
|
} else if p.isEmpty(data[beg:]) > 0 &&
|
2011-07-05 22:22:21 +02:00
|
|
|
(end >= len(data) ||
|
2011-07-07 19:56:45 +02:00
|
|
|
(p.quotePrefix(data[end:]) == 0 && p.isEmpty(data[end:]) == 0)) {
|
2011-06-30 04:15:58 +02:00
|
|
|
// blockquote ends with at least one blank line
|
|
|
|
// followed by something without a blockquote prefix
|
2011-07-05 22:22:21 +02:00
|
|
|
break
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
// this line is part of the blockquote
|
|
|
|
raw.Write(data[beg:end])
|
2011-05-30 01:00:31 +02:00
|
|
|
beg = end
|
|
|
|
}
|
|
|
|
|
2011-06-30 04:15:58 +02:00
|
|
|
var cooked bytes.Buffer
|
2011-07-07 19:56:45 +02:00
|
|
|
p.block(&cooked, raw.Bytes())
|
|
|
|
p.r.BlockQuote(out, cooked.Bytes())
|
2011-05-30 01:00:31 +02:00
|
|
|
return end
|
|
|
|
}
|
|
|
|
|
|
|
|
// returns prefix length for block code
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) codePrefix(data []byte) int {
|
2011-07-01 22:13:26 +02:00
|
|
|
if data[0] == ' ' && data[1] == ' ' && data[2] == ' ' && data[3] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 4
|
|
|
|
}
|
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) code(out *bytes.Buffer, data []byte) int {
|
2011-05-31 19:11:04 +02:00
|
|
|
var work bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
i := 0
|
|
|
|
for i < len(data) {
|
|
|
|
beg := i
|
|
|
|
for data[i] != '\n' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
blankline := p.isEmpty(data[beg:i]) > 0
|
|
|
|
if pre := p.codePrefix(data[beg:i]); pre > 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
beg += pre
|
2011-07-05 22:22:21 +02:00
|
|
|
} else if !blankline {
|
|
|
|
// non-empty, non-prefixed line breaks the pre
|
|
|
|
i = beg
|
|
|
|
break
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// verbatim copy to the working buffeu
|
|
|
|
if blankline {
|
|
|
|
work.WriteByte('\n')
|
|
|
|
} else {
|
|
|
|
work.Write(data[beg:i])
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
|
|
|
// trim all the \n off the end of work
|
|
|
|
workbytes := work.Bytes()
|
2011-07-05 02:56:29 +02:00
|
|
|
eol := len(workbytes)
|
|
|
|
for eol > 0 && workbytes[eol-1] == '\n' {
|
|
|
|
eol--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
if eol != len(workbytes) {
|
|
|
|
work.Truncate(eol)
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
work.WriteByte('\n')
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.BlockCode(out, work.Bytes(), "")
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
return i
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
|
|
|
// returns unordered list item prefix
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) uliPrefix(data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
i := 0
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// start with up to 3 spaces
|
2011-07-05 02:56:29 +02:00
|
|
|
for i < 3 && data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
// need a *, +, or - followed by a space
|
2011-07-05 02:56:29 +02:00
|
|
|
if (data[i] != '*' && data[i] != '+' && data[i] != '-') ||
|
2011-07-01 18:03:03 +02:00
|
|
|
data[i+1] != ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
return i + 2
|
|
|
|
}
|
|
|
|
|
|
|
|
// returns ordered list item prefix
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) oliPrefix(data []byte) int {
|
2011-05-30 01:00:31 +02:00
|
|
|
i := 0
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// start with up to 3 spaces
|
2011-07-05 02:56:29 +02:00
|
|
|
for i < 3 && data[i] == ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// count the digits
|
|
|
|
start := i
|
2011-07-05 02:56:29 +02:00
|
|
|
for data[i] >= '0' && data[i] <= '9' {
|
2011-05-30 01:00:31 +02:00
|
|
|
i++
|
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
// we need >= 1 digits followed by a dot and a space
|
2011-07-05 02:56:29 +02:00
|
|
|
if start == i || data[i] != '.' || data[i+1] != ' ' {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
return i + 2
|
|
|
|
}
|
|
|
|
|
|
|
|
// parse ordered or unordered list block
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) list(out *bytes.Buffer, data []byte, flags int) int {
|
2011-06-25 23:02:46 +02:00
|
|
|
i := 0
|
2011-07-01 19:19:42 +02:00
|
|
|
flags |= LIST_ITEM_BEGINNING_OF_LIST
|
2011-06-25 23:02:46 +02:00
|
|
|
work := func() bool {
|
|
|
|
for i < len(data) {
|
2011-07-07 19:56:45 +02:00
|
|
|
skip := p.listItem(out, data[i:], &flags)
|
2011-07-05 02:56:29 +02:00
|
|
|
i += skip
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
if skip == 0 || flags&LIST_ITEM_END_OF_LIST != 0 {
|
2011-06-25 23:02:46 +02:00
|
|
|
break
|
|
|
|
}
|
2011-07-01 19:19:42 +02:00
|
|
|
flags &= ^LIST_ITEM_BEGINNING_OF_LIST
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-25 23:02:46 +02:00
|
|
|
return true
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.List(out, work, flags)
|
2011-05-30 01:00:31 +02:00
|
|
|
return i
|
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// Parse a single list item.
|
|
|
|
// Assumes initial prefix is already removed if this is a sublist.
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) listItem(out *bytes.Buffer, data []byte, flags *int) int {
|
2011-07-05 02:56:29 +02:00
|
|
|
// keep track of the indentation of the first line
|
|
|
|
itemIndent := 0
|
|
|
|
for itemIndent < 3 && data[itemIndent] == ' ' {
|
|
|
|
itemIndent++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
i := p.uliPrefix(data)
|
2011-07-05 02:56:29 +02:00
|
|
|
if i == 0 {
|
2011-07-07 19:56:45 +02:00
|
|
|
i = p.oliPrefix(data)
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
if i == 0 {
|
2011-05-30 01:00:31 +02:00
|
|
|
return 0
|
|
|
|
}
|
|
|
|
|
|
|
|
// skip leading whitespace on first line
|
2011-07-05 02:56:29 +02:00
|
|
|
for data[i] == ' ' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// find the end of the line
|
|
|
|
line := i
|
|
|
|
for data[i-1] != '\n' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// get working buffer
|
|
|
|
var raw bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
// put the first line into the working buffer
|
2011-07-05 02:56:29 +02:00
|
|
|
raw.Write(data[line:i])
|
|
|
|
line = i
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
// process the following lines
|
2011-07-05 02:56:29 +02:00
|
|
|
containsBlankLine := false
|
|
|
|
sublist := 0
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
gatherlines:
|
2011-07-05 02:56:29 +02:00
|
|
|
for line < len(data) {
|
|
|
|
i++
|
|
|
|
|
|
|
|
// find the end of this line
|
|
|
|
for data[i-1] != '\n' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// if it is an empty line, guess that it is part of this item
|
|
|
|
// and move on to the next line
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.isEmpty(data[line:i]) > 0 {
|
2011-06-29 00:02:12 +02:00
|
|
|
containsBlankLine = true
|
2011-07-05 02:56:29 +02:00
|
|
|
line = i
|
2011-05-30 01:00:31 +02:00
|
|
|
continue
|
|
|
|
}
|
|
|
|
|
|
|
|
// calculate the indentation
|
2011-07-05 02:56:29 +02:00
|
|
|
indent := 0
|
|
|
|
for indent < 4 && line+indent < i && data[line+indent] == ' ' {
|
|
|
|
indent++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
chunk := data[line+indent : i]
|
|
|
|
|
|
|
|
// evaluate how this line fits in
|
|
|
|
switch {
|
|
|
|
// is this a nested list item?
|
2011-07-07 19:56:45 +02:00
|
|
|
case (p.uliPrefix(chunk) > 0 && !p.isHRule(chunk)) ||
|
|
|
|
p.oliPrefix(chunk) > 0:
|
2011-06-27 18:13:13 +02:00
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
if containsBlankLine {
|
2011-07-05 02:56:29 +02:00
|
|
|
*flags |= LIST_ITEM_CONTAINS_BLOCK
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// to be a nested list, it must be indented more
|
|
|
|
// if not, it is the next item in the same list
|
|
|
|
if indent <= itemIndent {
|
2011-07-07 19:56:45 +02:00
|
|
|
break gatherlines
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// is this the first item in the the nested list?
|
2011-05-30 01:00:31 +02:00
|
|
|
if sublist == 0 {
|
2011-07-05 02:56:29 +02:00
|
|
|
sublist = raw.Len()
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
|
|
|
|
// is this a nested prefix header?
|
2011-07-07 19:56:45 +02:00
|
|
|
case p.isPrefixHeader(chunk):
|
2011-07-05 02:56:29 +02:00
|
|
|
// 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
|
2011-07-07 19:56:45 +02:00
|
|
|
break gatherlines
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
*flags |= LIST_ITEM_CONTAINS_BLOCK
|
|
|
|
|
|
|
|
// anything following an empty line is only part
|
|
|
|
// of this item if it is indented 4 spaces
|
|
|
|
// (regardless of the indentation of the beginning of the item)
|
|
|
|
case containsBlankLine && indent < 4:
|
|
|
|
*flags |= LIST_ITEM_END_OF_LIST
|
2011-07-07 19:56:45 +02:00
|
|
|
break gatherlines
|
2011-07-05 02:56:29 +02:00
|
|
|
|
|
|
|
// a blank line means this should be parsed as a block
|
|
|
|
case containsBlankLine:
|
|
|
|
raw.WriteByte('\n')
|
|
|
|
*flags |= LIST_ITEM_CONTAINS_BLOCK
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-29 00:02:12 +02:00
|
|
|
containsBlankLine = false
|
2011-05-30 01:00:31 +02:00
|
|
|
|
|
|
|
// add the line into the working buffer without prefix
|
2011-07-05 02:56:29 +02:00
|
|
|
raw.Write(data[line+indent : i])
|
|
|
|
line = i
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// render the contents of the list item
|
|
|
|
rawBytes := raw.Bytes()
|
|
|
|
var cooked bytes.Buffer
|
2011-05-30 01:00:31 +02:00
|
|
|
if *flags&LIST_ITEM_CONTAINS_BLOCK != 0 {
|
|
|
|
// intermediate render of block li
|
2011-07-05 02:56:29 +02:00
|
|
|
if sublist > 0 {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.block(&cooked, rawBytes[:sublist])
|
|
|
|
p.block(&cooked, rawBytes[sublist:])
|
2011-05-30 01:00:31 +02:00
|
|
|
} else {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.block(&cooked, rawBytes)
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
} else {
|
|
|
|
// intermediate render of inline li
|
2011-07-05 02:56:29 +02:00
|
|
|
if sublist > 0 {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.inline(&cooked, rawBytes[:sublist])
|
|
|
|
p.block(&cooked, rawBytes[sublist:])
|
2011-05-30 01:00:31 +02:00
|
|
|
} else {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.inline(&cooked, rawBytes)
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// render the actual list item
|
|
|
|
cookedBytes := cooked.Bytes()
|
|
|
|
parsedEnd := len(cookedBytes)
|
|
|
|
|
|
|
|
// strip trailing newlines
|
|
|
|
for parsedEnd > 0 && cookedBytes[parsedEnd-1] == '\n' {
|
2011-06-29 23:38:35 +02:00
|
|
|
parsedEnd--
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.ListItem(out, cookedBytes[:parsedEnd], *flags)
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
return line
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// render a single paragraph that has already been parsed out
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) renderParagraph(out *bytes.Buffer, data []byte) {
|
2011-07-05 02:56:29 +02:00
|
|
|
if len(data) == 0 {
|
|
|
|
return
|
|
|
|
}
|
|
|
|
|
|
|
|
// trim leading spaces
|
2011-06-25 23:18:34 +02:00
|
|
|
beg := 0
|
2011-07-05 02:56:29 +02:00
|
|
|
for data[beg] == ' ' {
|
2011-06-25 23:18:34 +02:00
|
|
|
beg++
|
|
|
|
}
|
|
|
|
|
2011-07-05 02:56:29 +02:00
|
|
|
// trim trailing newline
|
|
|
|
end := len(data) - 1
|
|
|
|
|
|
|
|
// trim trailing spaces
|
|
|
|
for end > beg && data[end-1] == ' ' {
|
2011-06-01 00:07:15 +02:00
|
|
|
end--
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-06-27 01:21:11 +02:00
|
|
|
work := func() bool {
|
2011-07-07 19:56:45 +02:00
|
|
|
p.inline(out, data[beg:end])
|
2011-06-27 01:21:11 +02:00
|
|
|
return true
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
p.r.Paragraph(out, work)
|
2011-06-01 00:07:15 +02:00
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
|
2011-06-01 00:07:15 +02:00
|
|
|
// prev: index of 1st char of previous line
|
|
|
|
// line: index of 1st char of current line
|
|
|
|
// i: index of cursor/end of current line
|
|
|
|
var prev, line, i int
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// keep going until we find something to mark the end of the paragraph
|
|
|
|
for i < len(data) {
|
|
|
|
// mark the beginning of the current line
|
|
|
|
prev = line
|
|
|
|
current := data[i:]
|
|
|
|
line = i
|
|
|
|
|
|
|
|
// did we find a blank line marking the end of the paragraph?
|
2011-07-07 19:56:45 +02:00
|
|
|
if n := p.isEmpty(current); n > 0 {
|
|
|
|
p.renderParagraph(out, data[:i])
|
2011-06-01 00:07:15 +02:00
|
|
|
return i + n
|
|
|
|
}
|
|
|
|
|
|
|
|
// an underline under some text marks a header, so our paragraph ended on prev line
|
2011-06-29 19:13:17 +02:00
|
|
|
if i > 0 {
|
2011-07-07 19:56:45 +02:00
|
|
|
if level := p.isUnderlinedHeader(current); level > 0 {
|
2011-06-01 00:07:15 +02:00
|
|
|
// render the paragraph
|
2011-07-07 19:56:45 +02:00
|
|
|
p.renderParagraph(out, data[:prev])
|
2011-06-01 00:07:15 +02:00
|
|
|
|
2011-06-27 19:42:38 +02:00
|
|
|
// ignore leading and trailing whitespace
|
|
|
|
eol := i - 1
|
2011-07-01 18:03:03 +02:00
|
|
|
for prev < eol && data[prev] == ' ' {
|
2011-06-27 19:42:38 +02:00
|
|
|
prev++
|
|
|
|
}
|
2011-07-01 18:03:03 +02:00
|
|
|
for eol > prev && data[eol-1] == ' ' {
|
2011-06-27 19:42:38 +02:00
|
|
|
eol--
|
|
|
|
}
|
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// render the header
|
2011-06-27 19:42:38 +02:00
|
|
|
// this ugly double closure avoids forcing variables onto the heap
|
2011-07-07 19:56:45 +02:00
|
|
|
work := func(o *bytes.Buffer, pp *parser, d []byte) func() bool {
|
2011-06-25 16:20:08 +02:00
|
|
|
return func() bool {
|
2011-07-07 19:56:45 +02:00
|
|
|
pp.inline(o, d)
|
2011-06-25 16:20:08 +02:00
|
|
|
return true
|
|
|
|
}
|
2011-07-07 19:56:45 +02:00
|
|
|
}(out, p, data[prev:eol])
|
|
|
|
p.r.Header(out, work, level)
|
2011-06-01 00:07:15 +02:00
|
|
|
|
|
|
|
// find the end of the underline
|
2011-07-05 02:56:29 +02:00
|
|
|
for data[i] != '\n' {
|
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-06-01 00:07:15 +02:00
|
|
|
return i
|
|
|
|
}
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// if the next line starts a block of HTML, then the paragraph ends here
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
|
|
|
|
if data[i] == '<' && p.html(out, current, false) > 0 {
|
2011-06-01 00:07:15 +02:00
|
|
|
// rewind to before the HTML block
|
2011-07-07 19:56:45 +02:00
|
|
|
p.renderParagraph(out, data[:i])
|
2011-06-01 00:07:15 +02:00
|
|
|
return i
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
}
|
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// if there's a prefixed header or a horizontal rule after this, paragraph is over
|
2011-07-07 19:56:45 +02:00
|
|
|
if p.isPrefixHeader(current) || p.isHRule(current) {
|
|
|
|
p.renderParagraph(out, data[:i])
|
2011-06-01 00:07:15 +02:00
|
|
|
return i
|
|
|
|
}
|
2011-05-30 01:00:31 +02:00
|
|
|
|
2011-06-01 00:07:15 +02:00
|
|
|
// otherwise, scan to the beginning of the next line
|
2011-07-05 02:56:29 +02:00
|
|
|
for data[i] != '\n' {
|
2011-06-01 00:07:15 +02:00
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
2011-07-05 02:56:29 +02:00
|
|
|
i++
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|
|
|
|
|
2011-07-07 19:56:45 +02:00
|
|
|
p.renderParagraph(out, data[:i])
|
2011-06-01 00:07:15 +02:00
|
|
|
return i
|
2011-05-30 01:00:31 +02:00
|
|
|
}
|