Add support for markdown to HTML conversion (matrix). Closes #663 (#670)

This uses our own gomatrix lib with the SendHTML function which
adds HTML to formatted_body in matrix.
golang-commonmark is used to convert markdown into valid HTML.

49 files changed, 5966 insertions, 0 deletions

diff --git a/vendor/gitlab.com/golang-commonmark/markdown/.gitlab-ci.yml b/vendor/gitlab.com/golang-commonmark/markdown/.gitlab-ci.yml
new file mode 100644
index 00000000..58bdd5c4
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/.gitlab-ci.yml
@@ -0,0 +1,25 @@
+image: golang:1.11
+
+stages:
+  - build
+  - test
+
+before_script:
+  - go get github.com/russross/blackfriday
+  - go get gitlab.com/golang-commonmark/html
+  - go get gitlab.com/golang-commonmark/linkify
+  - go get gitlab.com/golang-commonmark/mdurl
+  - go get gitlab.com/golang-commonmark/puny
+  - go get gitlab.com/opennota/wd
+  - go get gopkg.in/russross/blackfriday.v2
+
+build:
+  stage: build
+  script:
+    - go build ./...
+
+test:
+  stage: test
+  script:
+    - test -z "$(gofmt -l . | tee /dev/stderr)"
+    - go test -cover ./...
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/AUTHORS b/vendor/gitlab.com/golang-commonmark/markdown/AUTHORS
new file mode 100644
index 00000000..041a96ce
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/AUTHORS
@@ -0,0 +1 @@
+opennota@gmail.com
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/LICENSE b/vendor/gitlab.com/golang-commonmark/markdown/LICENSE
new file mode 100644
index 00000000..9cdcd201
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/LICENSE
@@ -0,0 +1,10 @@
+Copyright (c) 2015, The Authors
+All rights reserved.
+
+Redistribution and use in source and binary forms, with or without modification, are permitted provided that the following conditions are met:
+
+1. Redistributions of source code must retain the above copyright notice, this list of conditions and the following disclaimer.
+
+2. Redistributions in binary form must reproduce the above copyright notice, this list of conditions and the following disclaimer in the documentation and/or other materials provided with the distribution.
+
+THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/README.md b/vendor/gitlab.com/golang-commonmark/markdown/README.md
new file mode 100644
index 00000000..ac2a1de8
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/README.md
@@ -0,0 +1,64 @@
+markdown [![GoDoc](http://godoc.org/gitlab.com/golang-commonmark/markdown?status.svg)](http://godoc.org/gitlab.com/golang-commonmark/markdown) [![License](https://img.shields.io/badge/licence-BSD--2--Clause-blue.svg)](https://opensource.org/licenses/BSD-2-Clause) [![Pipeline status](https://gitlab.com/golang-commonmark/markdown/badges/master/pipeline.svg)](https://gitlab.com/golang-commonmark/markdown/commits/master) [![Coverage report](https://gitlab.com/golang-commonmark/markdown/badges/master/coverage.svg)](https://gitlab.com/golang-commonmark/markdown/commits/master)
+========
+
+Package golang-commonmark/markdown provides a CommonMark-compliant markdown parser and renderer, written in Go.
+
+## Installation
+
+    go get -u gitlab.com/golang-commonmark/markdown
+
+You can also go get [mdtool](https://gitlab.com/golang-commonmark/mdtool), an example command-line tool:
+
+    go get -u gitlab.com/golang-commonmark/mdtool
+
+## Standards support
+
+Currently supported CommonMark spec: [v0.28](http://spec.commonmark.org/0.28/).
+
+## Extensions
+
+Besides the features required by CommonMark, golang-commonmark/markdown supports:
+
+  * Tables (GFM)
+  * Strikethrough (GFM)
+  * Autoconverting plain-text URLs to links
+  * Typographic replacements (smart quotes and other)
+
+## Usage
+
+``` go
+md := markdown.New(markdown.XHTMLOutput(true))
+fmt.Println(md.RenderToString([]byte("Header\n===\nText")))
+```
+
+Check out [the source of mdtool](https://gitlab.com/golang-commonmark/mdtool/blob/master/main.go) for a more complete example.
+
+The following options are currently supported:
+
+  Name            |  Type     |                        Description                          | Default
+  --------------- | --------- | ----------------------------------------------------------- | ---------
+  HTML            | bool      | whether to enable raw HTML                                  | false
+  Tables          | bool      | whether to enable GFM tables                                | true
+  Linkify         | bool      | whether to autoconvert plain-text URLs to links             | true
+  Typographer     | bool      | whether to enable typographic replacements                  | true
+  Quotes          | string / []string | double + single quote replacement pairs for the typographer | “”‘’
+  MaxNesting      | int       | maximum nesting level                                       | 20
+  LangPrefix      | string    | CSS language prefix for fenced blocks                       | language-
+  Breaks          | bool      | whether to convert newlines inside paragraphs into `<br>`   | false
+  XHTMLOutput     | bool      | whether to output XHTML instead of HTML                     | false
+
+## Benchmarks
+
+Rendering spec/spec-0.28.txt on a Intel(R) Core(TM) i5-2400 CPU @ 3.10GHz
+
+    BenchmarkRenderSpecNoHTML         100    10254720 ns/op    2998037 B/op    18225 allocs/op
+    BenchmarkRenderSpec               100    10180241 ns/op    2997307 B/op    18214 allocs/op
+    BenchmarkRenderSpecBlackFriday    200     7241749 ns/op    2834340 B/op    17101 allocs/op
+    BenchmarkRenderSpecBlackFriday2   200     7448256 ns/op    2991202 B/op    16705 allocs/op
+
+## See also
+
+https://github.com/jgm/CommonMark — the reference CommonMark implementations in C and JavaScript,
+  also contains the latest spec and an online demo.
+
+http://talk.commonmark.org — the CommonMark forum, a good place to join together the efforts of the developers.
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/align.go b/vendor/gitlab.com/golang-commonmark/markdown/align.go
new file mode 100644
index 00000000..bcb63600
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/align.go
@@ -0,0 +1,26 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+type Align byte
+
+const (
+	AlignNone = iota
+	AlignLeft
+	AlignCenter
+	AlignRight
+)
+
+func (a Align) String() string {
+	switch a {
+	case AlignLeft:
+		return "left"
+	case AlignCenter:
+		return "center"
+	case AlignRight:
+		return "right"
+	}
+	return ""
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/autolink.go b/vendor/gitlab.com/golang-commonmark/markdown/autolink.go
new file mode 100644
index 00000000..ab51c71b
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/autolink.go
@@ -0,0 +1,70 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"regexp"
+	"strings"
+)
+
+var (
+	rAutolink = regexp.MustCompile(`^<([a-zA-Z][a-zA-Z0-9+.\-]{1,31}):([^<>\x00-\x20]*)>`)
+	rEmail    = regexp.MustCompile(`^<([a-zA-Z0-9.!#$%&'*+/=?^_{|}~-]+@[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?(?:\.[a-zA-Z0-9](?:[a-zA-Z0-9-]{0,61}[a-zA-Z0-9])?)*)>`)
+)
+
+func ruleAutolink(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	src := s.Src
+
+	if src[pos] != '<' {
+		return false
+	}
+
+	tail := src[pos:]
+
+	if strings.IndexByte(tail, '>') < 0 {
+		return false
+	}
+
+	link := rAutolink.FindString(tail)
+	if link != "" {
+		link = link[1 : len(link)-1]
+		href := normalizeLink(link)
+		if !validateLink(href) {
+			return false
+		}
+
+		if !silent {
+			s.PushOpeningToken(&LinkOpen{Href: href})
+			s.PushToken(&Text{Content: normalizeLinkText(link)})
+			s.PushClosingToken(&LinkClose{})
+		}
+
+		s.Pos += len(link) + 2
+
+		return true
+	}
+
+	email := rEmail.FindString(tail)
+	if email != "" {
+		email = email[1 : len(email)-1]
+		href := normalizeLink("mailto:" + email)
+		if !validateLink(href) {
+			return false
+		}
+
+		if !silent {
+			s.PushOpeningToken(&LinkOpen{Href: href})
+			s.PushToken(&Text{Content: normalizeLinkText(email)})
+			s.PushClosingToken(&LinkClose{})
+		}
+
+		s.Pos += len(email) + 2
+
+		return true
+	}
+
+	return false
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/backticks.go b/vendor/gitlab.com/golang-commonmark/markdown/backticks.go
new file mode 100644
index 00000000..07f268dc
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/backticks.go
@@ -0,0 +1,60 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func ruleBackticks(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	src := s.Src
+
+	if src[pos] != '`' {
+		return false
+	}
+
+	start := pos
+	pos++
+	max := s.PosMax
+
+	for pos < max && src[pos] == '`' {
+		pos++
+	}
+
+	marker := src[start:pos]
+
+	matchStart := pos
+	matchEnd := pos
+	for {
+		matchStart = strings.IndexByte(src[matchEnd:], '`')
+		if matchStart == -1 {
+			break
+		}
+		matchStart += matchEnd
+
+		matchEnd = matchStart + 1
+
+		for matchEnd < max && src[matchEnd] == '`' {
+			matchEnd++
+		}
+
+		if matchEnd-matchStart == len(marker) {
+			if !silent {
+				s.PushToken(&CodeInline{
+					Content: normalizeInlineCode(src[pos:matchStart]),
+				})
+			}
+			s.Pos = matchEnd
+			return true
+		}
+	}
+
+	if !silent {
+		s.Pending.WriteString(marker)
+	}
+
+	s.Pos += len(marker)
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/balance_pairs.go b/vendor/gitlab.com/golang-commonmark/markdown/balance_pairs.go
new file mode 100644
index 00000000..d5830634
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/balance_pairs.go
@@ -0,0 +1,43 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleBalancePairs(s *StateInline) {
+	delimiters := s.Delimiters
+	max := len(delimiters)
+
+	for i := 0; i < max; i++ {
+		lastDelim := delimiters[i]
+
+		if !lastDelim.Close {
+			continue
+		}
+
+		j := i - lastDelim.Jump - 1
+
+		for j >= 0 {
+			currDelim := delimiters[j]
+
+			if currDelim.Open &&
+				currDelim.Marker == lastDelim.Marker &&
+				currDelim.End < 0 &&
+				currDelim.Level == lastDelim.Level {
+				oddMatch := (currDelim.Close || lastDelim.Open) &&
+					currDelim.Length != -1 &&
+					lastDelim.Length != -1 &&
+					(currDelim.Length+lastDelim.Length)%3 == 0
+				if !oddMatch {
+					delimiters[i].Jump = i - j
+					delimiters[i].Open = false
+					delimiters[j].End = i
+					delimiters[j].Jump = 0
+					break
+				}
+			}
+
+			j -= currDelim.Jump + 1
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/blockquote.go b/vendor/gitlab.com/golang-commonmark/markdown/blockquote.go
new file mode 100644
index 00000000..f8ad3626
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/blockquote.go
@@ -0,0 +1,233 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "unicode/utf8"
+
+var blockquoteTerminatedBy []BlockRule
+
+func ruleBlockQuote(s *StateBlock, startLine, endLine int, silent bool) bool {
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+	src := s.Src
+
+	if pos >= max {
+		return false
+	}
+	if src[pos] != '>' {
+		return false
+	}
+	pos++
+
+	if silent {
+		return true
+	}
+
+	initial := s.SCount[startLine] + pos - (s.BMarks[startLine] + s.TShift[startLine])
+	offset := initial
+
+	spaceAfterMarker := false
+	adjustTab := false
+	if pos < max {
+		if src[pos] == ' ' {
+			pos++
+			initial++
+			offset++
+			spaceAfterMarker = true
+		} else if src[pos] == '\t' {
+			spaceAfterMarker = true
+			if (s.BSCount[startLine]+offset)%4 == 3 {
+				pos++
+				initial++
+				offset++
+			} else {
+				adjustTab = true
+			}
+		}
+	}
+
+	oldBMarks := []int{s.BMarks[startLine]}
+	s.BMarks[startLine] = pos
+
+	for pos < max {
+		r, size := utf8.DecodeRuneInString(src[pos:])
+		if runeIsSpace(r) {
+			if r == '\t' {
+				d := 0
+				if adjustTab {
+					d = 1
+				}
+				offset += 4 - (offset+s.BSCount[startLine]+d)%4
+			} else {
+				offset++
+			}
+		} else {
+			break
+		}
+		pos += size
+	}
+
+	oldBSCount := []int{s.BSCount[startLine]}
+	d := 0
+	if spaceAfterMarker {
+		d = 1
+	}
+	s.BSCount[startLine] = s.SCount[startLine] + 1 + d
+
+	lastLineEmpty := pos >= max
+
+	oldSCount := []int{s.SCount[startLine]}
+	s.SCount[startLine] = offset - initial
+
+	oldTShift := []int{s.TShift[startLine]}
+	s.TShift[startLine] = pos - s.BMarks[startLine]
+
+	oldParentType := s.ParentType
+	s.ParentType = ptBlockQuote
+	wasOutdented := false
+
+	oldLineMax := s.LineMax
+
+	nextLine := startLine + 1
+	for ; nextLine < endLine; nextLine++ {
+		if s.SCount[nextLine] < s.BlkIndent {
+			wasOutdented = true
+		}
+		pos = s.BMarks[nextLine] + s.TShift[nextLine]
+		max = s.EMarks[nextLine]
+
+		if pos >= max {
+			break
+		}
+
+		pos++
+		if src[pos-1] == '>' && !wasOutdented {
+			initial = s.SCount[nextLine] + pos + (s.BMarks[nextLine] + s.TShift[nextLine])
+			offset = initial
+
+			if pos >= len(src) || src[pos] != ' ' && src[pos] != '\t' {
+				spaceAfterMarker = true
+			} else if src[pos] == ' ' {
+				pos++
+				initial++
+				offset++
+				adjustTab = false
+				spaceAfterMarker = true
+			} else if src[pos] == '\t' {
+				spaceAfterMarker = true
+
+				if (s.BSCount[nextLine]+offset)%4 == 3 {
+					pos++
+					initial++
+					offset++
+					adjustTab = false
+				} else {
+					adjustTab = true
+				}
+			}
+
+			oldBMarks = append(oldBMarks, s.BMarks[nextLine])
+			s.BMarks[nextLine] = pos
+
+			for pos < max {
+				r, size := utf8.DecodeRuneInString(src[pos:])
+				if runeIsSpace(r) {
+					if r == '\t' {
+						d := 0
+						if adjustTab {
+							d = 1
+						}
+						offset += 4 - (offset+s.BSCount[startLine]+d)%4
+					} else {
+						offset++
+					}
+				} else {
+					break
+				}
+				pos += size
+			}
+
+			lastLineEmpty = pos >= max
+
+			oldBSCount = append(oldBSCount, s.BSCount[nextLine])
+			d := 0
+			if spaceAfterMarker {
+				d = 1
+			}
+			s.BSCount[nextLine] = s.SCount[nextLine] + 1 + d
+
+			oldSCount = append(oldSCount, s.SCount[nextLine])
+			s.SCount[nextLine] = offset - initial
+
+			oldTShift = append(oldTShift, s.TShift[nextLine])
+			s.TShift[nextLine] = pos - s.BMarks[nextLine]
+
+			continue
+		}
+
+		if lastLineEmpty {
+			break
+		}
+
+		terminate := false
+		for _, r := range blockquoteTerminatedBy {
+			if r(s, nextLine, endLine, true) {
+				terminate = true
+				break
+			}
+		}
+
+		if terminate {
+			s.LineMax = nextLine
+
+			if s.BlkIndent != 0 {
+				oldBMarks = append(oldBMarks, s.BMarks[nextLine])
+				oldBSCount = append(oldBSCount, s.BSCount[nextLine])
+				oldTShift = append(oldTShift, s.TShift[nextLine])
+				oldSCount = append(oldSCount, s.SCount[nextLine])
+				s.SCount[nextLine] -= s.BlkIndent
+			}
+
+			break
+		}
+
+		oldBMarks = append(oldBMarks, s.BMarks[nextLine])
+		oldBSCount = append(oldBSCount, s.BSCount[nextLine])
+		oldTShift = append(oldTShift, s.TShift[nextLine])
+		oldSCount = append(oldSCount, s.SCount[nextLine])
+
+		s.SCount[nextLine] = -1
+	}
+
+	oldIndent := s.BlkIndent
+	s.BlkIndent = 0
+
+	tok := &BlockquoteOpen{
+		Map: [2]int{startLine, 0},
+	}
+	s.PushOpeningToken(tok)
+
+	s.Md.Block.Tokenize(s, startLine, nextLine)
+
+	s.PushClosingToken(&BlockquoteClose{})
+
+	s.LineMax = oldLineMax
+	s.ParentType = oldParentType
+	tok.Map[1] = s.Line
+
+	for i := 0; i < len(oldTShift); i++ {
+		s.BMarks[startLine+i] = oldBMarks[i]
+		s.TShift[startLine+i] = oldTShift[i]
+		s.SCount[startLine+i] = oldSCount[i]
+		s.BSCount[startLine+i] = oldBSCount[i]
+	}
+	s.BlkIndent = oldIndent
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/code.go b/vendor/gitlab.com/golang-commonmark/markdown/code.go
new file mode 100644
index 00000000..c6dc574e
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/code.go
@@ -0,0 +1,37 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleCode(s *StateBlock, startLine, endLine int, _ bool) bool {
+	if s.SCount[startLine]-s.BlkIndent < 4 {
+		return false
+	}
+
+	nextLine := startLine + 1
+	last := nextLine
+
+	for nextLine < endLine {
+		if s.IsLineEmpty(nextLine) {
+			nextLine++
+			continue
+		}
+
+		if s.SCount[nextLine]-s.BlkIndent >= 4 {
+			nextLine++
+			last = nextLine
+			continue
+		}
+
+		break
+	}
+
+	s.Line = last
+	s.PushToken(&CodeBlock{
+		Content: s.Lines(startLine, last, 4+s.BlkIndent, true),
+		Map:     [2]int{startLine, s.Line},
+	})
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/emphasis.go b/vendor/gitlab.com/golang-commonmark/markdown/emphasis.go
new file mode 100644
index 00000000..ea9cda21
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/emphasis.go
@@ -0,0 +1,89 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+type Delimiter struct {
+	Length int
+	Jump   int
+	Token  int
+	Level  int
+	End    int
+	Open   bool
+	Close  bool
+	Marker byte
+}
+
+func ruleEmphasis(s *StateInline, silent bool) bool {
+	src := s.Src
+	start := s.Pos
+	marker := src[start]
+
+	if silent {
+		return false
+	}
+
+	if marker != '_' && marker != '*' {
+		return false
+	}
+
+	canOpen, canClose, length := s.scanDelims(s.Pos, marker == '*')
+	for i := 0; i < length; i++ {
+		s.PushToken(&Text{Content: string(marker)})
+
+		s.Delimiters = append(s.Delimiters, Delimiter{
+			Marker: marker,
+			Length: length,
+			Jump:   i,
+			Token:  len(s.Tokens) - 1,
+			Level:  s.Level,
+			End:    -1,
+			Open:   canOpen,
+			Close:  canClose,
+		})
+	}
+
+	s.Pos += length
+
+	return true
+}
+
+func ruleEmphasisPostprocess(s *StateInline) {
+	delimiters := s.Delimiters
+	max := len(delimiters)
+	for i := max - 1; i >= 0; i-- {
+		startDelim := delimiters[i]
+		if startDelim.Marker != '_' && startDelim.Marker != '*' {
+			continue
+		}
+
+		if startDelim.End == -1 {
+			continue
+		}
+
+		endDelim := delimiters[startDelim.End]
+
+		isStrong := i > 0 &&
+			delimiters[i-1].End == startDelim.End+1 &&
+			delimiters[i-1].Token == startDelim.Token-1 &&
+			delimiters[startDelim.End+1].Token == endDelim.Token+1 &&
+			delimiters[i-1].Marker == startDelim.Marker
+
+		if isStrong {
+			s.Tokens[startDelim.Token] = &StrongOpen{}
+			s.Tokens[endDelim.Token] = &StrongClose{}
+
+			if text, ok := s.Tokens[delimiters[i-1].Token].(*Text); ok {
+				text.Content = ""
+			}
+			if text, ok := s.Tokens[delimiters[startDelim.End+1].Token].(*Text); ok {
+				text.Content = ""
+			}
+			i--
+		} else {
+			s.Tokens[startDelim.Token] = &EmphasisOpen{}
+			s.Tokens[endDelim.Token] = &EmphasisClose{}
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/entity.go b/vendor/gitlab.com/golang-commonmark/markdown/entity.go
new file mode 100644
index 00000000..fd5fba9c
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/entity.go
@@ -0,0 +1,35 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "gitlab.com/golang-commonmark/html"
+
+func ruleEntity(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	src := s.Src
+
+	if src[pos] != '&' {
+		return false
+	}
+
+	max := s.PosMax
+
+	if pos+1 < max {
+		if e, n := html.ParseEntity(src[pos:]); n > 0 {
+			if !silent {
+				s.Pending.WriteString(e)
+			}
+			s.Pos += n
+			return true
+		}
+	}
+
+	if !silent {
+		s.Pending.WriteByte('&')
+	}
+	s.Pos++
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/escape.go b/vendor/gitlab.com/golang-commonmark/markdown/escape.go
new file mode 100644
index 00000000..634b9735
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/escape.go
@@ -0,0 +1,61 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func escaped(b byte) bool {
+	return strings.IndexByte("\\!\"#$%&'()*+,./:;<=>?@[]^_`{|}~-", b) != -1
+}
+
+func ruleEscape(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	src := s.Src
+
+	if src[pos] != '\\' {
+		return false
+	}
+
+	pos++
+	max := s.PosMax
+
+	if pos < max {
+		b := src[pos]
+
+		if b < 0x7f && escaped(b) {
+			if !silent {
+				s.Pending.WriteByte(b)
+			}
+			s.Pos += 2
+			return true
+		}
+
+		if b == '\n' {
+			if !silent {
+				s.PushToken(&Hardbreak{})
+			}
+
+			pos++
+
+			for pos < max {
+				b := src[pos]
+				if !byteIsSpace(b) {
+					break
+				}
+				pos++
+			}
+
+			s.Pos = pos
+			return true
+		}
+	}
+
+	if !silent {
+		s.Pending.WriteByte('\\')
+	}
+	s.Pos++
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/fence.go b/vendor/gitlab.com/golang-commonmark/markdown/fence.go
new file mode 100644
index 00000000..04924dc4
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/fence.go
@@ -0,0 +1,102 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func ruleFence(s *StateBlock, startLine, endLine int, silent bool) bool {
+	haveEndMarker := false
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+	if pos+3 > max {
+		return false
+	}
+
+	src := s.Src
+
+	marker := src[pos]
+	if marker != '~' && marker != '`' {
+		return false
+	}
+
+	mem := pos
+	pos = s.SkipBytes(pos, marker)
+
+	len := pos - mem
+
+	if len < 3 {
+		return false
+	}
+
+	params := strings.TrimSpace(src[pos:max])
+
+	if strings.IndexByte(params, marker) >= 0 {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	nextLine := startLine
+
+	for {
+		nextLine++
+		if nextLine >= endLine {
+			break
+		}
+
+		mem = s.BMarks[nextLine] + s.TShift[nextLine]
+		pos = mem
+		max = s.EMarks[nextLine]
+
+		if pos < max && s.SCount[nextLine] < s.BlkIndent {
+			break
+		}
+
+		if pos >= max || src[pos] != marker {
+			continue
+		}
+
+		if s.SCount[nextLine]-s.BlkIndent >= 4 {
+			continue
+		}
+
+		pos = s.SkipBytes(pos, marker)
+
+		if pos-mem < len {
+			continue
+		}
+
+		pos = s.SkipSpaces(pos)
+
+		if pos < max {
+			continue
+		}
+
+		haveEndMarker = true
+
+		break
+	}
+
+	len = s.SCount[startLine]
+
+	s.Line = nextLine
+	if haveEndMarker {
+		s.Line++
+	}
+
+	s.PushToken(&Fence{
+		Params:  params,
+		Content: s.Lines(startLine+1, nextLine, len, true),
+		Map:     [2]int{startLine, s.Line},
+	})
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/fuzz.go b/vendor/gitlab.com/golang-commonmark/markdown/fuzz.go
new file mode 100644
index 00000000..66ac7697
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/fuzz.go
@@ -0,0 +1,9 @@
+//+build gofuzz
+
+package markdown
+
+func Fuzz(data []byte) int {
+	md := New(HTML(true))
+	md.Parse(data)
+	return 1
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/heading.go b/vendor/gitlab.com/golang-commonmark/markdown/heading.go
new file mode 100644
index 00000000..a09c4f72
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/heading.go
@@ -0,0 +1,59 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func ruleHeading(s *StateBlock, startLine, _ int, silent bool) bool {
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+	src := s.Src
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	if pos >= max || src[pos] != '#' {
+		return false
+	}
+
+	level := 1
+	pos++
+	for pos < max && src[pos] == '#' && level <= 6 {
+		level++
+		pos++
+	}
+
+	if level > 6 || (pos < max && !byteIsSpace(src[pos])) {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	max = s.SkipSpacesBack(max, pos)
+	tmp := s.SkipBytesBack(max, '#', pos)
+	if tmp > pos && byteIsSpace(src[tmp-1]) {
+		max = tmp
+	}
+
+	s.Line = startLine + 1
+
+	s.PushOpeningToken(&HeadingOpen{
+		HLevel: level,
+		Map:    [2]int{startLine, s.Line},
+	})
+
+	if pos < max {
+		s.PushToken(&Inline{
+			Content: strings.TrimSpace(src[pos:max]),
+			Map:     [2]int{startLine, s.Line},
+		})
+	}
+	s.PushClosingToken(&HeadingClose{HLevel: level})
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/helpers.go b/vendor/gitlab.com/golang-commonmark/markdown/helpers.go
new file mode 100644
index 00000000..d158dfcf
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/helpers.go
@@ -0,0 +1,164 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func parseLinkLabel(s *StateInline, start int, disableNested bool) int {
+	src := s.Src
+	labelEnd := -1
+	max := s.PosMax
+	oldPos := s.Pos
+
+	s.Pos = start + 1
+	level := 1
+	found := false
+
+	for s.Pos < max {
+		marker := src[s.Pos]
+
+		if marker == ']' {
+			level--
+			if level == 0 {
+				found = true
+				break
+			}
+		}
+
+		prevPos := s.Pos
+
+		s.Md.Inline.SkipToken(s)
+
+		if marker == '[' {
+			if prevPos == s.Pos-1 {
+				level++
+			} else if disableNested {
+				s.Pos = oldPos
+				return -1
+			}
+		}
+	}
+
+	if found {
+		labelEnd = s.Pos
+	}
+
+	s.Pos = oldPos
+
+	return labelEnd
+}
+
+func parseLinkDestination(s string, pos, max int) (url string, lines, endpos int, ok bool) {
+	start := pos
+	if pos < max && s[pos] == '<' {
+		pos++
+		for pos < max {
+			b := s[pos]
+			if b == '\n' || byteIsSpace(b) {
+				return
+			}
+			if b == '>' {
+				endpos = pos + 1
+				url = unescapeAll(s[start+1 : pos])
+				ok = true
+				return
+			}
+			if b == '\\' && pos+1 < max {
+				pos += 2
+				continue
+			}
+
+			pos++
+		}
+
+		return
+	}
+
+	level := 0
+	for pos < max {
+		b := s[pos]
+
+		if b == ' ' {
+			break
+		}
+
+		if b < 0x20 || b == 0x7f {
+			break
+		}
+
+		if b == '\\' && pos+1 < max {
+			pos += 2
+			continue
+		}
+
+		if b == '(' {
+			level++
+		}
+
+		if b == ')' {
+			if level == 0 {
+				break
+			}
+			level--
+		}
+
+		pos++
+	}
+
+	if start == pos {
+		return
+	}
+	if level != 0 {
+		return
+	}
+
+	url = unescapeAll(s[start:pos])
+	endpos = pos
+	ok = true
+
+	return
+}
+
+func parseLinkTitle(s string, pos, max int) (title string, nlines, endpos int, ok bool) {
+	lines := 0
+	start := pos
+
+	if pos >= max {
+		return
+	}
+
+	marker := s[pos]
+
+	if marker != '"' && marker != '\'' && marker != '(' {
+		return
+	}
+
+	pos++
+
+	if marker == '(' {
+		marker = ')'
+	}
+
+	for pos < max {
+		switch s[pos] {
+		case marker:
+			endpos = pos + 1
+			nlines = lines
+			title = unescapeAll(s[start+1 : pos])
+			ok = true
+			return
+		case '\n':
+			lines++
+		case '\\':
+			if pos+1 < max {
+				pos++
+				if s[pos] == '\n' {
+					lines++
+				}
+			}
+		}
+		pos++
+	}
+
+	return
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/hr.go b/vendor/gitlab.com/golang-commonmark/markdown/hr.go
new file mode 100644
index 00000000..fe25681d
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/hr.go
@@ -0,0 +1,54 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleHR(s *StateBlock, startLine, endLine int, silent bool) bool {
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+
+	if pos >= max {
+		return false
+	}
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	src := s.Src
+	marker := src[pos]
+	pos++
+
+	if marker != '*' && marker != '-' && marker != '_' {
+		return false
+	}
+
+	cnt := 1
+	for pos < max {
+		ch := src[pos]
+		pos++
+		if ch != marker && !byteIsSpace(ch) {
+			return false
+		}
+		if ch == marker {
+			cnt++
+		}
+	}
+
+	if cnt < 3 {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	s.Line = startLine + 1
+
+	s.PushToken(&Hr{
+		Map: [2]int{startLine, s.Line},
+	})
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/html_block.go b/vendor/gitlab.com/golang-commonmark/markdown/html_block.go
new file mode 100644
index 00000000..98a89499
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/html_block.go
@@ -0,0 +1,222 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"regexp"
+	"strings"
+)
+
+var (
+	htmlBlocks = []string{
+		"address",
+		"article",
+		"aside",
+		"base",
+		"basefont",
+		"blockquote",
+		"body",
+		"caption",
+		"center",
+		"col",
+		"colgroup",
+		"dd",
+		"details",
+		"dialog",
+		"dir",
+		"div",
+		"dl",
+		"dt",
+		"fieldset",
+		"figcaption",
+		"figure",
+		"footer",
+		"form",
+		"frame",
+		"frameset",
+		"h1",
+		"h2",
+		"h3",
+		"h4",
+		"h5",
+		"h6",
+		"head",
+		"header",
+		"hr",
+		"html",
+		"iframe",
+		"legend",
+		"li",
+		"link",
+		"main",
+		"menu",
+		"menuitem",
+		"meta",
+		"nav",
+		"noframes",
+		"ol",
+		"optgroup",
+		"option",
+		"p",
+		"param",
+		"section",
+		"source",
+		"summary",
+		"table",
+		"tbody",
+		"td",
+		"tfoot",
+		"th",
+		"thead",
+		"title",
+		"tr",
+		"track",
+		"ul",
+	}
+
+	htmlBlocksSet = make(map[string]bool)
+
+	rStartCond1 = regexp.MustCompile(`(?i)^(pre|script|style)([\n\t >]|$)`)
+	rEndCond1   = regexp.MustCompile(`(?i)</(pre|script|style)>`)
+	rStartCond6 = regexp.MustCompile(`(?i)^/?(` + strings.Join(htmlBlocks, "|") + `)(\s|$|>|/>)`)
+	rStartCond7 = regexp.MustCompile(`(?i)^(/[a-z][a-z0-9-]*|[a-z][a-z0-9-]*(\s+[a-z_:][a-z0-9_.:-]*\s*=\s*("[^"]*"|'[^']*'|[ "'=<>\x60]))*\s*/?)>\s*$`)
+)
+
+func init() {
+	for _, tag := range htmlBlocks {
+		htmlBlocksSet[tag] = true
+	}
+}
+
+func min(a, b int) int {
+	if a < b {
+		return a
+	}
+	return b
+}
+
+func matchTagName(s string) string {
+	if len(s) < 2 {
+		return ""
+	}
+
+	i := 0
+	if s[0] == '/' {
+		i++
+	}
+	start := i
+	max := min(15+i, len(s))
+	for i < max && isLetter(s[i]) {
+		i++
+	}
+	if i >= len(s) {
+		return ""
+	}
+
+	switch s[i] {
+	case ' ', '\n', '/', '>':
+		return strings.ToLower(s[start:i])
+	default:
+		return ""
+	}
+}
+
+func ruleHTMLBlock(s *StateBlock, startLine, endLine int, silent bool) bool {
+	if !s.Md.HTML {
+		return false
+	}
+
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+
+	if pos+1 >= max {
+		return false
+	}
+
+	src := s.Src
+
+	if src[pos] != '<' {
+		return false
+	}
+
+	pos++
+	b := src[pos]
+	if !htmlSecond(b) {
+		return false
+	}
+
+	nextLine := startLine + 1
+
+	var endCond func(string) bool
+
+	if pos+2 < max && isLetter(b) && rStartCond1.MatchString(src[pos:]) {
+		endCond = func(s string) bool {
+			return rEndCond1.MatchString(s)
+		}
+	} else if strings.HasPrefix(src[pos:], "!--") {
+		endCond = func(s string) bool {
+			return strings.Contains(s, "-->")
+		}
+	} else if b == '?' {
+		endCond = func(s string) bool {
+			return strings.Contains(s, "?>")
+		}
+	} else if b == '!' && pos+1 < max && isUppercaseLetter(src[pos+1]) {
+		endCond = func(s string) bool {
+			return strings.Contains(s, ">")
+		}
+	} else if strings.HasPrefix(src[pos:], "![CDATA[") {
+		endCond = func(s string) bool {
+			return strings.Contains(s, "]]>")
+		}
+	} else if pos+2 < max && (isLetter(b) || b == '/' && isLetter(src[pos+1])) {
+		terminator := true
+		if rStartCond6.MatchString(src[pos:max]) {
+		} else if rStartCond7.MatchString(src[pos:max]) {
+			terminator = false
+		} else {
+			return false
+		}
+		if silent {
+			return terminator
+		}
+		endCond = func(s string) bool {
+			return s == ""
+		}
+	} else {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	if !endCond(src[pos:max]) {
+		for nextLine < endLine {
+			if s.SCount[nextLine] < s.BlkIndent {
+				break
+			}
+
+			pos := s.BMarks[nextLine] + s.TShift[nextLine]
+			max := s.EMarks[nextLine]
+			lineText := src[pos:max]
+			if endCond(lineText) {
+				if pos != max {
+					nextLine++
+				}
+				break
+			}
+			nextLine++
+		}
+	}
+
+	s.Line = nextLine
+	s.PushToken(&HTMLBlock{
+		Content: s.Lines(startLine, nextLine, s.BlkIndent, true),
+		Map:     [2]int{startLine, nextLine},
+	})
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/html_inline.go b/vendor/gitlab.com/golang-commonmark/markdown/html_inline.go
new file mode 100644
index 00000000..0de54b20
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/html_inline.go
@@ -0,0 +1,57 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "regexp"
+
+var (
+	attrName     = `[a-zA-Z_:][a-zA-Z0-9:._-]*`
+	unquoted     = "[^\"'=<>`\\x00-\\x20]+"
+	singleQuoted = `'[^']*'`
+	doubleQuoted = `"[^"]*"`
+	attrValue    = `(?:` + unquoted + `|` + singleQuoted + `|` + doubleQuoted + `)`
+	attribute    = `(?:\s+` + attrName + `(?:\s*=\s*` + attrValue + `)?)`
+	openTag      = `<[A-Za-z][A-Za-z0-9-]*` + attribute + `*\s*/?>`
+	closeTag     = `</[A-Za-z][A-Za-z0-9-]*\s*>`
+	comment      = `<!---->|<!--(?:-?[^>-])(?:-?[^-])*-->`
+	processing   = `<[?].*?[?]>`
+	declaration  = `<![A-Z]+\s+[^>]*>`
+	cdata        = `<!\[CDATA\[[\s\S]*?\]\]>`
+	rHTMLTag     = regexp.MustCompile(`^(?:` + openTag + `|` + closeTag + `|` + comment +
+		`|` + processing + `|` + declaration + `|` + cdata + `)`)
+)
+
+func htmlSecond(b byte) bool {
+	return b == '!' || b == '/' || b == '?' || isLetter(b)
+}
+
+func ruleHTMLInline(s *StateInline, silent bool) bool {
+	if !s.Md.HTML {
+		return false
+	}
+
+	pos := s.Pos
+	src := s.Src
+	if pos+2 >= s.PosMax || src[pos] != '<' {
+		return false
+	}
+
+	if !htmlSecond(src[pos+1]) {
+		return false
+	}
+
+	match := rHTMLTag.FindString(src[pos:])
+	if match == "" {
+		return false
+	}
+
+	if !silent {
+		s.PushToken(&HTMLInline{Content: match})
+	}
+
+	s.Pos += len(match)
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/image.go b/vendor/gitlab.com/golang-commonmark/markdown/image.go
new file mode 100644
index 00000000..7f2e67d3
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/image.go
@@ -0,0 +1,131 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleImage(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	max := s.PosMax
+
+	if pos+2 >= max {
+		return false
+	}
+
+	src := s.Src
+	if src[pos] != '!' || src[pos+1] != '[' {
+		return false
+	}
+
+	labelStart := pos + 2
+	labelEnd := parseLinkLabel(s, pos+1, false)
+	if labelEnd < 0 {
+		return false
+	}
+
+	var href, title, label string
+	oldPos := pos
+	pos = labelEnd + 1
+	if pos < max && src[pos] == '(' {
+		pos++
+		for pos < max {
+			b := src[pos]
+			if !byteIsSpace(b) && b != '\n' {
+				break
+			}
+			pos++
+		}
+		if pos >= max {
+			return false
+		}
+
+		start := pos
+		url, _, endpos, ok := parseLinkDestination(src, pos, s.PosMax)
+		if ok {
+			url = normalizeLink(url)
+			if validateLink(url) {
+				href = url
+				pos = endpos
+			}
+		}
+
+		start = pos
+		for pos < max {
+			b := src[pos]
+			if !byteIsSpace(b) && b != '\n' {
+				break
+			}
+			pos++
+		}
+		if pos >= max {
+			return false
+		}
+
+		title, _, endpos, ok = parseLinkTitle(src, pos, s.PosMax)
+		if pos < max && start != pos && ok {
+			pos = endpos
+			for pos < max {
+				b := src[pos]
+				if !byteIsSpace(b) && b != '\n' {
+					break
+				}
+				pos++
+			}
+		}
+
+		if pos >= max || src[pos] != ')' {
+			s.Pos = oldPos
+			return false
+		}
+
+		pos++
+
+	} else {
+		if s.Env.References == nil {
+			return false
+		}
+
+		if pos < max && src[pos] == '[' {
+			start := pos + 1
+			pos = parseLinkLabel(s, pos, false)
+			if pos >= 0 {
+				label = src[start:pos]
+				pos++
+			} else {
+				pos = labelEnd + 1
+			}
+		} else {
+			pos = labelEnd + 1
+		}
+
+		if label == "" {
+			label = src[labelStart:labelEnd]
+		}
+
+		ref, ok := s.Env.References[normalizeReference(label)]
+		if !ok {
+			s.Pos = oldPos
+			return false
+		}
+
+		href = ref["href"]
+		title = ref["title"]
+	}
+
+	if !silent {
+		content := src[labelStart:labelEnd]
+
+		tokens := s.Md.Inline.Parse(content, s.Md, s.Env)
+
+		s.PushToken(&Image{
+			Src:    href,
+			Title:  title,
+			Tokens: tokens,
+		})
+	}
+
+	s.Pos = pos
+	s.PosMax = max
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/inline.go b/vendor/gitlab.com/golang-commonmark/markdown/inline.go
new file mode 100644
index 00000000..0bed18ac
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/inline.go
@@ -0,0 +1,13 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleInline(s *StateCore) {
+	for _, tok := range s.Tokens {
+		if tok, ok := tok.(*Inline); ok {
+			tok.Children = s.Md.Inline.Parse(tok.Content, s.Md, s.Env)
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/lheading.go b/vendor/gitlab.com/golang-commonmark/markdown/lheading.go
new file mode 100644
index 00000000..b404066d
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/lheading.go
@@ -0,0 +1,80 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func ruleLHeading(s *StateBlock, startLine, endLine int, silent bool) bool {
+	nextLine := startLine + 1
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	oldParentType := s.ParentType
+	s.ParentType = ptParagraph
+	src := s.Src
+
+	var pos int
+	var hLevel int
+outer:
+	for ; nextLine < endLine && !s.IsLineEmpty(nextLine); nextLine++ {
+		if s.SCount[nextLine]-s.BlkIndent > 3 {
+			continue
+		}
+
+		if s.SCount[nextLine] >= s.BlkIndent {
+			pos = s.BMarks[nextLine] + s.TShift[nextLine]
+			max := s.EMarks[nextLine]
+
+			if pos < max {
+				marker := src[pos]
+
+				if marker == '-' || marker == '=' {
+					pos = s.SkipBytes(pos, marker)
+					pos = s.SkipSpaces(pos)
+
+					if pos >= max {
+						hLevel = 1
+						if marker == '-' {
+							hLevel++
+						}
+						break
+					}
+				}
+			}
+		}
+
+		if s.SCount[nextLine] < 0 {
+			continue
+		}
+
+		for _, r := range paragraphTerminatedBy {
+			if r(s, nextLine, endLine, true) {
+				break outer
+			}
+		}
+	}
+
+	if hLevel == 0 {
+		return false
+	}
+
+	s.Line = nextLine + 1
+
+	s.PushOpeningToken(&HeadingOpen{
+		HLevel: hLevel,
+		Map:    [2]int{startLine, s.Line},
+	})
+	s.PushToken(&Inline{
+		Content: strings.TrimSpace(s.Lines(startLine, nextLine, s.BlkIndent, false)),
+		Map:     [2]int{startLine, s.Line - 1},
+	})
+	s.PushClosingToken(&HeadingClose{HLevel: hLevel})
+
+	s.ParentType = oldParentType
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/link.go b/vendor/gitlab.com/golang-commonmark/markdown/link.go
new file mode 100644
index 00000000..8ab2c04b
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/link.go
@@ -0,0 +1,132 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleLink(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	oldPos := s.Pos
+	max := s.PosMax
+	start := s.Pos
+	parseReference := true
+	src := s.Src
+
+	if src[pos] != '[' {
+		return false
+	}
+
+	labelStart := pos + 1
+	labelEnd := parseLinkLabel(s, pos, true)
+
+	if labelEnd < 0 {
+		return false
+	}
+
+	pos = labelEnd + 1
+
+	var title, href, label string
+	if pos < max && src[pos] == '(' {
+		parseReference = false
+
+		pos++
+		for pos < max {
+			code := src[pos]
+			if !byteIsSpace(code) && code != '\n' {
+				break
+			}
+			pos++
+		}
+		if pos >= max {
+			return false
+		}
+
+		start = pos
+		url, _, endpos, ok := parseLinkDestination(src, pos, s.PosMax)
+		if ok {
+			url = normalizeLink(url)
+			if validateLink(url) {
+				pos = endpos
+				href = url
+			}
+		}
+
+		start = pos
+		for pos < max {
+			code := src[pos]
+			if !byteIsSpace(code) && code != '\n' {
+				break
+			}
+			pos++
+		}
+
+		title, _, endpos, ok = parseLinkTitle(src, pos, s.PosMax)
+		if pos < max && start != pos && ok {
+			pos = endpos
+			for pos < max {
+				code := src[pos]
+				if !byteIsSpace(code) && code != '\n' {
+					break
+				}
+				pos++
+			}
+		}
+
+		if pos >= max || src[pos] != ')' {
+			parseReference = true
+		}
+
+		pos++
+	}
+
+	if parseReference {
+		if s.Env.References == nil {
+			return false
+		}
+
+		if pos < max && src[pos] == '[' {
+			start := pos + 1
+			pos = parseLinkLabel(s, pos, false)
+			if pos >= 0 {
+				label = src[start:pos]
+				pos++
+			} else {
+				pos = labelEnd + 1
+			}
+		} else {
+			pos = labelEnd + 1
+		}
+
+		if label == "" {
+			label = src[labelStart:labelEnd]
+		}
+
+		ref, ok := s.Env.References[normalizeReference(label)]
+		if !ok {
+			s.Pos = oldPos
+			return false
+		}
+
+		href = ref["href"]
+		title = ref["title"]
+	}
+
+	if !silent {
+		s.Pos = labelStart
+		s.PosMax = labelEnd
+
+		s.PushOpeningToken(&LinkOpen{
+			Href:  href,
+			Title: title,
+		})
+
+		s.Md.Inline.Tokenize(s)
+
+		s.PushClosingToken(&LinkClose{})
+	}
+
+	s.Pos = pos
+	s.PosMax = max
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/linkify.go b/vendor/gitlab.com/golang-commonmark/markdown/linkify.go
new file mode 100644
index 00000000..5935af1b
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/linkify.go
@@ -0,0 +1,131 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"strings"
+
+	"gitlab.com/golang-commonmark/linkify"
+)
+
+func isLinkOpen(s string) bool { return isLetter(s[1]) }
+
+func isLinkClose(s string) bool { return s[1] == '/' }
+
+func ruleLinkify(s *StateCore) {
+	blockTokens := s.Tokens
+
+	if !s.Md.Linkify {
+		return
+	}
+
+	for _, tok := range blockTokens {
+		if tok, ok := tok.(*Inline); ok {
+			tokens := tok.Children
+
+			htmlLinkLevel := 0
+
+			for i := len(tokens) - 1; i >= 0; i-- {
+				currentTok := tokens[i]
+
+				if _, ok := currentTok.(*LinkClose); ok {
+					i--
+					for tokens[i].Level() != currentTok.Level() {
+						if _, ok := tokens[i].(*LinkOpen); ok {
+							break
+						}
+						i--
+					}
+					continue
+				}
+
+				if currentTok, ok := currentTok.(*HTMLInline); ok {
+					if isLinkOpen(currentTok.Content) && htmlLinkLevel > 0 {
+						htmlLinkLevel--
+					}
+					if isLinkClose(currentTok.Content) {
+						htmlLinkLevel++
+					}
+				}
+				if htmlLinkLevel > 0 {
+					continue
+				}
+
+				if currentTok, ok := currentTok.(*Text); ok {
+					text := currentTok.Content
+					links := linkify.Links(text)
+					if len(links) == 0 {
+						continue
+					}
+
+					var nodes []Token
+					level := currentTok.Lvl
+					lastPos := 0
+
+					for _, ln := range links {
+						urlText := text[ln.Start:ln.End]
+						url := urlText
+						if ln.Scheme == "" {
+							url = "http://" + url
+						} else if ln.Scheme == "mailto:" && !strings.HasPrefix(url, "mailto:") {
+							url = "mailto:" + url
+						}
+						url = normalizeLink(url)
+						if !validateLink(url) {
+							continue
+						}
+
+						if ln.Scheme == "" {
+							urlText = strings.TrimPrefix(normalizeLinkText("http://"+urlText), "http://")
+						} else if ln.Scheme == "mailto:" && !strings.HasPrefix(urlText, "mailto:") {
+							urlText = strings.TrimPrefix(normalizeLinkText("mailto:"+urlText), "mailto:")
+						} else {
+							urlText = normalizeLinkText(urlText)
+						}
+
+						pos := ln.Start
+
+						if pos > lastPos {
+							tok := Text{
+								Content: text[lastPos:pos],
+								Lvl:     level,
+							}
+							nodes = append(nodes, &tok)
+						}
+
+						nodes = append(nodes, &LinkOpen{
+							Href: url,
+							Lvl:  level,
+						})
+						nodes = append(nodes, &Text{
+							Content: urlText,
+							Lvl:     level + 1,
+						})
+						nodes = append(nodes, &LinkClose{
+							Lvl: level,
+						})
+
+						lastPos = ln.End
+					}
+
+					if lastPos < len(text) {
+						tok := Text{
+							Content: text[lastPos:],
+							Lvl:     level,
+						}
+						nodes = append(nodes, &tok)
+					}
+
+					children := make([]Token, len(tokens)+len(nodes)-1)
+					copy(children, tokens[:i])
+					copy(children[i:], nodes)
+					copy(children[i+len(nodes):], tokens[i+1:])
+					tok.Children = children
+					tokens = children
+				}
+			}
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/list.go b/vendor/gitlab.com/golang-commonmark/markdown/list.go
new file mode 100644
index 00000000..3b2e0230
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/list.go
@@ -0,0 +1,285 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strconv"
+
+var listTerminatedBy []BlockRule
+
+func skipBulletListMarker(s *StateBlock, startLine int) int {
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+	src := s.Src
+
+	if pos >= max {
+		return -1
+	}
+
+	marker := src[pos]
+	if marker != '*' && marker != '-' && marker != '+' {
+		return -1
+	}
+	pos++
+
+	if pos < max && !byteIsSpace(src[pos]) {
+		return -1
+	}
+
+	return pos
+}
+
+func skipOrderedListMarker(s *StateBlock, startLine int) int {
+	start := s.BMarks[startLine] + s.TShift[startLine]
+	pos := start
+	max := s.EMarks[startLine]
+
+	if pos+1 >= max {
+		return -1
+	}
+
+	src := s.Src
+	ch := src[pos]
+	if ch < '0' || ch > '9' {
+		return -1
+	}
+	pos++
+
+	for {
+		if pos >= max {
+			return -1
+		}
+
+		ch = src[pos]
+		pos++
+
+		if ch >= '0' && ch <= '9' {
+			if pos-start >= 10 {
+				return -1
+			}
+			continue
+		}
+
+		if ch == ')' || ch == '.' {
+			break
+		}
+
+		return -1
+	}
+
+	if pos < max && !byteIsSpace(src[pos]) {
+		return -1
+	}
+
+	return pos
+}
+
+func markParagraphsTight(s *StateBlock, idx int) {
+	level := s.Level + 2
+	tokens := s.Tokens
+
+	for i := idx + 2; i < len(tokens)-2; i++ {
+		if tokens[i].Level() == level {
+			if tok, ok := tokens[i].(*ParagraphOpen); ok {
+				tok.Hidden = true
+				i += 2
+				tokens[i].(*ParagraphClose).Hidden = true
+			}
+		}
+	}
+}
+
+func ruleList(s *StateBlock, startLine, endLine int, silent bool) bool {
+	isTerminatingParagraph := false
+	tight := true
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+	src := s.Src
+
+	if silent && s.ParentType == ptParagraph {
+		if s.TShift[startLine] >= s.BlkIndent {
+			isTerminatingParagraph = true
+		}
+	}
+
+	var start int
+	var markerValue int
+
+	isOrdered := false
+	posAfterMarker := skipOrderedListMarker(s, startLine)
+	if posAfterMarker > 0 {
+		isOrdered = true
+		start = s.BMarks[startLine] + s.TShift[startLine]
+		markerValue, _ = strconv.Atoi(src[start : posAfterMarker-1])
+
+		if isTerminatingParagraph && markerValue != 1 {
+			return false
+		}
+	} else {
+		posAfterMarker = skipBulletListMarker(s, startLine)
+		if posAfterMarker < 0 {
+			return false
+		}
+	}
+
+	if isTerminatingParagraph {
+		if s.SkipSpaces(posAfterMarker) >= s.EMarks[startLine] {
+			return false
+		}
+	}
+
+	markerChar := src[posAfterMarker-1]
+
+	if silent {
+		return true
+	}
+
+	tokenIdx := len(s.Tokens)
+
+	var listMap *[2]int
+	if isOrdered {
+		tok := &OrderedListOpen{
+			Order: markerValue,
+			Map:   [2]int{startLine, 0},
+		}
+		s.PushOpeningToken(tok)
+		listMap = &tok.Map
+	} else {
+		tok := &BulletListOpen{
+			Map: [2]int{startLine, 0},
+		}
+		s.PushOpeningToken(tok)
+		listMap = &tok.Map
+	}
+
+	nextLine := startLine
+	prevEmptyEnd := false
+
+	oldParentType := s.ParentType
+	s.ParentType = ptList
+
+	var pos int
+	var contentStart int
+
+outer:
+	for nextLine < endLine {
+		pos = posAfterMarker
+		max := s.EMarks[nextLine]
+
+		initial := s.SCount[nextLine] + posAfterMarker - (s.BMarks[startLine] + s.TShift[startLine])
+		offset := initial
+
+	loop:
+		for pos < max {
+			switch src[pos] {
+			case '\t':
+				offset += 4 - (offset+s.BSCount[nextLine])%4
+			case ' ':
+				offset++
+			default:
+				break loop
+			}
+			pos++
+		}
+
+		contentStart = pos
+
+		indentAfterMarker := 1
+		if contentStart < max {
+			if iam := offset - initial; iam <= 4 {
+				indentAfterMarker = iam
+			}
+		}
+
+		indent := initial + indentAfterMarker
+
+		tok := &ListItemOpen{
+			Map: [2]int{startLine, 0},
+		}
+		s.PushOpeningToken(tok)
+		itemMap := &tok.Map
+
+		oldIndent := s.BlkIndent
+		oldTight := s.Tight
+		oldTShift := s.TShift[startLine]
+		oldLIndent := s.SCount[startLine]
+		s.BlkIndent = indent
+		s.Tight = true
+		s.TShift[startLine] = contentStart - s.BMarks[startLine]
+		s.SCount[startLine] = offset
+
+		if contentStart >= max && s.IsLineEmpty(startLine+1) {
+			s.Line = min(s.Line+2, endLine)
+		} else {
+			s.Md.Block.Tokenize(s, startLine, endLine)
+		}
+
+		if !s.Tight || prevEmptyEnd {
+			tight = false
+		}
+
+		prevEmptyEnd = s.Line-startLine > 1 && s.IsLineEmpty(s.Line-1)
+
+		s.BlkIndent = oldIndent
+		s.TShift[startLine] = oldTShift
+		s.SCount[startLine] = oldLIndent
+		s.Tight = oldTight
+
+		s.PushClosingToken(&ListItemClose{})
+
+		startLine = s.Line
+		nextLine = startLine
+		(*itemMap)[1] = nextLine
+
+		if nextLine >= endLine {
+			break
+		}
+
+		contentStart = s.BMarks[startLine]
+
+		if s.SCount[nextLine] < s.BlkIndent {
+			break
+		}
+
+		for _, r := range listTerminatedBy {
+			if r(s, nextLine, endLine, true) {
+				break outer
+			}
+		}
+
+		if isOrdered {
+			posAfterMarker = skipOrderedListMarker(s, nextLine)
+			if posAfterMarker < 0 {
+				break
+			}
+		} else {
+			posAfterMarker = skipBulletListMarker(s, nextLine)
+			if posAfterMarker < 0 {
+				break
+			}
+		}
+
+		if markerChar != src[posAfterMarker-1] {
+			break
+		}
+	}
+
+	if isOrdered {
+		s.PushClosingToken(&OrderedListClose{})
+	} else {
+		s.PushClosingToken(&BulletListClose{})
+	}
+
+	(*listMap)[1] = nextLine
+	s.Line = nextLine
+	s.ParentType = oldParentType
+
+	if tight {
+		markParagraphsTight(s, tokenIdx)
+	}
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/markdown.go b/vendor/gitlab.com/golang-commonmark/markdown/markdown.go
new file mode 100644
index 00000000..1e37b3ae
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/markdown.go
@@ -0,0 +1,112 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+// Package markdown provides CommonMark-compliant markdown parser and renderer.
+package markdown
+
+import (
+	"bytes"
+	"io"
+)
+
+type Markdown struct {
+	options
+	Block         ParserBlock
+	Inline        ParserInline
+	renderOptions RenderOptions
+}
+
+type RenderOptions struct {
+	LangPrefix string // CSS language class prefix for fenced blocks
+	XHTML      bool   // render as XHTML instead of HTML
+	Breaks     bool   // convert \n in paragraphs into <br>
+	Nofollow   bool   // add rel="nofollow" to the links
+}
+
+type options struct {
+	HTML        bool      // allow raw HTML in the markup
+	Tables      bool      // GFM tables
+	Linkify     bool      // autoconvert URL-like text to links
+	Typographer bool      // enable some typographic replacements
+	Quotes      [4]string // double/single quotes replacement pairs
+	MaxNesting  int       // maximum nesting level
+}
+
+type Environment struct {
+	References map[string]map[string]string
+}
+
+type CoreRule func(*StateCore)
+
+var coreRules []CoreRule
+
+func New(opts ...option) *Markdown {
+	m := &Markdown{
+		options: options{
+			Tables:      true,
+			Linkify:     true,
+			Typographer: true,
+			Quotes:      [4]string{"“", "”", "‘", "’"},
+			MaxNesting:  20,
+		},
+		renderOptions: RenderOptions{LangPrefix: "language-"},
+	}
+	for _, opt := range opts {
+		opt(m)
+	}
+	return m
+}
+
+func (m *Markdown) Parse(src []byte) []Token {
+	if len(src) == 0 {
+		return nil
+	}
+
+	s := &StateCore{
+		Md:  m,
+		Env: &Environment{},
+	}
+	s.Tokens = m.Block.Parse(src, m, s.Env)
+
+	for _, r := range coreRules {
+		r(s)
+	}
+	return s.Tokens
+}
+
+func (m *Markdown) Render(w io.Writer, src []byte) error {
+	if len(src) == 0 {
+		return nil
+	}
+
+	return NewRenderer(w).Render(m.Parse(src), m.renderOptions)
+}
+
+func (m *Markdown) RenderTokens(w io.Writer, tokens []Token) error {
+	if len(tokens) == 0 {
+		return nil
+	}
+
+	return NewRenderer(w).Render(tokens, m.renderOptions)
+}
+
+func (m *Markdown) RenderToString(src []byte) string {
+	if len(src) == 0 {
+		return ""
+	}
+
+	var buf bytes.Buffer
+	NewRenderer(&buf).Render(m.Parse(src), m.renderOptions)
+	return buf.String()
+}
+
+func (m *Markdown) RenderTokensToString(tokens []Token) string {
+	if len(tokens) == 0 {
+		return ""
+	}
+
+	var buf bytes.Buffer
+	NewRenderer(&buf).Render(tokens, m.renderOptions)
+	return buf.String()
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/newline.go b/vendor/gitlab.com/golang-commonmark/markdown/newline.go
new file mode 100644
index 00000000..c69fa3d9
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/newline.go
@@ -0,0 +1,46 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleNewline(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	src := s.Src
+
+	if src[pos] != '\n' {
+		return false
+	}
+
+	pending := s.Pending.Bytes()
+	pmax := len(pending) - 1
+	max := s.PosMax
+
+	if !silent {
+		if pmax >= 0 && pending[pmax] == ' ' {
+			if pmax >= 1 && pending[pmax-1] == ' ' {
+				pmax -= 2
+				for pmax >= 0 && pending[pmax] == ' ' {
+					pmax--
+				}
+				s.Pending.Truncate(pmax + 1)
+				s.PushToken(&Hardbreak{})
+			} else {
+				s.Pending.Truncate(pmax)
+				s.PushToken(&Softbreak{})
+			}
+		} else {
+			s.PushToken(&Softbreak{})
+		}
+	}
+
+	pos++
+
+	for pos < max && byteIsSpace(src[pos]) {
+		pos++
+	}
+
+	s.Pos = pos
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/options.go b/vendor/gitlab.com/golang-commonmark/markdown/options.go
new file mode 100644
index 00000000..7e5b2cf3
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/options.go
@@ -0,0 +1,77 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+type option func(m *Markdown)
+
+func HTML(b bool) option {
+	return func(m *Markdown) {
+		m.HTML = b
+	}
+}
+
+func Linkify(b bool) option {
+	return func(m *Markdown) {
+		m.Linkify = b
+	}
+}
+
+func Typographer(b bool) option {
+	return func(m *Markdown) {
+		m.Typographer = b
+	}
+}
+
+func Quotes(stringOrArray interface{}) option {
+	if s, ok := stringOrArray.(string); ok {
+		return func(m *Markdown) {
+			for i, r := range []rune(s) {
+				m.Quotes[i] = string(r)
+			}
+		}
+	}
+	a := stringOrArray.([]string)
+	return func(m *Markdown) {
+		for i, s := range a {
+			m.Quotes[i] = s
+		}
+	}
+}
+
+func MaxNesting(n int) option {
+	return func(m *Markdown) {
+		m.MaxNesting = n
+	}
+}
+
+func XHTMLOutput(b bool) option {
+	return func(m *Markdown) {
+		m.renderOptions.XHTML = b
+	}
+}
+
+func Breaks(b bool) option {
+	return func(m *Markdown) {
+		m.renderOptions.Breaks = b
+	}
+}
+
+func LangPrefix(p string) option {
+	return func(m *Markdown) {
+		m.renderOptions.LangPrefix = p
+	}
+}
+
+func Nofollow(b bool) option {
+	return func(m *Markdown) {
+		m.renderOptions.Nofollow = b
+	}
+}
+
+func Tables(b bool) option {
+	return func(m *Markdown) {
+		m.Tables = b
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/paragraph.go b/vendor/gitlab.com/golang-commonmark/markdown/paragraph.go
new file mode 100644
index 00000000..a3003ef6
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/paragraph.go
@@ -0,0 +1,51 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+var paragraphTerminatedBy []BlockRule
+
+func ruleParagraph(s *StateBlock, startLine, _ int, _ bool) bool {
+	nextLine := startLine + 1
+	endLine := s.LineMax
+
+	oldParentType := s.ParentType
+	s.ParentType = ptParagraph
+
+outer:
+	for ; nextLine < endLine && !s.IsLineEmpty(nextLine); nextLine++ {
+		if s.SCount[nextLine]-s.BlkIndent > 3 {
+			continue
+		}
+
+		if s.SCount[nextLine] < 0 {
+			continue
+		}
+
+		for _, r := range paragraphTerminatedBy {
+			if r(s, nextLine, endLine, true) {
+				break outer
+			}
+		}
+	}
+
+	content := strings.TrimSpace(s.Lines(startLine, nextLine, s.BlkIndent, false))
+
+	s.Line = nextLine
+
+	s.PushOpeningToken(&ParagraphOpen{
+		Map: [2]int{startLine, s.Line},
+	})
+	s.PushToken(&Inline{
+		Content: content,
+		Map:     [2]int{startLine, s.Line},
+	})
+	s.PushClosingToken(&ParagraphClose{})
+
+	s.ParentType = oldParentType
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/parser_block.go b/vendor/gitlab.com/golang-commonmark/markdown/parser_block.go
new file mode 100644
index 00000000..8a6ee1ad
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/parser_block.go
@@ -0,0 +1,159 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"bytes"
+	"unicode/utf8"
+)
+
+type ParserBlock struct{}
+
+type BlockRule func(*StateBlock, int, int, bool) bool
+
+var blockRules []BlockRule
+
+var nl = []byte{'\n'}
+
+func normalizeNewlines(src []byte) ([]byte, int) {
+	if bytes.IndexByte(src, '\r') == -1 {
+		return src, bytes.Count(src, nl)
+	}
+	n := 0
+	buf := make([]byte, 0, len(src))
+	for i := 0; i < len(src); i++ {
+		switch ch := src[i]; ch {
+		case '\n':
+			n++
+			buf = append(buf, '\n')
+		case '\r':
+			buf = append(buf, '\n')
+			n++
+			if i < len(src)-1 && src[i+1] == '\n' {
+				i++
+			}
+		default:
+			buf = append(buf, ch)
+		}
+	}
+	return buf, n
+}
+
+func (b ParserBlock) Parse(src []byte, md *Markdown, env *Environment) []Token {
+	src, n := normalizeNewlines(src)
+	if len(src) == 0 || src[len(src)-1] != '\n' {
+		n++
+	}
+	n++
+
+	indentFound := false
+	start := 0
+	indent := 0
+	offset := 0
+
+	mem := make([]int, 0, n*5)
+	bMarks := mem[0:0:n]
+	eMarks := mem[n : n : n*2]
+	tShift := mem[n*2 : n*2 : n*3]
+	sCount := mem[n*3 : n*3 : n*4]
+	bsCount := mem[n*4 : n*4 : n*5]
+
+	_, lastRuneLen := utf8.DecodeLastRune(src)
+	lastRunePos := len(src) - lastRuneLen
+	for pos, r := range string(src) {
+		if !indentFound {
+			if runeIsSpace(r) {
+				indent++
+				if r == '\t' {
+					offset += 4 - offset%4
+				} else {
+					offset++
+				}
+				continue
+			}
+			indentFound = true
+		}
+
+		if r == '\n' || pos == lastRunePos {
+			if r != '\n' {
+				pos = len(src)
+			}
+			bMarks = append(bMarks, start)
+			eMarks = append(eMarks, pos)
+			tShift = append(tShift, indent)
+			sCount = append(sCount, offset)
+			bsCount = append(bsCount, 0)
+
+			indentFound = false
+			indent = 0
+			offset = 0
+			start = pos + 1
+		}
+	}
+
+	bMarks = append(bMarks, len(src))
+	eMarks = append(eMarks, len(src))
+	tShift = append(tShift, 0)
+	sCount = append(sCount, 0)
+	bsCount = append(bsCount, 0)
+
+	var s StateBlock
+	s.BMarks = bMarks
+	s.EMarks = eMarks
+	s.TShift = tShift
+	s.SCount = sCount
+	s.BSCount = bsCount
+	s.LineMax = n - 1
+	s.Src = string(src)
+	s.Md = md
+	s.Env = env
+
+	b.Tokenize(&s, s.Line, s.LineMax)
+
+	return s.Tokens
+}
+
+func (ParserBlock) Tokenize(s *StateBlock, startLine, endLine int) {
+	line := startLine
+	hasEmptyLines := false
+	maxNesting := s.Md.MaxNesting
+
+	for line < endLine {
+		line = s.SkipEmptyLines(line)
+		s.Line = line
+		if line >= endLine {
+			break
+		}
+
+		if s.SCount[line] < s.BlkIndent {
+			break
+		}
+
+		if s.Level >= maxNesting {
+			s.Line = endLine
+			break
+		}
+
+		for _, r := range blockRules {
+			if r(s, line, endLine, false) {
+				break
+			}
+		}
+
+		s.Tight = !hasEmptyLines
+
+		if s.IsLineEmpty(s.Line - 1) {
+			hasEmptyLines = true
+		}
+
+		line = s.Line
+
+		if line < endLine && s.IsLineEmpty(line) {
+			hasEmptyLines = true
+			line++
+			s.Line = line
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/parser_inline.go b/vendor/gitlab.com/golang-commonmark/markdown/parser_inline.go
new file mode 100644
index 00000000..5e976ebd
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/parser_inline.go
@@ -0,0 +1,106 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "unicode/utf8"
+
+type ParserInline struct {
+}
+
+type (
+	InlineRule      func(*StateInline, bool) bool
+	PostprocessRule func(*StateInline)
+)
+
+var (
+	inlineRules      []InlineRule
+	postprocessRules []PostprocessRule
+)
+
+func (i ParserInline) Parse(src string, md *Markdown, env *Environment) []Token {
+	if src == "" {
+		return nil
+	}
+
+	var s StateInline
+	s.Src = src
+	s.Md = md
+	s.Env = env
+	s.PosMax = len(src)
+	s.Tokens = s.bootstrap[:0]
+
+	i.Tokenize(&s)
+
+	for _, r := range postprocessRules {
+		r(&s)
+	}
+
+	return s.Tokens
+}
+
+func (ParserInline) Tokenize(s *StateInline) {
+	end := s.PosMax
+	src := s.Src
+	maxNesting := s.Md.MaxNesting
+	ok := false
+
+	for s.Pos < end {
+		if s.Level < maxNesting {
+			for _, rule := range inlineRules {
+				ok = rule(s, false)
+				if ok {
+					break
+				}
+			}
+		}
+
+		if ok {
+			if s.Pos >= end {
+				break
+			}
+			continue
+		}
+
+		r, size := utf8.DecodeRuneInString(src[s.Pos:])
+		s.Pending.WriteRune(r)
+		s.Pos += size
+	}
+
+	if s.Pending.Len() > 0 {
+		s.PushPending()
+	}
+}
+
+func (ParserInline) SkipToken(s *StateInline) {
+	pos := s.Pos
+	if s.Cache != nil {
+		if newPos, ok := s.Cache[pos]; ok {
+			s.Pos = newPos
+			return
+		}
+	} else {
+		s.Cache = make(map[int]int)
+	}
+
+	ok := false
+	if s.Level < s.Md.MaxNesting {
+		for _, r := range inlineRules {
+			s.Level++
+			ok = r(s, true)
+			s.Level--
+			if ok {
+				break
+			}
+		}
+	} else {
+		s.Pos = s.PosMax
+	}
+
+	if !ok {
+		_, size := utf8.DecodeRuneInString(s.Src[s.Pos:])
+		s.Pos += size
+	}
+	s.Cache[pos] = s.Pos
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/plugins.go b/vendor/gitlab.com/golang-commonmark/markdown/plugins.go
new file mode 100644
index 00000000..5ba818cd
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/plugins.go
@@ -0,0 +1,158 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "sort"
+
+type registeredCoreRule struct {
+	id   int
+	rule CoreRule
+}
+
+var registeredCoreRules []registeredCoreRule
+
+type registeredBlockRule struct {
+	id         int
+	rule       BlockRule
+	terminates []int
+}
+
+var registeredBlockRules []registeredBlockRule
+
+type registeredInlineRule struct {
+	id   int
+	rule InlineRule
+}
+
+var registeredInlineRules []registeredInlineRule
+
+type registeredPostprocessRule struct {
+	id   int
+	rule PostprocessRule
+}
+
+var registeredPostprocessRules []registeredPostprocessRule
+
+func indexInt(a []int, n int) int {
+	for i, m := range a {
+		if m == n {
+			return i
+		}
+	}
+	return -1
+}
+
+func RegisterCoreRule(id int, rule CoreRule) {
+	registeredCoreRules = append(registeredCoreRules, registeredCoreRule{
+		id:   id,
+		rule: rule,
+	})
+	sort.Slice(registeredCoreRules, func(i, j int) bool {
+		return registeredCoreRules[i].id < registeredCoreRules[j].id
+	})
+
+	coreRules = coreRules[:0]
+	for _, r := range registeredCoreRules {
+		coreRules = append(coreRules, r.rule)
+	}
+}
+
+func RegisterBlockRule(id int, rule BlockRule, terminates []int) {
+	registeredBlockRules = append(registeredBlockRules, registeredBlockRule{
+		id:         id,
+		rule:       rule,
+		terminates: terminates,
+	})
+	sort.Slice(registeredBlockRules, func(i, j int) bool {
+		return registeredBlockRules[i].id < registeredBlockRules[j].id
+	})
+
+	blockRules = blockRules[:0]
+	blockquoteTerminatedBy = blockquoteTerminatedBy[:0]
+	listTerminatedBy = listTerminatedBy[:0]
+	referenceTerminatedBy = referenceTerminatedBy[:0]
+	paragraphTerminatedBy = paragraphTerminatedBy[:0]
+	for _, r := range registeredBlockRules {
+		blockRules = append(blockRules, r.rule)
+		if indexInt(r.terminates, 400) != -1 {
+			blockquoteTerminatedBy = append(blockquoteTerminatedBy, r.rule)
+		}
+		if indexInt(r.terminates, 600) != -1 {
+			listTerminatedBy = append(listTerminatedBy, r.rule)
+		}
+		if indexInt(r.terminates, 700) != -1 {
+			referenceTerminatedBy = append(referenceTerminatedBy, r.rule)
+		}
+		if indexInt(r.terminates, 1100) != -1 {
+			paragraphTerminatedBy = append(paragraphTerminatedBy, r.rule)
+		}
+	}
+}
+
+func RegisterInlineRule(id int, rule InlineRule) {
+	registeredInlineRules = append(registeredInlineRules, registeredInlineRule{
+		id:   id,
+		rule: rule,
+	})
+	sort.Slice(registeredInlineRules, func(i, j int) bool {
+		return registeredInlineRules[i].id < registeredInlineRules[j].id
+	})
+
+	inlineRules = inlineRules[:0]
+	for _, r := range registeredInlineRules {
+		inlineRules = append(inlineRules, r.rule)
+	}
+}
+
+func RegisterPostprocessRule(id int, rule PostprocessRule) {
+	registeredPostprocessRules = append(registeredPostprocessRules, registeredPostprocessRule{
+		id:   id,
+		rule: rule,
+	})
+	sort.Slice(registeredPostprocessRules, func(i, j int) bool {
+		return registeredPostprocessRules[i].id < registeredPostprocessRules[j].id
+	})
+
+	postprocessRules = postprocessRules[:0]
+	for _, r := range registeredPostprocessRules {
+		postprocessRules = append(postprocessRules, r.rule)
+	}
+}
+
+func init() {
+	RegisterCoreRule(100, ruleInline)
+	RegisterCoreRule(200, ruleLinkify)
+	RegisterCoreRule(300, ruleReplacements)
+	RegisterCoreRule(400, ruleSmartQuotes)
+
+	RegisterBlockRule(100, ruleTable, []int{1100, 700})
+	RegisterBlockRule(200, ruleCode, nil)
+	RegisterBlockRule(300, ruleFence, []int{1100, 700, 400, 600})
+	RegisterBlockRule(400, ruleBlockQuote, []int{1100, 700, 400, 600})
+	RegisterBlockRule(500, ruleHR, []int{1100, 700, 400, 600})
+	RegisterBlockRule(600, ruleList, []int{1100, 700, 400})
+	RegisterBlockRule(700, ruleReference, nil)
+	RegisterBlockRule(800, ruleHeading, []int{1100, 700, 400})
+	RegisterBlockRule(900, ruleLHeading, nil)
+	RegisterBlockRule(1000, ruleHTMLBlock, []int{1100, 700, 400})
+	RegisterBlockRule(1100, ruleParagraph, nil)
+
+	RegisterInlineRule(100, ruleText)
+	RegisterInlineRule(200, ruleNewline)
+	RegisterInlineRule(300, ruleEscape)
+	RegisterInlineRule(400, ruleBackticks)
+	RegisterInlineRule(500, ruleStrikeThrough)
+	RegisterInlineRule(600, ruleEmphasis)
+	RegisterInlineRule(700, ruleLink)
+	RegisterInlineRule(800, ruleImage)
+	RegisterInlineRule(900, ruleAutolink)
+	RegisterInlineRule(1000, ruleHTMLInline)
+	RegisterInlineRule(1100, ruleEntity)
+
+	RegisterPostprocessRule(100, ruleBalancePairs)
+	RegisterPostprocessRule(200, ruleStrikethroughPostprocess)
+	RegisterPostprocessRule(300, ruleEmphasisPostprocess)
+	RegisterPostprocessRule(400, ruleTextCollapse)
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/reference.go b/vendor/gitlab.com/golang-commonmark/markdown/reference.go
new file mode 100644
index 00000000..6e97809e
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/reference.go
@@ -0,0 +1,173 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+var referenceTerminatedBy []BlockRule
+
+func ruleReference(s *StateBlock, startLine, _ int, silent bool) bool {
+	lines := 0
+	pos := s.BMarks[startLine] + s.TShift[startLine]
+	max := s.EMarks[startLine]
+	nextLine := startLine + 1
+
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	src := s.Src
+
+	if src[pos] != '[' {
+		return false
+	}
+
+	pos++
+	for pos < max {
+		if src[pos] == ']' && src[pos-1] != '\\' {
+			if pos+1 == max {
+				return false
+			}
+			if src[pos+1] != ':' {
+				return false
+			}
+			break
+		}
+		pos++
+	}
+
+	endLine := s.LineMax
+
+	oldParentType := s.ParentType
+	s.ParentType = ptReference
+outer:
+	for ; nextLine < endLine && !s.IsLineEmpty(nextLine); nextLine++ {
+		if s.SCount[nextLine]-s.BlkIndent > 3 {
+			continue
+		}
+
+		if s.SCount[nextLine] < 0 {
+			continue
+		}
+
+		for _, r := range referenceTerminatedBy {
+			if r(s, nextLine, endLine, true) {
+				break outer
+			}
+		}
+	}
+
+	str := strings.TrimSpace(s.Lines(startLine, nextLine, s.BlkIndent, false))
+	max = len(str)
+
+	var labelEnd int
+	for pos = 1; pos < max; pos++ {
+		b := str[pos]
+		if b == '[' {
+			return false
+		} else if b == ']' {
+			labelEnd = pos
+			break
+		} else if b == '\n' {
+			lines++
+		} else if b == '\\' {
+			pos++
+			if pos < max && str[pos] == '\n' {
+				lines++
+			}
+		}
+	}
+
+	if labelEnd <= 0 || labelEnd+1 >= max || str[labelEnd+1] != ':' {
+		return false
+	}
+
+	for pos = labelEnd + 2; pos < max; pos++ {
+		b := str[pos]
+		if b == '\n' {
+			lines++
+		} else if !byteIsSpace(b) {
+			break
+		}
+	}
+
+	href, nlines, endpos, ok := parseLinkDestination(str, pos, max)
+	if !ok {
+		return false
+	}
+	href = normalizeLink(href)
+	if !validateLink(href) {
+		return false
+	}
+
+	pos = endpos
+	lines += nlines
+
+	destEndPos := pos
+	destEndLineNo := lines
+
+	start := pos
+	for ; pos < max; pos++ {
+		b := str[pos]
+		if b == '\n' {
+			lines++
+		} else if !byteIsSpace(b) {
+			break
+		}
+	}
+
+	title, nlines, endpos, ok := parseLinkTitle(str, pos, max)
+	if pos < max && start != pos && ok {
+		pos = endpos
+		lines += nlines
+	} else {
+		pos = destEndPos
+		lines = destEndLineNo
+	}
+
+	for pos < max && byteIsSpace(str[pos]) {
+		pos++
+	}
+
+	if pos < max && str[pos] != '\n' {
+		if title != "" {
+			title = ""
+			pos = destEndPos
+			lines = destEndLineNo
+			for pos < max && byteIsSpace(src[pos]) {
+				pos++
+			}
+		}
+	}
+
+	if pos < max && str[pos] != '\n' {
+		return false
+	}
+
+	label := normalizeReference(str[1:labelEnd])
+	if label == "" {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	if s.Env.References == nil {
+		s.Env.References = make(map[string]map[string]string)
+	}
+	if _, ok := s.Env.References[label]; !ok {
+		s.Env.References[label] = map[string]string{
+			"title": title,
+			"href":  href,
+		}
+	}
+
+	s.ParentType = oldParentType
+
+	s.Line = startLine + lines + 1
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/render.go b/vendor/gitlab.com/golang-commonmark/markdown/render.go
new file mode 100644
index 00000000..3955f051
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/render.go
@@ -0,0 +1,333 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"io"
+	"regexp"
+	"strconv"
+	"strings"
+
+	"gitlab.com/golang-commonmark/html"
+)
+
+type Renderer struct {
+	w *monadicWriter
+}
+
+func NewRenderer(w io.Writer) *Renderer {
+	return &Renderer{newMonadicWriter(w)}
+}
+
+func (r *Renderer) Render(tokens []Token, options RenderOptions) error {
+	for i, tok := range tokens {
+		if tok, ok := tok.(*Inline); ok {
+			r.renderInline(tok.Children, options)
+		} else {
+			r.renderToken(tokens, i, options)
+		}
+	}
+
+	r.w.Flush()
+
+	return r.w.err
+}
+
+func (r *Renderer) renderInline(tokens []Token, o RenderOptions) {
+	for i := range tokens {
+		r.renderToken(tokens, i, o)
+	}
+}
+
+func (r *Renderer) renderInlineAsText(tokens []Token) {
+	for _, tok := range tokens {
+		if text, ok := tok.(*Text); ok {
+			html.WriteEscapedString(r.w, text.Content)
+		} else if img, ok := tok.(*Image); ok {
+			r.renderInlineAsText(img.Tokens)
+		}
+	}
+}
+
+var rNotSpace = regexp.MustCompile(`^\S+`)
+
+func (r *Renderer) renderToken(tokens []Token, idx int, options RenderOptions) {
+	tok := tokens[idx]
+
+	if idx > 0 && tok.Block() && !tok.Closing() {
+		switch t := tokens[idx-1].(type) {
+		case *ParagraphOpen:
+			if t.Hidden {
+				r.w.WriteByte('\n')
+			}
+		case *ParagraphClose:
+			if t.Hidden {
+				r.w.WriteByte('\n')
+			}
+		}
+	}
+
+	switch tok := tok.(type) {
+	case *BlockquoteClose:
+		r.w.WriteString("</blockquote>")
+
+	case *BlockquoteOpen:
+		r.w.WriteString("<blockquote>")
+
+	case *BulletListClose:
+		r.w.WriteString("</ul>")
+
+	case *BulletListOpen:
+		r.w.WriteString("<ul>")
+
+	case *CodeBlock:
+		r.w.WriteString("<pre><code>")
+		html.WriteEscapedString(r.w, tok.Content)
+		r.w.WriteString("</code></pre>")
+
+	case *CodeInline:
+		r.w.WriteString("<code>")
+		html.WriteEscapedString(r.w, tok.Content)
+		r.w.WriteString("</code>")
+
+	case *EmphasisClose:
+		r.w.WriteString("</em>")
+
+	case *EmphasisOpen:
+		r.w.WriteString("<em>")
+
+	case *Fence:
+		r.w.WriteString("<pre><code")
+		if tok.Params != "" {
+			langName := strings.SplitN(unescapeAll(tok.Params), " ", 2)[0]
+			langName = rNotSpace.FindString(langName)
+			if langName != "" {
+				r.w.WriteString(` class="`)
+				r.w.WriteString(options.LangPrefix)
+				html.WriteEscapedString(r.w, langName)
+				r.w.WriteByte('"')
+			}
+		}
+		r.w.WriteByte('>')
+		html.WriteEscapedString(r.w, tok.Content)
+		r.w.WriteString("</code></pre>")
+
+	case *Hardbreak:
+		if options.XHTML {
+			r.w.WriteString("<br />\n")
+		} else {
+			r.w.WriteString("<br>\n")
+		}
+
+	case *HeadingClose:
+		r.w.WriteString("</h")
+		r.w.WriteByte("0123456789"[tok.HLevel])
+		r.w.WriteString(">")
+
+	case *HeadingOpen:
+		r.w.WriteString("<h")
+		r.w.WriteByte("0123456789"[tok.HLevel])
+		r.w.WriteByte('>')
+
+	case *Hr:
+		if options.XHTML {
+			r.w.WriteString("<hr />")
+		} else {
+			r.w.WriteString("<hr>")
+		}
+
+	case *HTMLBlock:
+		r.w.WriteString(tok.Content)
+		return // no newline
+
+	case *HTMLInline:
+		r.w.WriteString(tok.Content)
+
+	case *Image:
+		r.w.WriteString(`<img src="`)
+		html.WriteEscapedString(r.w, tok.Src)
+		r.w.WriteString(`" alt="`)
+		r.renderInlineAsText(tok.Tokens)
+		r.w.WriteByte('"')
+
+		if tok.Title != "" {
+			r.w.WriteString(` title="`)
+			html.WriteEscapedString(r.w, tok.Title)
+			r.w.WriteByte('"')
+		}
+		if options.XHTML {
+			r.w.WriteString(" />")
+		} else {
+			r.w.WriteByte('>')
+		}
+
+	case *LinkClose:
+		r.w.WriteString("</a>")
+
+	case *LinkOpen:
+		r.w.WriteString(`<a href="`)
+		html.WriteEscapedString(r.w, tok.Href)
+		r.w.WriteByte('"')
+		if tok.Title != "" {
+			r.w.WriteString(` title="`)
+			html.WriteEscapedString(r.w, (tok.Title))
+			r.w.WriteByte('"')
+		}
+		if tok.Target != "" {
+			r.w.WriteString(` target="`)
+			html.WriteEscapedString(r.w, tok.Target)
+			r.w.WriteByte('"')
+		}
+		if options.Nofollow {
+			r.w.WriteString(` rel="nofollow"`)
+		}
+		r.w.WriteByte('>')
+
+	case *ListItemClose:
+		r.w.WriteString("</li>")
+
+	case *ListItemOpen:
+		r.w.WriteString("<li>")
+
+	case *OrderedListClose:
+		r.w.WriteString("</ol>")
+
+	case *OrderedListOpen:
+		if tok.Order != 1 {
+			r.w.WriteString(`<ol start="`)
+			r.w.WriteString(strconv.Itoa(tok.Order))
+			r.w.WriteString(`">`)
+		} else {
+			r.w.WriteString("<ol>")
+		}
+
+	case *ParagraphClose:
+		if tok.Hidden {
+			return
+		}
+		if !tok.Tight {
+			r.w.WriteString("</p>")
+		} else if tokens[idx+1].Closing() {
+			return // no newline
+		}
+
+	case *ParagraphOpen:
+		if tok.Hidden {
+			return
+		}
+		if !tok.Tight {
+			r.w.WriteString("<p>")
+		}
+
+	case *Softbreak:
+		if options.Breaks {
+			if options.XHTML {
+				r.w.WriteString("<br />\n")
+			} else {
+				r.w.WriteString("<br>\n")
+			}
+		} else {
+			r.w.WriteByte('\n')
+		}
+		return
+
+	case *StrongClose:
+		r.w.WriteString("</strong>")
+
+	case *StrongOpen:
+		r.w.WriteString("<strong>")
+
+	case *StrikethroughClose:
+		r.w.WriteString("</s>")
+
+	case *StrikethroughOpen:
+		r.w.WriteString("<s>")
+
+	case *TableClose:
+		r.w.WriteString("</table>")
+
+	case *TableOpen:
+		r.w.WriteString("<table>")
+
+	case *TbodyClose:
+		r.w.WriteString("</tbody>")
+
+	case *TbodyOpen:
+		r.w.WriteString("<tbody>")
+
+	case *TdClose:
+		r.w.WriteString("</td>")
+
+	case *TdOpen:
+		if tok.Align != AlignNone {
+			r.w.WriteString(`<td style="text-align:`)
+			r.w.WriteString(tok.Align.String())
+			r.w.WriteString(`">`)
+		} else {
+			r.w.WriteString("<td>")
+		}
+
+	case *Text:
+		html.WriteEscapedString(r.w, tok.Content)
+
+	case *TheadClose:
+		r.w.WriteString("</thead>")
+
+	case *TheadOpen:
+		r.w.WriteString("<thead>")
+
+	case *ThClose:
+		r.w.WriteString("</th>")
+
+	case *ThOpen:
+		if align := tok.Align; align != AlignNone {
+			r.w.WriteString(`<th style="text-align:`)
+			r.w.WriteString(align.String())
+			r.w.WriteString(`">`)
+		} else {
+			r.w.WriteString("<th>")
+		}
+
+	case *TrClose:
+		r.w.WriteString("</tr>")
+
+	case *TrOpen:
+		r.w.WriteString("<tr>")
+
+	default:
+		panic("unknown token type")
+	}
+
+	needLf := false
+	if tok.Block() {
+		needLf = true
+
+		if tok.Opening() {
+			nextTok := tokens[idx+1]
+			blockquote := false
+			switch nextTok := nextTok.(type) {
+			case *Inline:
+				needLf = false
+			case *ParagraphOpen:
+				if nextTok.Tight || nextTok.Hidden {
+					needLf = false
+				}
+			case *ParagraphClose:
+				if nextTok.Tight || nextTok.Hidden {
+					needLf = false
+				}
+			case *BlockquoteClose:
+				blockquote = true
+			}
+			if !blockquote && needLf && nextTok.Closing() && nextTok.Tag() == tok.Tag() {
+				needLf = false
+			}
+		}
+	}
+
+	if needLf {
+		r.w.WriteByte('\n')
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/replacements.go b/vendor/gitlab.com/golang-commonmark/markdown/replacements.go
new file mode 100644
index 00000000..8d67af6e
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/replacements.go
@@ -0,0 +1,229 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+func exclquest(b byte) bool {
+	return b == '!' || b == '?'
+}
+
+func byteToLower(b byte) byte {
+	if b >= 'A' && b <= 'Z' {
+		return b - 'A' + 'a'
+	}
+	return b
+}
+
+var replChar = [256]bool{
+	'(': true,
+	'!': true,
+	'+': true,
+	',': true,
+	'-': true,
+	'.': true,
+	'?': true,
+}
+
+func performReplacements(s string) string {
+	var ss []string
+
+	start := 0
+	for i := 0; i < len(s); i++ {
+		b := s[i]
+
+		if replChar[b] {
+
+		outer:
+			switch b {
+			case '(':
+				if i+2 >= len(s) {
+					break
+				}
+
+				b2 := s[i+1]
+
+				b2 = byteToLower(b2)
+				switch b2 {
+				case 'c', 'r', 'p':
+					if s[i+2] != ')' {
+						break outer
+					}
+					switch b2 {
+					case 'c':
+						if start < i {
+							ss = append(ss, s[start:i])
+						}
+						ss = append(ss, "©")
+					case 'r':
+						if start < i {
+							ss = append(ss, s[start:i])
+						}
+						ss = append(ss, "®")
+					case 'p':
+						if start < i {
+							ss = append(ss, s[start:i])
+						}
+						ss = append(ss, "§")
+					}
+					i += 2
+					start = i + 1
+					continue
+
+				case 't':
+					if i+3 >= len(s) {
+						break outer
+					}
+					if s[i+3] != ')' || byteToLower(s[i+2]) != 'm' {
+						break outer
+					}
+					if start < i {
+						ss = append(ss, s[start:i])
+					}
+					ss = append(ss, "™")
+					i += 3
+					start = i + 1
+					continue
+				default:
+					break outer
+				}
+
+			case '+':
+				if i+1 >= len(s) || s[i+1] != '-' {
+					break
+				}
+				if start < i {
+					ss = append(ss, s[start:i])
+				}
+				ss = append(ss, "±")
+				i++
+				start = i + 1
+				continue
+
+			case '.':
+				if i+1 >= len(s) || s[i+1] != '.' {
+					break
+				}
+
+				j := i + 2
+				for j < len(s) && s[j] == '.' {
+					j++
+				}
+				if start < i {
+					ss = append(ss, s[start:i])
+				}
+				if i == 0 || !(s[i-1] == '?' || s[i-1] == '!') {
+					ss = append(ss, "…")
+				} else {
+					ss = append(ss, "..")
+				}
+				i = j - 1
+				start = i + 1
+				continue
+
+			case '?', '!':
+				if i+3 >= len(s) {
+					break
+				}
+				if !(exclquest(s[i+1]) && exclquest(s[i+2]) && exclquest(s[i+3])) {
+					break
+				}
+				if start < i {
+					ss = append(ss, s[start:i])
+				}
+				ss = append(ss, s[i:i+3])
+				j := i + 3
+				for j < len(s) && exclquest(s[j]) {
+					j++
+				}
+				i = j - 1
+				start = i + 1
+				continue
+
+			case ',':
+				if i+1 >= len(s) || s[i+1] != ',' {
+					break
+				}
+				if start < i {
+					ss = append(ss, s[start:i])
+				}
+				ss = append(ss, ",")
+				j := i + 2
+				for j < len(s) && s[j] == ',' {
+					j++
+				}
+				i = j - 1
+				start = i + 1
+				continue
+
+			case '-':
+				if i+1 >= len(s) || s[i+1] != '-' {
+					break
+				}
+				if i+2 >= len(s) || s[i+2] != '-' {
+					if start < i {
+						ss = append(ss, s[start:i])
+					}
+					ss = append(ss, "–")
+					i++
+					start = i + 1
+					continue
+				}
+				if i+3 >= len(s) || s[i+3] != '-' {
+					if start < i {
+						ss = append(ss, s[start:i])
+					}
+					ss = append(ss, "—")
+					i += 2
+					start = i + 1
+					continue
+				}
+
+				j := i + 3
+				for j < len(s) && s[j] == '-' {
+					j++
+				}
+				if start < i {
+					ss = append(ss, s[start:i])
+				}
+				ss = append(ss, s[i:j])
+				i = j - 1
+				start = i + 1
+				continue
+			}
+		}
+	}
+	if ss == nil {
+		return s
+	}
+	if start < len(s) {
+		ss = append(ss, s[start:])
+	}
+	return strings.Join(ss, "")
+}
+
+func ruleReplacements(s *StateCore) {
+	if !s.Md.Typographer {
+		return
+	}
+
+	insideLink := false
+	for _, tok := range s.Tokens {
+		if tok, ok := tok.(*Inline); ok {
+			for _, itok := range tok.Children {
+				switch itok := itok.(type) {
+				case *LinkOpen:
+					insideLink = true
+				case *LinkClose:
+					insideLink = false
+				case *Text:
+					if !insideLink {
+						itok.Content = performReplacements(itok.Content)
+					}
+				}
+			}
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/smartquotes.go b/vendor/gitlab.com/golang-commonmark/markdown/smartquotes.go
new file mode 100644
index 00000000..cbfa9c58
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/smartquotes.go
@@ -0,0 +1,256 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"strings"
+	"unicode"
+	"unicode/utf8"
+)
+
+func nextQuoteIndex(s []rune, from int) int {
+	for i := from; i < len(s); i++ {
+		r := s[i]
+		if r == '\'' || r == '"' {
+			return i
+		}
+	}
+	return -1
+}
+
+func firstRune(s string) rune {
+	for _, r := range s {
+		return r
+	}
+	return utf8.RuneError
+}
+
+func replaceQuotes(tokens []Token, s *StateCore) {
+	type stackItem struct {
+		token  int
+		text   []rune
+		pos    int
+		single bool
+		level  int
+	}
+	var stack []stackItem
+	var changed map[int][]rune
+
+	for i, tok := range tokens {
+		thisLevel := tok.Level()
+
+		j := len(stack) - 1
+		for j >= 0 {
+			if stack[j].level <= thisLevel {
+				break
+			}
+			j--
+		}
+		stack = stack[:j+1]
+
+		tok, ok := tok.(*Text)
+		if !ok || !strings.ContainsAny(tok.Content, `"'`) {
+			continue
+		}
+
+		text := []rune(tok.Content)
+		pos := 0
+		max := len(text)
+
+	loop:
+		for pos < max {
+			index := nextQuoteIndex(text, pos)
+			if index < 0 {
+				break
+			}
+
+			canOpen := true
+			canClose := true
+			pos = index + 1
+			isSingle := text[index] == '\''
+
+			lastChar := ' '
+			if index-1 > 0 {
+				lastChar = text[index-1]
+			} else {
+			loop1:
+				for j := i - 1; j >= 0; j-- {
+					switch tok := tokens[j].(type) {
+					case *Softbreak:
+						break loop1
+					case *Hardbreak:
+						break loop1
+					case *Text:
+						lastChar, _ = utf8.DecodeLastRuneInString(tok.Content)
+						break loop1
+					default:
+						continue
+					}
+				}
+			}
+
+			nextChar := ' '
+			if pos < max {
+				nextChar = text[pos]
+			} else {
+			loop2:
+				for j := i + 1; j < len(tokens); j++ {
+					switch tok := tokens[j].(type) {
+					case *Softbreak:
+						break loop2
+					case *Hardbreak:
+						break loop2
+					case *Text:
+						nextChar, _ = utf8.DecodeRuneInString(tok.Content)
+						break loop2
+					default:
+						continue
+					}
+				}
+			}
+
+			isLastPunct := isMdAsciiPunct(lastChar) || unicode.IsPunct(lastChar)
+			isNextPunct := isMdAsciiPunct(nextChar) || unicode.IsPunct(nextChar)
+			isLastWhiteSpace := unicode.IsSpace(lastChar)
+			isNextWhiteSpace := unicode.IsSpace(nextChar)
+
+			if isNextWhiteSpace {
+				canOpen = false
+			} else if isNextPunct {
+				if !(isLastWhiteSpace || isLastPunct) {
+					canOpen = false
+				}
+			}
+
+			if isLastWhiteSpace {
+				canClose = false
+			} else if isLastPunct {
+				if !(isNextWhiteSpace || isNextPunct) {
+					canClose = false
+				}
+			}
+
+			if nextChar == '"' && text[index] == '"' {
+				if lastChar >= '0' && lastChar <= '9' {
+					canClose = false
+					canOpen = false
+				}
+			}
+
+			if canOpen && canClose {
+				canOpen = false
+				canClose = isNextPunct
+			}
+
+			if !canOpen && !canClose {
+				if isSingle {
+					text[index] = '’'
+					if changed == nil {
+						changed = make(map[int][]rune)
+					}
+					changed[i] = text
+				}
+				continue
+			}
+
+			if canClose {
+				for j := len(stack) - 1; j >= 0; j-- {
+					item := stack[j]
+					if item.level < thisLevel {
+						break
+					}
+					if item.single == isSingle && item.level == thisLevel {
+						if changed == nil {
+							changed = make(map[int][]rune)
+						}
+
+						var q1, q2 string
+						if isSingle {
+							q1 = s.Md.options.Quotes[2]
+							q2 = s.Md.options.Quotes[3]
+						} else {
+							q1 = s.Md.options.Quotes[0]
+							q2 = s.Md.options.Quotes[1]
+						}
+
+						if utf8.RuneCountInString(q1) == 1 && utf8.RuneCountInString(q2) == 1 {
+							item.text[item.pos] = firstRune(q1)
+							text[index] = firstRune(q2)
+						} else if tok == tokens[item.token] {
+							newText := make([]rune, 0, len(text)-2+len(q1)+len(q2))
+							newText = append(newText, text[:item.pos]...)
+							newText = append(newText, []rune(q1)...)
+							newText = append(newText, text[item.pos+1:index]...)
+							newText = append(newText, []rune(q2)...)
+							newText = append(newText, text[index+1:]...)
+
+							text = newText
+							item.text = newText
+						} else {
+							newText := make([]rune, 0, len(item.text)-1+len(q1))
+							newText = append(newText, item.text[:item.pos]...)
+							newText = append(newText, []rune(q1)...)
+							newText = append(newText, item.text[item.pos+1:]...)
+							item.text = newText
+
+							newText = make([]rune, 0, len(text)-1+len(q2))
+							newText = append(newText, text[:index]...)
+							newText = append(newText, []rune(q2)...)
+							newText = append(newText, text[index+1:]...)
+
+							text = newText
+						}
+
+						max = len(text)
+
+						if changed == nil {
+							changed = make(map[int][]rune)
+						}
+						changed[i] = text
+						changed[item.token] = item.text
+						stack = stack[:j]
+						continue loop
+					}
+				}
+			}
+
+			if canOpen {
+				stack = append(stack, stackItem{
+					token:  i,
+					text:   text,
+					pos:    index,
+					single: isSingle,
+					level:  thisLevel,
+				})
+			} else if canClose && isSingle {
+				text[index] = '’'
+				if changed == nil {
+					changed = make(map[int][]rune)
+				}
+				changed[i] = text
+			}
+		}
+	}
+
+	if changed != nil {
+		for i, text := range changed {
+			tokens[i].(*Text).Content = string(text)
+		}
+	}
+}
+
+func ruleSmartQuotes(s *StateCore) {
+	if !s.Md.Typographer {
+		return
+	}
+
+	tokens := s.Tokens
+	for i := len(tokens) - 1; i >= 0; i-- {
+		tok := tokens[i]
+		if tok, ok := tok.(*Inline); ok {
+			replaceQuotes(tok.Children, s)
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/state_block.go b/vendor/gitlab.com/golang-commonmark/markdown/state_block.go
new file mode 100644
index 00000000..e1142416
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/state_block.go
@@ -0,0 +1,141 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import "strings"
+
+const (
+	ptRoot = iota
+	ptList
+	ptBlockQuote
+	ptParagraph
+	ptReference
+)
+
+type StateBlock struct {
+	StateCore
+
+	BMarks     []int // offsets of the line beginnings
+	EMarks     []int // offsets of the line endings
+	TShift     []int // indents for each line
+	SCount     []int
+	BSCount    []int
+	BlkIndent  int  // required block content indent (in a list etc.)
+	Line       int  // line index in the source string
+	LineMax    int  // number of lines
+	Tight      bool // loose or tight mode for lists
+	ParentType byte // parent block type
+	Level      int
+}
+
+func (s *StateBlock) IsLineEmpty(n int) bool {
+	return s.BMarks[n]+s.TShift[n] >= s.EMarks[n]
+}
+
+func (s *StateBlock) SkipEmptyLines(from int) int {
+	for from < s.LineMax && s.IsLineEmpty(from) {
+		from++
+	}
+	return from
+}
+
+func (s *StateBlock) SkipSpaces(pos int) int {
+	src := s.Src
+	for pos < len(src) && byteIsSpace(src[pos]) {
+		pos++
+	}
+	return pos
+}
+
+func (s *StateBlock) SkipBytes(pos int, b byte) int {
+	src := s.Src
+	for pos < len(src) && src[pos] == b {
+		pos++
+	}
+	return pos
+}
+
+func (s *StateBlock) SkipBytesBack(pos int, b byte, min int) int {
+	for pos > min {
+		pos--
+		if s.Src[pos] != b {
+			return pos + 1
+		}
+	}
+	return pos
+}
+
+func (s *StateBlock) SkipSpacesBack(pos int, min int) int {
+	for pos > min {
+		pos--
+		if !byteIsSpace(s.Src[pos]) {
+			return pos + 1
+		}
+	}
+	return pos
+}
+
+func (s *StateBlock) Lines(begin, end, indent int, keepLastLf bool) string {
+	if begin == end {
+		return ""
+	}
+
+	src := s.Src
+
+	queue := make([]string, end-begin)
+
+	for i, line := 0, begin; line < end; i, line = i+1, line+1 {
+		lineIndent := 0
+		lineStart := s.BMarks[line]
+		first := lineStart
+		last := s.EMarks[line]
+		if (line+1 < end || keepLastLf) && last < len(src) {
+			last++
+		}
+
+		for first < last && lineIndent < indent {
+			ch := src[first]
+
+			if byteIsSpace(ch) {
+				if ch == '\t' {
+					lineIndent += 4 - (lineIndent+s.BSCount[line])%4
+				} else {
+					lineIndent++
+				}
+			} else if first-lineStart < s.TShift[line] {
+				lineIndent++
+			} else {
+				break
+			}
+
+			first++
+		}
+
+		if lineIndent > indent {
+			queue[i] = strings.Repeat(" ", lineIndent-indent) + src[first:last]
+		} else {
+			queue[i] = src[first:last]
+		}
+	}
+
+	return strings.Join(queue, "")
+}
+
+func (s *StateBlock) PushToken(tok Token) {
+	tok.SetLevel(s.Level)
+	s.Tokens = append(s.Tokens, tok)
+}
+
+func (s *StateBlock) PushOpeningToken(tok Token) {
+	tok.SetLevel(s.Level)
+	s.Level++
+	s.Tokens = append(s.Tokens, tok)
+}
+
+func (s *StateBlock) PushClosingToken(tok Token) {
+	s.Level--
+	tok.SetLevel(s.Level)
+	s.Tokens = append(s.Tokens, tok)
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/state_core.go b/vendor/gitlab.com/golang-commonmark/markdown/state_core.go
new file mode 100644
index 00000000..24e51756
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/state_core.go
@@ -0,0 +1,13 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+type StateCore struct {
+	Src       string
+	Tokens    []Token
+	bootstrap [3]Token
+	Md        *Markdown
+	Env       *Environment
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/state_inline.go b/vendor/gitlab.com/golang-commonmark/markdown/state_inline.go
new file mode 100644
index 00000000..4bc08cf5
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/state_inline.go
@@ -0,0 +1,116 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"bytes"
+	"unicode"
+	"unicode/utf8"
+)
+
+type StateInline struct {
+	StateCore
+
+	Pos          int
+	PosMax       int
+	Level        int
+	Pending      bytes.Buffer
+	PendingLevel int
+	Delimiters   []Delimiter
+
+	Cache map[int]int
+}
+
+func (s *StateInline) PushToken(tok Token) {
+	if s.Pending.Len() > 0 {
+		s.PushPending()
+	}
+	tok.SetLevel(s.Level)
+	s.PendingLevel = s.Level
+	s.Tokens = append(s.Tokens, tok)
+}
+
+func (s *StateInline) PushOpeningToken(tok Token) {
+	if s.Pending.Len() > 0 {
+		s.PushPending()
+	}
+	tok.SetLevel(s.Level)
+	s.Level++
+	s.PendingLevel = s.Level
+	s.Tokens = append(s.Tokens, tok)
+}
+
+func (s *StateInline) PushClosingToken(tok Token) {
+	if s.Pending.Len() > 0 {
+		s.PushPending()
+	}
+	s.Level--
+	tok.SetLevel(s.Level)
+	s.PendingLevel = s.Level
+	s.Tokens = append(s.Tokens, tok)
+}
+
+func (s *StateInline) PushPending() {
+	s.Tokens = append(s.Tokens, &Text{
+		Content: s.Pending.String(),
+		Lvl:     s.PendingLevel,
+	})
+	s.Pending.Reset()
+}
+
+func (s *StateInline) scanDelims(start int, canSplitWord bool) (canOpen bool, canClose bool, length int) {
+	pos := start
+	max := s.PosMax
+	src := s.Src
+	marker := src[start]
+	leftFlanking, rightFlanking := true, true
+
+	lastChar := ' '
+	if start > 0 {
+		lastChar, _ = utf8.DecodeLastRuneInString(src[:start])
+	}
+
+	for pos < max && src[pos] == marker {
+		pos++
+	}
+	length = pos - start
+
+	nextChar := ' '
+	if pos < max {
+		nextChar, _ = utf8.DecodeRuneInString(src[pos:])
+	}
+
+	isLastPunct := isMdAsciiPunct(lastChar) || unicode.IsPunct(lastChar)
+	isNextPunct := isMdAsciiPunct(nextChar) || unicode.IsPunct(nextChar)
+
+	isLastWhiteSpace := unicode.IsSpace(lastChar)
+	isNextWhiteSpace := unicode.IsSpace(nextChar)
+
+	if isNextWhiteSpace {
+		leftFlanking = false
+	} else if isNextPunct {
+		if !(isLastWhiteSpace || isLastPunct) {
+			leftFlanking = false
+		}
+	}
+
+	if isLastWhiteSpace {
+		rightFlanking = false
+	} else if isLastPunct {
+		if !(isNextWhiteSpace || isNextPunct) {
+			rightFlanking = false
+		}
+	}
+
+	if !canSplitWord {
+		canOpen = leftFlanking && (!rightFlanking || isLastPunct)
+		canClose = rightFlanking && (!leftFlanking || isNextPunct)
+	} else {
+		canOpen = leftFlanking
+		canClose = rightFlanking
+	}
+
+	return
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/strikethrough.go b/vendor/gitlab.com/golang-commonmark/markdown/strikethrough.go
new file mode 100644
index 00000000..18079945
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/strikethrough.go
@@ -0,0 +1,101 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleStrikeThrough(s *StateInline, silent bool) bool {
+	src := s.Src
+	start := s.Pos
+	marker := src[start]
+
+	if silent {
+		return false
+	}
+
+	if src[start] != '~' {
+		return false
+	}
+
+	canOpen, canClose, length := s.scanDelims(start, true)
+	origLength := length
+	ch := string(marker)
+	if length < 2 {
+		return false
+	}
+
+	if length%2 != 0 {
+		s.PushToken(&Text{
+			Content: ch,
+		})
+		length--
+	}
+
+	for i := 0; i < length; i += 2 {
+		s.PushToken(&Text{
+			Content: ch + ch,
+		})
+
+		s.Delimiters = append(s.Delimiters, Delimiter{
+			Marker: marker,
+			Length: -1,
+			Jump:   i,
+			Token:  len(s.Tokens) - 1,
+			Level:  s.Level,
+			End:    -1,
+			Open:   canOpen,
+			Close:  canClose,
+		})
+	}
+
+	s.Pos += origLength
+
+	return true
+
+}
+
+func ruleStrikethroughPostprocess(s *StateInline) {
+	var loneMarkers []int
+	delimiters := s.Delimiters
+	max := len(delimiters)
+
+	for i := 0; i < max; i++ {
+		startDelim := delimiters[i]
+
+		if startDelim.Marker != '~' {
+			continue
+		}
+
+		if startDelim.End == -1 {
+			continue
+		}
+
+		endDelim := delimiters[startDelim.End]
+
+		s.Tokens[startDelim.Token] = &StrikethroughOpen{}
+		s.Tokens[endDelim.Token] = &StrikethroughClose{}
+
+		if text, ok := s.Tokens[endDelim.Token-1].(*Text); ok && text.Content == "~" {
+			loneMarkers = append(loneMarkers, endDelim.Token-1)
+		}
+	}
+
+	for len(loneMarkers) > 0 {
+		i := loneMarkers[len(loneMarkers)-1]
+		loneMarkers = loneMarkers[:len(loneMarkers)-1]
+		j := i + 1
+
+		for j < len(s.Tokens) {
+			if _, ok := s.Tokens[j].(*StrikethroughClose); !ok {
+				break
+			}
+			j++
+		}
+
+		j--
+
+		if i != j {
+			s.Tokens[i], s.Tokens[j] = s.Tokens[j], s.Tokens[i]
+		}
+	}
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/table.go b/vendor/gitlab.com/golang-commonmark/markdown/table.go
new file mode 100644
index 00000000..52ee176e
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/table.go
@@ -0,0 +1,229 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"regexp"
+	"strings"
+)
+
+func getLine(s *StateBlock, line int) string {
+	pos := s.BMarks[line] + s.BlkIndent
+	max := s.EMarks[line]
+	if pos >= max {
+		return ""
+	}
+	return s.Src[pos:max]
+}
+
+func escapedSplit(s string) (result []string) {
+	pos := 0
+	escapes := 0
+	lastPos := 0
+	backTicked := false
+	lastBackTick := 0
+
+	for pos < len(s) {
+		ch := s[pos]
+		if ch == '`' {
+			if backTicked {
+				backTicked = false
+				lastBackTick = pos
+			} else if escapes%2 == 0 {
+				backTicked = true
+				lastBackTick = pos
+			}
+		} else if ch == '|' && (escapes%2 == 0) && !backTicked {
+			result = append(result, s[lastPos:pos])
+			lastPos = pos + 1
+		}
+
+		if ch == '\\' {
+			escapes++
+		} else {
+			escapes = 0
+		}
+
+		pos++
+
+		if pos == len(s) && backTicked {
+			backTicked = false
+			pos = lastBackTick + 1
+		}
+	}
+
+	return append(result, s[lastPos:])
+}
+
+var rColumn = regexp.MustCompile("^:?-+:?$")
+
+func ruleTable(s *StateBlock, startLine, endLine int, silent bool) bool {
+	if !s.Md.Tables {
+		return false
+	}
+
+	if startLine+2 > endLine {
+		return false
+	}
+
+	nextLine := startLine + 1
+
+	if s.SCount[nextLine] < s.BlkIndent {
+		return false
+	}
+
+	if s.SCount[nextLine]-s.BlkIndent >= 4 {
+		return false
+	}
+
+	pos := s.BMarks[nextLine] + s.TShift[nextLine]
+	if pos >= s.EMarks[nextLine] {
+		return false
+	}
+
+	src := s.Src
+	ch := src[pos]
+	pos++
+
+	if ch != '|' && ch != '-' && ch != ':' {
+		return false
+	}
+
+	for pos < s.EMarks[nextLine] {
+		ch = src[pos]
+		if ch != '|' && ch != '-' && ch != ':' && !byteIsSpace(ch) {
+			return false
+		}
+		pos++
+	}
+
+	//
+
+	lineText := getLine(s, startLine+1)
+
+	columns := strings.Split(lineText, "|")
+	var aligns []Align
+	for i := 0; i < len(columns); i++ {
+		t := strings.TrimSpace(columns[i])
+		if t == "" {
+			if i == 0 || i == len(columns)-1 {
+				continue
+			}
+			return false
+		}
+
+		if !rColumn.MatchString(t) {
+			return false
+		}
+
+		if t[len(t)-1] == ':' {
+			if t[0] == ':' {
+				aligns = append(aligns, AlignCenter)
+			} else {
+				aligns = append(aligns, AlignRight)
+			}
+		} else if t[0] == ':' {
+			aligns = append(aligns, AlignLeft)
+		} else {
+			aligns = append(aligns, AlignNone)
+		}
+	}
+
+	lineText = strings.TrimSpace(getLine(s, startLine))
+	if strings.IndexByte(lineText, '|') == -1 {
+		return false
+	}
+	if s.SCount[startLine]-s.BlkIndent >= 4 {
+		return false
+	}
+	columns = escapedSplit(strings.TrimSuffix(strings.TrimPrefix(lineText, "|"), "|"))
+	columnCount := len(columns)
+	if columnCount > len(aligns) {
+		return false
+	}
+
+	if silent {
+		return true
+	}
+
+	tableTok := &TableOpen{
+		Map: [2]int{startLine, 0},
+	}
+	s.PushOpeningToken(tableTok)
+	s.PushOpeningToken(&TheadOpen{
+		Map: [2]int{startLine, startLine + 1},
+	})
+	s.PushOpeningToken(&TrOpen{
+		Map: [2]int{startLine, startLine + 1},
+	})
+
+	for i := 0; i < len(columns); i++ {
+		s.PushOpeningToken(&ThOpen{
+			Align: aligns[i],
+			Map:   [2]int{startLine, startLine + 1},
+		})
+		s.PushToken(&Inline{
+			Content: strings.TrimSpace(columns[i]),
+			Map:     [2]int{startLine, startLine + 1},
+		})
+		s.PushClosingToken(&ThClose{})
+	}
+
+	s.PushClosingToken(&TrClose{})
+	s.PushClosingToken(&TheadClose{})
+
+	tbodyTok := &TbodyOpen{
+		Map: [2]int{startLine + 2, 0},
+	}
+	s.PushOpeningToken(tbodyTok)
+
+	for nextLine = startLine + 2; nextLine < endLine; nextLine++ {
+		if s.SCount[nextLine] < s.BlkIndent {
+			break
+		}
+
+		lineText = strings.TrimSpace(getLine(s, nextLine))
+		if strings.IndexByte(lineText, '|') == -1 {
+			break
+		}
+		if s.SCount[nextLine]-s.BlkIndent >= 4 {
+			break
+		}
+		columns = escapedSplit(strings.TrimPrefix(strings.TrimSuffix(lineText, "|"), "|"))
+
+		if len(columns) < len(aligns) {
+			columns = append(columns, make([]string, len(aligns)-len(columns))...)
+		} else if len(columns) > len(aligns) {
+			columns = columns[:len(aligns)]
+		}
+
+		s.PushOpeningToken(&TrOpen{})
+		for i := 0; i < columnCount; i++ {
+			tdOpen := TdOpen{}
+			if i < len(aligns) {
+				tdOpen.Align = aligns[i]
+			}
+			s.PushOpeningToken(&tdOpen)
+
+			inline := Inline{}
+			if i < len(columns) {
+				inline.Content = strings.TrimSpace(columns[i])
+			}
+			s.PushToken(&inline)
+
+			s.PushClosingToken(&TdClose{})
+		}
+		s.PushClosingToken(&TrClose{})
+	}
+
+	s.PushClosingToken(&TbodyClose{})
+	s.PushClosingToken(&TableClose{})
+
+	tableTok.Map[1] = nextLine
+	tbodyTok.Map[1] = nextLine
+	s.Line = nextLine
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/table_fsm.go b/vendor/gitlab.com/golang-commonmark/markdown/table_fsm.go
new file mode 100644
index 00000000..b605e45c
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/table_fsm.go
@@ -0,0 +1,108 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func isHeaderLine(s string) bool {
+	if s == "" {
+		return false
+	}
+
+	st := 0
+	n := 0
+	for i := 0; i < len(s); i++ {
+		b := s[i]
+		switch st {
+		case 0: // initial state
+			switch b {
+			case '|':
+				st = 1
+			case ':':
+				st = 2
+			case '-':
+				st = 3
+				n++
+			case ' ':
+				break
+			default:
+				return false
+			}
+
+		case 1: // |
+			switch b {
+			case ' ':
+				break
+			case ':':
+				st = 2
+			case '-':
+				st = 3
+				n++
+			default:
+				return false
+			}
+
+		case 2: // |:
+			switch b {
+			case ' ':
+				break
+			case '-':
+				st = 3
+				n++
+			default:
+				return false
+			}
+
+		case 3: // |:-
+			switch b {
+			case '-':
+				break
+			case ':':
+				st = 4
+			case '|':
+				st = 5
+			case ' ':
+				st = 6
+			default:
+				return false
+			}
+
+		case 4: // |:---:
+			switch b {
+			case ' ':
+				break
+			case '|':
+				st = 5
+			default:
+				return false
+			}
+
+		case 5: // |:---:|
+			switch b {
+			case ' ':
+				break
+			case ':':
+				st = 2
+			case '-':
+				st = 3
+				n++
+			default:
+				return false
+			}
+
+		case 6: // |:--- SPACE
+			switch b {
+			case ' ':
+				break
+			case ':':
+				st = 4
+			case '|':
+				st = 5
+			default:
+				return false
+			}
+		}
+	}
+
+	return n >= 1
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/text.go b/vendor/gitlab.com/golang-commonmark/markdown/text.go
new file mode 100644
index 00000000..ae98e743
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/text.go
@@ -0,0 +1,52 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+var terminatorCharTable = [256]bool{
+	'\n': true,
+	'!':  true,
+	'#':  true,
+	'$':  true,
+	'%':  true,
+	'&':  true,
+	'*':  true,
+	'+':  true,
+	'-':  true,
+	':':  true,
+	'<':  true,
+	'=':  true,
+	'>':  true,
+	'@':  true,
+	'[':  true,
+	'\\': true,
+	']':  true,
+	'^':  true,
+	'_':  true,
+	'`':  true,
+	'{':  true,
+	'}':  true,
+	'~':  true,
+}
+
+func ruleText(s *StateInline, silent bool) bool {
+	pos := s.Pos
+	max := s.PosMax
+	src := s.Src
+
+	for pos < max && !terminatorCharTable[src[pos]] {
+		pos++
+	}
+	if pos == s.Pos {
+		return false
+	}
+
+	if !silent {
+		s.Pending.WriteString(src[s.Pos:pos])
+	}
+
+	s.Pos = pos
+
+	return true
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/text_collapse.go b/vendor/gitlab.com/golang-commonmark/markdown/text_collapse.go
new file mode 100644
index 00000000..25b98b80
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/text_collapse.go
@@ -0,0 +1,42 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+func ruleTextCollapse(s *StateInline) {
+	level := 0
+	tokens := s.Tokens
+	max := len(tokens)
+
+	curr := 0
+	last := 0
+	for ; curr < max; curr++ {
+		tok := tokens[curr]
+		if tok.Opening() {
+			level++
+		} else if tok.Closing() {
+			level--
+		}
+
+		tok.SetLevel(level)
+
+		if text, ok := tok.(*Text); ok && curr+1 < max {
+			if text2, ok := tokens[curr+1].(*Text); ok {
+				text2.Content = text.Content + text2.Content
+				continue
+			}
+		}
+
+		if curr != last {
+			tokens[last] = tokens[curr]
+		}
+		last++
+	}
+
+	if curr != last {
+		tokens = tokens[:last]
+	}
+
+	s.Tokens = tokens
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/token.go b/vendor/gitlab.com/golang-commonmark/markdown/token.go
new file mode 100644
index 00000000..1b14e5fe
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/token.go
@@ -0,0 +1,753 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+type Token interface {
+	Tag() string
+	Opening() bool
+	Closing() bool
+	Block() bool
+	Level() int
+	SetLevel(lvl int)
+}
+
+type BlockquoteOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type BlockquoteClose struct {
+	Lvl int
+}
+
+type BulletListOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type BulletListClose struct {
+	Lvl int
+}
+
+type OrderedListOpen struct {
+	Order int
+	Map   [2]int
+	Lvl   int
+}
+
+type OrderedListClose struct {
+	Lvl int
+}
+
+type ListItemOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type ListItemClose struct {
+	Lvl int
+}
+
+type CodeBlock struct {
+	Content string
+	Map     [2]int
+	Lvl     int
+}
+
+type CodeInline struct {
+	Content string
+	Lvl     int
+}
+
+type EmphasisOpen struct {
+	Lvl int
+}
+
+type EmphasisClose struct {
+	Lvl int
+}
+
+type StrongOpen struct {
+	Lvl int
+}
+
+type StrongClose struct {
+	Lvl int
+}
+
+type StrikethroughOpen struct {
+	Lvl int
+}
+
+type StrikethroughClose struct {
+	Lvl int
+}
+
+type Fence struct {
+	Params  string
+	Content string
+	Map     [2]int
+	Lvl     int
+}
+
+type Softbreak struct {
+	Lvl int
+}
+
+type Hardbreak struct {
+	Lvl int
+}
+
+type HeadingOpen struct {
+	HLevel int
+	Map    [2]int
+	Lvl    int
+}
+
+type HeadingClose struct {
+	HLevel int
+	Lvl    int
+}
+
+type HTMLBlock struct {
+	Content string
+	Map     [2]int
+	Lvl     int
+}
+
+type HTMLInline struct {
+	Content string
+	Lvl     int
+}
+
+type Hr struct {
+	Map [2]int
+	Lvl int
+}
+
+type Image struct {
+	Src    string
+	Title  string
+	Tokens []Token
+	Lvl    int
+}
+
+type Inline struct {
+	Content  string
+	Map      [2]int
+	Children []Token
+	Lvl      int
+}
+
+type LinkOpen struct {
+	Href   string
+	Title  string
+	Target string
+	Lvl    int
+}
+
+type LinkClose struct {
+	Lvl int
+}
+
+type ParagraphOpen struct {
+	Tight  bool
+	Hidden bool
+	Map    [2]int
+	Lvl    int
+}
+
+type ParagraphClose struct {
+	Tight  bool
+	Hidden bool
+	Lvl    int
+}
+
+type TableOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type TableClose struct {
+	Lvl int
+}
+
+type TheadOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type TheadClose struct {
+	Lvl int
+}
+
+type TrOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type TrClose struct {
+	Lvl int
+}
+
+type ThOpen struct {
+	Align Align
+	Map   [2]int
+	Lvl   int
+}
+
+type ThClose struct {
+	Lvl int
+}
+
+type TbodyOpen struct {
+	Map [2]int
+	Lvl int
+}
+
+type TbodyClose struct {
+	Lvl int
+}
+
+type TdOpen struct {
+	Align Align
+	Map   [2]int
+	Lvl   int
+}
+
+type TdClose struct {
+	Lvl int
+}
+
+type Text struct {
+	Content string
+	Lvl     int
+}
+
+var htags = []string{
+	"",
+	"h1",
+	"h2",
+	"h3",
+	"h4",
+	"h5",
+	"h6",
+}
+
+func (t *BlockquoteOpen) Level() int { return t.Lvl }
+
+func (t *BlockquoteClose) Level() int { return t.Lvl }
+
+func (t *BulletListOpen) Level() int { return t.Lvl }
+
+func (t *BulletListClose) Level() int { return t.Lvl }
+
+func (t *OrderedListOpen) Level() int { return t.Lvl }
+
+func (t *OrderedListClose) Level() int { return t.Lvl }
+
+func (t *ListItemOpen) Level() int { return t.Lvl }
+
+func (t *ListItemClose) Level() int { return t.Lvl }
+
+func (t *CodeBlock) Level() int { return t.Lvl }
+
+func (t *CodeInline) Level() int { return t.Lvl }
+
+func (t *EmphasisOpen) Level() int { return t.Lvl }
+
+func (t *EmphasisClose) Level() int { return t.Lvl }
+
+func (t *StrongOpen) Level() int { return t.Lvl }
+
+func (t *StrongClose) Level() int { return t.Lvl }
+
+func (t *StrikethroughOpen) Level() int { return t.Lvl }
+
+func (t *StrikethroughClose) Level() int { return t.Lvl }
+
+func (t *Fence) Level() int { return t.Lvl }
+
+func (t *Softbreak) Level() int { return t.Lvl }
+
+func (t *Hardbreak) Level() int { return t.Lvl }
+
+func (t *HeadingOpen) Level() int { return t.Lvl }
+
+func (t *HeadingClose) Level() int { return t.Lvl }
+
+func (t *HTMLBlock) Level() int { return t.Lvl }
+
+func (t *HTMLInline) Level() int { return t.Lvl }
+
+func (t *Hr) Level() int { return t.Lvl }
+
+func (t *Image) Level() int { return t.Lvl }
+
+func (t *Inline) Level() int { return t.Lvl }
+
+func (t *LinkOpen) Level() int { return t.Lvl }
+
+func (t *LinkClose) Level() int { return t.Lvl }
+
+func (t *ParagraphOpen) Level() int { return t.Lvl }
+
+func (t *ParagraphClose) Level() int { return t.Lvl }
+
+func (t *TableOpen) Level() int { return t.Lvl }
+
+func (t *TableClose) Level() int { return t.Lvl }
+
+func (t *TheadOpen) Level() int { return t.Lvl }
+
+func (t *TheadClose) Level() int { return t.Lvl }
+
+func (t *TrOpen) Level() int { return t.Lvl }
+
+func (t *TrClose) Level() int { return t.Lvl }
+
+func (t *ThOpen) Level() int { return t.Lvl }
+
+func (t *ThClose) Level() int { return t.Lvl }
+
+func (t *TbodyOpen) Level() int { return t.Lvl }
+
+func (t *TbodyClose) Level() int { return t.Lvl }
+
+func (t *TdOpen) Level() int { return t.Lvl }
+
+func (t *TdClose) Level() int { return t.Lvl }
+
+func (t *Text) Level() int { return t.Lvl }
+
+func (t *BlockquoteOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *BlockquoteClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *BulletListOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *BulletListClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *OrderedListOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *OrderedListClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ListItemOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ListItemClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *CodeBlock) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *CodeInline) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *EmphasisOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *EmphasisClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *StrongOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *StrongClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *StrikethroughOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *StrikethroughClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Fence) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Softbreak) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Hardbreak) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *HeadingOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *HeadingClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *HTMLBlock) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *HTMLInline) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Hr) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Image) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Inline) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *LinkOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *LinkClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ParagraphOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ParagraphClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TableOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TableClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TheadOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TheadClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TrOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TrClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ThOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *ThClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TbodyOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TbodyClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TdOpen) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *TdClose) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *Text) SetLevel(lvl int) { t.Lvl = lvl }
+
+func (t *BlockquoteOpen) Opening() bool { return true }
+
+func (t *BlockquoteClose) Opening() bool { return false }
+
+func (t *BulletListOpen) Opening() bool { return true }
+
+func (t *BulletListClose) Opening() bool { return false }
+
+func (t *OrderedListOpen) Opening() bool { return true }
+
+func (t *OrderedListClose) Opening() bool { return false }
+
+func (t *ListItemOpen) Opening() bool { return true }
+
+func (t *ListItemClose) Opening() bool { return false }
+
+func (t *CodeBlock) Opening() bool { return false }
+
+func (t *CodeInline) Opening() bool { return false }
+
+func (t *EmphasisOpen) Opening() bool { return true }
+
+func (t *EmphasisClose) Opening() bool { return false }
+
+func (t *StrongOpen) Opening() bool { return true }
+
+func (t *StrongClose) Opening() bool { return false }
+
+func (t *StrikethroughOpen) Opening() bool { return true }
+
+func (t *StrikethroughClose) Opening() bool { return false }
+
+func (t *Fence) Opening() bool { return false }
+
+func (t *Softbreak) Opening() bool { return false }
+
+func (t *Hardbreak) Opening() bool { return false }
+
+func (t *HeadingOpen) Opening() bool { return true }
+
+func (t *HeadingClose) Opening() bool { return false }
+
+func (t *HTMLBlock) Opening() bool { return false }
+
+func (t *HTMLInline) Opening() bool { return false }
+
+func (t *Hr) Opening() bool { return false }
+
+func (t *Image) Opening() bool { return false }
+
+func (t *Inline) Opening() bool { return false }
+
+func (t *LinkOpen) Opening() bool { return true }
+
+func (t *LinkClose) Opening() bool { return false }
+
+func (t *ParagraphOpen) Opening() bool { return true }
+
+func (t *ParagraphClose) Opening() bool { return false }
+
+func (t *TableOpen) Opening() bool { return true }
+
+func (t *TableClose) Opening() bool { return false }
+
+func (t *TheadOpen) Opening() bool { return true }
+
+func (t *TheadClose) Opening() bool { return false }
+
+func (t *TrOpen) Opening() bool { return true }
+
+func (t *TrClose) Opening() bool { return false }
+
+func (t *ThOpen) Opening() bool { return true }
+
+func (t *ThClose) Opening() bool { return false }
+
+func (t *TbodyOpen) Opening() bool { return true }
+
+func (t *TbodyClose) Opening() bool { return false }
+
+func (t *TdOpen) Opening() bool { return true }
+
+func (t *TdClose) Opening() bool { return false }
+
+func (t *Text) Opening() bool { return false }
+
+func (t *BlockquoteOpen) Closing() bool { return false }
+
+func (t *BlockquoteClose) Closing() bool { return true }
+
+func (t *BulletListOpen) Closing() bool { return false }
+
+func (t *BulletListClose) Closing() bool { return true }
+
+func (t *OrderedListOpen) Closing() bool { return false }
+
+func (t *OrderedListClose) Closing() bool { return true }
+
+func (t *ListItemOpen) Closing() bool { return false }
+
+func (t *ListItemClose) Closing() bool { return true }
+
+func (t *CodeBlock) Closing() bool { return false }
+
+func (t *CodeInline) Closing() bool { return false }
+
+func (t *EmphasisOpen) Closing() bool { return false }
+
+func (t *EmphasisClose) Closing() bool { return true }
+
+func (t *StrongOpen) Closing() bool { return false }
+
+func (t *StrongClose) Closing() bool { return true }
+
+func (t *StrikethroughOpen) Closing() bool { return false }
+
+func (t *StrikethroughClose) Closing() bool { return true }
+
+func (t *Fence) Closing() bool { return false }
+
+func (t *Softbreak) Closing() bool { return false }
+
+func (t *Hardbreak) Closing() bool { return false }
+
+func (t *HeadingOpen) Closing() bool { return false }
+
+func (t *HeadingClose) Closing() bool { return true }
+
+func (t *HTMLBlock) Closing() bool { return false }
+
+func (t *HTMLInline) Closing() bool { return false }
+
+func (t *Hr) Closing() bool { return false }
+
+func (t *Image) Closing() bool { return false }
+
+func (t *Inline) Closing() bool { return false }
+
+func (t *LinkOpen) Closing() bool { return false }
+
+func (t *LinkClose) Closing() bool { return true }
+
+func (t *ParagraphOpen) Closing() bool { return false }
+
+func (t *ParagraphClose) Closing() bool { return true }
+
+func (t *TableOpen) Closing() bool { return false }
+
+func (t *TableClose) Closing() bool { return true }
+
+func (t *TheadOpen) Closing() bool { return false }
+
+func (t *TheadClose) Closing() bool { return true }
+
+func (t *TrOpen) Closing() bool { return false }
+
+func (t *TrClose) Closing() bool { return true }
+
+func (t *ThOpen) Closing() bool { return false }
+
+func (t *ThClose) Closing() bool { return true }
+
+func (t *TbodyOpen) Closing() bool { return false }
+
+func (t *TbodyClose) Closing() bool { return true }
+
+func (t *TdOpen) Closing() bool { return false }
+
+func (t *TdClose) Closing() bool { return true }
+
+func (t *Text) Closing() bool { return false }
+
+func (t *BlockquoteOpen) Block() bool { return true }
+
+func (t *BlockquoteClose) Block() bool { return true }
+
+func (t *BulletListOpen) Block() bool { return true }
+
+func (t *BulletListClose) Block() bool { return true }
+
+func (t *OrderedListOpen) Block() bool { return true }
+
+func (t *OrderedListClose) Block() bool { return true }
+
+func (t *ListItemOpen) Block() bool { return true }
+
+func (t *ListItemClose) Block() bool { return true }
+
+func (t *CodeBlock) Block() bool { return true }
+
+func (t *CodeInline) Block() bool { return false }
+
+func (t *EmphasisOpen) Block() bool { return false }
+
+func (t *EmphasisClose) Block() bool { return false }
+
+func (t *StrongOpen) Block() bool { return false }
+
+func (t *StrongClose) Block() bool { return false }
+
+func (t *StrikethroughOpen) Block() bool { return false }
+
+func (t *StrikethroughClose) Block() bool { return false }
+
+func (t *Fence) Block() bool { return true }
+
+func (t *Softbreak) Block() bool { return false }
+
+func (t *Hardbreak) Block() bool { return false }
+
+func (t *HeadingOpen) Block() bool { return true }
+
+func (t *HeadingClose) Block() bool { return true }
+
+func (t *HTMLBlock) Block() bool { return true }
+
+func (t *HTMLInline) Block() bool { return false }
+
+func (t *Hr) Block() bool { return true }
+
+func (t *Image) Block() bool { return false }
+
+func (t *Inline) Block() bool { return false }
+
+func (t *LinkOpen) Block() bool { return false }
+
+func (t *LinkClose) Block() bool { return false }
+
+func (t *ParagraphOpen) Block() bool { return true }
+
+func (t *ParagraphClose) Block() bool { return true }
+
+func (t *TableOpen) Block() bool { return true }
+
+func (t *TableClose) Block() bool { return true }
+
+func (t *TheadOpen) Block() bool { return true }
+
+func (t *TheadClose) Block() bool { return true }
+
+func (t *TrOpen) Block() bool { return true }
+
+func (t *TrClose) Block() bool { return true }
+
+func (t *ThOpen) Block() bool { return true }
+
+func (t *ThClose) Block() bool { return true }
+
+func (t *TbodyOpen) Block() bool { return true }
+
+func (t *TbodyClose) Block() bool { return true }
+
+func (t *TdOpen) Block() bool { return true }
+
+func (t *TdClose) Block() bool { return true }
+
+func (t *Text) Block() bool { return false }
+
+func (t *BlockquoteOpen) Tag() string { return "blockquote" }
+
+func (t *BlockquoteClose) Tag() string { return "blockquote" }
+
+func (t *BulletListOpen) Tag() string { return "ul" }
+
+func (t *BulletListClose) Tag() string { return "ul" }
+
+func (t *OrderedListOpen) Tag() string { return "ol" }
+
+func (t *OrderedListClose) Tag() string { return "ol" }
+
+func (t *ListItemOpen) Tag() string { return "li" }
+
+func (t *ListItemClose) Tag() string { return "li" }
+
+func (t *CodeBlock) Tag() string { return "code" }
+
+func (t *CodeInline) Tag() string { return "code" }
+
+func (t *EmphasisOpen) Tag() string { return "em" }
+
+func (t *EmphasisClose) Tag() string { return "em" }
+
+func (t *StrongOpen) Tag() string { return "strong" }
+
+func (t *StrongClose) Tag() string { return "strong" }
+
+func (t *StrikethroughOpen) Tag() string { return "s" }
+
+func (t *StrikethroughClose) Tag() string { return "s" }
+
+func (t *Fence) Tag() string { return "code" }
+
+func (t *Softbreak) Tag() string { return "br" }
+
+func (t *Hardbreak) Tag() string { return "br" }
+
+func (t *HeadingOpen) Tag() string { return htags[t.HLevel] }
+
+func (t *HeadingClose) Tag() string { return htags[t.HLevel] }
+
+func (t *HTMLBlock) Tag() string { return "" }
+
+func (t *HTMLInline) Tag() string { return "" }
+
+func (t *Hr) Tag() string { return "hr" }
+
+func (t *Image) Tag() string { return "img" }
+
+func (t *Inline) Tag() string { return "" }
+
+func (t *LinkOpen) Tag() string { return "a" }
+
+func (t *LinkClose) Tag() string { return "a" }
+
+func (t *ParagraphOpen) Tag() string { return "p" }
+
+func (t *ParagraphClose) Tag() string { return "p" }
+
+func (t *TableOpen) Tag() string { return "table" }
+
+func (t *TableClose) Tag() string { return "table" }
+
+func (t *TheadOpen) Tag() string { return "thead" }
+
+func (t *TheadClose) Tag() string { return "thead" }
+
+func (t *TrOpen) Tag() string { return "tr" }
+
+func (t *TrClose) Tag() string { return "tr" }
+
+func (t *ThOpen) Tag() string { return "th" }
+
+func (t *ThClose) Tag() string { return "th" }
+
+func (t *TbodyOpen) Tag() string { return "tbody" }
+
+func (t *TbodyClose) Tag() string { return "tbody" }
+
+func (t *TdOpen) Tag() string { return "td" }
+
+func (t *TdClose) Tag() string { return "td" }
+
+func (t *Text) Tag() string { return "" }
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/urlschema.go b/vendor/gitlab.com/golang-commonmark/markdown/urlschema.go
new file mode 100644
index 00000000..06b9e6a1
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/urlschema.go
@@ -0,0 +1,185 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+var urlSchemas = []string{
+	"aaa",
+	"aaas",
+	"about",
+	"acap",
+	"adiumxtra",
+	"afp",
+	"afs",
+	"aim",
+	"apt",
+	"attachment",
+	"aw",
+	"beshare",
+	"bitcoin",
+	"bolo",
+	"callto",
+	"cap",
+	"chrome",
+	"chrome-extension",
+	"cid",
+	"coap",
+	"com-eventbrite-attendee",
+	"content",
+	"crid",
+	"cvs",
+	"data",
+	"dav",
+	"dict",
+	"dlna-playcontainer",
+	"dlna-playsingle",
+	"dns",
+	"doi",
+	"dtn",
+	"dvb",
+	"ed2k",
+	"facetime",
+	"feed",
+	"file",
+	"finger",
+	"fish",
+	"ftp",
+	"geo",
+	"gg",
+	"git",
+	"gizmoproject",
+	"go",
+	"gopher",
+	"gtalk",
+	"h323",
+	"hcp",
+	"http",
+	"https",
+	"iax",
+	"icap",
+	"icon",
+	"im",
+	"imap",
+	"info",
+	"ipn",
+	"ipp",
+	"irc",
+	"irc6",
+	"ircs",
+	"iris",
+	"iris.beep",
+	"iris.lwz",
+	"iris.xpc",
+	"iris.xpcs",
+	"itms",
+	"jar",
+	"javascript",
+	"jms",
+	"keyparc",
+	"lastfm",
+	"ldap",
+	"ldaps",
+	"magnet",
+	"mailto",
+	"maps",
+	"market",
+	"message",
+	"mid",
+	"mms",
+	"ms-help",
+	"msnim",
+	"msrp",
+	"msrps",
+	"mtqp",
+	"mumble",
+	"mupdate",
+	"mvn",
+	"news",
+	"nfs",
+	"ni",
+	"nih",
+	"nntp",
+	"notes",
+	"oid",
+	"opaquelocktoken",
+	"palm",
+	"paparazzi",
+	"platform",
+	"pop",
+	"pres",
+	"proxy",
+	"psyc",
+	"query",
+	"res",
+	"resource",
+	"rmi",
+	"rsync",
+	"rtmp",
+	"rtsp",
+	"secondlife",
+	"service",
+	"session",
+	"sftp",
+	"sgn",
+	"shttp",
+	"sieve",
+	"sip",
+	"sips",
+	"skype",
+	"smb",
+	"sms",
+	"snmp",
+	"soap.beep",
+	"soap.beeps",
+	"soldat",
+	"spotify",
+	"ssh",
+	"steam",
+	"svn",
+	"tag",
+	"teamspeak",
+	"tel",
+	"telnet",
+	"tftp",
+	"things",
+	"thismessage",
+	"tip",
+	"tn3270",
+	"tv",
+	"udp",
+	"unreal",
+	"urn",
+	"ut2004",
+	"vemmi",
+	"ventrilo",
+	"view-source",
+	"webcal",
+	"ws",
+	"wss",
+	"wtai",
+	"wyciwyg",
+	"xcon",
+	"xcon-userid",
+	"xfire",
+	"xmlrpc.beep",
+	"xmlrpc.beeps",
+	"xmpp",
+	"xri",
+	"ymsgr",
+	"z39.50r",
+	"z39.50s",
+}
+
+var urlSchemasSet = make(map[string]bool)
+
+func init() {
+	for _, s := range urlSchemas {
+		urlSchemasSet[s] = true
+	}
+}
+
+func matchSchema(s string) bool {
+	_, ok := urlSchemasSet[s]
+	return ok
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/util.go b/vendor/gitlab.com/golang-commonmark/markdown/util.go
new file mode 100644
index 00000000..d821202c
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/util.go
@@ -0,0 +1,264 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"bytes"
+	"regexp"
+	"strings"
+	"unicode"
+
+	"gitlab.com/golang-commonmark/html"
+	"gitlab.com/golang-commonmark/mdurl"
+	"gitlab.com/golang-commonmark/puny"
+)
+
+func runeIsSpace(r rune) bool {
+	return r == ' ' || r == '\t'
+}
+
+func byteIsSpace(b byte) bool {
+	return b == ' ' || b == '\t'
+}
+
+func isLetter(b byte) bool {
+	return b >= 'a' && b <= 'z' || b >= 'A' && b <= 'Z'
+}
+
+func isUppercaseLetter(b byte) bool {
+	return b >= 'A' && b <= 'Z'
+}
+
+func mdpunct(b byte) bool {
+	return strings.IndexByte("!\"#$%&'()*+,-./:;<=>?@[\\]^_`{|}~", b) != -1
+}
+
+func isMdAsciiPunct(r rune) bool {
+	if r > 0x7e {
+		return false
+	}
+	return mdpunct(byte(r))
+}
+
+func recodeHostnameFor(proto string) bool {
+	switch proto {
+	case "http", "https", "mailto":
+		return true
+	}
+	return false
+}
+
+func normalizeLink(url string) string {
+	parsed, err := mdurl.Parse(url)
+	if err != nil {
+		return ""
+	}
+
+	if parsed.Host != "" && (parsed.Scheme == "" || recodeHostnameFor(parsed.Scheme)) {
+		parsed.Host = puny.ToASCII(parsed.Host)
+	}
+
+	parsed.Scheme = parsed.RawScheme
+
+	return mdurl.Encode(parsed.String())
+}
+
+func normalizeLinkText(url string) string {
+	parsed, err := mdurl.Parse(url)
+	if err != nil {
+		return ""
+	}
+
+	if parsed.Host != "" && (parsed.Scheme == "" || recodeHostnameFor(parsed.Scheme)) {
+		parsed.Host = puny.ToUnicode(parsed.Host)
+	}
+
+	parsed.Scheme = parsed.RawScheme
+
+	return mdurl.Decode(parsed.String())
+}
+
+var badProtos = []string{"file", "javascript", "vbscript"}
+
+var rGoodData = regexp.MustCompile(`^data:image/(gif|png|jpeg|webp);`)
+
+func removeSpecial(s string) string {
+	i := 0
+	for i < len(s) && !(s[i] <= 0x20 || s[i] == 0x7f) {
+		i++
+	}
+	if i >= len(s) {
+		return s
+	}
+	buf := make([]byte, len(s))
+	j := 0
+	for i := 0; i < len(s); i++ {
+		if !(s[i] <= 0x20 || s[i] == 0x7f) {
+			buf[j] = s[i]
+			j++
+		}
+	}
+	return string(buf[:j])
+}
+
+func validateLink(url string) bool {
+	str := strings.TrimSpace(url)
+	str = strings.ToLower(str)
+
+	if strings.IndexByte(str, ':') >= 0 {
+		proto := strings.SplitN(str, ":", 2)[0]
+		proto = removeSpecial(proto)
+		for _, p := range badProtos {
+			if proto == p {
+				return false
+			}
+		}
+		if proto == "data" && !rGoodData.MatchString(str) {
+			return false
+		}
+	}
+
+	return true
+}
+
+func unescapeAll(s string) string {
+	anyChanges := false
+	i := 0
+	for i < len(s)-1 {
+		b := s[i]
+		if b == '\\' {
+			if mdpunct(s[i+1]) {
+				anyChanges = true
+				break
+			}
+		} else if b == '&' {
+			if e, n := html.ParseEntity(s[i:]); n > 0 && e != html.BadEntity {
+				anyChanges = true
+				break
+			}
+		}
+		i++
+	}
+
+	if !anyChanges {
+		return s
+	}
+
+	buf := make([]byte, len(s))
+	copy(buf[:i], s)
+	j := i
+	for i < len(s) {
+		b := s[i]
+		if b == '\\' {
+			if i+1 < len(s) {
+				b = s[i+1]
+				if mdpunct(b) {
+					buf[j] = b
+					j++
+				} else {
+					buf[j] = '\\'
+					j++
+					buf[j] = b
+					j++
+				}
+				i += 2
+				continue
+			}
+		} else if b == '&' {
+			if e, n := html.ParseEntity(s[i:]); n > 0 && e != html.BadEntity {
+				if len(e) > n && len(buf) == len(s) {
+					newBuf := make([]byte, cap(buf)*2)
+					copy(newBuf[:j], buf)
+					buf = newBuf
+				}
+				j += copy(buf[j:], e)
+				i += n
+				continue
+			}
+		}
+		buf[j] = b
+		j++
+		i++
+	}
+
+	return string(buf[:j])
+}
+
+func normalizeInlineCode(s string) string {
+	if s == "" {
+		return ""
+	}
+
+	byteFuckery := false
+	i := 0
+	for i < len(s)-1 {
+		b := s[i]
+		if b == '\n' {
+			byteFuckery = true
+			break
+		}
+		if b == ' ' {
+			i++
+			b = s[i]
+			if b == ' ' || b == '\n' {
+				i--
+				byteFuckery = true
+				break
+			}
+		}
+		i++
+	}
+
+	if !byteFuckery {
+		return strings.TrimSpace(s)
+	}
+
+	buf := make([]byte, len(s))
+	copy(buf[:i], s)
+	buf[i] = ' '
+	i++
+	j := i
+	lastSpace := true
+	for i < len(s) {
+		b := s[i]
+		switch b {
+		case ' ', '\n':
+			if lastSpace {
+				break
+			}
+
+			buf[j] = ' '
+			lastSpace = true
+			j++
+		default:
+			buf[j] = b
+			lastSpace = false
+			j++
+		}
+
+		i++
+	}
+
+	return string(bytes.TrimSpace(buf[:j]))
+}
+
+func normalizeReference(s string) string {
+	var buf bytes.Buffer
+	lastSpace := false
+	for _, r := range s {
+		if unicode.IsSpace(r) {
+			if !lastSpace {
+				buf.WriteByte(' ')
+				lastSpace = true
+			}
+			continue
+		}
+
+		buf.WriteRune(unicode.ToLower(r))
+		lastSpace = false
+	}
+
+	return string(bytes.TrimSpace(buf.Bytes()))
+}
diff --git a/vendor/gitlab.com/golang-commonmark/markdown/writer.go b/vendor/gitlab.com/golang-commonmark/markdown/writer.go
new file mode 100644
index 00000000..886a1de4
--- /dev/null
+++ b/vendor/gitlab.com/golang-commonmark/markdown/writer.go
@@ -0,0 +1,69 @@
+// Copyright 2015 The Authors. All rights reserved.
+// Use of this source code is governed by a BSD-style
+// license that can be found in the LICENSE file.
+
+package markdown
+
+import (
+	"bufio"
+	"io"
+)
+
+type writer interface {
+	Write([]byte) (int, error)
+	WriteByte(byte) error
+	WriteString(string) (int, error)
+	Flush() error
+}
+
+type monadicWriter struct {
+	writer
+	err error
+}
+
+func newMonadicWriter(w io.Writer) *monadicWriter {
+	if w, ok := w.(writer); ok {
+		return &monadicWriter{writer: w}
+	}
+	return &monadicWriter{writer: bufio.NewWriter(w)}
+}
+
+func (w *monadicWriter) Write(p []byte) (n int, err error) {
+	if w.err != nil {
+		return
+	}
+
+	n, err = w.writer.Write(p)
+	w.err = err
+	return
+}
+
+func (w *monadicWriter) WriteByte(b byte) (err error) {
+	if w.err != nil {
+		return
+	}
+
+	err = w.writer.WriteByte(b)
+	w.err = err
+	return
+}
+
+func (w *monadicWriter) WriteString(s string) (n int, err error) {
+	if w.err != nil {
+		return
+	}
+
+	n, err = w.writer.WriteString(s)
+	w.err = err
+	return
+}
+
+func (w *monadicWriter) Flush() (err error) {
+	if w.err != nil {
+		return
+	}
+
+	err = w.writer.Flush()
+	w.err = err
+	return
+}