diff --git a/core/api/convert/md/md.go b/core/api/convert/md/md.go
index 1c83de4a..eaad1bed 100644
--- a/core/api/convert/md/md.go
+++ b/core/api/convert/md/md.go
@@ -24,5 +24,5 @@ import (
// PagesHTML set to the given (*api.DocumentConversionRequest).Filedata converted by the blackfriday lib.
func Convert(ctx context.Context, in interface{}) (interface{}, error) {
return &api.DocumentConversionResponse{
- PagesHTML: blackfriday.MarkdownCommon(in.(*api.DocumentConversionRequest).Filedata)}, nil
+ PagesHTML: blackfriday.Run(in.(*api.DocumentConversionRequest).Filedata)}, nil
}
diff --git a/domain/section/markdown/markdown.go b/domain/section/markdown/markdown.go
index bf7539da..ef45ab5b 100644
--- a/domain/section/markdown/markdown.go
+++ b/domain/section/markdown/markdown.go
@@ -47,7 +47,7 @@ func (*Provider) Command(ctx *provider.Context, w http.ResponseWriter, r *http.R
// Render converts markdown data into HTML suitable for browser rendering.
func (*Provider) Render(ctx *provider.Context, config, data string) string {
- result := blackfriday.MarkdownCommon([]byte(data))
+ result := blackfriday.Run([]byte(data))
return string(result)
}
diff --git a/vendor/github.com/documize/blackfriday/README.md b/vendor/github.com/documize/blackfriday/README.md
index 52e3b25b..2e0db355 100644
--- a/vendor/github.com/documize/blackfriday/README.md
+++ b/vendor/github.com/documize/blackfriday/README.md
@@ -8,7 +8,7 @@ punctuation substitutions, etc.), and it is safe for all utf-8
(unicode) input.
HTML output is currently supported, along with Smartypants
-extensions. An experimental LaTeX output engine is also included.
+extensions.
It started as a translation from C of [Sundown][3].
@@ -16,63 +16,87 @@ It started as a translation from C of [Sundown][3].
Installation
------------
-Blackfriday is compatible with Go 1. If you are using an older
-release of Go, consider using v1.1 of blackfriday, which was based
-on the last stable release of Go prior to Go 1. You can find it as a
-tagged commit on github.
+Blackfriday is compatible with any modern Go release. With Go 1.7 and git
+installed:
-With Go 1 and git installed:
-
- go get github.com/russross/blackfriday
+ go get gopkg.in/russross/blackfriday.v2
will download, compile, and install the package into your `$GOPATH`
directory hierarchy. Alternatively, you can achieve the same if you
import it into a project:
- import "github.com/russross/blackfriday"
+ import "gopkg.in/russross/blackfriday.v2"
and `go get` without parameters.
+
+Versions
+--------
+
+Currently maintained and recommended version of Blackfriday is `v2`. It's being
+developed on its own branch: https://github.com/russross/blackfriday/v2. You
+should install and import it via [gopkg.in][6] at
+`gopkg.in/russross/blackfriday.v2`.
+
+Version 2 offers a number of improvements over v1:
+
+* Cleaned up API
+* A separate call to [`Parse`][4], which produces an abstract syntax tree for
+ the document
+* Latest bug fixes
+* Flexibility to easily add your own rendering extensions
+
+Potential drawbacks:
+
+* Our benchmarks show v2 to be slightly slower than v1. Currently in the
+ ballpark of around 15%.
+* API breakage. If you can't afford modifying your code to adhere to the new API
+ and don't care too much about the new features, v2 is probably not for you.
+* Several bug fixes are trailing behind and still need to be forward-ported to
+ v2. See issue [#348](https://github.com/russross/blackfriday/issues/348) for
+ tracking.
+
Usage
-----
-For basic usage, it is as simple as getting your input into a byte
-slice and calling:
+For the most sensible markdown processing, it is as simple as getting your input
+into a byte slice and calling:
- output := blackfriday.MarkdownBasic(input)
+```go
+output := blackfriday.Run(input)
+```
-This renders it with no extensions enabled. To get a more useful
-feature set, use this instead:
+Your input will be parsed and the output rendered with a set of most popular
+extensions enabled. If you want the most basic feature set, corresponding with
+the bare Markdown specification, use:
- output := blackfriday.MarkdownCommon(input)
+```go
+output := blackfriday.Run(input, blackfriday.WithNoExtensions())
+```
### Sanitize untrusted content
Blackfriday itself does nothing to protect against malicious content. If you are
-dealing with user-supplied markdown, we recommend running blackfriday's output
-through HTML sanitizer such as
-[Bluemonday](https://github.com/microcosm-cc/bluemonday).
+dealing with user-supplied markdown, we recommend running Blackfriday's output
+through HTML sanitizer such as [Bluemonday][5].
-Here's an example of simple usage of blackfriday together with bluemonday:
+Here's an example of simple usage of Blackfriday together with Bluemonday:
-``` go
+```go
import (
"github.com/microcosm-cc/bluemonday"
"github.com/russross/blackfriday"
)
// ...
-unsafe := blackfriday.MarkdownCommon(input)
+unsafe := blackfriday.Run(input)
html := bluemonday.UGCPolicy().SanitizeBytes(unsafe)
```
### Custom options
-If you want to customize the set of options, first get a renderer
-(currently either the HTML or LaTeX output engines), then use it to
-call the more general `Markdown` function. For examples, see the
-implementations of `MarkdownBasic` and `MarkdownCommon` in
-`markdown.go`.
+If you want to customize the set of options, use `blackfriday.WithExtensions`,
+`blackfriday.WithRenderer` and `blackfriday.WithRefOverride`.
You can also check out `blackfriday-tool` for a more complete example
of how to use it. Download and install it using:
@@ -114,7 +138,7 @@ All features of Sundown are supported, including:
know and send me the input that does it.
NOTE: "safety" in this context means *runtime safety only*. In order to
- protect yourself agains JavaScript injection in untrusted content, see
+ protect yourself against JavaScript injection in untrusted content, see
[this example](https://github.com/russross/blackfriday#sanitize-untrusted-content).
* **Fast processing**. It is fast enough to render on-demand in
@@ -160,7 +184,7 @@ implements the following extensions:
and supply a language (to make syntax highlighting simple). Just
mark it like this:
- ``` go
+ ```go
func getTrue() bool {
return true
}
@@ -169,16 +193,33 @@ implements the following extensions:
You can use 3 or more backticks to mark the beginning of the
block, and the same number to mark the end of the block.
+* **Definition lists**. A simple definition list is made of a single-line
+ term followed by a colon and the definition for that term.
+
+ Cat
+ : Fluffy animal everyone likes
+
+ Internet
+ : Vector of transmission for pictures of cats
+
+ Terms must be separated from the previous definition by a blank line.
+
+* **Footnotes**. A marker in the text that will become a superscript number;
+ a footnote definition that will be placed in a list of footnotes at the
+ end of the document. A footnote looks like this:
+
+ This is a footnote.[^1]
+
+ [^1]: the footnote text.
+
* **Autolinking**. Blackfriday can find URLs that have not been
explicitly marked as links and turn them into links.
* **Strikethrough**. Use two tildes (`~~`) to mark text that
should be crossed out.
-* **Hard line breaks**. With this extension enabled (it is off by
- default in the `MarkdownBasic` and `MarkdownCommon` convenience
- functions), newlines in the input translate into line breaks in
- the output.
+* **Hard line breaks**. With this extension enabled newlines in the input
+ translate into line breaks in the output. This extension is off by default.
* **Smart quotes**. Smartypants-style punctuation substitution is
supported, turning normal double- and single-quote marks into
@@ -205,7 +246,7 @@ are a few of note:
* [github_flavored_markdown](https://godoc.org/github.com/shurcooL/github_flavored_markdown):
provides a GitHub Flavored Markdown renderer with fenced code block
- highlighting, clickable header anchor links.
+ highlighting, clickable heading anchor links.
It's not customizable, and its goal is to produce HTML output
equivalent to the [GitHub Markdown API endpoint](https://developer.github.com/v3/markdown/#render-a-markdown-document-in-raw-mode),
@@ -214,15 +255,8 @@ are a few of note:
* [markdownfmt](https://github.com/shurcooL/markdownfmt): like gofmt,
but for markdown.
-* LaTeX output: renders output as LaTeX. This is currently part of the
- main Blackfriday repository, but may be split into its own project
- in the future. If you are interested in owning and maintaining the
- LaTeX output component, please be in touch.
-
- It renders some basic documents, but is only experimental at this
- point. In particular, it does not do any inline escaping, so input
- that happens to look like LaTeX code will be passed through without
- modification.
+* [LaTeX output](https://bitbucket.org/ambrevar/blackfriday-latex):
+ renders output as LaTeX.
Todo
@@ -241,6 +275,9 @@ License
[Blackfriday is distributed under the Simplified BSD License](LICENSE.txt)
- [1]: http://daringfireball.net/projects/markdown/ "Markdown"
- [2]: http://golang.org/ "Go Language"
+ [1]: https://daringfireball.net/projects/markdown/ "Markdown"
+ [2]: https://golang.org/ "Go Language"
[3]: https://github.com/vmg/sundown "Sundown"
+ [4]: https://godoc.org/gopkg.in/russross/blackfriday.v2#Parse "Parse func"
+ [5]: https://github.com/microcosm-cc/bluemonday "Bluemonday"
+ [6]: https://labix.org/gopkg.in "gopkg.in"
diff --git a/vendor/github.com/documize/blackfriday/block.go b/vendor/github.com/documize/blackfriday/block.go
index 1f300b33..d7da33f2 100644
--- a/vendor/github.com/documize/blackfriday/block.go
+++ b/vendor/github.com/documize/blackfriday/block.go
@@ -15,18 +15,26 @@ package blackfriday
import (
"bytes"
+ "html"
+ "regexp"
"github.com/shurcooL/sanitized_anchor_name"
)
+const (
+ charEntity = "&(?:#x[a-f0-9]{1,8}|#[0-9]{1,8}|[a-z][a-z0-9]{1,31});"
+ escapable = "[!\"#$%&'()*+,./:;<=>?@[\\\\\\]^_`{|}~-]"
+)
+
+var (
+ reBackslashOrAmp = regexp.MustCompile("[\\&]")
+ reEntityOrEscapedChar = regexp.MustCompile("(?i)\\\\" + escapable + "|" + charEntity)
+)
+
// Parse block-level data.
// Note: this function and many that it calls assume that
// the input buffer ends with a newline.
-func (p *parser) block(out *bytes.Buffer, data []byte) {
- if len(data) == 0 || data[len(data)-1] != '\n' {
- panic("block input is missing terminating newline")
- }
-
+func (p *Markdown) block(data []byte) {
// this is called recursively: enforce a maximum depth
if p.nesting >= p.maxNesting {
return
@@ -35,14 +43,14 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// parse out one block-level construct at a time
for len(data) > 0 {
- // prefixed header:
+ // prefixed heading:
//
- // # Header 1
- // ## Header 2
+ // # Heading 1
+ // ## Heading 2
// ...
- // ###### Header 6
- if p.isPrefixHeader(data) {
- data = data[p.prefixHeader(out, data):]
+ // ###### Heading 6
+ if p.isPrefixHeading(data) {
+ data = data[p.prefixHeading(data):]
continue
}
@@ -52,7 +60,7 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// ...
//
if data[0] == '<' {
- if i := p.html(out, data, true); i > 0 {
+ if i := p.html(data, true); i > 0 {
data = data[i:]
continue
}
@@ -63,9 +71,9 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// % stuff
// % more stuff
// % even more stuff
- if p.flags&EXTENSION_TITLEBLOCK != 0 {
+ if p.extensions&Titleblock != 0 {
if data[0] == '%' {
- if i := p.titleBlock(out, data, true); i > 0 {
+ if i := p.titleBlock(data, true); i > 0 {
data = data[i:]
continue
}
@@ -87,7 +95,7 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// return b
// }
if p.codePrefix(data) > 0 {
- data = data[p.code(out, data):]
+ data = data[p.code(data):]
continue
}
@@ -101,8 +109,8 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// return n * fact(n-1)
// }
// ```
- if p.flags&EXTENSION_FENCED_CODE != 0 {
- if i := p.fencedCode(out, data, true); i > 0 {
+ if p.extensions&FencedCode != 0 {
+ if i := p.fencedCodeBlock(data, true); i > 0 {
data = data[i:]
continue
}
@@ -116,9 +124,9 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// or
// ______
if p.isHRule(data) {
- p.r.HRule(out)
+ p.addBlock(HorizontalRule, nil)
var i int
- for i = 0; data[i] != '\n'; i++ {
+ for i = 0; i < len(data) && data[i] != '\n'; i++ {
}
data = data[i:]
continue
@@ -129,7 +137,7 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// > A big quote I found somewhere
// > on the web
if p.quotePrefix(data) > 0 {
- data = data[p.quote(out, data):]
+ data = data[p.quote(data):]
continue
}
@@ -139,8 +147,8 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// ------|-----|---------
// Bob | 31 | 555-1234
// Alice | 27 | 555-4321
- if p.flags&EXTENSION_TABLES != 0 {
- if i := p.table(out, data); i > 0 {
+ if p.extensions&Tables != 0 {
+ if i := p.table(data); i > 0 {
data = data[i:]
continue
}
@@ -153,7 +161,7 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
//
// also works with + or -
if p.uliPrefix(data) > 0 {
- data = data[p.list(out, data, 0):]
+ data = data[p.list(data, 0):]
continue
}
@@ -162,7 +170,7 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
// 1. Item 1
// 2. Item 2
if p.oliPrefix(data) > 0 {
- data = data[p.list(out, data, LIST_TYPE_ORDERED):]
+ data = data[p.list(data, ListTypeOrdered):]
continue
}
@@ -174,55 +182,62 @@ func (p *parser) block(out *bytes.Buffer, data []byte) {
//
// Term 2
// : Definition c
- if p.flags&EXTENSION_DEFINITION_LISTS != 0 {
+ if p.extensions&DefinitionLists != 0 {
if p.dliPrefix(data) > 0 {
- data = data[p.list(out, data, LIST_TYPE_DEFINITION):]
+ data = data[p.list(data, ListTypeDefinition):]
continue
}
}
// anything else must look like a normal paragraph
- // note: this finds underlined headers, too
- data = data[p.paragraph(out, data):]
+ // note: this finds underlined headings, too
+ data = data[p.paragraph(data):]
}
p.nesting--
}
-func (p *parser) isPrefixHeader(data []byte) bool {
+func (p *Markdown) addBlock(typ NodeType, content []byte) *Node {
+ p.closeUnmatchedBlocks()
+ container := p.addChild(typ, 0)
+ container.content = content
+ return container
+}
+
+func (p *Markdown) isPrefixHeading(data []byte) bool {
if data[0] != '#' {
return false
}
- if p.flags&EXTENSION_SPACE_HEADERS != 0 {
+ if p.extensions&SpaceHeadings != 0 {
level := 0
- for level < 6 && data[level] == '#' {
+ for level < 6 && level < len(data) && data[level] == '#' {
level++
}
- if data[level] != ' ' {
+ if level == len(data) || data[level] != ' ' {
return false
}
}
return true
}
-func (p *parser) prefixHeader(out *bytes.Buffer, data []byte) int {
+func (p *Markdown) prefixHeading(data []byte) int {
level := 0
- for level < 6 && data[level] == '#' {
+ for level < 6 && level < len(data) && data[level] == '#' {
level++
}
i := skipChar(data, level, ' ')
end := skipUntilChar(data, i, '\n')
skip := end
id := ""
- if p.flags&EXTENSION_HEADER_IDS != 0 {
+ if p.extensions&HeadingIDs != 0 {
j, k := 0, 0
- // find start/end of header id
+ // find start/end of heading id
for j = i; j < end-1 && (data[j] != '{' || data[j+1] != '#'); j++ {
}
for k = j + 1; k < end && data[k] != '}'; k++ {
}
- // extract header id iff found
+ // extract heading id iff found
if j < end && k < end {
id = string(data[j+2 : k])
end = j
@@ -242,45 +257,41 @@ func (p *parser) prefixHeader(out *bytes.Buffer, data []byte) int {
end--
}
if end > i {
- if id == "" && p.flags&EXTENSION_AUTO_HEADER_IDS != 0 {
+ if id == "" && p.extensions&AutoHeadingIDs != 0 {
id = sanitized_anchor_name.Create(string(data[i:end]))
}
- work := func() bool {
- p.inline(out, data[i:end])
- return true
- }
- p.r.Header(out, work, level, id)
+ block := p.addBlock(Heading, data[i:end])
+ block.HeadingID = id
+ block.Level = level
}
return skip
}
-func (p *parser) isUnderlinedHeader(data []byte) int {
- // test of level 1 header
+func (p *Markdown) isUnderlinedHeading(data []byte) int {
+ // test of level 1 heading
if data[0] == '=' {
i := skipChar(data, 1, '=')
i = skipChar(data, i, ' ')
- if data[i] == '\n' {
+ if i < len(data) && data[i] == '\n' {
return 1
- } else {
- return 0
}
+ return 0
}
- // test of level 2 header
+ // test of level 2 heading
if data[0] == '-' {
i := skipChar(data, 1, '-')
i = skipChar(data, i, ' ')
- if data[i] == '\n' {
+ if i < len(data) && data[i] == '\n' {
return 2
- } else {
- return 0
}
+ return 0
}
return 0
}
-func (p *parser) titleBlock(out *bytes.Buffer, data []byte, doRender bool) int {
+func (p *Markdown) titleBlock(data []byte, doRender bool) int {
if data[0] != '%' {
return 0
}
@@ -294,12 +305,17 @@ func (p *parser) titleBlock(out *bytes.Buffer, data []byte, doRender bool) int {
}
data = bytes.Join(splitData[0:i], []byte("\n"))
- p.r.TitleBlock(out, data)
+ consumed := len(data)
+ data = bytes.TrimPrefix(data, []byte("% "))
+ data = bytes.Replace(data, []byte("\n% "), []byte("\n"), -1)
+ block := p.addBlock(Heading, data)
+ block.Level = 1
+ block.IsTitleblock = true
- return len(data)
+ return consumed
}
-func (p *parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
+func (p *Markdown) html(data []byte, doRender bool) int {
var i, j int
// identify the opening tag
@@ -311,12 +327,12 @@ func (p *parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
// handle special cases
if !tagfound {
// check for an HTML comment
- if size := p.htmlComment(out, data, doRender); size > 0 {
+ if size := p.htmlComment(data, doRender); size > 0 {
return size
}
// check for an tag
- if size := p.htmlHr(out, data, doRender); size > 0 {
+ if size := p.htmlHr(data, doRender); size > 0 {
return size
}
@@ -391,31 +407,20 @@ func (p *parser) html(out *bytes.Buffer, data []byte, doRender bool) int {
for end > 0 && data[end-1] == '\n' {
end--
}
- p.r.BlockHtml(out, data[:end])
+ finalizeHTMLBlock(p.addBlock(HTMLBlock, data[:end]))
}
return i
}
+func finalizeHTMLBlock(block *Node) {
+ block.Literal = block.content
+ block.content = nil
+}
+
// HTML comment, lax form
-func (p *parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool) int {
- if data[0] != '<' || data[1] != '!' || data[2] != '-' || data[3] != '-' {
- return 0
- }
-
- i := 5
-
- // scan for an end-of-comment marker, across lines if necessary
- for i < len(data) && !(data[i-2] == '-' && data[i-1] == '-' && data[i] == '>') {
- i++
- }
- i++
-
- // no end-of-comment marker
- if i >= len(data) {
- return 0
- }
-
+func (p *Markdown) htmlComment(data []byte, doRender bool) int {
+ i := p.inlineHTMLComment(data)
// needs to end with a blank line
if j := p.isEmpty(data[i:]); j > 0 {
size := i + j
@@ -425,16 +430,19 @@ func (p *parser) htmlComment(out *bytes.Buffer, data []byte, doRender bool) int
for end > 0 && data[end-1] == '\n' {
end--
}
- p.r.BlockHtml(out, data[:end])
+ block := p.addBlock(HTMLBlock, data[:end])
+ finalizeHTMLBlock(block)
}
return size
}
-
return 0
}
// HR, which is the only self-closing block tag considered
-func (p *parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
+func (p *Markdown) htmlHr(data []byte, doRender bool) int {
+ if len(data) < 4 {
+ return 0
+ }
if data[0] != '<' || (data[1] != 'h' && data[1] != 'H') || (data[2] != 'r' && data[2] != 'R') {
return 0
}
@@ -442,13 +450,11 @@ func (p *parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
// not an tag after all; at least not a valid one
return 0
}
-
i := 3
- for data[i] != '>' && data[i] != '\n' {
+ for i < len(data) && data[i] != '>' && data[i] != '\n' {
i++
}
-
- if data[i] == '>' {
+ if i < len(data) && data[i] == '>' {
i++
if j := p.isEmpty(data[i:]); j > 0 {
size := i + j
@@ -458,30 +464,31 @@ func (p *parser) htmlHr(out *bytes.Buffer, data []byte, doRender bool) int {
for end > 0 && data[end-1] == '\n' {
end--
}
- p.r.BlockHtml(out, data[:end])
+ finalizeHTMLBlock(p.addBlock(HTMLBlock, data[:end]))
}
return size
}
}
-
return 0
}
-func (p *parser) htmlFindTag(data []byte) (string, bool) {
+func (p *Markdown) htmlFindTag(data []byte) (string, bool) {
i := 0
- for isalnum(data[i]) {
+ for i < len(data) && isalnum(data[i]) {
i++
}
key := string(data[:i])
- if blockTags[key] {
+ if _, ok := blockTags[key]; ok {
return key, true
}
return "", false
}
-func (p *parser) htmlFindEnd(tag string, data []byte) int {
+func (p *Markdown) htmlFindEnd(tag string, data []byte) int {
// assume data[0] == '<' && data[1] == '/' already tested
-
+ if tag == "hr" {
+ return 2
+ }
// check if tag is a match
closetag := []byte("")
if !bytes.HasPrefix(data, closetag) {
@@ -501,7 +508,7 @@ func (p *parser) htmlFindEnd(tag string, data []byte) int {
return i
}
- if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
+ if p.extensions&LaxHTMLBlocks != 0 {
return i
}
if skip = p.isEmpty(data[i:]); skip == 0 {
@@ -512,7 +519,7 @@ func (p *parser) htmlFindEnd(tag string, data []byte) int {
return i + skip
}
-func (p *parser) isEmpty(data []byte) int {
+func (*Markdown) isEmpty(data []byte) int {
// it is okay to call isEmpty on an empty buffer
if len(data) == 0 {
return 0
@@ -524,10 +531,13 @@ func (p *parser) isEmpty(data []byte) int {
return 0
}
}
- return i + 1
+ if i < len(data) && data[i] == '\n' {
+ i++
+ }
+ return i
}
-func (p *parser) isHRule(data []byte) bool {
+func (*Markdown) isHRule(data []byte) bool {
i := 0
// skip up to three spaces
@@ -543,7 +553,7 @@ func (p *parser) isHRule(data []byte) bool {
// the whole line must be the char or whitespace
n := 0
- for data[i] != '\n' {
+ for i < len(data) && data[i] != '\n' {
switch {
case data[i] == c:
n++
@@ -556,21 +566,23 @@ func (p *parser) isHRule(data []byte) bool {
return n >= 3
}
-func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (skip int, marker string) {
+// isFenceLine checks if there's a fence line (e.g., ``` or ``` go) at the beginning of data,
+// and returns the end index if so, or 0 otherwise. It also returns the marker found.
+// If syntax is not nil, it gets set to the syntax specified in the fence line.
+func isFenceLine(data []byte, syntax *string, oldmarker string) (end int, marker string) {
i, size := 0, 0
- skip = 0
// skip up to three spaces
for i < len(data) && i < 3 && data[i] == ' ' {
i++
}
- if i >= len(data) {
- return
- }
// check for the marker characters: ~ or `
+ if i >= len(data) {
+ return 0, ""
+ }
if data[i] != '~' && data[i] != '`' {
- return
+ return 0, ""
}
c := data[i]
@@ -581,27 +593,28 @@ func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (s
i++
}
- if i >= len(data) {
- return
- }
-
// the marker char must occur at least 3 times
if size < 3 {
- return
+ return 0, ""
}
marker = string(data[i-size : i])
// if this is the end marker, it must match the beginning marker
if oldmarker != "" && marker != oldmarker {
- return
+ return 0, ""
}
+ // TODO(shurcooL): It's probably a good idea to simplify the 2 code paths here
+ // into one, always get the syntax, and discard it if the caller doesn't care.
if syntax != nil {
syn := 0
i = skipChar(data, i, ' ')
if i >= len(data) {
- return
+ if i == len(data) {
+ return i, marker
+ }
+ return 0, ""
}
syntaxStart := i
@@ -616,7 +629,7 @@ func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (s
}
if i >= len(data) || data[i] != '}' {
- return
+ return 0, ""
}
// strip all whitespace at the beginning and the end
@@ -638,33 +651,38 @@ func (p *parser) isFencedCode(data []byte, syntax **string, oldmarker string) (s
}
}
- language := string(data[syntaxStart : syntaxStart+syn])
- *syntax = &language
+ *syntax = string(data[syntaxStart : syntaxStart+syn])
}
i = skipChar(data, i, ' ')
if i >= len(data) || data[i] != '\n' {
- return
+ if i == len(data) {
+ return i, marker
+ }
+ return 0, ""
}
-
- skip = i + 1
- return
+ return i + 1, marker // Take newline into account.
}
-func (p *parser) fencedCode(out *bytes.Buffer, data []byte, doRender bool) int {
- var lang *string
- beg, marker := p.isFencedCode(data, &lang, "")
+// fencedCodeBlock returns the end index if data contains a fenced code block at the beginning,
+// or 0 otherwise. It writes to out if doRender is true, otherwise it has no side effects.
+// If doRender is true, a final newline is mandatory to recognize the fenced code block.
+func (p *Markdown) fencedCodeBlock(data []byte, doRender bool) int {
+ var syntax string
+ beg, marker := isFenceLine(data, &syntax, "")
if beg == 0 || beg >= len(data) {
return 0
}
var work bytes.Buffer
+ work.Write([]byte(syntax))
+ work.WriteByte('\n')
for {
// safe to assume beg < len(data)
// check for the end of the code block
- fenceEnd, _ := p.isFencedCode(data[beg:], nil, marker)
+ fenceEnd, _ := isFenceLine(data[beg:], nil, marker)
if fenceEnd != 0 {
beg += fenceEnd
break
@@ -685,30 +703,56 @@ func (p *parser) fencedCode(out *bytes.Buffer, data []byte, doRender bool) int {
beg = end
}
- syntax := ""
- if lang != nil {
- syntax = *lang
- }
-
if doRender {
- p.r.BlockCode(out, work.Bytes(), syntax)
+ block := p.addBlock(CodeBlock, work.Bytes()) // TODO: get rid of temp buffer
+ block.IsFenced = true
+ finalizeCodeBlock(block)
}
return beg
}
-func (p *parser) table(out *bytes.Buffer, data []byte) int {
- var header bytes.Buffer
- i, columns := p.tableHeader(&header, data)
+func unescapeChar(str []byte) []byte {
+ if str[0] == '\\' {
+ return []byte{str[1]}
+ }
+ return []byte(html.UnescapeString(string(str)))
+}
+
+func unescapeString(str []byte) []byte {
+ if reBackslashOrAmp.Match(str) {
+ return reEntityOrEscapedChar.ReplaceAllFunc(str, unescapeChar)
+ }
+ return str
+}
+
+func finalizeCodeBlock(block *Node) {
+ if block.IsFenced {
+ newlinePos := bytes.IndexByte(block.content, '\n')
+ firstLine := block.content[:newlinePos]
+ rest := block.content[newlinePos+1:]
+ block.Info = unescapeString(bytes.Trim(firstLine, "\n"))
+ block.Literal = rest
+ } else {
+ block.Literal = block.content
+ }
+ block.content = nil
+}
+
+func (p *Markdown) table(data []byte) int {
+ table := p.addBlock(Table, nil)
+ i, columns := p.tableHeader(data)
if i == 0 {
+ p.tip = table.Parent
+ table.Unlink()
return 0
}
- var body bytes.Buffer
+ p.addBlock(TableBody, nil)
for i < len(data) {
pipes, rowStart := 0, i
- for ; data[i] != '\n'; i++ {
+ for ; i < len(data) && data[i] != '\n'; i++ {
if data[i] == '|' {
pipes++
}
@@ -720,12 +764,12 @@ func (p *parser) table(out *bytes.Buffer, data []byte) int {
}
// include the newline in data sent to tableRow
- i++
- p.tableRow(&body, data[rowStart:i], columns, false)
+ if i < len(data) && data[i] == '\n' {
+ i++
+ }
+ p.tableRow(data[rowStart:i], columns, false)
}
- p.r.Table(out, header.Bytes(), body.Bytes(), columns)
-
return i
}
@@ -738,10 +782,10 @@ func isBackslashEscaped(data []byte, i int) bool {
return backslashes&1 == 1
}
-func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns []int) {
+func (p *Markdown) tableHeader(data []byte) (size int, columns []CellAlignFlags) {
i := 0
colCount := 1
- for i = 0; data[i] != '\n'; i++ {
+ for i = 0; i < len(data) && data[i] != '\n'; i++ {
if data[i] == '|' && !isBackslashEscaped(data, i) {
colCount++
}
@@ -753,7 +797,11 @@ func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns
}
// include the newline in the data sent to tableRow
- header := data[:i+1]
+ j := i
+ if j < len(data) && data[j] == '\n' {
+ j++
+ }
+ header := data[:j]
// column count ignores pipes at beginning or end of line
if data[0] == '|' {
@@ -763,7 +811,7 @@ func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns
colCount--
}
- columns = make([]int, colCount)
+ columns = make([]CellAlignFlags, colCount)
// move on to the header underline
i++
@@ -779,27 +827,29 @@ func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns
// each column header is of form: / *:?-+:? *|/ with # dashes + # colons >= 3
// and trailing | optional on last column
col := 0
- for data[i] != '\n' {
+ for i < len(data) && data[i] != '\n' {
dashes := 0
if data[i] == ':' {
i++
- columns[col] |= TABLE_ALIGNMENT_LEFT
+ columns[col] |= TableAlignmentLeft
dashes++
}
- for data[i] == '-' {
+ for i < len(data) && data[i] == '-' {
i++
dashes++
}
- if data[i] == ':' {
+ if i < len(data) && data[i] == ':' {
i++
- columns[col] |= TABLE_ALIGNMENT_RIGHT
+ columns[col] |= TableAlignmentRight
dashes++
}
- for data[i] == ' ' {
+ for i < len(data) && data[i] == ' ' {
i++
}
-
+ if i == len(data) {
+ return
+ }
// end of column test is messy
switch {
case dashes < 3:
@@ -810,12 +860,12 @@ func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns
// marker found, now skip past trailing whitespace
col++
i++
- for data[i] == ' ' {
+ for i < len(data) && data[i] == ' ' {
i++
}
// trailing junk found after last column
- if col >= colCount && data[i] != '\n' {
+ if col >= colCount && i < len(data) && data[i] != '\n' {
return
}
@@ -836,27 +886,31 @@ func (p *parser) tableHeader(out *bytes.Buffer, data []byte) (size int, columns
return
}
- p.tableRow(out, header, columns, true)
- size = i + 1
+ p.addBlock(TableHead, nil)
+ p.tableRow(header, columns, true)
+ size = i
+ if size < len(data) && data[size] == '\n' {
+ size++
+ }
return
}
-func (p *parser) tableRow(out *bytes.Buffer, data []byte, columns []int, header bool) {
+func (p *Markdown) tableRow(data []byte, columns []CellAlignFlags, header bool) {
+ p.addBlock(TableRow, nil)
i, col := 0, 0
- var rowWork bytes.Buffer
if data[i] == '|' && !isBackslashEscaped(data, i) {
i++
}
for col = 0; col < len(columns) && i < len(data); col++ {
- for data[i] == ' ' {
+ for i < len(data) && data[i] == ' ' {
i++
}
cellStart := i
- for (data[i] != '|' || isBackslashEscaped(data, i)) && data[i] != '\n' {
+ for i < len(data) && (data[i] != '|' || isBackslashEscaped(data, i)) && data[i] != '\n' {
i++
}
@@ -865,42 +919,33 @@ func (p *parser) tableRow(out *bytes.Buffer, data []byte, columns []int, header
// skip the end-of-cell marker, possibly taking us past end of buffer
i++
- for cellEnd > cellStart && data[cellEnd-1] == ' ' {
+ for cellEnd > cellStart && cellEnd-1 < len(data) && data[cellEnd-1] == ' ' {
cellEnd--
}
- var cellWork bytes.Buffer
- p.inline(&cellWork, data[cellStart:cellEnd])
-
- if header {
- p.r.TableHeaderCell(&rowWork, cellWork.Bytes(), columns[col])
- } else {
- p.r.TableCell(&rowWork, cellWork.Bytes(), columns[col])
- }
+ cell := p.addBlock(TableCell, data[cellStart:cellEnd])
+ cell.IsHeader = header
+ cell.Align = columns[col]
}
// pad it out with empty columns to get the right number
for ; col < len(columns); col++ {
- if header {
- p.r.TableHeaderCell(&rowWork, nil, columns[col])
- } else {
- p.r.TableCell(&rowWork, nil, columns[col])
- }
+ cell := p.addBlock(TableCell, nil)
+ cell.IsHeader = header
+ cell.Align = columns[col]
}
// silently ignore rows with too many cells
-
- p.r.TableRow(out, rowWork.Bytes())
}
// returns blockquote prefix length
-func (p *parser) quotePrefix(data []byte) int {
+func (p *Markdown) quotePrefix(data []byte) int {
i := 0
- for i < 3 && data[i] == ' ' {
+ for i < 3 && i < len(data) && data[i] == ' ' {
i++
}
- if data[i] == '>' {
- if data[i+1] == ' ' {
+ if i < len(data) && data[i] == '>' {
+ if i+1 < len(data) && data[i+1] == ' ' {
return i + 2
}
return i + 1
@@ -908,57 +953,79 @@ func (p *parser) quotePrefix(data []byte) int {
return 0
}
+// blockquote ends with at least one blank line
+// followed by something without a blockquote prefix
+func (p *Markdown) terminateBlockquote(data []byte, beg, end int) bool {
+ if p.isEmpty(data[beg:]) <= 0 {
+ return false
+ }
+ if end >= len(data) {
+ return true
+ }
+ return p.quotePrefix(data[end:]) == 0 && p.isEmpty(data[end:]) == 0
+}
+
// parse a blockquote fragment
-func (p *parser) quote(out *bytes.Buffer, data []byte) int {
+func (p *Markdown) quote(data []byte) int {
+ block := p.addBlock(BlockQuote, nil)
var raw bytes.Buffer
beg, end := 0, 0
for beg < len(data) {
end = beg
- for data[end] != '\n' {
+ // Step over whole lines, collecting them. While doing that, check for
+ // fenced code and if one's found, incorporate it altogether,
+ // irregardless of any contents inside it
+ for end < len(data) && data[end] != '\n' {
+ if p.extensions&FencedCode != 0 {
+ if i := p.fencedCodeBlock(data[end:], false); i > 0 {
+ // -1 to compensate for the extra end++ after the loop:
+ end += i - 1
+ break
+ }
+ }
+ end++
+ }
+ if end < len(data) && data[end] == '\n' {
end++
}
- end++
-
if pre := p.quotePrefix(data[beg:]); pre > 0 {
// skip the prefix
beg += pre
- } else if p.isEmpty(data[beg:]) > 0 &&
- (end >= len(data) ||
- (p.quotePrefix(data[end:]) == 0 && p.isEmpty(data[end:]) == 0)) {
- // blockquote ends with at least one blank line
- // followed by something without a blockquote prefix
+ } else if p.terminateBlockquote(data, beg, end) {
break
}
-
// this line is part of the blockquote
raw.Write(data[beg:end])
beg = end
}
-
- var cooked bytes.Buffer
- p.block(&cooked, raw.Bytes())
- p.r.BlockQuote(out, cooked.Bytes())
+ p.block(raw.Bytes())
+ p.finalize(block)
return end
}
// returns prefix length for block code
-func (p *parser) codePrefix(data []byte) int {
- if data[0] == ' ' && data[1] == ' ' && data[2] == ' ' && data[3] == ' ' {
+func (p *Markdown) codePrefix(data []byte) int {
+ if len(data) >= 1 && data[0] == '\t' {
+ return 1
+ }
+ if len(data) >= 4 && data[0] == ' ' && data[1] == ' ' && data[2] == ' ' && data[3] == ' ' {
return 4
}
return 0
}
-func (p *parser) code(out *bytes.Buffer, data []byte) int {
+func (p *Markdown) code(data []byte) int {
var work bytes.Buffer
i := 0
for i < len(data) {
beg := i
- for data[i] != '\n' {
+ for i < len(data) && data[i] != '\n' {
+ i++
+ }
+ if i < len(data) && data[i] == '\n' {
i++
}
- i++
blankline := p.isEmpty(data[beg:i]) > 0
if pre := p.codePrefix(data[beg:i]); pre > 0 {
@@ -969,7 +1036,7 @@ func (p *parser) code(out *bytes.Buffer, data []byte) int {
break
}
- // verbatim copy to the working buffeu
+ // verbatim copy to the working buffer
if blankline {
work.WriteByte('\n')
} else {
@@ -989,122 +1056,183 @@ func (p *parser) code(out *bytes.Buffer, data []byte) int {
work.WriteByte('\n')
- p.r.BlockCode(out, work.Bytes(), "")
+ block := p.addBlock(CodeBlock, work.Bytes()) // TODO: get rid of temp buffer
+ block.IsFenced = false
+ finalizeCodeBlock(block)
return i
}
// returns unordered list item prefix
-func (p *parser) uliPrefix(data []byte) int {
+func (p *Markdown) uliPrefix(data []byte) int {
i := 0
-
// start with up to 3 spaces
- for i < 3 && data[i] == ' ' {
+ for i < len(data) && i < 3 && data[i] == ' ' {
i++
}
-
- // need a *, +, or - followed by a space
+ if i >= len(data)-1 {
+ return 0
+ }
+ // need one of {'*', '+', '-'} followed by a space or a tab
if (data[i] != '*' && data[i] != '+' && data[i] != '-') ||
- data[i+1] != ' ' {
+ (data[i+1] != ' ' && data[i+1] != '\t') {
return 0
}
return i + 2
}
// returns ordered list item prefix
-func (p *parser) oliPrefix(data []byte) int {
+func (p *Markdown) oliPrefix(data []byte) int {
i := 0
// start with up to 3 spaces
- for i < 3 && data[i] == ' ' {
+ for i < 3 && i < len(data) && data[i] == ' ' {
i++
}
// count the digits
start := i
- for data[i] >= '0' && data[i] <= '9' {
+ for i < len(data) && data[i] >= '0' && data[i] <= '9' {
i++
}
+ if start == i || i >= len(data)-1 {
+ return 0
+ }
- // we need >= 1 digits followed by a dot and a space
- if start == i || data[i] != '.' || data[i+1] != ' ' {
+ // we need >= 1 digits followed by a dot and a space or a tab
+ if data[i] != '.' || !(data[i+1] == ' ' || data[i+1] == '\t') {
return 0
}
return i + 2
}
// returns definition list item prefix
-func (p *parser) dliPrefix(data []byte) int {
- i := 0
-
- // need a : followed by a spaces
- if data[i] != ':' || data[i+1] != ' ' {
+func (p *Markdown) dliPrefix(data []byte) int {
+ if len(data) < 2 {
return 0
}
- for data[i] == ' ' {
+ i := 0
+ // need a ':' followed by a space or a tab
+ if data[i] != ':' || !(data[i+1] == ' ' || data[i+1] == '\t') {
+ return 0
+ }
+ for i < len(data) && data[i] == ' ' {
i++
}
return i + 2
}
// parse ordered or unordered list block
-func (p *parser) list(out *bytes.Buffer, data []byte, flags int) int {
+func (p *Markdown) list(data []byte, flags ListType) int {
i := 0
- flags |= LIST_ITEM_BEGINNING_OF_LIST
- work := func() bool {
- for i < len(data) {
- skip := p.listItem(out, data[i:], &flags)
- i += skip
+ flags |= ListItemBeginningOfList
+ block := p.addBlock(List, nil)
+ block.ListFlags = flags
+ block.Tight = true
- if skip == 0 || flags&LIST_ITEM_END_OF_LIST != 0 {
- break
- }
- flags &= ^LIST_ITEM_BEGINNING_OF_LIST
+ for i < len(data) {
+ skip := p.listItem(data[i:], &flags)
+ if flags&ListItemContainsBlock != 0 {
+ block.ListData.Tight = false
}
- return true
+ i += skip
+ if skip == 0 || flags&ListItemEndOfList != 0 {
+ break
+ }
+ flags &= ^ListItemBeginningOfList
}
- p.r.List(out, work, flags)
+ above := block.Parent
+ finalizeList(block)
+ p.tip = above
return i
}
+// Returns true if block ends with a blank line, descending if needed
+// into lists and sublists.
+func endsWithBlankLine(block *Node) bool {
+ // TODO: figure this out. Always false now.
+ for block != nil {
+ //if block.lastLineBlank {
+ //return true
+ //}
+ t := block.Type
+ if t == List || t == Item {
+ block = block.LastChild
+ } else {
+ break
+ }
+ }
+ return false
+}
+
+func finalizeList(block *Node) {
+ block.open = false
+ item := block.FirstChild
+ for item != nil {
+ // check for non-final list item ending with blank line:
+ if endsWithBlankLine(item) && item.Next != nil {
+ block.ListData.Tight = false
+ break
+ }
+ // recurse into children of list item, to see if there are spaces
+ // between any of them:
+ subItem := item.FirstChild
+ for subItem != nil {
+ if endsWithBlankLine(subItem) && (item.Next != nil || subItem.Next != nil) {
+ block.ListData.Tight = false
+ break
+ }
+ subItem = subItem.Next
+ }
+ item = item.Next
+ }
+}
+
// Parse a single list item.
// Assumes initial prefix is already removed if this is a sublist.
-func (p *parser) listItem(out *bytes.Buffer, data []byte, flags *int) int {
+func (p *Markdown) listItem(data []byte, flags *ListType) int {
// keep track of the indentation of the first line
itemIndent := 0
- for itemIndent < 3 && data[itemIndent] == ' ' {
- itemIndent++
+ if data[0] == '\t' {
+ itemIndent += 4
+ } else {
+ for itemIndent < 3 && data[itemIndent] == ' ' {
+ itemIndent++
+ }
}
+ var bulletChar byte = '*'
i := p.uliPrefix(data)
if i == 0 {
i = p.oliPrefix(data)
+ } else {
+ bulletChar = data[i-2]
}
if i == 0 {
i = p.dliPrefix(data)
// reset definition term flag
if i > 0 {
- *flags &= ^LIST_TYPE_TERM
+ *flags &= ^ListTypeTerm
}
}
if i == 0 {
- // if in defnition list, set term flag and continue
- if *flags&LIST_TYPE_DEFINITION != 0 {
- *flags |= LIST_TYPE_TERM
+ // if in definition list, set term flag and continue
+ if *flags&ListTypeDefinition != 0 {
+ *flags |= ListTypeTerm
} else {
return 0
}
}
// skip leading whitespace on first line
- for data[i] == ' ' {
+ for i < len(data) && data[i] == ' ' {
i++
}
// find the end of the line
line := i
- for i > 0 && data[i-1] != '\n' {
+ for i > 0 && i < len(data) && data[i-1] != '\n' {
i++
}
@@ -1124,7 +1252,7 @@ gatherlines:
i++
// find the end of this line
- for data[i-1] != '\n' {
+ for i < len(data) && data[i-1] != '\n' {
i++
}
@@ -1138,11 +1266,18 @@ gatherlines:
// calculate the indentation
indent := 0
- for indent < 4 && line+indent < i && data[line+indent] == ' ' {
- indent++
+ indentIndex := 0
+ if data[line] == '\t' {
+ indentIndex++
+ indent += 4
+ } else {
+ for indent < 4 && line+indent < i && data[line+indent] == ' ' {
+ indent++
+ indentIndex++
+ }
}
- chunk := data[line+indent : i]
+ chunk := data[line+indentIndex : i]
// evaluate how this line fits in
switch {
@@ -1152,7 +1287,7 @@ gatherlines:
p.dliPrefix(chunk) > 0:
if containsBlankLine {
- *flags |= LIST_ITEM_CONTAINS_BLOCK
+ *flags |= ListItemContainsBlock
}
// to be a nested list, it must be indented more
@@ -1166,94 +1301,89 @@ gatherlines:
sublist = raw.Len()
}
- // is this a nested prefix header?
- case p.isPrefixHeader(chunk):
- // if the header is not indented, it is not nested in the list
+ // is this a nested prefix heading?
+ case p.isPrefixHeading(chunk):
+ // if the heading is not indented, it is not nested in the list
// and thus ends the list
if containsBlankLine && indent < 4 {
- *flags |= LIST_ITEM_END_OF_LIST
+ *flags |= ListItemEndOfList
break gatherlines
}
- *flags |= LIST_ITEM_CONTAINS_BLOCK
+ *flags |= ListItemContainsBlock
// anything following an empty line is only part
// of this item if it is indented 4 spaces
// (regardless of the indentation of the beginning of the item)
case containsBlankLine && indent < 4:
- if *flags&LIST_TYPE_DEFINITION != 0 && i < len(data)-1 {
+ if *flags&ListTypeDefinition != 0 && i < len(data)-1 {
// is the next item still a part of this list?
next := i
- for data[next] != '\n' {
+ for next < len(data) && data[next] != '\n' {
next++
}
for next < len(data)-1 && data[next] == '\n' {
next++
}
if i < len(data)-1 && data[i] != ':' && data[next] != ':' {
- *flags |= LIST_ITEM_END_OF_LIST
+ *flags |= ListItemEndOfList
}
} else {
- *flags |= LIST_ITEM_END_OF_LIST
+ *flags |= ListItemEndOfList
}
break gatherlines
// a blank line means this should be parsed as a block
case containsBlankLine:
raw.WriteByte('\n')
- *flags |= LIST_ITEM_CONTAINS_BLOCK
+ *flags |= ListItemContainsBlock
}
- // if this line was preceeded by one or more blanks,
+ // if this line was preceded by one or more blanks,
// re-introduce the blank into the buffer
if containsBlankLine {
containsBlankLine = false
raw.WriteByte('\n')
-
}
// add the line into the working buffer without prefix
- raw.Write(data[line+indent : i])
+ raw.Write(data[line+indentIndex : i])
line = i
}
rawBytes := raw.Bytes()
+ block := p.addBlock(Item, nil)
+ block.ListFlags = *flags
+ block.Tight = false
+ block.BulletChar = bulletChar
+ block.Delimiter = '.' // Only '.' is possible in Markdown, but ')' will also be possible in CommonMark
+
// render the contents of the list item
- var cooked bytes.Buffer
- if *flags&LIST_ITEM_CONTAINS_BLOCK != 0 && *flags&LIST_TYPE_TERM == 0 {
+ if *flags&ListItemContainsBlock != 0 && *flags&ListTypeTerm == 0 {
// intermediate render of block item, except for definition term
if sublist > 0 {
- p.block(&cooked, rawBytes[:sublist])
- p.block(&cooked, rawBytes[sublist:])
+ p.block(rawBytes[:sublist])
+ p.block(rawBytes[sublist:])
} else {
- p.block(&cooked, rawBytes)
+ p.block(rawBytes)
}
} else {
// intermediate render of inline item
if sublist > 0 {
- p.inline(&cooked, rawBytes[:sublist])
- p.block(&cooked, rawBytes[sublist:])
+ child := p.addChild(Paragraph, 0)
+ child.content = rawBytes[:sublist]
+ p.block(rawBytes[sublist:])
} else {
- p.inline(&cooked, rawBytes)
+ child := p.addChild(Paragraph, 0)
+ child.content = rawBytes
}
}
-
- // render the actual list item
- cookedBytes := cooked.Bytes()
- parsedEnd := len(cookedBytes)
-
- // strip trailing newlines
- for parsedEnd > 0 && cookedBytes[parsedEnd-1] == '\n' {
- parsedEnd--
- }
- p.r.ListItem(out, cookedBytes[:parsedEnd], *flags)
-
return line
}
// render a single paragraph that has already been parsed out
-func (p *parser) renderParagraph(out *bytes.Buffer, data []byte) {
+func (p *Markdown) renderParagraph(data []byte) {
if len(data) == 0 {
return
}
@@ -1264,27 +1394,29 @@ func (p *parser) renderParagraph(out *bytes.Buffer, data []byte) {
beg++
}
+ end := len(data)
// trim trailing newline
- end := len(data) - 1
+ if data[len(data)-1] == '\n' {
+ end--
+ }
// trim trailing spaces
for end > beg && data[end-1] == ' ' {
end--
}
- work := func() bool {
- p.inline(out, data[beg:end])
- return true
- }
- p.r.Paragraph(out, work)
+ p.addBlock(Paragraph, data[beg:end])
}
-func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
+func (p *Markdown) paragraph(data []byte) int {
// prev: index of 1st char of previous line
// line: index of 1st char of current line
// i: index of cursor/end of current line
var prev, line, i int
-
+ tabSize := TabSizeDefault
+ if p.extensions&TabSizeEight != 0 {
+ tabSize = TabSizeDouble
+ }
// keep going until we find something to mark the end of the paragraph
for i < len(data) {
// mark the beginning of the current line
@@ -1292,24 +1424,32 @@ func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
current := data[i:]
line = i
+ // did we find a reference or a footnote? If so, end a paragraph
+ // preceding it and report that we have consumed up to the end of that
+ // reference:
+ if refEnd := isReference(p, current, tabSize); refEnd > 0 {
+ p.renderParagraph(data[:i])
+ return i + refEnd
+ }
+
// did we find a blank line marking the end of the paragraph?
if n := p.isEmpty(current); n > 0 {
// did this blank line followed by a definition list item?
- if p.flags&EXTENSION_DEFINITION_LISTS != 0 {
+ if p.extensions&DefinitionLists != 0 {
if i < len(data)-1 && data[i+1] == ':' {
- return p.list(out, data[prev:], LIST_TYPE_DEFINITION)
+ return p.list(data[prev:], ListTypeDefinition)
}
}
- p.renderParagraph(out, data[:i])
+ p.renderParagraph(data[:i])
return i + n
}
- // an underline under some text marks a header, so our paragraph ended on prev line
+ // an underline under some text marks a heading, so our paragraph ended on prev line
if i > 0 {
- if level := p.isUnderlinedHeader(current); level > 0 {
+ if level := p.isUnderlinedHeading(current); level > 0 {
// render the paragraph
- p.renderParagraph(out, data[:prev])
+ p.renderParagraph(data[:prev])
// ignore leading and trailing whitespace
eol := i - 1
@@ -1320,24 +1460,17 @@ func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
eol--
}
- // render the header
- // this ugly double closure avoids forcing variables onto the heap
- work := func(o *bytes.Buffer, pp *parser, d []byte) func() bool {
- return func() bool {
- pp.inline(o, d)
- return true
- }
- }(out, p, data[prev:eol])
-
id := ""
- if p.flags&EXTENSION_AUTO_HEADER_IDS != 0 {
+ if p.extensions&AutoHeadingIDs != 0 {
id = sanitized_anchor_name.Create(string(data[prev:eol]))
}
- p.r.Header(out, work, level, id)
+ block := p.addBlock(Heading, data[prev:eol])
+ block.Level = level
+ block.HeadingID = id
// find the end of the underline
- for data[i] != '\n' {
+ for i < len(data) && data[i] != '\n' {
i++
}
return i
@@ -1345,45 +1478,72 @@ func (p *parser) paragraph(out *bytes.Buffer, data []byte) int {
}
// if the next line starts a block of HTML, then the paragraph ends here
- if p.flags&EXTENSION_LAX_HTML_BLOCKS != 0 {
- if data[i] == '<' && p.html(out, current, false) > 0 {
+ if p.extensions&LaxHTMLBlocks != 0 {
+ if data[i] == '<' && p.html(current, false) > 0 {
// rewind to before the HTML block
- p.renderParagraph(out, data[:i])
+ p.renderParagraph(data[:i])
return i
}
}
- // if there's a prefixed header or a horizontal rule after this, paragraph is over
- if p.isPrefixHeader(current) || p.isHRule(current) {
- p.renderParagraph(out, data[:i])
+ // if there's a prefixed heading or a horizontal rule after this, paragraph is over
+ if p.isPrefixHeading(current) || p.isHRule(current) {
+ p.renderParagraph(data[:i])
return i
}
+ // if there's a fenced code block, paragraph is over
+ if p.extensions&FencedCode != 0 {
+ if p.fencedCodeBlock(current, false) > 0 {
+ p.renderParagraph(data[:i])
+ return i
+ }
+ }
+
// if there's a definition list item, prev line is a definition term
- if p.flags&EXTENSION_DEFINITION_LISTS != 0 {
+ if p.extensions&DefinitionLists != 0 {
if p.dliPrefix(current) != 0 {
- return p.list(out, data[prev:], LIST_TYPE_DEFINITION)
+ ret := p.list(data[prev:], ListTypeDefinition)
+ return ret
}
}
// if there's a list after this, paragraph is over
- if p.flags&EXTENSION_NO_EMPTY_LINE_BEFORE_BLOCK != 0 {
+ if p.extensions&NoEmptyLineBeforeBlock != 0 {
if p.uliPrefix(current) != 0 ||
p.oliPrefix(current) != 0 ||
p.quotePrefix(current) != 0 ||
p.codePrefix(current) != 0 {
- p.renderParagraph(out, data[:i])
+ p.renderParagraph(data[:i])
return i
}
}
// otherwise, scan to the beginning of the next line
- for data[i] != '\n' {
- i++
+ nl := bytes.IndexByte(data[i:], '\n')
+ if nl >= 0 {
+ i += nl + 1
+ } else {
+ i += len(data[i:])
}
- i++
}
- p.renderParagraph(out, data[:i])
+ p.renderParagraph(data[:i])
+ return i
+}
+
+func skipChar(data []byte, start int, char byte) int {
+ i := start
+ for i < len(data) && data[i] == char {
+ i++
+ }
+ return i
+}
+
+func skipUntilChar(text []byte, start int, char byte) int {
+ i := start
+ for i < len(text) && text[i] != char {
+ i++
+ }
return i
}
diff --git a/vendor/github.com/documize/blackfriday/block_test.go b/vendor/github.com/documize/blackfriday/block_test.go
index f52506fd..0a2a4d84 100644
--- a/vendor/github.com/documize/blackfriday/block_test.go
+++ b/vendor/github.com/documize/blackfriday/block_test.go
@@ -14,68 +14,10 @@
package blackfriday
import (
+ "strings"
"testing"
)
-func runMarkdownBlockWithRenderer(input string, extensions int, renderer Renderer) string {
- return string(Markdown([]byte(input), renderer, extensions))
-}
-
-func runMarkdownBlock(input string, extensions int) string {
- htmlFlags := 0
- htmlFlags |= HTML_USE_XHTML
-
- renderer := HtmlRenderer(htmlFlags, "", "")
-
- return runMarkdownBlockWithRenderer(input, extensions, renderer)
-}
-
-func runnerWithRendererParameters(parameters HtmlRendererParameters) func(string, int) string {
- return func(input string, extensions int) string {
- htmlFlags := 0
- htmlFlags |= HTML_USE_XHTML
-
- renderer := HtmlRendererWithParameters(htmlFlags, "", "", parameters)
-
- return runMarkdownBlockWithRenderer(input, extensions, renderer)
- }
-}
-
-func doTestsBlock(t *testing.T, tests []string, extensions int) {
- doTestsBlockWithRunner(t, tests, extensions, runMarkdownBlock)
-}
-
-func doTestsBlockWithRunner(t *testing.T, tests []string, extensions int, runner func(string, int) string) {
- // catch and report panics
- var candidate string
- defer func() {
- if err := recover(); err != nil {
- t.Errorf("\npanic while processing [%#v]: %s\n", candidate, err)
- }
- }()
-
- for i := 0; i+1 < len(tests); i += 2 {
- input := tests[i]
- candidate = input
- expected := tests[i+1]
- actual := runner(candidate, extensions)
- if actual != expected {
- t.Errorf("\nInput [%#v]\nExpected[%#v]\nActual [%#v]",
- candidate, expected, actual)
- }
-
- // now test every substring to stress test bounds checking
- if !testing.Short() {
- for start := 0; start < len(input); start++ {
- for end := start + 1; end <= len(input); end++ {
- candidate = input[start:end]
- _ = runMarkdownBlock(candidate, extensions)
- }
- }
- }
- }
-}
-
func TestPrefixHeaderNoExtensions(t *testing.T) {
var tests = []string{
"# Header 1\n",
@@ -202,7 +144,7 @@ func TestPrefixHeaderSpaceExtension(t *testing.T) {
"
\n",
}
- doTestsBlock(t, tests, EXTENSION_LAX_HTML_BLOCKS)
+ doTestsBlock(t, tests, LaxHTMLBlocks)
}
func TestFencedCodeBlock(t *testing.T) {
@@ -1061,8 +1011,126 @@ func TestFencedCodeBlock(t *testing.T) {
"Some text before a fenced code block\n``` oz\ncode blocks breakup paragraphs\n```\nSome text in between\n``` oz\nmultiple code blocks work okay\n```\nAnd some text after a fenced code block",
"
+`,
+ }
+
+ // These 2 alternative forms of blockquoted fenced code blocks should produce same output.
+ forms := [2]string{
+ cat("> plain quoted text",
+ "> ```fenced",
+ "code",
+ " with leading single space correctly preserved",
+ "okay",
+ "```",
+ "> rest of quoted text"),
+ cat("> plain quoted text",
+ "> ```fenced",
+ "> code",
+ "> with leading single space correctly preserved",
+ "> okay",
+ "> ```",
+ "> rest of quoted text"),
+ }
+ want := `
+
plain quoted text
+
+
code
+ with leading single space correctly preserved
+okay
+
+
+
+
+`,
+ }
+ doTestsParam(t, tests, TestParams{HTMLFlags: UseXHTML | CompletePage})
+}
+
+func TestIsFenceLine(t *testing.T) {
+ tests := []struct {
+ data []byte
+ syntaxRequested bool
+ wantEnd int
+ wantMarker string
+ wantSyntax string
+ }{
+ {
+ data: []byte("```"),
+ wantEnd: 3,
+ wantMarker: "```",
+ },
+ {
+ data: []byte("```\nstuff here\n"),
+ wantEnd: 4,
+ wantMarker: "```",
+ },
+ {
+ data: []byte("```\nstuff here\n"),
+ syntaxRequested: true,
+ wantEnd: 4,
+ wantMarker: "```",
+ },
+ {
+ data: []byte("stuff here\n```\n"),
+ wantEnd: 0,
+ },
+ {
+ data: []byte("```"),
+ syntaxRequested: true,
+ wantEnd: 3,
+ wantMarker: "```",
+ },
+ {
+ data: []byte("``` go"),
+ syntaxRequested: true,
+ wantEnd: 6,
+ wantMarker: "```",
+ wantSyntax: "go",
+ },
}
- doTestsBlock(t, tests, EXTENSION_TITLEBLOCK)
-
+ for _, test := range tests {
+ var syntax *string
+ if test.syntaxRequested {
+ syntax = new(string)
+ }
+ end, marker := isFenceLine(test.data, syntax, "```")
+ if got, want := end, test.wantEnd; got != want {
+ t.Errorf("got end %v, want %v", got, want)
+ }
+ if got, want := marker, test.wantMarker; got != want {
+ t.Errorf("got marker %q, want %q", got, want)
+ }
+ if test.syntaxRequested {
+ if got, want := *syntax, test.wantSyntax; got != want {
+ t.Errorf("got syntax %q, want %q", got, want)
+ }
+ }
+ }
}
diff --git a/vendor/github.com/documize/blackfriday/doc.go b/vendor/github.com/documize/blackfriday/doc.go
new file mode 100644
index 00000000..5b3fa987
--- /dev/null
+++ b/vendor/github.com/documize/blackfriday/doc.go
@@ -0,0 +1,18 @@
+// Package blackfriday is a markdown processor.
+//
+// It translates plain text with simple formatting rules into an AST, which can
+// then be further processed to HTML (provided by Blackfriday itself) or other
+// formats (provided by the community).
+//
+// The simplest way to invoke Blackfriday is to call the Run function. It will
+// take a text input and produce a text output in HTML (or other format).
+//
+// A slightly more sophisticated way to use Blackfriday is to create a Markdown
+// processor and to call Parse, which returns a syntax tree for the input
+// document. You can leverage Blackfriday's parsing for content extraction from
+// markdown documents. You can assign a custom renderer and set various options
+// to the Markdown processor.
+//
+// If you're interested in calling Blackfriday from command line, see
+// https://github.com/russross/blackfriday-tool.
+package blackfriday
diff --git a/vendor/github.com/documize/blackfriday/esc.go b/vendor/github.com/documize/blackfriday/esc.go
new file mode 100644
index 00000000..6385f27c
--- /dev/null
+++ b/vendor/github.com/documize/blackfriday/esc.go
@@ -0,0 +1,34 @@
+package blackfriday
+
+import (
+ "html"
+ "io"
+)
+
+var htmlEscaper = [256][]byte{
+ '&': []byte("&"),
+ '<': []byte("<"),
+ '>': []byte(">"),
+ '"': []byte("""),
+}
+
+func escapeHTML(w io.Writer, s []byte) {
+ var start, end int
+ for end < len(s) {
+ escSeq := htmlEscaper[s[end]]
+ if escSeq != nil {
+ w.Write(s[start:end])
+ w.Write(escSeq)
+ start = end + 1
+ }
+ end++
+ }
+ if start < len(s) && end <= len(s) {
+ w.Write(s[start:end])
+ }
+}
+
+func escLink(w io.Writer, text []byte) {
+ unesc := html.UnescapeString(string(text))
+ escapeHTML(w, []byte(unesc))
+}
diff --git a/vendor/github.com/documize/blackfriday/esc_test.go b/vendor/github.com/documize/blackfriday/esc_test.go
new file mode 100644
index 00000000..ff67d546
--- /dev/null
+++ b/vendor/github.com/documize/blackfriday/esc_test.go
@@ -0,0 +1,48 @@
+package blackfriday
+
+import (
+ "bytes"
+ "testing"
+)
+
+func TestEsc(t *testing.T) {
+ tests := []string{
+ "abc", "abc",
+ "a&c", "a&c",
+ "<", "<",
+ "[]:<", "[]:<",
+ "Hello |"
+ processingInstruction = "[<][?].*?[?][>]"
+ singleQuotedValue = "'[^']*'"
+ tagName = "[A-Za-z][A-Za-z0-9-]*"
+ unquotedValue = "[^\"'=<>`\\x00-\\x20]+"
+)
+
+// HTMLRendererParameters is a collection of supplementary parameters tweaking
+// the behavior of various parts of HTML renderer.
+type HTMLRendererParameters struct {
// Prepend this text to each relative URL.
AbsolutePrefix string
// Add this text to each footnote anchor, to ensure uniqueness.
@@ -64,34 +82,34 @@ type HtmlRendererParameters struct {
// HTML_FOOTNOTE_RETURN_LINKS flag is enabled. If blank, the string
// [return] is used.
FootnoteReturnLinkContents string
- // If set, add this text to the front of each Header ID, to ensure
+ // If set, add this text to the front of each Heading ID, to ensure
// uniqueness.
- HeaderIDPrefix string
- // If set, add this text to the back of each Header ID, to ensure uniqueness.
- HeaderIDSuffix string
+ HeadingIDPrefix string
+ // If set, add this text to the back of each Heading ID, to ensure uniqueness.
+ HeadingIDSuffix string
+
+ Title string // Document title (used if CompletePage is set)
+ CSS string // Optional CSS file URL (used if CompletePage is set)
+ Icon string // Optional icon file URL (used if CompletePage is set)
+
+ Flags HTMLFlags // Flags allow customizing this renderer's behavior
}
-// Html is a type that implements the Renderer interface for HTML output.
+// HTMLRenderer is a type that implements the Renderer interface for HTML output.
//
-// Do not create this directly, instead use the HtmlRenderer function.
-type Html struct {
- flags int // HTML_* options
+// Do not create this directly, instead use the NewHTMLRenderer function.
+type HTMLRenderer struct {
+ HTMLRendererParameters
+
closeTag string // how to end singleton tags: either " />" or ">"
- title string // document title
- css string // optional css file url (used with HTML_COMPLETE_PAGE)
- parameters HtmlRendererParameters
+ // Track heading IDs to prevent ID collision in a single generation.
+ headingIDs map[string]int
- // table of contents data
- tocMarker int
- headerCount int
- currentLevel int
- toc *bytes.Buffer
+ lastOutputLen int
+ disableTags int
- // Track header IDs to prevent ID collision in a single generation.
- headerIDs map[string]int
-
- smartypants *smartypantsRenderer
+ sr *SPRenderer
}
const (
@@ -99,715 +117,31 @@ const (
htmlClose = ">"
)
-// HtmlRenderer creates and configures an Html object, which
+// NewHTMLRenderer creates and configures an HTMLRenderer object, which
// satisfies the Renderer interface.
-//
-// flags is a set of HTML_* options ORed together.
-// title is the title of the document, and css is a URL for the document's
-// stylesheet.
-// title and css are only used when HTML_COMPLETE_PAGE is selected.
-func HtmlRenderer(flags int, title string, css string) Renderer {
- return HtmlRendererWithParameters(flags, title, css, HtmlRendererParameters{})
-}
-
-func HtmlRendererWithParameters(flags int, title string,
- css string, renderParameters HtmlRendererParameters) Renderer {
+func NewHTMLRenderer(params HTMLRendererParameters) *HTMLRenderer {
// configure the rendering engine
closeTag := htmlClose
- if flags&HTML_USE_XHTML != 0 {
+ if params.Flags&UseXHTML != 0 {
closeTag = xhtmlClose
}
- if renderParameters.FootnoteReturnLinkContents == "" {
- renderParameters.FootnoteReturnLinkContents = `[return]`
+ if params.FootnoteReturnLinkContents == "" {
+ params.FootnoteReturnLinkContents = `[return]`
}
- return &Html{
- flags: flags,
+ return &HTMLRenderer{
+ HTMLRendererParameters: params,
+
closeTag: closeTag,
- title: title,
- css: css,
- parameters: renderParameters,
+ headingIDs: make(map[string]int),
- headerCount: 0,
- currentLevel: 0,
- toc: new(bytes.Buffer),
-
- headerIDs: make(map[string]int),
-
- smartypants: smartypants(flags),
+ sr: NewSmartypantsRenderer(params.Flags),
}
}
-// Using if statements is a bit faster than a switch statement. As the compiler
-// improves, this should be unnecessary this is only worthwhile because
-// attrEscape is the single largest CPU user in normal use.
-// Also tried using map, but that gave a ~3x slowdown.
-func escapeSingleChar(char byte) (string, bool) {
- if char == '"' {
- return """, true
- }
- if char == '&' {
- return "&", true
- }
- if char == '<' {
- return "<", true
- }
- if char == '>' {
- return ">", true
- }
- return "", false
-}
-
-func attrEscape(out *bytes.Buffer, src []byte) {
- org := 0
- for i, ch := range src {
- if entity, ok := escapeSingleChar(ch); ok {
- if i > org {
- // copy all the normal characters since the last escape
- out.Write(src[org:i])
- }
- org = i + 1
- out.WriteString(entity)
- }
- }
- if org < len(src) {
- out.Write(src[org:])
- }
-}
-
-func entityEscapeWithSkip(out *bytes.Buffer, src []byte, skipRanges [][]int) {
- end := 0
- for _, rang := range skipRanges {
- attrEscape(out, src[end:rang[0]])
- out.Write(src[rang[0]:rang[1]])
- end = rang[1]
- }
- attrEscape(out, src[end:])
-}
-
-func (options *Html) GetFlags() int {
- return options.flags
-}
-
-func (options *Html) TitleBlock(out *bytes.Buffer, text []byte) {
- text = bytes.TrimPrefix(text, []byte("% "))
- text = bytes.Replace(text, []byte("\n% "), []byte("\n"), -1)
- out.WriteString("
")
- out.Write(text)
- out.WriteString("\n
")
-}
-
-func (options *Html) Header(out *bytes.Buffer, text func() bool, level int, id string) {
- marker := out.Len()
- doubleSpace(out)
-
- if id == "" && options.flags&HTML_TOC != 0 {
- id = fmt.Sprintf("toc_%d", options.headerCount)
- }
-
- if id != "" {
- id = options.ensureUniqueHeaderID(id)
-
- if options.parameters.HeaderIDPrefix != "" {
- id = options.parameters.HeaderIDPrefix + id
- }
-
- if options.parameters.HeaderIDSuffix != "" {
- id = id + options.parameters.HeaderIDSuffix
- }
-
- out.WriteString(fmt.Sprintf("", level, id))
- } else {
- out.WriteString(fmt.Sprintf("", level))
- }
-
- tocMarker := out.Len()
- if !text() {
- out.Truncate(marker)
- return
- }
-
- // are we building a table of contents?
- if options.flags&HTML_TOC != 0 {
- options.TocHeaderWithAnchor(out.Bytes()[tocMarker:], level, id)
- }
-
- out.WriteString(fmt.Sprintf("\n", level))
-}
-
-func (options *Html) BlockHtml(out *bytes.Buffer, text []byte) {
- if options.flags&HTML_SKIP_HTML != 0 {
- return
- }
-
- doubleSpace(out)
- out.Write(text)
- out.WriteByte('\n')
-}
-
-func (options *Html) HRule(out *bytes.Buffer) {
- doubleSpace(out)
- out.WriteString("")
- } else {
- out.WriteString("\">")
- }
-
- attrEscape(out, text)
- out.WriteString("\n")
-}
-
-func (options *Html) BlockQuote(out *bytes.Buffer, text []byte) {
- doubleSpace(out)
- out.WriteString("
![\"")](\""))
0 {
- attrEscape(out, alt)
- }
- if len(title) > 0 {
- out.WriteString("\" title=\"")
- attrEscape(out, title)
- }
-
- out.WriteByte('"')
- out.WriteString(options.closeTag)
-}
-
-func (options *Html) LineBreak(out *bytes.Buffer) {
- out.WriteString("