#Markdown parser written in Clojure
You can try out the parser here.
A markdown parser that compiles to both Clojure and ClojureScript.
Note: markdown-clj
requires Clojure 1.2+ to run.
Markdown-clj can be invoked either by calling md-to-html
or md-to-html-string
functions.
The md-to-html
function accepts an input containing Markdown markup and an output where
the resulting HTML will be written. The input and output parameters will be passed to a reader
and a writer respectively:
(ns foo
(:use markdown.core))
(md-to-html "input.md" "output.html")
(md-to-html (input-stream "input.md") (output-stream "test.txt"))
The md-to-html-string
function accepts a string with markdown content and returns a string with the resulting HTML:
(md-to-html-string "# This is a test\nsome code follows\n```clojure\n(defn foo [])\n```")
<h1> This is a test</h1>some code follows<pre><code class="clojure">(defn foo [])
</code></pre>
Both md-to-html
and md-to-html-string
can accept optional parameters:
Specifying :heading-anchors
will create anchors for the heading tags, eg:
(markdown/md-to-html-string "###foo bar BAz" :heading-anchors true)
<h3><a name=\"heading\" class=\"anchor\" href=\"#foo_bar_baz\"></a>foo bar BAz</h3>
The code blocks default to a highlight.js compatible format of:
<pre><code class="clojure">some code</code></pre>
Specifying :code-style
will override the default code class formatting for code blocks, eg:
(md-to-html-string "# This is a test\nsome code follows\n```clojure\n(defn foo [])\n```"
:code-style #(str "class=\"brush: " % "\""))
<h1> This is a test</h1>some code follows<pre><code class="brush: clojure">
(defn foo [])
</code></pre>
Additional transformers can be specified using the :custom-transformers
key.
A transformer function must accept two arguments.
First argument is the string representing the current line and the second is the map representing the current state.
The default state keys are:
:code
- inside a code section:codeblock
- inside a code block:eof
- end of file:heading
- in a heading:hr
- in a horizontal line:lists
- inside a list:blockquote
- inside a blockquote:paragraph?
- in a paragraph?:last-line-empty?
- was last line an empty line?
For example, if we wanted to add a transformer that would capitalize all text we could do the following:
(defn capitalize [text state]
[(.toUpperCase text) state])
(markdown/md-to-html-string "#foo" :custom-transformers [capitalize])
<H1>FOO</H1>
Alternatively, you could provide a custom set of transformers to replace the default transformers using the :replacement-transformers
key.
(markdown/md-to-html-string "#foo" :replacement-transformers [capitalize])
This can also be used to add preprocessor transformers. For example, if we wanted to sanitize any image links we could do the following:
(use 'markdown.transformers 'markdown.core)
(defn escape-images [text state]
[(clojure.string/replace text #"(!\[.*?\]\()(.+?)(\))" "") state])
(markdown/md-to-html-string
"foo ![Alt text](/path/to/img.jpg \"Optional Title\") bar [text](http://test)"
:replacement-transformers (cons escape-images transformer-vector))
"<p>foo bar <a href='http://test'>text</a></p>"
The ClojureScript portion works the same as above except that the entry function is called md->html
. It accepts
a string followed by the options as its input, and returns the resulting HTML string:
(ns myscript
(:require [markdown.core :refer [md->html]]))
(.log js/console
(md->html "##This is a heading\nwith a paragraph following it"))
(.log js/console
(md->html "# This is a test\nsome code follows\n```clojure\n(defn foo [])\n```"
:code-style #(str "class=\"" % "\"")))
console.log(markdown.core.mdToHtml("##This is a heading\nwith a paragraph following it"));
Control characters can be escaped using \
\\ backslash
\` backtick
\* asterisk
\_ underscore
\{ curly braces
\}
\[ square brackets
\]
\( parentheses
\)
\# hash mark
\+ plus sign
\- minus sign (hyphen)
\. dot
\! exclamation mark
Blockquote, Strong, Bold, Emphasis, Italics, Heading, Line, Linebreak, Paragraph, Strikethrough
This is a shortcut style for creating “automatic” links for URLs and email addresses:
<http://example.com/>
will be turned this into:
<a href="http://example.com/">http://example.com/</a>
Automatic links for email addresses work similarly, except that they are hex encoded:
will be turned into:
<a href=\"address@example.com\">address@example.com</a>
Code Block, Indented Code, Inline Code
the number of hashes indicates the level of the heading
# Heading
##Sub-heading
### Sub-sub-heading
headings can also be defined using =
and -
for h1
and h2
respectively
Heading 1
=========
Heading 2
---------
***
* * *
*****
- - -
______
If a line ends with two or more spaces a <br />
tag will be inserted at the end.
*foo*
_foo_
**foo**
__foo__
>
prefixes regular blockquote paragraphs. >-
prefixes a
blockquote footer that can be used for author attribution.
>This is a blockquote
with some content
>this is another blockquote
> Everyone thinks of changing the world,
but no one thinks of changing himself.
>- Leo Tolstoy
This is a paragraph, it's
split into separate lines.
This is another paragraph.
indenting an item makes it into a sublist of the item above it, ordered and unordered lists can be nested within one another. List items can be split over multiple lines.
* Foo
* Bar
* Baz
* foo
* bar
* baz
1. foo
2. bar
more content
## subheading
***
**strong text** in the list
* fuzz
* blah
* blue
* brass
1. Foo
2. Bar
3. Baz
Any special characters in code will be escaped with their corresponding HTML codes.
Here's some code `x + y = z` that's inlined.
Using three backquotes indicates a start of a code block, the next three backquotes ends the code block section. Optionally, the language name can be put after the backquotes to produce a tag compatible with highlight.js, eg:
```clojure
(defn foo [bar] "baz")
```
indenting by at least 4 spaces creates a code block
some
code
here
note: XML is escaped in code sections
~~foo~~
a^2 + b^2 = c^2
[github](http://github.com)
![Alt text](http://server/path/to/img.jpg)
![Alt text](/path/to/img.jpg "Optional Title")
[![Continuous Integration status](https://secure.travis-ci.org/yogthos/markdown-clj.png)](http://travis-ci.org/yogthos/markdown-clj)
The parser reads the content line by line, this means that tag content is not allowed to span multiple lines.
Copyright (C) 2012 Yogthos
Distributed under the Eclipse Public License, the same as Clojure.