This tiny project is aimed to investigate ways of processing of Markdown markup language for Scala projects. It was inspired by the article of Brian Carper "Clojure and Markdown (and Javascript and Java and...)"
The project provides a trivial Scala API to process Markdown markup based on three implementations:
- MarkdownJ is a pure Java library
- Showdown is a JavaScript implementation of Markdown, so it's possible to call
it from Java using JavaScript implementations (this is the way Brian Carper proposed).
I chose
- Mozilla Rhino. Rhino permits to run JS on-the-fly and to compile it. As Showdown JavaScript is static, I compile it on the first run and use compiled function on the subsequent runs.
- Java6 Scripting (which is basically Rhino, according to this article)
Benchmark
Along with the library you may find a simple benchmark comparing all three methods described above. Do not trust absolute numbers blindly, as every run does Markdown source processing, transformation to XML and then compares resulting XML to the reference one.
Average time to process long Markdown document (Markdown syntax description):
- Java6 = 322 ms
- Rhino = 312 ms
- MarkdownJ = 183 ms
You may notice there is almost no difference between "compiled function" performance of Rhino and usual Java6 scripting.
Tests run with JDK 1.6 on MacBookPro 2.53GHz
Special note for Apple VM
If you have trouble running JavaScript engine in Java6 under Mac OSX, "How to include JavaScript engine in Apple's Java 6 VM" has a good recipe.