[
<source srcset="https://cdn3.brettterpstra.com/uploads/2011/02/MarkdownRules-1.7394.jpg 1x, https://cdn3.brettterpstra.com/uploads/2011/02/MarkdownRules-1@2x.7394.jpg 2x" />
<img class="alignright" data-original="https://cdn3.brettterpstra.com/uploads/2011/02/MarkdownRules-1.7394.jpg" data-at2x="https://cdn3.brettterpstra.com/uploads/2011/02/MarkdownRules-1@2x.7394.jpg" src="https://cdn3.brettterpstra.com/uploads/2011/02/MarkdownRules-1.jpg" />
</noscript>
</picture>][marky]Marky the Markdownifier, my web tool that extracts web articles and converts them to Markdown, has received an update that resolves quite a few standing issues. It’s a major update and partial rewrite, so there may still be some bugs. Please feel free to drop me a line on Twitter or contact me directly if you run into problems.
The biggest changes are that the “readability” functions that strip ads, menus, comments, etc. out of the post are now handled by an updated version of PHP Readability, and the “markdownifying” part is handled by html2text instead of PHP Markdown Extra. The previews are also rendered by MultiMarkdown 4 now.
The most obvious improvement in all of this is that if a post contains images within links (i.e. hyperlinked images), it won’t throw off the reference counter for the links. Previously, this circumstance pretty much entirely broke the output. There are still some edge cases that are breaking it, but I’m narrowing them down one at a time. Twitter timeline capture is broken, but I’m not planning to fix that one. A few articles are coming up blank; if you run into one of those, please shoot me the url.
All of the API features should still be intact, and the bookmarklets should work as they always have. Head over to MarkdownRules.com and check out the new, improved version in your choice of flavors.