FWIW, this is the source code for a
sed script I use on my Mac OS X system to convert HTML output generated by MacDown into a format I need. MacDown generates some extra “cruft” that I don’t need, so I use these
sed commands to clean up that HTML output:
# clean h1 tags. also add newlines before each h1. s?<h1 id="toc_.*">\(.*\)</h1>?\ \ \ <h1>\1</h1>? # clean h2 tags s?<h2 id="toc_.*">\(.*\)</h2>?<h2>\1</h2>? # clean h3 tags s?<h3 id="toc_.*">\(.*\)</h3>?<h3>\1</h3>? # clean pre tags s?<pre><code>?<pre>? s?</code></pre>?</pre>? # these next two lines are unique to modifying "scala>" content # inside pre tags (something i do while converting the scala cookbook) s?^scala> \(.*\)$?scala\> <strong>\1</strong>? s?^<pre>scala> \(.*\)$?<pre>scala\> <strong>\1</strong>?
I won’t explain those
sed commands, I just wanted to put it out here in case I needed it again. Okay, one thing to say is that the “header” patterns use the
sed “search and replace” functionality.