# HTML CDATA sections
Single-line CDATA:
<![CDATA[foo bar baz]]>
Multi-line CDATA with markdown-looking content inside (must stay literal):
< and **not strong**
]]>
CDATA with blank line inside (block should NOT terminate):
<![CDATA[
first
second
]]>
CDATA with angle brackets and ampersands (the whole point of CDATA):
<![CDATA[a < b && c > d]]>
CDATA followed by a paragraph (the paragraph IS parsed as markdown):
<![CDATA[
body
]]>
*now emphasis* here.
Partial terminator `]]` inside body should NOT close (need full `]]>`):
<![CDATA[contains ]] and ] but no closer until]]>
Mid-paragraph CDATA becomes inline raw HTML:
The data <![CDATA[>&<]]> is preserved verbatim.