REST and content wrapped inside !CDATA in xml feeds

Question

I would like some assistance in extracting jpg images for a gallery inside !CDATA in REST feeds like
http://saturn.jpl.nasa.gov/news/latest-cassini-huygens.xml - source looks like:
<item>
<title><![CDATA[NASA Cassini Images May Reveal Birth of a Saturn Moon]]></title>
<description><![CDATA[<table border="0" cellpadding="5"><tr><td><img src="http://saturn.jpl.nasa.gov/multimedia/images/rings/images/IMG005008-th100-75.jpg" alt="Commotion at Ring's Edge May Be Effect of Small Icy Object" width="100" height="75" border="0" align="left"></td><td valign="top">Cassini has been documenting a process that may be replaying, in miniature, the birth process of about two dozen moons of Saturn.</td></tr></table>]]></description>
<link>http://saturn.jpl.nasa.gov/news/newsreleases/newsrelease20140414/</link>
</item>
-- Barb Bowman

Answer 1

Barb,
Try this:
1) Create a custom gallery
2) Set the custom gallery data source to latest_cassini_huygens_xml!channel!item
3) Add a image visual into the gallery
4) Set the image visual data source to Mid(ThisItem!description!'$text',Find("http",ThisItem!description!'$text'),Find("jpg",ThisItem!description!'$text')+3-Find("http",ThisItem!description!'$text'))
Note: The Mid function extracts out the img src url from the description text.
 Yes! Thank you! Is there more documentation on this functionality (other than what is on blob)? So the +3 does in relation to extracting the image in combination with the word 'jpg' following the word "http' pulls the whole string and since
it is an image field, it works (?). I sure didn't get that from the docs on blob...
2.34 
Mid
Usage: Mid(text, start_position, num_chars)
Mid(textColumn, start_positions, num_chars)
Returns the characters from the middle of a string, given the starting position and the number of characters to extract.
Examples:
Mid(“ABCDEFG”, 3, 3) returns “CDE”.
Mid( [“ABCDEFG”, “Sienna”], [2, 3], [3, 3]) returns [“BCD”, “enn”].
I'd like to also extract the text without HTML markup (which happens when I use the description value (obviously) but I don't think the Mid function helps here. Is there something else that might work to extract the text I want?
<description><![CDATA[<table border="0" cellpadding="5"><tr><td><img src="http://saturn.jpl.nasa.gov/multimedia/images/rings/images/IMG005008-th100-75.jpg"
alt="Commotion at Ring's Edge May Be Effect of Small Icy Object" width="100" height="75" border="0" align="left"></td><td valign="top">Cassini has been
documenting a process that may be replaying, in miniature, the birth process of about two dozen moons of Saturn.</td></tr></table>]]></description>
-- Barb Bowman