Leap2A importer: improve robustness of <content> importing

[Nigel McNie]

In import/leap/lib.php, function fix_artefact_reference.

If a <content> tag in a Leap2A import has more than one child node, the importer gets confused and only imports one of them (either the first or the last, not sure which).

E.g.:

        <content type="html">&lt;p&gt;&amp;nbsp;&lt;/p&gt; &lt;p&gt;a paragraph&lt;/p&gt; &lt;p&gt;&amp;nbsp;&lt;/p&gt</content>

This is:

<content>
  <p>
  <p>
  <p>

Which isn't handled properly. I think the leap2A spec mentions that content like this is not a good idea, but the fix seems pretty easy. Patch attached.

[Hugh Davenport (hugh-davenport)]

https://reviews.mahara.org/1299

[Kristina Hoeppner (kris-hoeppner)]

It's actually https://reviews.mahara.org/#/c/1162/

[Son Nguyen (ngson2000)]

I have done several tests for this patch. However, it did not work as expected.
Some child HTML tags of <content> tag did not import such as <p>, <a>, <img>, <table>, <ol>, and <ul>.

See my import file in the attached file?field.comment=I have done several tests for this patch. However, it did not work as expected.
Some child HTML tags of <content> tag did not import such as <p>, <a>, <img>, <table>, <ol>, and <ul>.

See my import file in the attached file

[Robert Lyon (robertl-9)]

According to the documentation here: http://www.leapspecs.org/2A/literals#content_or_description

We should not be trying to import <content></content> as HTML (all escaped) but rather as XHTML (not escaped)

eg, instead of this:
<content type="html">&lt;p&gt;&amp;nbsp;&lt;/p&gt; &lt;p&gt;a paragraph&lt;/p&gt; &lt;p&gt;&amp;nbsp;&lt;/p&gt</content>

it should be more like this:

<content type="xhtml">
<div xmlns="http://www.w3.org/1999/xhtml">
<p> </p> <p>a paragrapgh</p> <p> </p>
</div>
</content>

where the containing <div> gets stripped on import

So on import should we take the contents of <content type="html"> unescape it, then run it thru html tidy and import it as xhtml?