CDATA elements cannot contain sequence ]]>. When that sequence is appeared in the data, it should be split into two pieces and each piece should be put in their own CDATA container:

]]>

become

<![CDATA[]]]>
<![CDATA[]>]]>

or

<![CDATA[]]]]>
<![CDATA[>]]>

Currently in serde deserializer only one CDATA event processed at time, that means, that deserialization

<root>
  <string><![CDATA[]]]]><![CDATA[>]]></string>
</root>

into

struct AnyName {
  string: String,
}

would fail or wrongly return ]] instead of ]]>.

To fix that we should merge CDATA events, that there are some ambiguities that should be investigated:

1. should we merge CDATA and text events:

   <![CDATA[one]]>two

   should return onetwo?

   Judging from that and that, we should do that

2. should we ignore comments between CDATA events? Between CDATA and text events?

   <![CDATA[one]]><!--comment--><![CDATA[two]]>

   should return onetwo?

   <![CDATA[one]]><!--comment-->two

   should return onetwo?
   Currently all comments are skips at very early stage and deserializer sees

   <![CDATA[one]]><!--comment--><![CDATA[two]]>

   as

   <![CDATA[one]]><![CDATA[two]]>

3. should we ignore processing instructions between CDATA events? Between CDATA and text events?

   <![CDATA[one]]><?pi?><![CDATA[two]]>

   should return onetwo?

   <![CDATA[one]]><?pi?>two

   should return onetwo?
   Currently all processing instructions are skips at very early stage and deserializer sees

   <![CDATA[one]]><?pi?><![CDATA[two]]>

   as

   <![CDATA[one]]><![CDATA[two]]>

4. should we ignore whitespaces between CDATAs? Between CDATA and text?

   <![CDATA[one]]>
   <![CDATA[two]]>

   should return onetwo?

   <![CDATA[one]]>
   two

   should return onetwo?