TODO

* document Pod::POM::View and friends

* add more thorough testing

* fix link generation via L<...>

* handle =encoding

* more views for different styles, formats, etc.

* According to the new podspec: C<< o->foo >> is the same as
  C<o-E<gt>foo>. adjust Pod::POM to handle this kind of markup
  also check that this works: S<C<< $aryref->[$foo] >>> (Most likely
  the text parser needs to be rewritten) [stas]

* you need to escape < and > if they're inside of [A-Z]<...>. Outside
  of formatting code ([A-Z]<...>) the parser shouldn't complain and
  leave < and > as is.

* =begin/=end blocks are mishandled, I have a working version but it's
  imperfect. Parser's rewrite will provide a robust solution [stas]

* some pods may include sequences unrecognized by Pod::POM, currently
  it spits many undefined var warnings (e.g. add N<foo> to the tests).
  - should gracefull handle this situation and probably warn about the
    unknown sequence each time it's encountered with line number.
  - should probably ignore the sequence, but process what's inside the
    sequence like the unknown sequence wasn't there. [stas]

* the =over/=item can have very different usages, currently
  Pod::POM::View::Pod doesn't deparse correctly the following:

  =over
  
  =item
  
  Cat
  
  =item
  
  Sat
  
  =item
  
  Mat
  
  =back

  it loses the =item parts, simply copy the rest of the item tests I've
  recently added to t/htmlview.t to t/complete.t and you will see that
  the generated pod misses '=item' keys for this section. [stas]

* if the pod doesn't start from the top of the file, but there is some
  leading text, the parser will report warnings on the line numbers
  relative to the beginning of the pod and not the absolute linenumber
  in the file. To make the warnings useful, the parser probably should
  accomodate for skipped non-pod lines. Think of .pm files with
  inlined pod (before each function) where this will come very useful.