fix to dom elements truncated with limit

[sam2x]

Please review this patch, it fix a very annoying bug.

Bug Explanation: https://github.com/ismarslomic/angular-read-more/issues/6#issuecomment-229992773

Summary:

• It support plain text (ie: without html inside)
• Afaik, the patch doesnt break any feautre of the directive (transparent to user)
• Fix the bug where tags are truncated, which resulted to some content with html not parsed like " style='text-align:center'> Hey this is a long text that would not appear in less part!!!! </span>

Solution of this patch :

• I use DOMParser recursively parse the content of vm.hmText (ie: data passed to hm-text directive)
• If a text element is encountered, we take it in account to verify that the limit is not reached
• If the limit is reached, the parsing is stopped, and the current element text is truncated. The function return the html parsed until the last tag which reached the limit.
• Plain text are considered as a text element, so no edge case should appear.

[coveralls]

Coverage remained the same at 100.0% when pulling 8542c3f41ea66aff22b37249a5f05bda420a50ab on sam2x:fix_dom_elements_truncated into 6faff51cb036d6126b3133e59e0067eb7b546f15 on ismarslomic:master.

[ismarslomic]

Could you also update/add unit tests for this change?

[mtebele]

Just merge it, I need that fix!

[sam2x]

Note that if you use this fix you cannot have animation transition on text. @ismarslomic i don't know how to add unit test, sry.