Highlights

Added a setting for changing the output of the <img> tag
Added a method to replace html encoded unicode characters with their unicode counterparts
Changed the default for replace_unicode so the alt attribute is the unicode character that was made into an image. This to allow for proper behavior when copy-pasting.

Thanks to @pistos2 for the work on replace_html_entities and setting the unicode character as the alt value.

Changelog

Emoji.replace_html_entities replaces all unicode html entities
with their corresponding unicode character. This method is now by
default called before Emoji.replace_unicode. (@pistos2)
Emoji.replace_unicode will now set the <img alt> attribute to be
the unicode character being replaced. This change done to allow
marking and copying strings with emojis and the unicode character
being copied along correctly. (@pistos2)
Change the <img> tag to also have a title attribute that is the
text representation of the current character being encoded.
Added settings for new options:
- EMOJI_ALT_AS_UNICODE, default: True
- EMOJI_REPLACE_HTML_ENTITIES, default: True
- EMOJI_IMG_TAG, default:
  <img src="{0}" alt="{1}" title="{2}" class="emoji">
Added support for running on a narrow unicode build of Python.
The test suite passed on narrow builds but it has not been production
tested with a narrow build.