semantic history fails to identify unicode python strings (u prefix)
Here is a random example of Python output of an object:
{ u'generate_macs': { '__file__': u'/Users/ssbarnea/rdo/tripleo-quickstart/roles/libvirt/setup/overcloud/tasks/main.yml',
Unicode strings are printed using u'...'
format but iterm identifies the string as containing the full sequence u
thus clicking on it will try to open it in the browser.
Desired behavior is to identify only the text between single quotes so when we click, it will open the text editor.