;;; -*- mode: fundamental; coding: utf-8; indent-tabs-mode: t; -*-

;;;
;;; Copyright (c) 2012 -- 2012 Stephan Oepen (oe@ifi.uio.no); 
;;; see `LICENSE' for conditions.
;;;


;;;
;;; _fix_me_
;;; following are a set of `cheap and cheerful' REPP rules to make simplified
;;; HTML text palatable to parsing with the ERG.  for the time being, we are
;;; mostly just throwing out markup, with the exception of italics, emphasis,
;;; et al., which can signal use--mention distinctions.
;;;

!</?(?:a|b|code|h[0-9]|li|small|strong|title)>			
!<em[^>]*>((?:(?!</em>).)*)</em>				¦i \1 i¦
!<i[^>]*>((?:(?!</i>).)*)</i>					¦i \1 i¦
;;
;; in blogs, authors at times try to be witty and use HTML strike-through to
;; show changes of mind or otherwise inappropriate content.
;;
!<s[^>]*>((?!</s>).)*</s>					 

;;
;; _fix_me_
;; what about sub- and super-scripts?  a task for the GMLC.     (5-mar-12; oe)
;;
!</?su(?:b|p)>						

;;
;; as we do for wikipedia text, collapse pieces of non-English into a token
;; that the grammar treats much like a proper name.  for robustness, enforce
;; token boundaries around these.
;;
!<code[^>]*>(?:(?!</code>).)*</code>				 <code/> 
!<img[^>]*>(?:(?!</img>).)*</img>				 <img/> 
!<kbd[^>]*>(?:(?!</kbd>).)*</kbd>				 <kbd/> 

;;
;; finally, for peace of mind, normalize whitespace sequences to a single space
;; 
!  +