Skip to content

anonymous /gist:139987

Embed URL

HTTPS clone URL

Subversion checkout URL

You can clone with
or
.
Download ZIP
A ruby snippet for Parsing and cleaning Word HTML
#
# This function takes messy Word HTML pasted into a WYSIWYG and cleans it up
# It leaves the tags and attributes specified in the params
# Copyright (c) 2009, Radio New Zealand
# Released under the MIT license
require 'rubygems'
require 'sanitize'
def clean_up_word_html(html, elements = ['p', 'b', 'h1', 'h2', 'h3', 'h4', 'h5', 'h6'], attributes={})
email_regex = /<p>Email:\s+((\w|\-|\_|\.)+\@((\w|\-|\_)+\.)+[a-zA-Z]{2,})/i
html.gsub! /[\n|\r]/ , ''
# keep only the things we want.
html = Sanitize.clean( html, :elements => elements, :attributes => attributes )
# butt up any tags
html.gsub! /&nbsp;/ , ' '
html.gsub! />\s+</ , '><'
#remove email address lines
html.gsub! email_regex , '<p>'
# post sanitize cleanup of empty blocks
# the order of removal is import - this is the way word stacks these elements
html.gsub! /<i><\/i>/ , ''
html.gsub! /<b><\/b>/ , ''
html.gsub! /<\/b><b>/ , ''
html.gsub! /<p><\/p>/ , ''
html.gsub! /<p><b><\/b><\/p>/ , ''
# misc - fix butted times
html.gsub! /(\d)am / , '\1 am '
html.gsub! /(\d)pm / , '\1 pm '
# misc - remove multiple space that may cause doc specific regexs to fail (in dates for example)
html.gsub! /\s+/ , ' '
# add new lines at the end of lines
html.gsub! /<\/(p|h\d|dt|dd|dl)>/, '</\1>' + "\n"
html.gsub! /<dl>/ , '<dl>' + "\n"
html
end
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Something went wrong with that request. Please try again.