filters/remove-trackers-and-ads.plugin

   1 import sys
   2 import re
   3
   4 data = sys.stdin.read()
   5
   6 regexes = [
   7         # Blogger tracking images
   8         r"""<img [^>]*?src=["']https?://blogger\.googleusercontent\.com/tracker/[^'"]*['"][^>]*>""",
   9         # Feedburner ads
  10         r"""<p><a href=['"]https?://[a-zA-Z0-9\-\.]*/~a/[a-zA-Z0-9]*\?a=[a-zA-Z0-9]*['"]><img border=['"]0['"] src=['"]https?://[a-zA-Z0-9\.\-]*/~a/[a-zA-Z0-9/]*\?i=[a-zA-Z0-9]*['"]/></a></p>""",
  11         # Feedburner tracking images
  12         r"""<img [^>]*?src=['"]https?://feeds\.feedburner\.com/~r/[^"']*['"][^>]*>""",
  13         # Wordpress tracking images
  14         r"""<img [^>]*?src=['"]https?://stats\.wordpress\.com/b\.gif[^"']*['"][^>]*>""",
  15         # Amazon tracking images
  16         r"""<img [^>]*?src=['"]https?://www\.assoc-amazon\.com/e/ir[^"']*['"][^>]*>""",
  17         # Paypal tracking images
  18         r"""<img [^>]*?src=['"]https?://www\.paypalobjects\.com/[^/]*/i/scr/pixel.gif[^"']*['"][^>]*>""",
  19         # Other tracking images
  20         r"""<img [^>]*?width=["']1["'][^>]*?height=["']1["'][^>]*>""",
  21         r"""<img [^>]*?height=["']1["'][^>]*?width=["']1["'][^>]*>""",
  22 ]
  23
  24 for regex in regexes: data = re.sub(regex, '', data)
  25
  26 sys.stdout.write(data)