fixed bug with reading documents without id - temporary solution, whole DocumentReader class should be rebuilt

In other words, cut out poliqarp from build if this flag is False.
The flag remains forced to be True by default, though.
This is mostly to be able to cut poliqarp out of pure WIN32 build until
such time it actually builds (it doesn't atm).

import corpus2
corpus = corpus2.Corpus()
tagset = corpus2.get_named_tagset('nkjp')
cr = corpus2.CorpusReader(tagset, 'document')
crfile = '/home/rk/tmp/corpus_file.txt'
readed_corp = cr.read(crfile)
while True:
  doc = readed_corp.next_document()
  if not doc:
    break
  print doc.path()

fix equality and lt check in SWIG wrappers (the renames prevented the automagic SWIG wrapping from working)