This is a Python wrapper to the low-level API of CQP which allows you to access CQP corpora in the same way as Perl's CWB::CL
If you installed CQP in a non-standard location (which is the default for newer versions of CQP), point the setup in the right direction with, e.g.
- To install the module, use the standard
- python setup.py build sudo python setup.py install
- As a prerequisite, install Cython with
- pip install Cython
If you use an old version of CQP (CWB 2.99 and older), you need to change the value of the "extra_libs" variable in setup.py.
To give you an idea how to use the library, see the following sample:
--- 8< --- from CWB.CL import Corpus
# open the corpus corpus=Corpus('TUEPP') # get sentences and words sentences=corpus.attribute('s','s') words=corpus.attribute('word','p') postags=corpus.attribute('pos','p') # retrieve offsets of the 1235th sentence (0-based) s_1234=sentences
- for w,p in zip(words[s_1234:s_1234+1],postags[s_1234:s_1234+1]):
- print "%s/%s"%(w,p)
--- 8< ---
In order to test the CWB.CL module's correct installation independently of any CQP corpora, you can do a
python -m doctest tests/idlist.txt
which should terminate with no output when everything is well.