Add Cython version of MmReader (piskvorky#1825)

* simple cython version of mmreader

* confirm MmCorpus read correct values from corpus file

* remove unnecessary include dirs for _mmreader.c

* add numpy-style docstrings

* update conversion specifier for doubles for sscanf

* drop plain python version of mmreader

* additional unittests for mmcorpus

* add gzipped mmcorpus unittest

* add bzip compressed test corpus

* remove tox test for non-compiled word2vec

* add back tox test for non-compiled word2vec

* add dummy classes for tox's non-compiled word2vec check

* fix ordering of imports for gensim.corpora

* remove tox test for non-compiled code

* update MANIFEST to include cython mmreader

* add back python only version of MmReader

* mend

* mend

* move python version of MmReader back to matutils

* fix import problems

* fix import order

MANIFEST.in

@@ -12,3 +12,5 @@ include gensim/models/doc2vec_inner.c
 include gensim/models/doc2vec_inner.pyx
 include gensim/models/fasttext_inner.c
 include gensim/models/fasttext_inner.pyx
+include gensim/corpora/_mmreader.c
+include gensim/corpora/_mmreader.pyx

gensim/__init__.py

@@ -1,8 +1,8 @@
 """This package contains interfaces and functionality to compute pair-wise document similarities within a corpus
 of documents.
-
 """
-from gensim import parsing, matutils, interfaces, corpora, models, similarities, summarization, utils  # noqa:F401
+
+from gensim import parsing, corpora, matutils, interfaces, models, similarities, summarization, utils  # noqa:F401
 import logging
 
 __version__ = '3.3.0'