1. Mikhail Korobov
  2. opencorpora-tools

Source

opencorpora-tools / README.rst

opencorpora-tools

This package provides Python interface to http://opencorpora.org/

Installation

pip install opencorpora-tools

If you have python < 2.7 then argparse and ordereddict packages are required:

pip install argparse
pip install ordereddict

Usage

Obtaining corpora

Opencorpora-tools works with XML from http://opencorpora.org/.

You can download and unpack the XML manually (from 'Downloads' page) or just use the provided command-line util:

$ opencorpora download

Run opencorpora download --help for more options.

Using corpora

Initialize:

>>> import opencorpora
>>> corpus = opencorpora.Corpora('annot.opcorpora.xml')

Get a list of documents:

>>> catalog = corpus.catalog()
>>> doc_id, doc_title = catalog[1590]
>>> print doc_id
1610

>>> doc_title
24105 Герман Греф советует россиянам «не суетиться» с валютой

Work with a document:

>>> doc = corpus[1610]
>>> print doc.title()
24105 Герман Греф советует россиянам «не суетиться» с валютой

>>> print doc.words()[11]
Сбербанка

>>> doc.sents()[0]
<class 'opencorpora.Sentence'>: Герман Греф советует россиянам «не суетиться» с валютой

>>> print doc.paras()[0]
Герман Греф советует россиянам «не суетиться» с валютой Президент Сбербанка уверен, что в ближайшее время на валютных рынках сохранится высокая волатильность и «шараханье».

Corpora, Document, Paragraph and Sentence classes support the following methods (when it make sense, e.g. sentence doesn't have paragraphs):

  • words() - returns a list of words and other tokens;
  • sents() - returns a list of Sentence instances;
  • paras() - returns a list of Paragraph instances;
  • documents() - returns a list of Document instances (this is memory hog!);
  • tagged_words() - returns a list of (str, str);
  • tagged_sents() - returns a list of (list of (str, str));
  • tagged_paras() - returns a list of (list of (list of (str, str)));
  • iterwords(), itersents(), iterparas(), iterdocuments(), iter_tagged_words, iter_tagged_sents, iter_tagged_paras - return iterators over words, sentences, paragraphs or documents;

You can also iterate over Corpora, Document, Paragraph and Sentence (this yields documents, paragraphs, sentences and words), e.g.:

>>> sent = doc.sents()[0]
>>> for word in sent:
...     print word
...
Герман
Греф
советует
россиянам
«
не
суетиться
»
с
валютой

The API is modelled after NLTK's CorpusReader API.

It it not exactly the same, but is very similar. E.g. sents() in opencorpora-tools returns a list of Sentence instances and sents() in NLTK returns a list of list of strings, but Sentence instances quacks like a list of strings (it can be indexed, iterated, etc.) so opencorpora.Corpora API may be seen as a superset of NLTK CorpusReader API.

Performance

OpenCorpora XML is huge (>250MB) so building full DOM tree requires a lot of memory (several GB) and should be avoided.

opencorpora-tools handles it this way:

  1. corpus[doc_id] or corpus.get_document(doc_id) don't load the original XML to memory and don't parse it. They use precomputed offset information to slice the XML instead. The offset information is computed on first access and saved to "<name>.~" file.

    Consider document loading O(1) regarding XML size. Individual documents are not huge so they and loaded and parsed as usual.

  2. There are iterator methods for all corpora API.

Development

Development happens at github and bitbucket:

The main issue tracker is at github.

Feel free to submit ideas, bugs, pull requests (git or hg) or regular patches.

Running tests

Make sure tox is installed and run

$ tox

from the source checkout. Tests should pass under python 2.6..3.2 and pypy > 1.8.