root/pyndexter/trunk/pyndexter/indexers/_hyperestraier.py

# -*- coding: utf-8 -*-
#
# Copyright (C) 2006 Alec Thomas <alec@swapoff.org>
#
# This software is licensed as described in the file COPYING, which
# you should have received as part of this distribution.
#

"""
Hyperestraier
-------------

Adapter for Hyperestraier_ using the swigged bindings.

.. _Hyperestraier: http://hyperestraier.sourceforge.net/

Usage
~~~~~

::

    hyperestraier://<path>?hype_mode=<int>

``hype_mode`` (default: auto)
    Override the default ``READONLY``/``READWRITE`` modes in Pyndexter and use
    Hyperestraier database open modes. See the Hyperestraier docs for details.

Installation
~~~~~~~~~~~~

Install your distributions Hyperestraier package (typically the package
``hyperestraier``).

If your distribution also includes the SWIG bindings as packages, install
these, otherwise:

::

    wget http://hyperestraier.sourceforge.net/binding/hyper_estraier_wrappers-0.0.15.tar.gz
    tar xfzv hyper_estraier_wrappers-0.0.15.tar.gz
    cd hyper_estraier_wrappers-0.0.15
    make
    make install
"""

import os
import hyperestraier
from pyndexter import *


__all__ = ['HyperestraierIndexer', 'HyperestraierResult']


class HyperestraierIndexer(Indexer):
    """ Pyndexter adapter for the Hyperestraier indexer. """
    def __init__(self, framework, host, path, username='admin',
                 password='admin', port=1978):
        Indexer.__init__(self, framework)

        self.db = hyperestraier.Node()
        self.db.set_url('http://%s:%s/%s' % (host, port, path))
        self.db.set_auth(username, password)

    def index(self, document):
        hdoc = hyperestraier.Document()
        for k, v in document.attributes.iteritems():
            hdoc.add_attr(u'@' + k, v)
        for line in document.content.splitlines():
            hdoc.add_text(line)
        self.db.put_doc(hdoc, 1)

    def discard(self, uri):
        uri = unicode(uri)
        if not self.db.out_doc_by_uri(uri):
            raise DocumentNotFound(uri)

    def fetch(self, uri):
        uri = unicode(uri)
        doc = self.db.get_doc_by_uri(uri)
        attributes = self._translate_attributes(doc)
        return Document(content=u'\n'.join([t for t in doc.texts()]),
                        quality=0.99,
                        **attributes)

    def search(self, query):
        phrase = query.as_string(not_='ANDNOT ')
        return self.hype_search(phrase, query, simple=False)

    def optimise(self):
        self.db.optimize()

    def flush(self):
        self.db.sync()

    def close(self):
        self.db.close()
        self.db = None

    # Hyperestraier-specific methods
    def hype_search(self, phrase, query, simple=True, order=None):
        """ Full Hyperestraier search phrase. """
        cond = hyperestraier.Condition()
        cond.set_phrase(phrase)
        search = self.db.search(cond, 0)
        return HyperestraierResult(self, query, search or [])

    # Internal methods
    def _translate_attributes(self, hdoc):
        attributes = {}
        for k in hdoc.attr_names():
            k = k.encode('utf-8')
            if k[0] == '@':
                attributes[k[1:]] = hdoc.attr(k)
            else:
                attributes[k] = hdoc.attr(k)
        attributes['uri'] = URI(attributes['uri'])
        return attributes


indexer_factory = PluginFactory(HyperestraierIndexer)


class HyperestraierResult(Result):
    def __iter__(self):
        for doc in self.context.docs:
            yield self._translate(doc)

    def __len__(self):
        return self.context.doc_num()

    def __getitem__(self, index):
        return self._translate(self.context.get_doc[index])

    # Internal methods
    def _translate(self, doc):
        return Hit(current=self.indexer.framework.fetch,
                   indexed=self.indexer.fetch,
                   **self.indexer._translate_attributes(doc))