p01.elasticstub 0.5.4

Elasticsearch server stub setup

This package provides a elasticsearch server stub setup based on a real elasticsearch server.

Contents

• README

  • setup

  • testing

• CHANGES

  • 0.5.4 (2017-11-02)

  • 0.5.3 (2014-07-10)

  • 0.5.2 (2012-12-22)

  • 0.5.1 (2012-12-10)

  • 0.5.0 (2012-11-18)

README

setup

This test is using an elasticsearch server. The test setUp method used for this test is calling our startElasticSearchServer method which is starting an elasticsearch server. The first time this test get called a new elasticsearch server will get downloaded. The test setup looks like:

def test_suite():
    return unittest.TestSuite((
        doctest.DocFileSuite('README.txt',
            setUp=testing.doctestSetUp, tearDown=testing.doctestTearDown,
            optionflags=doctest.NORMALIZE_WHITESPACE|doctest.ELLIPSIS,
            encoding='utf-8'),
        ))

If you like to set some custom settings, you can use the confSource which must point to a config folder with elasticsearch.yml or elasticsearch.json and logging.yml and optional mapping definitions. Your custom doctest setUp and tearDown method could look like:

def mySetUp(test):
    # use default elasticsearch with our server and conf source dir
    here = os.path.dirname(__file__)
    serverDir = os.path.join(here, 'server')
    confSource = os.path.join(here, 'config')
    p01.elasticstub.testing.startElasticSearchServer(serverDir=serverDir,
        confSource=confSource)

def myTearDown(test):
    p01.elasticstub.testing.stopElasticSearchServer()
    # do some custom teardown stuff here

testing

Let’s setup a python httplib connection:

>>> import httplib
>>> conn = httplib.HTTPConnection('localhost', 45200)

and test the cluster state:

>>> conn.request('GET', '_cluster/state')
>>> response = conn.getresponse()
>>> response.status
200

>>> import json
>>> from pprint import pprint
>>> body = response.read()
>>> pprint(json.loads(body))
{u'blocks': {},
 u'cluster_name': u'p01_elasticstub_testing',
 u'master_node': u'...',
 u'metadata': {u'cluster_uuid': u'...',
               u'index-graveyard': {u'...': []},
               u'indices': {},
               u'templates': {}},
 u'nodes': {u'...': {u'attributes': {},
                                        u'ephemeral_id': u'...',
                                        u'name': u'...',
                                        u'transport_address': u'...'}},
 u'routing_nodes': {u'nodes': {u'...': []},
                    u'unassigned': []},
 u'routing_table': {u'indices': {}},
 u'state_uuid': u'...',
 u'version': 2}

As you can see our mapping is empty:

>>> conn.request('GET', '/testing/test/_mapping')
>>> response = conn.getresponse()
>>> body = response.read()
>>> pprint(json.loads(body))
{u'error': {u'index': u'testing',
            u'index_uuid': u'_na_',
            u'reason': u'no such index',
            u'resource.id': u'testing',
            u'resource.type': u'index_or_alias',
            u'root_cause': [{u'index': u'testing',
                             u'index_uuid': u'_na_',
                             u'reason': u'no such index',
                             u'resource.id': u'testing',
                             u'resource.type': u'index_or_alias',
                             u'type': u'index_not_found_exception'}],
            u'type': u'index_not_found_exception'},
 u'status': 404}

Let’s index a simple item:

>>> body = json.dumps({u'title': u'Title'})
>>> conn.request('POST', '/testing/test/1', body)
>>> response = conn.getresponse()
>>> body = response.read()
>>> pprint(json.loads(body))
{u'_id': u'1',
 u'_index': u'testing',
 u'_shards': {u'failed': 0, u'successful': 1, u'total': 2},
 u'_type': u'test',
 u'_version': 1,
 u'created': True,
 u'result': u'created'}

refresh:

>>> conn.request('GET', '/testing/test/_refresh')
>>> response = conn.getresponse()
>>> body = response.read()
>>> pprint(json.loads(body))
{u'_id': u'_refresh',
 u'_index': u'testing',
 u'_type': u'test',
 u'found': False}

Let’s set a mapping:

>>> body = json.dumps({'test': {'properties': {'title': {'type': 'string'}}}})
>>> conn.request('POST', '/testing/test/_mapping', body)
>>> response = conn.getresponse()
>>> body = response.read()
>>> pprint(json.loads(body))
{u'acknowledged': True}

and test our mapping again:

>>> conn.request('GET', '/testing/test/_mapping')
>>> response = conn.getresponse()
>>> body = response.read()
>>> pprint(json.loads(body))
{u'testing': {u'mappings': {u'test': {u'properties': {u'title': {u'fields': {u'keyword': {u'ignore_above': 256,
                                                                                          u'type': u'keyword'}},
                                                                 u'type': u'text'}}}}}}

CHANGES

0.5.4 (2017-11-02)

• switch to elasticsearch 5.6.4 and adjust startup script environment setup

• removed thrift plugin support

• adjust default elasticsearch config, added log4j config file

0.5.3 (2014-07-10)

• switch to elasticsearch version 1.2.1. Note; version 1.2.2 does not work at least on windows. Startup quits with en error like {1.2.2}: Initialization Failed … ExecutionError[java.lang.IncompatibleClassChangeError: Implementing class] IncompatibleClassChangeError[Implementing class]

0.5.2 (2012-12-22)

• switch to elasticsearch 0.20.1. Note; there is no need to switch to newer releases just for get the latest version. Simply use the downloadURL attribute for set a newer elasticsearch version download url

• It seems that the auto mapping doesn’t work anymore by default. Added explicit mapping setup after inserting first item. This seems to wrk as it should.

• bugfix: adjust base url, if used with version, from git to elasticsearch

0.5.1 (2012-12-10)

• switch to elasticsearch 0.20.0

• added version argument to startElasticSearchServer

• cleanup imports

0.5.0 (2012-11-18)

• initial release tested on win 32bit and posix 32bit. NOT tested on win 64bit, posix 64bit and mac 32/64bit systems.