def test_add_document(self, session, index, job):
session.add(job)
session.commit()
index = Index()
index.add_document(job.to_document())
hits = index.search(job.title)
assert len(hits) == 1
assert int(hits[0]['id']) == job.id
def test_add_document(self, session, index, job):
session.add(job)
session.commit()
index = Index()
index.add_document(job.to_document())
hits = index.search(job.title)
assert len(hits) == 1
assert int(hits[0]['id']) == job.id
def main(query, session):
name = settings.SEARCH_INDEX_NAME
directory = settings.SEARCH_INDEX_DIRECTORY
if not IndexManager.exists(name, directory):
die('Search index does not exist!')
if isinstance(query, str):
query = unicode(query, 'utf-8')
index = Index()
for result in index.search(query):
job = session.query(Job).get(result['id'])
print job
def index(app):
# By creating the index we are actually clearing it, so any tests that
# require a clean index, should require this feature.
name = app.config['SEARCH_INDEX_NAME']
directory = app.config['SEARCH_INDEX_DIRECTORY']
IndexManager.create(Schema, name, directory)
return Index(name, directory)
def search_jobs(self, query, sort=None, limit=None):
index = Index()
hits = []
kwargs = dict()
if sort:
kwargs['sort'] = sort
if limit:
kwargs['limit'] = limit
for hit in index.search(query, **kwargs):
job = db.session.query(Job).get(hit['id'])
# Make sure that we don't accidentally return an unpublished job
# that happened to be in the search index.
if job and job.published:
hits.append(job)
return hits
def search_jobs(self, query, sort=None, limit=None):
index = Index()
hits = []
kwargs = dict()
if sort:
kwargs['sort'] = sort
if limit:
kwargs['limit'] = limit
for hit in index.search(query, **kwargs):
job = db.session.query(Job).get(hit['id'])
# Make sure that we don't accidentally return an unpublished job
# that happened to be in the search index.
if job and job.published:
hits.append(job)
return hits
def test_search_limit(self, session, index, job):
doc = job.to_document()
timestamp = doc['created']
bulk = []
for i in range(15):
doc = copy.deepcopy(doc)
doc['id'] = unicode(i)
doc['created'] = timestamp - timedelta(days=i)
bulk.append(doc)
index = Index()
index.add_document_bulk(bulk)
# Search with ascending sort, should return the ids in reverse order.
hits = index.search(job.title, sort=('created', 'asc'))
assert [int(hit['id']) for hit in hits] == range(15)[::-1]
# Search with descending sort.
hits = index.search(job.title, sort=('created', 'desc'))
assert [int(hit['id']) for hit in hits] == range(15)
def main(should_create, index_all, session):
name = settings.SEARCH_INDEX_NAME
directory = settings.SEARCH_INDEX_DIRECTORY
if should_create:
print blue("You've asked to (re)create index '{}'.".format(name))
IndexManager.create(Schema, name, directory)
if not IndexManager.exists(name, directory):
die('Search index does not exist!')
index = Index()
start = time.time()
kwargs = {} if index_all else {'published': True}
jobs = session.query(Job).filter_by(**kwargs).all()
index.add_document_bulk([job.to_document() for job in jobs])
duration = time.time() - start
print green("{0} documents added okay in {1:.2f} ms.".format(len(jobs), duration))
def test_delete_document(self, session, index, job):
session.add(job)
session.commit()
doc = job.to_document()
index = Index()
index.add_document(doc)
hits = index.search(job.title)
assert len(hits) == 1
index.delete_document(doc['id'])
hits = index.search(job.title)
assert len(hits) == 0
def main(should_create, index_all, session):
name = settings.SEARCH_INDEX_NAME
directory = settings.SEARCH_INDEX_DIRECTORY
if should_create:
print blue("You've asked to (re)create index '{}'.".format(name))
IndexManager.create(Schema, name, directory)
if not IndexManager.exists(name, directory):
die('Search index does not exist!')
index = Index()
start = time.time()
kwargs = {} if index_all else {'published': True}
jobs = session.query(Job).filter_by(**kwargs).all()
index.add_document_bulk([job.to_document() for job in jobs])
duration = time.time() - start
print green("{0} documents added okay in {1:.2f} ms.".format(
len(jobs), duration))
def test_update_document(self, session, index, job):
session.add(job)
session.commit()
doc = job.to_document()
index = Index()
index.add_document(doc)
doc['job_type'] = u'updated'
index.update_document(doc)
hits = index.search(u'updated')
assert len(hits) == 1
assert int(hits[0]['id']) == job.id
def test_delete_document(self, session, index, job):
session.add(job)
session.commit()
doc = job.to_document()
index = Index()
index.add_document(doc)
hits = index.search(job.title)
assert len(hits) == 1
index.delete_document(doc['id'])
hits = index.search(job.title)
assert len(hits) == 0
def test_update_document(self, session, index, job):
session.add(job)
session.commit()
doc = job.to_document()
index = Index()
index.add_document(doc)
doc['job_type'] = u'updated'
index.update_document(doc)
hits = index.search(u'updated')
assert len(hits) == 1
assert int(hits[0]['id']) == job.id
def test_search_limit(self, session, index, job):
doc = job.to_document()
timestamp = doc['created']
bulk = []
for i in range(15):
doc = copy.deepcopy(doc)
doc['id'] = unicode(i)
doc['created'] = timestamp - timedelta(days=i)
bulk.append(doc)
index = Index()
index.add_document_bulk(bulk)
# Search with ascending sort, should return the ids in reverse order.
hits = index.search(job.title, sort=('created', 'asc'))
assert [int(hit['id']) for hit in hits] == range(15)[::-1]
# Search with descending sort.
hits = index.search(job.title, sort=('created', 'desc'))
assert [int(hit['id']) for hit in hits] == range(15)
def index(job):
index = Index()
document = job.to_document()
rv = index.add_document(document)
logger.info(u"Job ({}) added to index.".format(job.id))
return rv
def deindex(job):
index = Index()
document = job.to_document()
rv = index.delete_document(document['id'])
logger.info(u"Job ({}) deleted from index.".format(job.id))
return rv
def index(job):
index = Index()
document = job.to_document()
rv = index.add_document(document)
logger.info(u"Job ({}) added to index.".format(job.id))
return rv
def deindex(job):
index = Index()
document = job.to_document()
rv = index.delete_document(document['id'])
logger.info(u"Job ({}) deleted from index.".format(job.id))
return rv
def search(self, query):
index = Index()
return index.search(query)
def search(self, query):
index = Index()
return index.search(query)
