def test_create_skillset(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
skillset = SearchIndexerSkillset(name='test-ss',
skills=list([s]),
description="desc")
result = client.create_skillset(skillset)
assert isinstance(result, SearchIndexerSkillset)
assert result.name == "test-ss"
assert result.description == "desc"
assert result.e_tag
assert len(result.skills) == 1
assert isinstance(result.skills[0], EntityRecognitionSkill)
assert len(client.get_skillsets()) == 1
def test_create_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
result = client.create_indexer(indexer)
assert result.name == "sample-indexer"
assert result.target_index_name == "hotels"
assert result.data_source_name == "sample-datasource"
def _prepare_indexer(self,
endpoint,
api_key,
name="sample-indexer",
ds_name="sample-datasource",
id_name="hotels"):
con_str = self.settings.AZURE_STORAGE_CONNECTION_STRING
self.scrubber.register_name_pair(con_str, 'connection_string')
container = SearchIndexerDataContainer(name='searchcontainer')
data_source = SearchIndexerDataSourceConnection(
name=ds_name,
type="azureblob",
connection_string=con_str,
container=container)
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
ds = client.create_datasource(data_source)
index_name = id_name
fields = [{
"name": "hotelId",
"type": "Edm.String",
"key": True,
"searchable": False
}]
index = SearchIndex(name=index_name, fields=fields)
ind = SearchIndexClient(
endpoint, AzureKeyCredential(api_key)).create_index(index)
return SearchIndexer(name=name,
data_source_name=ds.name,
target_index_name=ind.name)
def test_list_datasource(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
data_source_connection1 = self._create_data_source_connection()
data_source_connection2 = self._create_data_source_connection(name="another-sample")
created1 = client.create_data_source_connection(data_source_connection1)
created2 = client.create_data_source_connection(data_source_connection2)
result = client.get_data_source_connections()
assert isinstance(result, list)
assert set(x.name for x in result) == {"sample-datasource", "another-sample"}
def _create_skillset():
client = SearchIndexerClient(service_endpoint, AzureKeyCredential(key))
inp = InputFieldMappingEntry(name="text", source="/document/lastRenovationDate")
output = OutputFieldMappingEntry(name="dateTimes", target_name="RenovatedDate")
s = EntityRecognitionSkill(name="merge-skill", inputs=[inp], outputs=[output])
skillset = SearchIndexerSkillset(name='hotel-data-skill', skills=[s], description="example skillset")
result = client.create_skillset(skillset)
return result
def test_list_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer1 = self._prepare_indexer(endpoint, api_key)
indexer2 = self._prepare_indexer(endpoint, api_key, name="another-indexer", ds_name="another-datasource", id_name="another-index")
created1 = client.create_indexer(indexer1)
created2 = client.create_indexer(indexer2)
result = client.get_indexers()
assert isinstance(result, list)
assert set(x.name for x in result) == {"sample-indexer", "another-indexer"}
def _create_datasource():
# Here we create a datasource. As mentioned in the description we have stored it in
# "searchcontainer"
ds_client = SearchIndexerClient(service_endpoint, AzureKeyCredential(key))
container = SearchIndexerDataContainer(name='searchcontainer')
data_source = ds_client.create_datasource(
name="hotel-datasource",
type="azureblob",
connection_string=connection_string,
container=container)
return data_source
def test_indexer_endpoint_https(self):
credential = AzureKeyCredential(key="old_api_key")
client = SearchIndexerClient("endpoint", credential)
assert client._endpoint.startswith('https')
client = SearchIndexerClient("https://endpoint", credential)
assert client._endpoint.startswith('https')
with pytest.raises(ValueError):
client = SearchIndexerClient("http://endpoint", credential)
with pytest.raises(ValueError):
client = SearchIndexerClient(12345, credential)
def test_create_or_update_skillset_if_unchanged(self, api_key, endpoint,
index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
ss = client.create_or_update_skillset(name='test-ss',
skills=[s],
description="desc1")
etag = ss.e_tag
client.create_or_update_skillset(name='test-ss',
skills=[s],
description="desc2",
skillset=ss)
assert len(client.get_skillsets()) == 1
ss.e_tag = etag
with pytest.raises(HttpResponseError):
client.create_or_update_skillset(
name='test-ss',
skills=[s],
skillset=ss,
match_condition=MatchConditions.IfNotModified)
def test_create_skillset_validation(self, **kwargs):
with pytest.raises(ValueError) as err:
client = SearchIndexerClient("fake_endpoint",
AzureKeyCredential("fake_key"))
name = "test-ss"
s1 = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizationsS1")
],
description="Skill Version 1",
model_version="1",
include_typeless_entities=True)
s2 = EntityRecognitionSkill(
inputs=[
InputFieldMappingEntry(name="text",
source="/document/content")
],
outputs=[
OutputFieldMappingEntry(name="organizations",
target_name="organizationsS2")
],
skill_version=EntityRecognitionSkillVersion.LATEST,
description="Skill Version 3",
model_version="3",
include_typeless_entities=True)
s3 = SentimentSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="score", target_name="scoreS3")
],
skill_version=SentimentSkillVersion.V1,
description="Sentiment V1",
include_opinion_mining=True)
skillset = SearchIndexerSkillset(name=name,
skills=list([s1, s2, s3]),
description="desc")
client.create_skillset(skillset)
assert 'include_typeless_entities' in str(err.value)
assert 'model_version' in str(err.value)
assert 'include_opinion_mining' in str(err.value)
def test_search_indexers(self, endpoint, api_key, **kwargs):
storage_cs = kwargs.get("search_storage_connection_string")
container_name = kwargs.get("search_storage_container_name")
client = SearchIndexerClient(endpoint, api_key)
index_client = SearchIndexClient(endpoint, api_key)
self._test_create_indexer(client, index_client, storage_cs,
container_name)
self._test_delete_indexer(client, index_client, storage_cs,
container_name)
self._test_get_indexer(client, index_client, storage_cs,
container_name)
self._test_list_indexer(client, index_client, storage_cs,
container_name)
self._test_create_or_update_indexer(client, index_client, storage_cs,
container_name)
self._test_reset_indexer(client, index_client, storage_cs,
container_name)
self._test_run_indexer(client, index_client, storage_cs,
container_name)
self._test_get_indexer_status(client, index_client, storage_cs,
container_name)
self._test_create_or_update_indexer_if_unchanged(
client, index_client, storage_cs, container_name)
self._test_delete_indexer_if_unchanged(client, index_client,
storage_cs, container_name)
def test_delete_datasource(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
data_source = self._create_datasource()
result = client.create_datasource(data_source)
assert len(client.get_datasources()) == 1
client.delete_datasource("sample-datasource")
assert len(client.get_datasources()) == 0
def test_delete_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
result = client.create_indexer(indexer)
assert len(client.get_indexers()) == 1
client.delete_indexer("sample-indexer")
assert len(client.get_indexers()) == 0
def test_create_or_update_datasource(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
data_source_connection = self._create_data_source_connection()
created = client.create_data_source_connection(data_source_connection)
assert len(client.get_data_source_connections()) == 1
data_source_connection.description = "updated"
client.create_or_update_data_source_connection(data_source_connection)
assert len(client.get_data_source_connections()) == 1
result = client.get_data_source_connection("sample-datasource")
assert result.name == "sample-datasource"
assert result.description == "updated"
def test_create_or_update_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
created = client.create_indexer(indexer)
assert len(client.get_indexers()) == 1
indexer.description = "updated"
client.create_or_update_indexer(indexer)
assert len(client.get_indexers()) == 1
result = client.get_indexer("sample-indexer")
assert result.name == "sample-indexer"
assert result.description == "updated"
def test_run_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
result = client.create_indexer(indexer)
assert len(client.get_indexers()) == 1
start = time.time()
client.run_indexer("sample-indexer")
assert client.get_indexer_status("sample-indexer").status == 'running'
def sample_indexer_workflow():
# Now that we have a datasource and an index, we can create an indexer.
skillset_name = _create_skillset().name
print("Skillset is created")
ds_name = _create_datasource().name
print("Data source is created")
ind_name = _create_index().name
print("Index is created")
# we pass the data source, skillsets and targeted index to build an indexer
parameters = IndexingParameters(configuration={"parsingMode": "jsonArray"})
indexer = SearchIndexer(
name="hotel-data-indexer",
data_source_name=ds_name,
target_index_name=ind_name,
skillset_name=skillset_name,
parameters=parameters
)
indexer_client = SearchIndexerClient(service_endpoint, AzureKeyCredential(key))
indexer_client.create_indexer(indexer) # create the indexer
# to get an indexer
result = indexer_client.get_indexer("hotel-data-indexer")
print(result)
# To run an indexer, we can use run_indexer()
indexer_client.run_indexer(result.name)
# Using create or update to schedule an indexer
schedule = IndexingSchedule(interval=datetime.timedelta(hours=24))
result.schedule = schedule
updated_indexer = indexer_client.create_or_update_indexer(result)
print(updated_indexer)
# get the status of an indexer
indexer_client.get_indexer_status(updated_indexer.name)
def test_data_source(self, endpoint, api_key, **kwargs):
storage_cs = kwargs.get("search_storage_connection_string")
client = SearchIndexerClient(endpoint, api_key)
self._test_create_datasource(client, storage_cs)
self._test_delete_datasource(client, storage_cs)
self._test_get_datasource(client, storage_cs)
self._test_list_datasources(client, storage_cs)
self._test_create_or_update_datasource(client, storage_cs)
self._test_create_or_update_datasource_if_unchanged(client, storage_cs)
self._test_delete_datasource_if_unchanged(client, storage_cs)
self._test_delete_datasource_string_if_unchanged(client, storage_cs)
def test_skillset_crud(self, api_key, endpoint):
client = SearchIndexerClient(endpoint, api_key)
self._test_create_skillset_validation()
self._test_create_skillset(client)
self._test_get_skillset(client)
self._test_get_skillsets(client)
self._test_create_or_update_skillset(client)
self._test_create_or_update_skillset_if_unchanged(client)
self._test_create_or_update_skillset_inplace(client)
self._test_delete_skillset_if_unchanged(client)
self._test_delete_skillset(client)
def test_indexer_credential_roll(self):
credential = AzureKeyCredential(key="old_api_key")
client = SearchIndexerClient("endpoint", credential)
assert client._headers == {
"api-key": "old_api_key",
"Accept": "application/json;odata.metadata=minimal",
}
credential.update("new_api_key")
assert client._headers == {
"api-key": "new_api_key",
"Accept": "application/json;odata.metadata=minimal",
}
def test_reset_indexer(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
result = client.create_indexer(indexer)
assert len(client.get_indexers()) == 1
result = client.reset_indexer("sample-indexer")
assert client.get_indexer_status("sample-indexer").last_result.status in ('InProgress', 'reset')
def test_delete_skillset_if_unchanged(self, api_key, endpoint, index_name,
**kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
skillset = SearchIndexerSkillset(name='test-ss',
skills=list([s]),
description="desc")
result = client.create_skillset(skillset)
etag = result.e_tag
skillset = SearchIndexerSkillset(name='test-ss',
skills=list([s]),
description="updated")
updated = client.create_or_update_skillset(skillset)
updated.e_tag = etag
with pytest.raises(HttpResponseError):
client.delete_skillset(
updated, match_condition=MatchConditions.IfNotModified)
def test_delete_datasource_if_unchanged(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
data_source_connection = self._create_data_source_connection()
created = client.create_data_source_connection(data_source_connection)
etag = created.e_tag
# Now update the data source connection
data_source_connection.description = "updated"
client.create_or_update_data_source_connection(data_source_connection)
# prepare data source connection
data_source_connection.e_tag = etag # reset to the original data source connection
with pytest.raises(HttpResponseError):
client.delete_data_source_connection(data_source_connection, match_condition=MatchConditions.IfNotModified)
assert len(client.get_data_source_connections()) == 1
def test_delete_indexer_if_unchanged(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
indexer = self._prepare_indexer(endpoint, api_key)
result = client.create_indexer(indexer)
etag = result.e_tag
indexer.description = "updated"
client.create_or_update_indexer(indexer)
indexer.e_tag = etag
with pytest.raises(HttpResponseError):
client.delete_indexer(indexer, match_condition=MatchConditions.IfNotModified)
def test_skillset_crud(self, api_key, endpoint):
client = SearchIndexerClient(endpoint, api_key)
self._test_create_skillset_validation()
self._test_create_skillset(client)
self._test_get_skillset(client)
self._test_get_skillsets(client)
# TODO: Disabled due to service regression. See #22769
#self._test_create_or_update_skillset(client)
self._test_create_or_update_skillset_if_unchanged(client)
# TODO: Disabled due to service regression. See #22769
#self._test_create_or_update_skillset_inplace(client)
# TODO: Disabled due to service regression. See #22769
#self._test_delete_skillset_if_unchanged(client)
self._test_delete_skillset(client)
def test_delete_datasource_string_if_unchanged(self, api_key, endpoint,
index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
data_source = self._create_datasource()
created = client.create_datasource(data_source)
etag = created.e_tag
# Now update the data source
data_source.description = "updated"
client.create_or_update_datasource(data_source)
# prepare data source
data_source.e_tag = etag # reset to the original datasource
with pytest.raises(ValueError):
client.delete_datasource(
data_source.name,
match_condition=MatchConditions.IfNotModified)
def test_delete_skillset(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
result = client.create_skillset(name='test-ss',
skills=[s],
description="desc")
assert len(client.get_skillsets()) == 1
client.delete_skillset("test-ss")
assert len(client.get_skillsets()) == 0
def test_get_skillsets(self, api_key, endpoint, index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
client.create_skillset(name='test-ss-1',
skills=[s],
description="desc1")
client.create_skillset(name='test-ss-2',
skills=[s],
description="desc2")
result = client.get_skillsets()
assert isinstance(result, list)
assert all(isinstance(x, SearchIndexerSkillset) for x in result)
assert set(x.name for x in result) == {"test-ss-1", "test-ss-2"}
def test_create_or_update_skillset_if_unchanged(self, api_key, endpoint,
index_name, **kwargs):
client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
s = EntityRecognitionSkill(inputs=[
InputFieldMappingEntry(name="text", source="/document/content")
],
outputs=[
OutputFieldMappingEntry(
name="organizations",
target_name="organizations")
])
skillset1 = SearchIndexerSkillset(name='test-ss',
skills=list([s]),
description="desc1")
ss = client.create_or_update_skillset(skillset1)
etag = ss.e_tag
skillset2 = SearchIndexerSkillset(name='test-ss',
skills=[s],
description="desc2",
skillset=ss)
client.create_or_update_skillset(skillset2)
assert len(client.get_skillsets()) == 1
def test_indexer_init(self):
client = SearchIndexerClient("endpoint", CREDENTIAL)
assert client._headers == {
"api-key": "test_api_key",
"Accept": "application/json;odata.metadata=minimal",
}
