def test_create_or_update_index(self, api_key, endpoint, index_name,
**kwargs):
name = "hotels"
fields = [
SimpleField(name="hotelId",
type=SearchFieldDataType.String,
key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double)
]
cors_options = CorsOptions(allowed_origins=["*"],
max_age_in_seconds=60)
scoring_profiles = []
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
client = SearchIndexClient(endpoint, AzureKeyCredential(api_key))
result = client.create_or_update_index(index=index)
assert len(result.scoring_profiles) == 0
assert result.cors_options.allowed_origins == cors_options.allowed_origins
assert result.cors_options.max_age_in_seconds == cors_options.max_age_in_seconds
scoring_profile = ScoringProfile(name="MyProfile")
scoring_profiles = []
scoring_profiles.append(scoring_profile)
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = client.create_or_update_index(index=index)
assert result.scoring_profiles[0].name == scoring_profile.name
assert result.cors_options.allowed_origins == cors_options.allowed_origins
assert result.cors_options.max_age_in_seconds == cors_options.max_age_in_seconds
async def _test_create_or_update_index(self, client):
name = "hotels-cou"
fields = fields = [
SimpleField(name="hotelId",
type=SearchFieldDataType.String,
key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double)
]
cors_options = CorsOptions(allowed_origins=["*"],
max_age_in_seconds=60)
scoring_profiles = []
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = await client.create_or_update_index(index=index)
assert len(result.scoring_profiles) == 0
assert result.cors_options.allowed_origins == cors_options.allowed_origins
assert result.cors_options.max_age_in_seconds == cors_options.max_age_in_seconds
scoring_profile = ScoringProfile(name="MyProfile")
scoring_profiles = []
scoring_profiles.append(scoring_profile)
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = await client.create_or_update_index(index=index)
assert result.scoring_profiles[0].name == scoring_profile.name
assert result.cors_options.allowed_origins == cors_options.allowed_origins
assert result.cors_options.max_age_in_seconds == cors_options.max_age_in_seconds
def test_pack_search_index():
pattern_analyzer = PatternAnalyzer(
name="test_analyzer",
flags=["CANON_EQ"]
)
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = PatternTokenizer(
name="test_tokenizer",
flags=["CANON_EQ"]
)
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(
name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers
)
result = pack_search_index(index)
assert isinstance(result.analyzers[0], _PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, str)
assert result.analyzers[0].flags == "CANON_EQ"
assert isinstance(result.tokenizers[0], _PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, str)
assert result.tokenizers[0].flags == "CANON_EQ"
def test_unpack_search_index_enum():
pattern_analyzer = _PatternAnalyzer(
name="test_analyzer",
flags=RegexFlags.canon_eq
)
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = _PatternTokenizer(
name="test_tokenizer",
flags=RegexFlags.canon_eq
)
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(
name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers
)
result = unpack_search_index(index)
assert isinstance(result.analyzers[0], PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, list)
assert result.analyzers[0].flags[0] == "CANON_EQ"
assert isinstance(result.tokenizers[0], PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, list)
assert result.tokenizers[0].flags[0] == "CANON_EQ"
def test_multi_unpack_search_index():
pattern_analyzer = _PatternAnalyzer(
name="test_analyzer",
flags="CANON_EQ|MULTILINE"
)
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = _PatternTokenizer(
name="test_tokenizer",
flags="CANON_EQ|MULTILINE"
)
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(
name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers
)
result = unpack_search_index(index)
assert isinstance(result.analyzers[0], PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, list)
assert result.analyzers[0].flags[0] == "CANON_EQ"
assert result.analyzers[0].flags[1] == "MULTILINE"
assert isinstance(result.tokenizers[0], PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, list)
assert result.tokenizers[0].flags[0] == "CANON_EQ"
assert result.tokenizers[0].flags[1] == "MULTILINE"
def create_index(name, endpoint, key):
# Create a service client
client = SearchIndexClient(endpoint, AzureKeyCredential(key))
fields = [
SimpleField(name='Id', type=SearchFieldDataType.String, key=True),
SearchableField(name='FileName', type=SearchFieldDataType.String),
SimpleField(name='FilePath', type=SearchFieldDataType.String),
SearchableField(name='KeyPhrases',
collection=True,
type=SearchFieldDataType.String,
analyzer_name="en.lucene"),
SearchableField(name='People',
collection=True,
type=SearchFieldDataType.String),
SearchableField(name='Organisation',
collection=True,
type=SearchFieldDataType.String),
SearchableField(name='Location',
collection=True,
type=SearchFieldDataType.String)
]
cors_options = CorsOptions(allowed_origins=["*"], max_age_in_seconds=60)
scoring_profiles = []
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = client.create_index(index)
async def _test_delete_indexes_if_unchanged(self, client):
# First create an index
name = "hotels-del-unchanged"
fields = [{
"name": "hotelId",
"type": "Edm.String",
"key": True,
"searchable": False
}, {
"name": "baseRate",
"type": "Edm.Double"
}]
scoring_profile = ScoringProfile(name="MyProfile")
scoring_profiles = []
scoring_profiles.append(scoring_profile)
cors_options = CorsOptions(allowed_origins=["*"],
max_age_in_seconds=60)
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = await client.create_index(index)
etag = result.e_tag
# get eTag and update
index.scoring_profiles = []
await client.create_or_update_index(index)
index.e_tag = etag
with pytest.raises(HttpResponseError):
await client.delete_index(
index, match_condition=MatchConditions.IfNotModified)
def test_delete_indexes_if_unchanged(self, api_key, endpoint, index_name,
**kwargs):
client = SearchIndexClient(endpoint, AzureKeyCredential(api_key))
# First create an index
name = "hotels"
fields = [{
"name": "hotelId",
"type": "Edm.String",
"key": True,
"searchable": False
}, {
"name": "baseRate",
"type": "Edm.Double"
}]
scoring_profile = ScoringProfile(name="MyProfile")
scoring_profiles = []
scoring_profiles.append(scoring_profile)
cors_options = CorsOptions(allowed_origins=["*"],
max_age_in_seconds=60)
index = SearchIndex(name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = client.create_index(index)
etag = result.e_tag
# get e tag and update
index.scoring_profiles = []
client.create_or_update_index(index)
index.e_tag = etag
with pytest.raises(HttpResponseError):
client.delete_index(index,
match_condition=MatchConditions.IfNotModified)
async def update_index():
# [START update_index_async]
name = "hotels"
fields = [
SimpleField(name="hotelId", type=SearchFieldDataType.String, key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double),
SearchableField(name="description", type=SearchFieldDataType.String, collection=True),
SearchableField(name="hotelName", type=SearchFieldDataType.String),
ComplexField(name="address", fields=[
SimpleField(name="streetAddress", type=SearchFieldDataType.String),
SimpleField(name="city", type=SearchFieldDataType.String),
SimpleField(name="state", type=SearchFieldDataType.String),
], collection=True)
]
cors_options = CorsOptions(allowed_origins=["*"], max_age_in_seconds=60)
scoring_profile = ScoringProfile(
name="MyProfile"
)
scoring_profiles = []
scoring_profiles.append(scoring_profile)
index = SearchIndex(
name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = await client.create_or_update_index(index=index)
async def create_indexer():
# create an index
index_name = "hotels"
fields = [
SimpleField(name="hotelId", type=SearchFieldDataType.String, key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double)
]
index = SearchIndex(name=index_name, fields=fields)
ind_client = SearchIndexerClient(service_endpoint, AzureKeyCredential(key))
async with ind_client:
await ind_client.create_index(index)
# [START create_indexer_async]
# create a datasource
container = SearchIndexerDataContainer(name='searchcontainer')
data_source_connection = SearchIndexerDataSourceConnection(
name="indexer-datasource",
type="azureblob",
connection_string=connection_string,
container=container)
async with ind_client:
data_source = await ind_client.create_data_source_connection(
data_source_connection)
# create an indexer
indexer = SearchIndexer(name="async-sample-indexer",
data_source_name="async-indexer-datasource",
target_index_name="indexer-hotels")
async with indexers_client:
result = await indexers_client.create_indexer(indexer)
print("Create new Indexer - async-sample-indexer")
async 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_connection = SearchIndexerDataSourceConnection(
name=ds_name,
type="azureblob",
connection_string=con_str,
container=container
)
ds_client = SearchIndexerClient(endpoint, AzureKeyCredential(api_key))
ds = await ds_client.create_data_source_connection(data_source_connection)
index_name = id_name
fields = [
{
"name": "hotelId",
"type": "Edm.String",
"key": True,
"searchable": False
}]
index = SearchIndex(name=index_name, fields=fields)
ind_client = SearchIndexClient(endpoint, AzureKeyCredential(api_key))
ind = await ind_client.create_index(index)
return SearchIndexer(name=name, data_source_name=ds.name, target_index_name=ind.name)
def _create_index():
name = "hotel-index"
# Here we create an index with listed fields.
fields = [
SimpleField(name="hotelId",
type=SearchFieldDataType.String,
filterable=True,
sortable=True,
key=True),
SearchableField(name="hotelName", type=SearchFieldDataType.String),
SimpleField(name="description", type=SearchFieldDataType.String),
SimpleField(name="descriptionFr", type=SearchFieldDataType.String),
SimpleField(name="category", type=SearchFieldDataType.String),
SimpleField(name="parkingIncluded",
type=SearchFieldDataType.Boolean,
filterable=True),
SimpleField(name="smokingAllowed",
type=SearchFieldDataType.Boolean,
filterable=True),
SimpleField(name="lastRenovationDate",
type=SearchFieldDataType.String),
SimpleField(name="rating",
type=SearchFieldDataType.Int64,
sortable=True),
SimpleField(name="location", type=SearchFieldDataType.GeographyPoint),
]
cors_options = CorsOptions(allowed_origins=["*"], max_age_in_seconds=60)
# pass in the name, fields and cors options and create the index
index = SearchIndex(name=name, fields=fields, cors_options=cors_options)
index_client = SearchIndexClient(service_endpoint, AzureKeyCredential(key))
result = index_client.create_index(index)
return result
def _create_index(self, client, index_name):
fields = [
SimpleField(name="hotelId",
type=SearchFieldDataType.String,
key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double),
]
scoring_profile = ScoringProfile(name="MyProfile")
scoring_profiles = []
scoring_profiles.append(scoring_profile)
cors_options = CorsOptions(allowed_origins=["*"],
max_age_in_seconds=60)
index = SearchIndex(name=index_name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = client.create_index(index)
def test_listize_flags_for_index():
pattern_analyzer = _PatternAnalyzer(name="test_analyzer", flags="CANON_EQ")
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = _PatternTokenizer(name="test_tokenizer",
flags="CANON_EQ")
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers)
result = listize_flags_for_index(index)
assert isinstance(result.analyzers[0], PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, list)
assert result.analyzers[0].flags[0] == "CANON_EQ"
assert isinstance(result.tokenizers[0], PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, list)
assert result.tokenizers[0].flags[0] == "CANON_EQ"
def test_delistize_multi_flags_for_index():
pattern_analyzer = PatternAnalyzer(name="test_analyzer",
flags=["CANON_EQ", "MULTILINE"])
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = PatternTokenizer(name="test_analyzer",
flags=["CANON_EQ", "MULTILINE"])
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers)
result = delistize_flags_for_index(index)
assert isinstance(result.analyzers[0], _PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, str)
assert result.analyzers[0].flags == "CANON_EQ|MULTILINE"
assert isinstance(result.tokenizers[0], _PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, str)
assert result.tokenizers[0].flags == "CANON_EQ|MULTILINE"
def _prepare_indexer(self, client, index_client, storage_cs, name,
container_name):
data_source_connection = SearchIndexerDataSourceConnection(
name=f"{name}-ds",
type="azureblob",
connection_string=storage_cs,
container=SearchIndexerDataContainer(name=container_name))
ds = client.create_data_source_connection(data_source_connection)
fields = [{
"name": "hotelId",
"type": "Edm.String",
"key": True,
"searchable": False
}]
index = SearchIndex(name=f"{name}-hotels", fields=fields)
ind = index_client.create_index(index)
return SearchIndexer(name=name,
data_source_name=ds.name,
target_index_name=ind.name)
def test_multi_pack_search_index():
pattern_analyzer = PatternAnalyzer(name="test_analyzer",
flags=["CANON_EQ", "MULTILINE"])
analyzers = []
analyzers.append(pattern_analyzer)
pattern_tokenizer = PatternTokenizer(name="test_analyzer",
flags=["CANON_EQ", "MULTILINE"])
tokenizers = []
tokenizers.append(pattern_tokenizer)
index = SearchIndex(name="test",
fields=None,
analyzers=analyzers,
tokenizers=tokenizers)
result = index._to_generated()
assert isinstance(result.analyzers[0], _PatternAnalyzer)
assert isinstance(result.analyzers[0].flags, str)
assert result.analyzers[0].flags == "CANON_EQ|MULTILINE"
assert isinstance(result.tokenizers[0], _PatternTokenizer)
assert isinstance(result.tokenizers[0].flags, str)
assert result.tokenizers[0].flags == "CANON_EQ|MULTILINE"
def create_index():
# [START create_index]
name = "hotels"
fields = [
SimpleField(name="hotelId", type=SearchFieldDataType.String, key=True),
SimpleField(name="baseRate", type=SearchFieldDataType.Double),
SearchableField(name="description", type=SearchFieldDataType.String, collection=True),
ComplexField(name="address", fields=[
SimpleField(name="streetAddress", type=SearchFieldDataType.String),
SimpleField(name="city", type=SearchFieldDataType.String),
], collection=True)
]
cors_options = CorsOptions(allowed_origins=["*"], max_age_in_seconds=60)
scoring_profiles = []
index = SearchIndex(
name=name,
fields=fields,
scoring_profiles=scoring_profiles,
cors_options=cors_options)
result = client.create_index(index)
def create_schema_from_json_and_upload(schema,
index_name,
admin_client,
url=False):
cors_options = CorsOptions(allowed_origins=["*"], max_age_in_seconds=60)
scoring_profiles = []
schema_data = get_schema_data(schema, url)
index = SearchIndex(name=index_name,
fields=schema_data['fields'],
scoring_profiles=scoring_profiles,
suggesters=schema_data['suggesters'],
cors_options=cors_options)
try:
upload_schema = admin_client.create_index(index)
if upload_schema:
print(f'Schema uploaded; Index created for {index_name}.')
else:
exit(0)
except:
print("Unexpected error:", sys.exc_info()[0])
