def test_add_empty_index():
"""test index add when index fields is empty string."""
schema_json = {"file_name": {"type": "string"}, "label": {"type": "number"}}
header = ShardHeader()
schema = header.build_schema(schema_json, ["data"], "img") # create schema
header.add_schema(schema) # add schema
with pytest.raises(Exception, match="incompatible"):
header.add_index_fields("")
def test_add_index_with_incorrect_field():
"""test index add with incorrect field(64)."""
header = ShardHeader()
mkv_schema_json = {"file_name": {"type": "string"},
"id": {"type": "number"}, "prelabel": {"type": "string"}}
schema = header.build_schema(mkv_schema_json, ["data"], "mkv_schema")
header.add_schema(schema)
with pytest.raises(Exception, match="incompatible function arguments"):
header.add_index_fields([(-1, "id")])
def test_file_writer_fail_add_index():
"""test file writer, read when failed on adding index."""
data_raw = get_data("../data/mindrecord/testImageNetData/")
schema_json = {
"file_name": {
"type": "string"
},
"label": {
"type": "number"
}
}
header = ShardHeader()
schema = header.build_schema(schema_json, ["data"], "img") # create schema
schema_id = header.add_schema(schema) # add schema
with pytest.raises(TypeError, match="missing 1 "):
ret = header.add_index_fields()
assert ret == FAILED
with pytest.raises(MRMAddIndexError):
index_fields = []
ret = header.add_index_fields(index_fields)
assert ret == FAILED
file_name = os.path.join(os.getcwd(),
"test_001.mindrecord") # set output filename
writer = ShardWriter() # test_file_writer
ret = writer.open([file_name])
assert ret == SUCCESS, 'failed on opening files.'
ret = writer.set_shard_header(header) # write header
assert ret == SUCCESS, 'failed on setting header.'
ret = writer.write_raw_cv_data({schema_id: data_raw})
assert ret == SUCCESS, 'failed on writing raw data.'
ret = writer.commit() # commit data
assert ret == SUCCESS, "commit failed"
# ShardIndexGenerator
generator = ShardIndexGenerator(os.path.realpath(file_name))
generator.build()
generator.write_to_db()
reader = ShardReader()
ret = reader.open(file_name)
reader.launch()
index = 0
_, blob_fields = reader.get_blob_fields()
iterator = reader.get_next()
while iterator:
for blob, raw in iterator:
raw[blob_fields[0]] = bytes(blob)
logger.info("#item{}: {}".format(index, raw))
index += 1
iterator = reader.get_next()
reader.finish()
reader.close()
os.remove("{}".format(file_name))
os.remove("{}.db".format(file_name))
def test_mindrecord_add_index_016():
"""test index add when index fields are incorrect."""
schema_json = {"id": {"type": "number"}, "name": {"type": "string"},
"label": {"type": "string"}, "key": {"type": "string"}}
header = ShardHeader()
schema = header.build_schema(schema_json, ["data"], "img")
header.add_schema(schema)
index_fields_list = [(0, "id")]
with pytest.raises(Exception):
header.add_index_fields(index_fields_list)
def test_add_index_with_string_list():
"""test index add with list of string(64)."""
header = ShardHeader()
schema_json = {"id": {"type": "number"}, "name": {"type": "string"},
"label": {"type": "string"}, "key": {"type": "string"}}
schema = header.build_schema(schema_json, ["key"], "schema_desc")
header.add_schema(schema)
ret = header.add_index_fields(["id", "label"])
assert ret == SUCCESS
def test_nlp_file_writer():
"""test nlp file writer using shard api"""
schema_json = {
"id": {
"type": "string"
},
"label": {
"type": "number"
},
"rating": {
"type": "number"
},
"input_ids": {
"type": "array",
"items": {
"type": "number"
}
},
"input_mask": {
"type": "array",
"items": {
"type": "number"
}
},
"segment_ids": {
"type": "array",
"items": {
"type": "number"
}
}
}
data = list(
get_nlp_data("../data/mindrecord/testAclImdbData/pos",
"../data/mindrecord/testAclImdbData/vocab.txt", 10))
header = ShardHeader()
schema = header.build_schema(schema_json, ["segment_ids"], "nlp_schema")
schema_id = header.add_schema(schema)
assert schema_id == 0, 'failed on adding schema'
index_fields_list = ["id", "rating"]
ret = header.add_index_fields(index_fields_list)
assert ret == SUCCESS, 'failed on adding index fields.'
writer = ShardWriter()
paths = ["{}{}".format(NLP_FILE_NAME, x) for x in range(FILES_NUM)]
ret = writer.open(paths)
assert ret == SUCCESS, 'failed on opening files.'
writer.set_header_size(1 << 14)
writer.set_page_size(1 << 15)
ret = writer.set_shard_header(header)
assert ret == SUCCESS, 'failed on setting header.'
ret = writer.write_raw_nlp_data({schema_id: data})
assert ret == SUCCESS, 'failed on writing raw data.'
ret = writer.commit()
assert ret == SUCCESS, 'failed on committing.'
generator = ShardIndexGenerator(os.path.realpath(paths[0]))
generator.build()
generator.write_to_db()
def test_mindrecord_add_index_011():
"""test index add"""
schema_json = {"id": {"type": "number"}, "name": {"type": "string"},
"label": {"type": "string"}, "key": {"type": "string"}}
header = ShardHeader()
schema = header.build_schema(schema_json, ["data"], "img") # create schema
header.add_schema(schema) # add schema
index_fields_list = ["id", "name", "label", "key"]
ret = header.add_index_fields(index_fields_list)
assert ret == 0, 'failed on adding index fields.'
def test_mkv_file_writer():
"""test mkv file writer using shard api"""
data = get_mkv_data("../data/mindrecord/testVehPerData/")
schema_json = {
"file_name": {
"type": "string"
},
"id": {
"type": "number"
},
"prelabel": {
"type": "string"
}
}
header = ShardHeader()
img_schema = header.build_schema(schema_json, ["data"], "img_schema")
schema_id = header.add_schema(img_schema)
assert schema_id == 0, 'failed on building schema.'
index_fields_list = ["id", "file_name"]
ret = header.add_index_fields(index_fields_list)
assert ret == SUCCESS, 'failed on adding index fields.'
writer = ShardWriter()
paths = ["{}{}".format(MKV_FILE_NAME, x) for x in range(FILES_NUM)]
ret = writer.open(paths)
assert ret == SUCCESS, 'failed on opening files.'
writer.set_header_size(1 << 24)
writer.set_page_size(1 << 25)
ret = writer.set_shard_header(header)
assert ret == SUCCESS, 'failed on setting header.'
ret = writer.write_raw_cv_data({schema_id: data})
assert ret == SUCCESS, 'failed on writing raw data.'
ret = writer.commit()
assert ret == SUCCESS, 'failed on committing.'
generator = ShardIndexGenerator(os.path.realpath(paths[0]))
generator.build()
generator.write_to_db()