File size: 80,927 Bytes
5980447 |
1 2 |
{"repo": "learningOrchestra/mlToolKits", "pull_number": 108, "instance_id": "learningOrchestra__mlToolKits-108", "issue_numbers": "", "base_commit": "6a83b4bcbdf5096b7ed855bcebab358511d19f2c", "patch": "diff --git a/microservices/data_type_handler_image/data_type_handler.py b/microservices/data_type_handler_image/data_type_handler.py\n--- a/microservices/data_type_handler_image/data_type_handler.py\n+++ b/microservices/data_type_handler_image/data_type_handler.py\n@@ -1,4 +1,7 @@\n from pymongo import MongoClient\n+from datetime import datetime\n+import pytz\n+from concurrent.futures import ThreadPoolExecutor\n \n \n class DataTypeConverter:\n@@ -7,8 +10,10 @@ class DataTypeConverter:\n STRING_TYPE = \"string\"\n NUMBER_TYPE = \"number\"\n \n- def __init__(self, database_connector):\n+ def __init__(self, database_connector, metadata_handler):\n self.database_connector = database_connector\n+ self.thread_pool = ThreadPoolExecutor()\n+ self.metadata_handler = metadata_handler\n \n def field_converter(self, filename, field, field_type):\n query = {}\n@@ -27,9 +32,9 @@ def field_converter(self, filename, field, field_type):\n \n elif field_type == self.NUMBER_TYPE:\n if (\n- document[field] == int\n- or document[field] == float\n- or document[field] is None\n+ document[field] == int\n+ or document[field] == float\n+ or document[field] is None\n ):\n continue\n if document[field] == \"\":\n@@ -42,15 +47,53 @@ def field_converter(self, filename, field, field_type):\n \n self.database_connector.update_one(filename, values, document)\n \n- def file_converter(self, filename, fields_dictionary):\n+ def convert_existent_file(self, filename, fields_dictionary):\n+\n+ self.metadata_handler.update_finished_metadata_file(filename, False)\n+\n+ self.thread_pool.submit(self.field_file_converter, filename,\n+ fields_dictionary)\n \n+ def field_file_converter(self, filename, fields_dictionary):\n for field in fields_dictionary:\n- self.field_converter(filename, field, fields_dictionary[field])\n+ self.field_converter(filename, field,\n+ fields_dictionary[field])\n+\n+ self.metadata_handler.update_finished_metadata_file(filename, True)\n+\n+\n+class FileMetadataHandler:\n+ def __init__(self, database_connector):\n+ self.database_connector = database_connector\n+\n+ def create_metadata_file(self, filename):\n+ timezone_london = pytz.timezone(\"Etc/Greenwich\")\n+ london_time = datetime.now(timezone_london)\n+\n+ metadata_file = {\n+ \"filename\": filename,\n+ \"time_created\": london_time.strftime(\"%Y-%m-%dT%H:%M:%S-00:00\"),\n+ \"_id\": 0,\n+ \"finished\": False,\n+ \"type\": \"dataType\"\n+ }\n+ self.database_connector.insert_one_in_file(filename, metadata_file)\n+\n+ def update_finished_metadata_file(self, filename, flag):\n+ metadata_new_value = {\n+ \"finished\": flag,\n+ }\n+ metadata_query = {\n+ \"_id\": 0\n+ }\n+ self.database_connector.update_one(filename, metadata_new_value,\n+ metadata_query)\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def find(self, filename, query):\n@@ -69,6 +112,10 @@ def find_one(self, filename, query):\n file_collection = self.database[filename]\n return file_collection.find_one(query)\n \n+ def insert_one_in_file(self, filename, json_object):\n+ file_collection = self.database[filename]\n+ file_collection.insert_one(json_object)\n+\n \n class DataTypeHandlerRequestValidator:\n MESSAGE_INVALID_FIELDS = \"invalid_fields\"\n@@ -92,11 +139,13 @@ def fields_validator(self, filename, fields):\n \n filename_metadata_query = {\"filename\": filename}\n \n- filename_metadata = self.database.find_one(filename, filename_metadata_query)\n+ filename_metadata = self.database.find_one(filename,\n+ filename_metadata_query)\n \n for field in fields:\n if field not in filename_metadata[\"fields\"]:\n raise Exception(self.MESSAGE_INVALID_FIELDS)\n \n- if fields[field] != self.NUMBER_TYPE and fields[field] != self.STRING_TYPE:\n+ if fields[field] != self.NUMBER_TYPE and \\\n+ fields[field] != self.STRING_TYPE:\n raise Exception(self.MESSAGE_INVALID_FIELDS)\ndiff --git a/microservices/data_type_handler_image/server.py b/microservices/data_type_handler_image/server.py\n--- a/microservices/data_type_handler_image/server.py\n+++ b/microservices/data_type_handler_image/server.py\n@@ -3,7 +3,8 @@\n from data_type_handler import (\n MongoOperations,\n DataTypeHandlerRequestValidator,\n- DataTypeConverter)\n+ DataTypeConverter,\n+ FileMetadataHandler)\n \n HTTP_STATUS_CODE_SUCESS = 200\n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n@@ -16,12 +17,12 @@\n MESSAGE_RESULT = \"result\"\n \n FILENAME_NAME = \"filename\"\n-\n+FIELD_TYPES_NAMES = \"types\"\n+PARENT_FILENAME_NAME = \"input_filename\"\n FIRST_ARGUMENT = 0\n \n MESSAGE_INVALID_URL = \"invalid_url\"\n MESSAGE_DUPLICATE_FILE = \"duplicate_file\"\n-MESSAGE_CHANGED_FILE = \"file_changed\"\n MESSAGE_DELETED_FILE = \"deleted_file\"\n \n DATABASE_URL = \"DATABASE_URL\"\n@@ -29,51 +30,49 @@\n DATABASE_NAME = \"DATABASE_NAME\"\n DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n \n-PATCH = \"PATCH\"\n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/dataset/\"\n+MICROSERVICE_URI_GET_PARAMS = \"?query={}&limit=20&skip=0\"\n \n app = Flask(__name__)\n \n \n-def collection_database_url(database_url, database_name, database_filename,\n- database_replica_set):\n- return database_url + '/' + \\\n- database_name + '.' + \\\n- database_filename + \"?replicaSet=\" + \\\n- database_replica_set + \\\n- \"&authSource=admin\"\n-\n-\n-@app.route('/fieldtypes/<filename>', methods=[PATCH])\n-def change_data_type(filename):\n+@app.route('/fieldTypes', methods=[\"PATCH\"])\n+def change_data_type():\n database = MongoOperations(\n- os.environ[DATABASE_URL] + '/?replicaSet=' +\n- os.environ[DATABASE_REPLICA_SET], os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n os.environ[DATABASE_NAME])\n \n request_validator = DataTypeHandlerRequestValidator(database)\n \n try:\n- request_validator.filename_validator(filename)\n+ parent_filename = request.json[PARENT_FILENAME_NAME]\n+\n+ request_validator.filename_validator(parent_filename)\n except Exception as invalid_filename:\n return jsonify(\n- {MESSAGE_RESULT:\n- invalid_filename.args[FIRST_ARGUMENT]}),\\\n- HTTP_STATUS_CODE_NOT_ACCEPTABLE\n+ {MESSAGE_RESULT: invalid_filename.args[FIRST_ARGUMENT]}), \\\n+ HTTP_STATUS_CODE_NOT_ACCEPTABLE\n \n try:\n request_validator.fields_validator(\n- filename, request.json)\n+ parent_filename, request.json[FIELD_TYPES_NAMES])\n except Exception as invalid_fields:\n return jsonify(\n- {MESSAGE_RESULT: invalid_fields.args[FIRST_ARGUMENT]}),\\\n- HTTP_STATUS_CODE_NOT_ACCEPTABLE\n-\n- data_type_converter = DataTypeConverter(database)\n- data_type_converter.file_converter(\n- filename, request.json)\n-\n- return jsonify({MESSAGE_RESULT: MESSAGE_CHANGED_FILE}), \\\n- HTTP_STATUS_CODE_SUCESS\n+ {MESSAGE_RESULT: invalid_fields.args[FIRST_ARGUMENT]}), \\\n+ HTTP_STATUS_CODE_NOT_ACCEPTABLE\n+\n+ metadata_handler = FileMetadataHandler(database)\n+ data_type_converter = DataTypeConverter(database, metadata_handler)\n+ data_type_converter.convert_existent_file(\n+ parent_filename, request.json[FIELD_TYPES_NAMES])\n+\n+ return jsonify({\n+ MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[PARENT_FILENAME_NAME] +\n+ MICROSERVICE_URI_GET_PARAMS}), HTTP_STATUS_CODE_SUCESS\n \n \n if __name__ == \"__main__\":\ndiff --git a/microservices/database_api_image/database.py b/microservices/database_api_image/database.py\n--- a/microservices/database_api_image/database.py\n+++ b/microservices/database_api_image/database.py\n@@ -1,4 +1,3 @@\n-import os\n from pymongo import MongoClient, errors, ASCENDING\n from bson.json_util import dumps\n import requests\n@@ -25,7 +24,8 @@ def __init__(self, database_object, file_manager_object):\n \n def add_file(self, url, filename):\n try:\n- self.file_manager_object.storage_file(filename, url, self.database_object)\n+ self.file_manager_object.storage_file(filename, url,\n+ self.database_object)\n \n except requests.exceptions.RequestException:\n raise Exception(self.MESSAGE_INVALID_URL)\n@@ -41,7 +41,7 @@ def read_file(self, filename, skip, limit, query):\n limit = int(limit)\n \n for file in self.database_object.find_in_file(\n- filename, query_object, skip, limit\n+ filename, query_object, skip, limit\n ):\n result.append(json.loads(dumps(file)))\n \n@@ -50,13 +50,15 @@ def read_file(self, filename, skip, limit, query):\n def delete_file(self, filename):\n self.database_object.delete_file(filename)\n \n- def get_files(self):\n+ def get_files(self, type):\n result = []\n \n for file in self.database_object.get_filenames():\n metadata_file = self.database_object.find_one_in_file(\n- file, {ROW_ID: METADATA_ROW_ID}\n+ file, {ROW_ID: METADATA_ROW_ID, \"type\": type}\n )\n+ if metadata_file == None:\n+ continue\n metadata_file.pop(ROW_ID)\n result.append(metadata_file)\n \n@@ -67,11 +69,10 @@ class MongoOperations:\n DATABASE_URL = \"DATABASE_URL\"\n DATABASE_PORT = \"DATABASE_PORT\"\n \n- def __init__(self):\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n self.mongo_client = MongoClient(\n- os.environ[self.DATABASE_URL], int(os.environ[self.DATABASE_PORT])\n- )\n- self.database = self.mongo_client.database\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n+ self.database = self.mongo_client[database_name]\n \n def connection(self, filename):\n return self.database[filename]\n@@ -79,7 +80,8 @@ def connection(self, filename):\n def find_in_file(self, filename, query, skip=0, limit=1):\n file_collection = self.database[filename]\n return (\n- file_collection.find(query).sort(ROW_ID, ASCENDING).skip(skip).limit(limit)\n+ file_collection.find(query).sort(ROW_ID, ASCENDING).skip(\n+ skip).limit(limit)\n )\n \n def delete_file(self, filename):\n@@ -109,7 +111,8 @@ class CsvDownloader:\n file_headers = None\n \n def __init__(self):\n- self.thread_pool = ThreadPoolExecutor(max_workers=self.MAX_NUMBER_THREADS)\n+ self.thread_pool = ThreadPoolExecutor(\n+ max_workers=self.MAX_NUMBER_THREADS)\n self.download_tratament_queue = Queue(maxsize=self.MAX_QUEUE_SIZE)\n self.tratament_save_queue = Queue(maxsize=self.MAX_QUEUE_SIZE)\n \n@@ -152,7 +155,8 @@ def save_file(self, database_connection, filename):\n {\"$set\": {self.FINISHED: True, \"fields\": self.file_headers}},\n )\n \n- def validate_csv_url(self, url):\n+ @staticmethod\n+ def validate_csv_url(url):\n with closing(requests.get(url, stream=True)) as r:\n reader = csv.reader(\n codecs.iterdecode(r.iter_lines(), encoding=\"utf-8\"),\n@@ -163,8 +167,8 @@ def validate_csv_url(self, url):\n first_symbol_html = \"<\"\n first_symbol_json = \"{\"\n if (\n- first_line[0][0] == first_symbol_html\n- or first_line[0][0] == first_symbol_json\n+ first_line[0][0] == first_symbol_html\n+ or first_line[0][0] == first_symbol_json\n ):\n raise requests.exceptions.RequestException\n \n@@ -181,6 +185,7 @@ def storage_file(self, filename, url, database_connection):\n ROW_ID: METADATA_ROW_ID,\n self.FINISHED: False,\n \"fields\": \"processing\",\n+ \"type\": \"dataset\"\n }\n database_connection.insert_one_in_file(filename, metadata_file)\n self.thread_pool.submit(self.download_file, url)\ndiff --git a/microservices/database_api_image/server.py b/microservices/database_api_image/server.py\n--- a/microservices/database_api_image/server.py\n+++ b/microservices/database_api_image/server.py\n@@ -1,6 +1,7 @@\n from flask import jsonify, request, Flask\n import os\n from database import CsvDownloader, DatabaseApi, MongoOperations\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS = 200\n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n@@ -12,32 +13,40 @@\n \n MESSAGE_RESULT = \"result\"\n \n+DATABASE_URL = \"DATABASE_URL\"\n+DATABASE_PORT = \"DATABASE_PORT\"\n+DATABASE_NAME = \"DATABASE_NAME\"\n+DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n+\n FILENAME = \"filename\"\n+URL_FIELD_NAME = \"url\"\n \n FIRST_ARGUMENT = 0\n \n MESSAGE_INVALID_URL = \"invalid_url\"\n MESSAGE_DUPLICATE_FILE = \"duplicate_file\"\n-MESSAGE_CREATED_FILE = \"file_created\"\n MESSAGE_DELETED_FILE = \"deleted_file\"\n \n-GET = \"GET\"\n-POST = \"POST\"\n-DELETE = \"DELETE\"\n-\n PAGINATE_FILE_LIMIT = 20\n \n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/dataset/\"\n+MICROSERVICE_URI_GET_PARAMS = \"?query={}&limit=10&skip=0\"\n+\n app = Flask(__name__)\n \n \n-@app.route(\"/files\", methods=[POST])\n+@app.route(\"/files\", methods=[\"POST\"])\n def create_file():\n- file_downloader_and_saver = CsvDownloader()\n- mongo_operations = MongoOperations()\n- database = DatabaseApi(mongo_operations, file_downloader_and_saver)\n+ file_downloader = CsvDownloader()\n+ mongo_operations = MongoOperations(\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME])\n+ database = DatabaseApi(mongo_operations, file_downloader)\n \n try:\n- database.add_file(request.json[\"url\"], request.json[FILENAME])\n+ database.add_file(request.json[URL_FIELD_NAME], request.json[FILENAME])\n \n except Exception as error_message:\n \n@@ -54,16 +63,24 @@ def create_file():\n )\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({\n+ MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[FILENAME] +\n+ MICROSERVICE_URI_GET_PARAMS}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n-@app.route(\"/files/<filename>\", methods=[GET])\n+@app.route(\"/files/<filename>\", methods=[\"GET\"])\n def read_files(filename):\n- file_downloader_and_saver = CsvDownloader()\n- mongo_operations = MongoOperations()\n- database = DatabaseApi(mongo_operations, file_downloader_and_saver)\n+ file_downloader = CsvDownloader()\n+ mongo_operations = MongoOperations(\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME])\n+ database = DatabaseApi(mongo_operations, file_downloader)\n \n limit = int(request.args.get(\"limit\"))\n if limit > PAGINATE_FILE_LIMIT:\n@@ -76,25 +93,38 @@ def read_files(filename):\n return jsonify({MESSAGE_RESULT: file_result}), HTTP_STATUS_CODE_SUCESS\n \n \n-@app.route(\"/files\", methods=[GET])\n+@app.route(\"/files\", methods=[\"GET\"])\n def read_files_descriptor():\n- file_downloader_and_saver = CsvDownloader()\n- mongo_operations = MongoOperations()\n- database = DatabaseApi(mongo_operations, file_downloader_and_saver)\n+ file_downloader = CsvDownloader()\n+ mongo_operations = MongoOperations(\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME])\n+ database = DatabaseApi(mongo_operations, file_downloader)\n \n- return jsonify({MESSAGE_RESULT: database.get_files()}), HTTP_STATUS_CODE_SUCESS\n+ return jsonify(\n+ {MESSAGE_RESULT: database.get_files(\n+ request.args.get(\"type\"))}), HTTP_STATUS_CODE_SUCESS\n \n \n-@app.route(\"/files/<filename>\", methods=[DELETE])\n+@app.route(\"/files/<filename>\", methods=[\"DELETE\"])\n def delete_file(filename):\n- file_downloader_and_saver = CsvDownloader()\n- mongo_operations = MongoOperations()\n- database = DatabaseApi(mongo_operations, file_downloader_and_saver)\n+ file_downloader = CsvDownloader()\n+ mongo_operations = MongoOperations(\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME])\n+ database = DatabaseApi(mongo_operations, file_downloader)\n \n- database.delete_file(filename)\n+ thread_pool = ThreadPoolExecutor()\n+ thread_pool.submit(database.delete_file, filename)\n \n- return jsonify({MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n+ return jsonify(\n+ {MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n \n \n if __name__ == \"__main__\":\n- app.run(host=os.environ[DATABASE_API_HOST], port=int(os.environ[DATABASE_API_PORT]))\n+ app.run(host=os.environ[DATABASE_API_HOST],\n+ port=int(os.environ[DATABASE_API_PORT]))\ndiff --git a/microservices/histogram_image/histogram.py b/microservices/histogram_image/histogram.py\n--- a/microservices/histogram_image/histogram.py\n+++ b/microservices/histogram_image/histogram.py\n@@ -1,4 +1,6 @@\n from pymongo import MongoClient\n+from datetime import datetime\n+import pytz\n \n \n class Histogram:\n@@ -9,11 +11,17 @@ def __init__(self, database_connector):\n self.database_connector = database_connector\n \n def create_histogram(self, filename, histogram_filename, fields):\n+ timezone_london = pytz.timezone(\"Etc/Greenwich\")\n+ london_time = datetime.now(timezone_london)\n+\n metadata_histogram_filename = {\n- \"filename_parent\": filename,\n+ \"parent_filename\": filename,\n \"fields\": fields,\n \"filename\": histogram_filename,\n- \"_id\": 0,\n+ \"type\": \"histogram\",\n+ self.DOCUMENT_ID_NAME: self.METADATA_DOCUMENT_ID,\n+ \"finished\": False,\n+ \"time_created\": london_time.strftime(\"%Y-%m-%dT%H:%M:%S-00:00\")\n }\n \n self.database_connector.insert_one_in_file(\n@@ -25,20 +33,30 @@ def create_histogram(self, filename, histogram_filename, fields):\n for field in fields:\n field_accumulator = \"$\" + field\n print(field_accumulator, flush=True)\n- pipeline = [{\"$group\": {\"_id\": field_accumulator, \"count\": {\"$sum\": 1}}}]\n+ pipeline = [\n+ {\"$group\": {\"_id\": field_accumulator, \"count\": {\"$sum\": 1}}}]\n \n field_result = {\n field: self.database_connector.aggregate(filename, pipeline),\n- \"_id\": document_id,\n+ self.DOCUMENT_ID_NAME: document_id,\n }\n document_id += 1\n \n- self.database_connector.insert_one_in_file(histogram_filename, field_result)\n+ self.database_connector.insert_one_in_file(histogram_filename,\n+ field_result)\n+\n+ metadata_finished_true_query = {\"finished\": True}\n+ metadata_id_query = {self.DOCUMENT_ID_NAME: self.METADATA_DOCUMENT_ID}\n+\n+ self.database_connector.update_one(filename,\n+ metadata_finished_true_query,\n+ metadata_id_query)\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def find(self, filename, query):\n@@ -93,7 +111,8 @@ def fields_validator(self, filename, fields):\n \n filename_metadata_query = {\"filename\": filename}\n \n- filename_metadata = self.database.find_one(filename, filename_metadata_query)\n+ filename_metadata = self.database.find_one(filename,\n+ filename_metadata_query)\n \n for field in fields:\n if field not in filename_metadata[\"fields\"]:\ndiff --git a/microservices/histogram_image/server.py b/microservices/histogram_image/server.py\n--- a/microservices/histogram_image/server.py\n+++ b/microservices/histogram_image/server.py\n@@ -1,6 +1,7 @@\n from flask import jsonify, Flask, request\n import os\n from histogram import MongoOperations, HistogramRequestValidator, Histogram\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n HTTP_STATUS_CODE_NOT_ACCEPTABLE = 406\n@@ -11,8 +12,9 @@\n \n MESSAGE_RESULT = \"result\"\n \n-FIELDS_NAME = \"fields\"\n-HISTOGRAM_FILENAME_NAME = \"histogram_filename\"\n+FIELDS_NAME = \"names\"\n+HISTOGRAM_FILENAME_NAME = \"output_filename\"\n+PARENT_FILENAME_NAME = \"input_filename\"\n \n FIRST_ARGUMENT = 0\n \n@@ -23,21 +25,19 @@\n DATABASE_NAME = \"DATABASE_NAME\"\n DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n \n-POST = \"POST\"\n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/explore/histogram/\"\n+MICROSERVICE_URI_GET_PARAMS = \"?query={}&limit=10&skip=0\"\n \n app = Flask(__name__)\n \n+thread_pool = ThreadPoolExecutor()\n \n-def collection_database_url(database_url, database_replica_set):\n- return database_url + \"/?replicaSet=\" + database_replica_set\n \n-\n-@app.route(\"/histograms/<parent_filename>\", methods=[POST])\n-def create_histogram(parent_filename):\n+@app.route(\"/histograms\", methods=[\"POST\"])\n+def create_histogram():\n database = MongoOperations(\n- collection_database_url(\n- os.environ[DATABASE_URL], os.environ[DATABASE_REPLICA_SET]\n- ),\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n os.environ[DATABASE_PORT],\n os.environ[DATABASE_NAME],\n )\n@@ -50,11 +50,13 @@ def create_histogram(parent_filename):\n )\n except Exception as invalid_histogram_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_histogram_filename.args[FIRST_ARGUMENT]}),\n+ jsonify({MESSAGE_RESULT: invalid_histogram_filename.args[\n+ FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_CONFLICT,\n )\n \n try:\n+ parent_filename = request.json[PARENT_FILENAME_NAME]\n request_validator.filename_validator(parent_filename)\n except Exception as invalid_filename:\n return (\n@@ -63,25 +65,40 @@ def create_histogram(parent_filename):\n )\n \n try:\n- request_validator.fields_validator(parent_filename, request.json[FIELDS_NAME])\n+ request_validator.fields_validator(parent_filename,\n+ request.json[FIELDS_NAME])\n except Exception as invalid_fields:\n return (\n jsonify({MESSAGE_RESULT: invalid_fields.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n- histogram = Histogram(database)\n- histogram.create_histogram(\n- parent_filename,\n- request.json[HISTOGRAM_FILENAME_NAME],\n- request.json[FIELDS_NAME],\n- )\n+ thread_pool.submit(histogram_async_processing,\n+ database,\n+ parent_filename,\n+ request.json[HISTOGRAM_FILENAME_NAME],\n+ request.json[FIELDS_NAME])\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({\n+ MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[HISTOGRAM_FILENAME_NAME] +\n+ MICROSERVICE_URI_GET_PARAMS}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n+def histogram_async_processing(database, parent_filename, histogram_filename,\n+ fields_name):\n+ histogram = Histogram(database)\n+ histogram.create_histogram(\n+ parent_filename,\n+ histogram_filename,\n+ fields_name,\n+ )\n+\n+\n if __name__ == \"__main__\":\n- app.run(host=os.environ[HISTOGRAM_HOST], port=int(os.environ[HISTOGRAM_PORT]))\n+ app.run(host=os.environ[HISTOGRAM_HOST],\n+ port=int(os.environ[HISTOGRAM_PORT]))\ndiff --git a/microservices/model_builder_image/model_builder.py b/microservices/model_builder_image/model_builder.py\n--- a/microservices/model_builder_image/model_builder.py\n+++ b/microservices/model_builder_image/model_builder.py\n@@ -1,9 +1,12 @@\n from pyspark.sql import SparkSession\n import os\n import time\n+import numpy as np\n from pyspark.ml.evaluation import MulticlassClassificationEvaluator\n from pymongo import MongoClient\n from concurrent.futures import ThreadPoolExecutor, wait\n+from datetime import datetime\n+import pytz\n from pyspark.ml.classification import (\n LogisticRegression,\n DecisionTreeClassifier,\n@@ -28,34 +31,37 @@ def __init__(self, database_connector):\n \n self.spark_session = (\n SparkSession.builder.appName(\"model_builder\")\n- .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n- .config(\"spark.driver.host\", os.environ[MODEL_BUILDER_HOST_NAME])\n- .config(\n+ .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n+ .config(\"spark.driver.host\",\n+ os.environ[MODEL_BUILDER_HOST_NAME])\n+ .config(\n \"spark.jars.packages\",\n \"org.mongodb.spark:mongo-spark\" + \"-connector_2.11:2.4.2\",\n )\n- .config(\"spark.memory.fraction\", 0.8)\n- .config(\"spark.executor.memory\", \"1g\")\n- .config(\"spark.sql.shuffle.partitions\", \"800\")\n- .config(\"spark.memory.offHeap.enabled\", \"true\")\n- .config(\"spark.memory.offHeap.size\", \"1g\")\n- .config(\"spark.scheduler.mode\", \"FAIR\")\n- .config(\"spark.scheduler.pool\", \"model_builder\")\n- .config(\"spark.scheduler.allocation.file\", \"./fairscheduler.xml\")\n- .master(\n+ .config(\"spark.memory.fraction\", 0.8)\n+ .config(\"spark.executor.memory\", \"1g\")\n+ .config(\"spark.sql.shuffle.partitions\", \"800\")\n+ .config(\"spark.memory.offHeap.enabled\", \"true\")\n+ .config(\"spark.memory.offHeap.size\", \"1g\")\n+ .config(\"spark.scheduler.mode\", \"FAIR\")\n+ .config(\"spark.scheduler.pool\", \"model_builder\")\n+ .config(\"spark.scheduler.allocation.file\",\n+ \"./fairscheduler.xml\")\n+ .master(\n \"spark://\"\n + os.environ[SPARKMASTER_HOST]\n + \":\"\n + str(os.environ[SPARKMASTER_PORT])\n )\n- .getOrCreate()\n+ .getOrCreate()\n )\n \n self.thread_pool = ThreadPoolExecutor()\n \n def file_processor(self, database_url):\n file = (\n- self.spark_session.read.format(\"mongo\").option(\"uri\", database_url).load()\n+ self.spark_session.read.format(\"mongo\").option(\"uri\",\n+ database_url).load()\n )\n \n file_without_metadata = file.filter(\n@@ -70,12 +76,14 @@ def file_processor(self, database_url):\n \"time_created\",\n \"url\",\n \"parent_filename\",\n+ \"type\"\n ]\n processed_file = file_without_metadata.drop(*metadata_fields)\n \n return processed_file\n \n- def fields_from_dataframe(self, dataframe, is_string):\n+ @staticmethod\n+ def fields_from_dataframe(dataframe, is_string):\n text_fields = []\n first_row = dataframe.first()\n \n@@ -91,12 +99,13 @@ def fields_from_dataframe(self, dataframe, is_string):\n return text_fields\n \n def build_model(\n- self,\n- database_url_training,\n- database_url_test,\n- preprocessor_code,\n- classificators_list,\n- prediction_filename,\n+ self,\n+ database_url_training,\n+ database_url_test,\n+ preprocessor_code,\n+ classificators_list,\n+ train_filename,\n+ test_filename,\n ):\n training_df = self.file_processor(database_url_training)\n testing_df = self.file_processor(database_url_test)\n@@ -118,40 +127,51 @@ def build_model(\n \n classificator_threads = []\n \n+ timezone_london = pytz.timezone(\"Etc/Greenwich\")\n+ london_time = datetime.now(timezone_london)\n+ now_time = london_time.strftime(\"%Y-%m-%dT%H:%M:%S-00:00\")\n+\n+ metadata_document = {\n+ \"parent_filename\": [train_filename, test_filename],\n+ \"time_created\": now_time,\n+ \"_id\": 0,\n+ \"type\": \"builder\",\n+ \"finished\": False\n+ }\n+\n for classificator_name in classificators_list:\n classificator = classificator_switcher[classificator_name]\n \n+ metadata_classifier = metadata_document.copy()\n+ metadata_classifier[\"classifier\"] = classificator_name\n+ metadata_classifier[\n+ \"filename\"] = test_filename + \"_\" + classificator_name\n+\n+ self.database.insert_one_in_file(\n+ metadata_classifier[\"filename\"],\n+ metadata_classifier)\n+\n classificator_threads.append(\n self.thread_pool.submit(\n self.classificator_handler,\n classificator,\n- classificator_name,\n features_training,\n features_testing,\n features_evaluation,\n- prediction_filename,\n+ metadata_classifier,\n )\n )\n wait(classificator_threads)\n self.spark_session.stop()\n \n def classificator_handler(\n- self,\n- classificator,\n- classificator_name,\n- features_training,\n- features_testing,\n- features_evaluation,\n- prediction_filename,\n+ self,\n+ classificator,\n+ features_training,\n+ features_testing,\n+ features_evaluation,\n+ metadata_document\n ):\n- prediction_filename_name = (\n- prediction_filename + \"_prediction_\" + classificator_name\n- )\n- metadata_document = {\n- \"filename\": prediction_filename_name,\n- \"classificator\": classificator_name,\n- \"_id\": 0,\n- }\n \n classificator.featuresCol = \"features\"\n \n@@ -163,7 +183,6 @@ def classificator_handler(\n metadata_document[\"fit_time\"] = fit_time\n \n if features_evaluation is not None:\n-\n evaluation_prediction = model.transform(features_evaluation)\n \n evaluator_f1 = MulticlassClassificationEvaluator(\n@@ -171,10 +190,10 @@ def classificator_handler(\n )\n \n evaluator_accuracy = MulticlassClassificationEvaluator(\n- labelCol=\"label\", predictionCol=\"prediction\", metricName=\"accuracy\"\n+ labelCol=\"label\", predictionCol=\"prediction\",\n+ metricName=\"accuracy\"\n )\n \n- print(classificator_name, flush=True)\n evaluation_prediction.select(\"label\", \"prediction\").show()\n \n model_f1 = evaluator_f1.evaluate(evaluation_prediction)\n@@ -186,13 +205,11 @@ def classificator_handler(\n testing_prediction = model.transform(features_testing)\n \n self.save_classificator_result(\n- prediction_filename_name, testing_prediction, metadata_document\n+ testing_prediction,\n+ metadata_document\n )\n \n- def save_classificator_result(self, filename_name, predicted_df, filename_metatada):\n- self.database.delete_file(filename_name)\n- self.database.insert_one_in_file(filename_name, filename_metatada)\n-\n+ def save_classificator_result(self, predicted_df, filename_metatada):\n document_id = 1\n for row in predicted_df.collect():\n row_dict = row.asDict()\n@@ -204,12 +221,19 @@ def save_classificator_result(self, filename_name, predicted_df, filename_metata\n del row_dict[\"features\"]\n del row_dict[\"rawPrediction\"]\n \n- self.database.insert_one_in_file(filename_name, row_dict)\n+ self.database.insert_one_in_file(filename_metatada[\"filename\"],\n+ row_dict)\n+\n+ flag_true_query = {\"finished\": True}\n+ metadata_file_query = {\"_id\": 0}\n+ self.database.update_one(filename_metatada[\"filename\"], flag_true_query,\n+ metadata_file_query)\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def get_filenames(self):\n@@ -219,6 +243,11 @@ def find_one(self, filename, query):\n file_collection = self.database[filename]\n return file_collection.find_one(query)\n \n+ def update_one(self, filename, new_value, query):\n+ new_values_query = {\"$set\": new_value}\n+ file_collection = self.database[filename]\n+ file_collection.update_one(query, new_values_query)\n+\n def insert_one_in_file(self, filename, json_object):\n file_collection = self.database[filename]\n file_collection.insert_one(json_object)\n@@ -227,11 +256,28 @@ def delete_file(self, filename):\n file_collection = self.database[filename]\n file_collection.drop()\n \n+ @staticmethod\n+ def collection_database_url(database_url, database_name,\n+ database_filename,\n+ database_replica_set\n+ ):\n+ return (\n+ database_url\n+ + \"/\"\n+ + database_name\n+ + \".\"\n+ + database_filename\n+ + \"?replicaSet=\"\n+ + database_replica_set\n+ + \"&authSource=admin\"\n+ )\n+\n \n class ModelBuilderRequestValidator:\n MESSAGE_INVALID_TRAINING_FILENAME = \"invalid_training_filename\"\n MESSAGE_INVALID_TEST_FILENAME = \"invalid_test_filename\"\n- MESSAGE_INVALID_CLASSIFICATOR = \"invalid_classificator_name\"\n+ MESSAGE_INVALID_CLASSIFICATOR = \"invalid_classifier_name\"\n+ MESSSAGE_INVALID_PREDICTION_NAME = \"prediction_filename_already_exists\"\n \n def __init__(self, database_connector):\n self.database = database_connector\n@@ -248,6 +294,14 @@ def test_filename_validator(self, test_filename):\n if test_filename not in filenames:\n raise Exception(self.MESSAGE_INVALID_TEST_FILENAME)\n \n+ def predictions_filename_validator(self, test_filename, classificator_list):\n+ filenames = self.database.get_filenames()\n+\n+ for classificator_name in classificator_list:\n+ prediction_filename = test_filename + \"_\" + classificator_name\n+ if prediction_filename in filenames:\n+ raise Exception(self.MESSSAGE_INVALID_PREDICTION_NAME)\n+\n def model_classificators_validator(self, classificators_list):\n classificator_names_list = [\"lr\", \"dt\", \"rf\", \"gb\", \"nb\"]\n for classificator_name in classificators_list:\ndiff --git a/microservices/model_builder_image/server.py b/microservices/model_builder_image/server.py\n--- a/microservices/model_builder_image/server.py\n+++ b/microservices/model_builder_image/server.py\n@@ -5,54 +5,41 @@\n MongoOperations,\n ModelBuilderRequestValidator,\n )\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n HTTP_STATUS_CODE_NOT_ACCEPTABLE = 406\n+HTTP_STATUS_CODE_CONFICLT = 409\n \n MODEL_BUILDER_HOST_IP = \"MODEL_BUILDER_HOST_IP\"\n MODEL_BUILDER_HOST_PORT = \"MODEL_BUILDER_HOST_PORT\"\n \n-GET = \"GET\"\n-POST = \"POST\"\n-DELETE = \"DELETE\"\n-\n MESSAGE_RESULT = \"result\"\n-MESSAGE_CREATED_FILE = \"created_file\"\n \n DATABASE_URL = \"DATABASE_URL\"\n DATABASE_PORT = \"DATABASE_PORT\"\n DATABASE_NAME = \"DATABASE_NAME\"\n DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n \n-\n-TRAINING_FILENAME = \"training_filename\"\n+TRAINING_FILENAME = \"train_filename\"\n TEST_FILENAME = \"test_filename\"\n-PREPROCESSOR_CODE_NAME = \"preprocessor_code\"\n-CLASSIFICATORS_NAME = \"classificators_list\"\n+MODELING_CODE_NAME = \"modeling_code\"\n+CLASSIFIERS_NAME = \"classifiers_list\"\n FIRST_ARGUMENT = 0\n \n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/builder/\"\n+MICROSERVICE_URI_GET_PARAMS = \"?query={}&limit=10&skip=0\"\n+\n app = Flask(__name__)\n \n-\n-def collection_database_url(\n- database_url, database_name, database_filename, database_replica_set\n-):\n- return (\n- database_url\n- + \"/\"\n- + database_name\n- + \".\"\n- + database_filename\n- + \"?replicaSet=\"\n- + database_replica_set\n- + \"&authSource=admin\"\n- )\n+thread_pool = ThreadPoolExecutor()\n \n \n-@app.route(\"/models\", methods=[POST])\n+@app.route(\"/models\", methods=[\"POST\"])\n def create_model():\n database = MongoOperations(\n- os.environ[DATABASE_URL] + \"/?replicaSet=\" + os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n os.environ[DATABASE_PORT],\n os.environ[DATABASE_NAME],\n )\n@@ -60,10 +47,12 @@ def create_model():\n request_validator = ModelBuilderRequestValidator(database)\n \n try:\n- request_validator.training_filename_validator(request.json[TRAINING_FILENAME])\n+ request_validator.training_filename_validator(\n+ request.json[TRAINING_FILENAME])\n except Exception as invalid_training_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_training_filename.args[FIRST_ARGUMENT]}),\n+ jsonify({MESSAGE_RESULT: invalid_training_filename.args[\n+ FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n@@ -71,53 +60,99 @@ def create_model():\n request_validator.test_filename_validator(request.json[TEST_FILENAME])\n except Exception as invalid_test_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_test_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_test_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n try:\n request_validator.model_classificators_validator(\n- request.json[CLASSIFICATORS_NAME]\n+ request.json[CLASSIFIERS_NAME]\n )\n- except Exception as invalid_classificator_name:\n+ except Exception as invalid_classifier_name:\n return (\n- jsonify({MESSAGE_RESULT: invalid_classificator_name.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_classifier_name.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n- database_url_training = collection_database_url(\n+ try:\n+ request_validator.predictions_filename_validator(\n+ request.json[TEST_FILENAME], request.json[CLASSIFIERS_NAME])\n+ except Exception as invalid_prediction_filename:\n+ return (\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_prediction_filename.args[\n+ FIRST_ARGUMENT]}),\n+ HTTP_STATUS_CODE_CONFICLT,\n+ )\n+\n+ database_url_training = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n request.json[TRAINING_FILENAME],\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- database_url_test = collection_database_url(\n+ database_url_test = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n request.json[TEST_FILENAME],\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- model_builder = SparkModelBuilder(database)\n-\n- model_builder.build_model(\n+ thread_pool.submit(\n+ model_builder_async_processing,\n+ database,\n database_url_training,\n database_url_test,\n- request.json[PREPROCESSOR_CODE_NAME],\n- request.json[CLASSIFICATORS_NAME],\n+ request.json[MODELING_CODE_NAME],\n+ request.json[CLASSIFIERS_NAME],\n+ request.json[TRAINING_FILENAME],\n request.json[TEST_FILENAME],\n )\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({\n+ MESSAGE_RESULT:\n+ create_prediction_files_uri(\n+ request.json[CLASSIFIERS_NAME],\n+ request.json[TEST_FILENAME])}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n+def create_prediction_files_uri(classifiers_list, test_filename):\n+ classifiers_uri = []\n+ for classifier in classifiers_list:\n+ classifiers_uri.append(\n+ MICROSERVICE_URI_GET +\n+ test_filename +\n+ \"_\" +\n+ classifier +\n+ MICROSERVICE_URI_GET_PARAMS)\n+\n+ return classifiers_uri\n+\n+\n+def model_builder_async_processing(database, database_url_training,\n+ database_url_test, modeling_code,\n+ classifiers_name, train_filename,\n+ test_filename):\n+ model_builder = SparkModelBuilder(database)\n+\n+ model_builder.build_model(\n+ database_url_training,\n+ database_url_test,\n+ modeling_code,\n+ classifiers_name,\n+ train_filename,\n+ test_filename,\n+ )\n+\n+\n if __name__ == \"__main__\":\n app.run(\n host=os.environ[MODEL_BUILDER_HOST_IP],\n- port=int(os.environ[MODEL_BUILDER_HOST_PORT]),\n- debug=True,\n+ port=int(os.environ[MODEL_BUILDER_HOST_PORT])\n )\ndiff --git a/microservices/pca_image/pca.py b/microservices/pca_image/pca.py\n--- a/microservices/pca_image/pca.py\n+++ b/microservices/pca_image/pca.py\n@@ -23,23 +23,23 @@ class PcaGenerator:\n def __init__(self, database_url_input):\n self.spark_session = (\n SparkSession.builder.appName(\"pca\")\n- .config(\"spark.mongodb.input.uri\", database_url_input)\n- .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n- .config(\"spark.driver.host\", os.environ[PCA_HOST_NAME])\n- .config(\n+ .config(\"spark.mongodb.input.uri\", database_url_input)\n+ .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n+ .config(\"spark.driver.host\", os.environ[PCA_HOST_NAME])\n+ .config(\n \"spark.jars.packages\",\n \"org.mongodb.spark:mongo-spark\" + \"-connector_2.11:2.4.2\",\n )\n- .master(\n+ .master(\n \"spark://\"\n + os.environ[SPARKMASTER_HOST]\n + \":\"\n + str(os.environ[SPARKMASTER_PORT])\n )\n- .getOrCreate()\n+ .getOrCreate()\n )\n \n- def create_image(self, filename, label_name, pca_filename):\n+ def create_image(self, label_name, pca_filename):\n dataframe = self.file_processor()\n dataframe = dataframe.dropna()\n string_fields = self.fields_from_dataframe(dataframe, is_string=True)\n@@ -59,12 +59,15 @@ def create_image(self, filename, label_name, pca_filename):\n \n if label_name is not None:\n embedded_array[label_name] = encoded_dataframe[label_name]\n- sns_plot = sns.scatterplot(x=0, y=1, data=embedded_array, hue=label_name)\n+ sns_plot = sns.scatterplot(x=0, y=1, data=embedded_array,\n+ hue=label_name)\n sns_plot.get_figure().savefig(image_path)\n else:\n sns_plot = sns.scatterplot(x=0, y=1, data=embedded_array)\n sns_plot.get_figure().savefig(image_path)\n \n+ self.spark_session.stop()\n+\n def file_processor(self):\n file = self.spark_session.read.format(self.MONGO_SPARK_SOURCE).load()\n \n@@ -80,6 +83,7 @@ def file_processor(self):\n \"time_created\",\n \"url\",\n \"parent_filename\",\n+ \"type\"\n ]\n processed_file = file_without_metadata.drop(*metadata_fields)\n \n@@ -103,8 +107,9 @@ def fields_from_dataframe(dataframe, is_string):\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def find_one(self, filename, query):\n@@ -114,6 +119,21 @@ def find_one(self, filename, query):\n def get_filenames(self):\n return self.database.list_collection_names()\n \n+ @staticmethod\n+ def collection_database_url(database_url, database_name, database_filename,\n+ database_replica_set\n+ ):\n+ return (\n+ database_url\n+ + \"/\"\n+ + database_name\n+ + \".\"\n+ + database_filename\n+ + \"?replicaSet=\"\n+ + database_replica_set\n+ + \"&authSource=admin\"\n+ )\n+\n \n class PcaRequestValidator:\n MESSAGE_INVALID_FILENAME = \"invalid_filename\"\n@@ -130,18 +150,20 @@ def parent_filename_validator(self, filename):\n if filename not in filenames:\n raise Exception(self.MESSAGE_INVALID_FILENAME)\n \n- def pca_filename_existence_validator(self, pca_filename):\n+ @staticmethod\n+ def pca_filename_existence_validator(pca_filename):\n images = os.listdir(os.environ[IMAGES_PATH])\n image_name = pca_filename + IMAGE_FORMAT\n if image_name in images:\n- raise Exception(self.MESSAGE_DUPLICATE_FILE)\n+ raise Exception(PcaRequestValidator.MESSAGE_DUPLICATE_FILE)\n \n- def no_pca_filename_existence_validator(self, pca_filename):\n+ @staticmethod\n+ def pca_filename_inexistence_validator(pca_filename):\n images = os.listdir(os.environ[IMAGES_PATH])\n image_name = pca_filename + IMAGE_FORMAT\n \n if image_name not in images:\n- raise Exception(self.MESSAGE_NOT_FOUND)\n+ raise Exception(PcaRequestValidator.MESSAGE_NOT_FOUND)\n \n def filename_label_validator(self, filename, label):\n if label is None:\n@@ -149,7 +171,8 @@ def filename_label_validator(self, filename, label):\n \n filename_metadata_query = {\"filename\": filename}\n \n- filename_metadata = self.database.find_one(filename, filename_metadata_query)\n+ filename_metadata = self.database.find_one(filename,\n+ filename_metadata_query)\n \n if label not in filename_metadata[\"fields\"]:\n raise Exception(self.MESSAGE_INVALID_LABEL)\ndiff --git a/microservices/pca_image/server.py b/microservices/pca_image/server.py\n--- a/microservices/pca_image/server.py\n+++ b/microservices/pca_image/server.py\n@@ -1,6 +1,7 @@\n from flask import jsonify, request, Flask, send_file\n import os\n from pca import PcaGenerator, MongoOperations, PcaRequestValidator\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS = 200\n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n@@ -18,46 +19,31 @@\n DATABASE_PORT = \"DATABASE_PORT\"\n DATABASE_NAME = \"DATABASE_NAME\"\n DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n-FULL_DATABASE_URL = (\n- os.environ[DATABASE_URL] + \"/?replicaSet=\" + os.environ[DATABASE_REPLICA_SET]\n-)\n-\n-GET = \"GET\"\n-POST = \"POST\"\n-DELETE = \"DELETE\"\n \n MESSAGE_RESULT = \"result\"\n-PCA_FILENAME_NAME = \"pca_filename\"\n-LABEL_NAME = \"label_name\"\n+PCA_FILENAME_NAME = \"output_filename\"\n+PARENT_FILENAME_NAME = \"input_filename\"\n+LABEL_NAME = \"label\"\n \n-MESSAGE_CREATED_FILE = \"created_file\"\n MESSAGE_DELETED_FILE = \"deleted_file\"\n MESSAGE_NOT_FOUND = \"not_found_file\"\n \n FIRST_ARGUMENT = 0\n \n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/explore/pca/\"\n+\n app = Flask(__name__)\n \n+thread_pool = ThreadPoolExecutor()\n \n-def collection_database_url(\n- database_url, database_name, database_filename, database_replica_set\n-):\n- return (\n- database_url\n- + \"/\"\n- + database_name\n- + \".\"\n- + database_filename\n- + \"?replicaSet=\"\n- + database_replica_set\n- + \"&authSource=admin\"\n- )\n \n-\n-@app.route(\"/images/<parent_filename>\", methods=[POST])\n-def create_pca(parent_filename):\n+@app.route(\"/images\", methods=[\"POST\"])\n+def pca_plot():\n database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME]\n )\n request_validator = PcaRequestValidator(database)\n \n@@ -67,12 +53,14 @@ def create_pca(parent_filename):\n )\n except Exception as invalid_pca_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_CONFLICT,\n )\n \n try:\n- request_validator.parent_filename_validator(parent_filename)\n+ request_validator.parent_filename_validator(\n+ request.json[PARENT_FILENAME_NAME])\n except Exception as invalid_filename:\n return (\n jsonify({MESSAGE_RESULT: invalid_filename.args[FIRST_ARGUMENT]}),\n@@ -81,7 +69,7 @@ def create_pca(parent_filename):\n \n try:\n request_validator.filename_label_validator(\n- parent_filename, request.json[LABEL_NAME]\n+ request.json[PARENT_FILENAME_NAME], request.json[LABEL_NAME]\n )\n except Exception as invalid_label:\n return (\n@@ -89,43 +77,51 @@ def create_pca(parent_filename):\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n- database_url_input = collection_database_url(\n+ database_url_input = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n- parent_filename,\n+ request.json[PARENT_FILENAME_NAME],\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- pca_generator = PcaGenerator(database_url_input)\n-\n- pca_generator.create_image(\n- parent_filename, request.json[LABEL_NAME], request.json[PCA_FILENAME_NAME]\n- )\n+ thread_pool.submit(pca_async_processing,\n+ database_url_input,\n+ request.json[LABEL_NAME],\n+ request.json[PCA_FILENAME_NAME])\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({\n+ MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[PCA_FILENAME_NAME]}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n-@app.route(\"/images\", methods=[GET])\n+def pca_async_processing(database_url_input, label_name,\n+ pca_filename):\n+ pca_generator = PcaGenerator(database_url_input)\n+\n+ pca_generator.create_image(\n+ label_name, pca_filename\n+ )\n+\n+\n+@app.route(\"/images\", methods=[\"GET\"])\n def get_images():\n images = os.listdir(os.environ[IMAGES_PATH])\n return jsonify({MESSAGE_RESULT: images}), HTTP_STATUS_CODE_SUCESS\n \n \n-@app.route(\"/images/<filename>\", methods=[GET])\n+@app.route(\"/images/<filename>\", methods=[\"GET\"])\n def get_image(filename):\n- database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n- )\n- request_validator = PcaRequestValidator(database)\n-\n try:\n- request_validator.no_pca_filename_existence_validator(filename)\n+ PcaRequestValidator.pca_filename_inexistence_validator(filename)\n+\n except Exception as invalid_pca_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_FOUND,\n )\n \n@@ -134,25 +130,23 @@ def get_image(filename):\n return send_file(image_path, mimetype=\"image/png\")\n \n \n-@app.route(\"/images/<filename>\", methods=[DELETE])\n+@app.route(\"/images/<filename>\", methods=[\"DELETE\"])\n def delete_image(filename):\n- database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n- )\n- request_validator = PcaRequestValidator(database)\n-\n try:\n- request_validator.no_pca_filename_existence_validator(filename)\n+ PcaRequestValidator.pca_filename_inexistence_validator(filename)\n except Exception as invalid_pca_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_pca_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_FOUND,\n )\n \n image_path = os.environ[IMAGES_PATH] + \"/\" + filename + IMAGE_FORMAT\n- os.remove(image_path)\n \n- return jsonify({MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n+ thread_pool.submit(os.remove, image_path)\n+\n+ return jsonify(\n+ {MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n \n \n if __name__ == \"__main__\":\ndiff --git a/microservices/projection_image/projection.py b/microservices/projection_image/projection.py\n--- a/microservices/projection_image/projection.py\n+++ b/microservices/projection_image/projection.py\n@@ -16,35 +16,41 @@ class SparkManager:\n MONGO_SPARK_SOURCE = \"com.mongodb.spark.sql.DefaultSource\"\n METADATA_FILE_ID = 0\n database_url_output = None\n+ MAX_NUMBER_THREADS = 3\n \n def __init__(self, database_url_input, database_url_output):\n self.database_url_output = database_url_output\n \n self.spark_session = (\n SparkSession.builder.appName(\"projection\")\n- .config(\"spark.mongodb.input.uri\", database_url_input)\n- .config(\"spark.mongodb.output.uri\", database_url_output)\n- .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n- .config(\"spark.driver.host\", os.environ[PROJECTION_HOST_NAME])\n- .config(\n+ .config(\"spark.mongodb.input.uri\", database_url_input)\n+ .config(\"spark.mongodb.output.uri\", database_url_output)\n+ .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n+ .config(\"spark.driver.host\", os.environ[PROJECTION_HOST_NAME])\n+ .config(\n \"spark.jars.packages\",\n \"org.mongodb.spark:mongo-spark\" + \"-connector_2.11:2.4.2\",\n )\n- .master(\n+ .master(\n \"spark://\"\n + os.environ[SPARKMASTER_HOST]\n + \":\"\n + str(os.environ[SPARKMASTER_PORT])\n )\n- .getOrCreate()\n+ .getOrCreate()\n )\n \n def projection(self, filename, projection_filename, fields):\n timezone_london = pytz.timezone(\"Etc/Greenwich\")\n london_time = datetime.now(timezone_london)\n \n+ fields.append(self.DOCUMENT_ID)\n fields_without_id = fields.copy()\n- fields_without_id.remove(self.DOCUMENT_ID)\n+\n+ try:\n+ fields_without_id.remove(self.DOCUMENT_ID)\n+ except Exception:\n+ pass\n \n metadata_content = (\n projection_filename,\n@@ -53,6 +59,7 @@ def projection(self, filename, projection_filename, fields):\n filename,\n self.METADATA_FILE_ID,\n fields_without_id,\n+ \"projection\"\n )\n \n metadata_fields = [\n@@ -62,6 +69,7 @@ def projection(self, filename, projection_filename, fields):\n \"parent_filename\",\n self.DOCUMENT_ID,\n \"fields\",\n+ \"type\",\n ]\n \n metadata_dataframe = self.spark_session.createDataFrame(\n@@ -70,16 +78,21 @@ def projection(self, filename, projection_filename, fields):\n \n metadata_dataframe.write.format(self.MONGO_SPARK_SOURCE).save()\n \n- self.submit_projection_job_spark(fields, metadata_content, metadata_fields)\n+ self.submit_projection_job_spark(fields,\n+ metadata_content,\n+ metadata_fields)\n \n- def submit_projection_job_spark(self, fields, metadata_content, metadata_fields):\n- dataframe = self.spark_session.read.format(self.MONGO_SPARK_SOURCE).load()\n+ def submit_projection_job_spark(self, fields, metadata_content,\n+ metadata_fields):\n+ dataframe = self.spark_session.read.format(\n+ self.MONGO_SPARK_SOURCE).load()\n dataframe = dataframe.filter(\n dataframe[self.DOCUMENT_ID] != self.METADATA_FILE_ID\n )\n \n projection_dataframe = dataframe.select(*fields)\n- projection_dataframe.write.format(self.MONGO_SPARK_SOURCE).mode(\"append\").save()\n+ projection_dataframe.write.format(self.MONGO_SPARK_SOURCE).mode(\n+ \"append\").save()\n \n metadata_content_list = list(metadata_content)\n metadata_content_list[metadata_content_list.index(False)] = True\n@@ -97,8 +110,9 @@ def submit_projection_job_spark(self, fields, metadata_content, metadata_fields)\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def find_one(self, filename, query):\n@@ -108,6 +122,22 @@ def find_one(self, filename, query):\n def get_filenames(self):\n return self.database.list_collection_names()\n \n+ @staticmethod\n+ def collection_database_url(\n+ database_url, database_name, database_filename,\n+ database_replica_set\n+ ):\n+ return (\n+ database_url\n+ + \"/\"\n+ + database_name\n+ + \".\"\n+ + database_filename\n+ + \"?replicaSet=\"\n+ + database_replica_set\n+ + \"&authSource=admin\"\n+ )\n+\n \n class ProjectionRequestValidator:\n MESSAGE_INVALID_FIELDS = \"invalid_fields\"\n@@ -136,7 +166,8 @@ def projection_fields_validator(self, filename, projection_fields):\n \n filename_metadata_query = {\"filename\": filename}\n \n- filename_metadata = self.database.find_one(filename, filename_metadata_query)\n+ filename_metadata = self.database.find_one(filename,\n+ filename_metadata_query)\n \n for field in projection_fields:\n if field not in filename_metadata[\"fields\"]:\ndiff --git a/microservices/projection_image/server.py b/microservices/projection_image/server.py\n--- a/microservices/projection_image/server.py\n+++ b/microservices/projection_image/server.py\n@@ -1,6 +1,8 @@\n from flask import jsonify, request, Flask\n import os\n-from projection import SparkManager, MongoOperations, ProjectionRequestValidator\n+from projection import SparkManager, MongoOperations, \\\n+ ProjectionRequestValidator\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n HTTP_STATUS_CODE_CONFLICT = 409\n@@ -17,40 +19,25 @@\n DOCUMENT_ID = \"_id\"\n METADATA_DOCUMENT_ID = 0\n \n-GET = \"GET\"\n-POST = \"POST\"\n-DELETE = \"DELETE\"\n-\n MESSAGE_RESULT = \"result\"\n-PROJECTION_FILENAME_NAME = \"projection_filename\"\n-FIELDS_NAME = \"fields\"\n+PROJECTION_FILENAME_NAME = \"output_filename\"\n+PARENT_FILENAME_NAME = \"input_filename\"\n+FIELDS_NAME = \"names\"\n \n-MESSAGE_CREATED_FILE = \"created_file\"\n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/transform/projection/\"\n+MICROSERVICE_URI_GET_PARAMS = \"?query={}&limit=20&skip=0\"\n \n FIRST_ARGUMENT = 0\n \n app = Flask(__name__)\n+thread_pool = ThreadPoolExecutor()\n \n \n-def collection_database_url(\n- database_url, database_name, database_filename, database_replica_set\n-):\n- return (\n- database_url\n- + \"/\"\n- + database_name\n- + \".\"\n- + database_filename\n- + \"?replicaSet=\"\n- + database_replica_set\n- + \"&authSource=admin\"\n- )\n-\n-\n-@app.route(\"/projections/<parent_filename>\", methods=[POST])\n-def create_projection(parent_filename):\n+@app.route(\"/projections\", methods=[\"POST\"])\n+def create_projection():\n database = MongoOperations(\n- os.environ[DATABASE_URL] + \"/?replicaSet=\" + os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n os.environ[DATABASE_PORT],\n os.environ[DATABASE_NAME],\n )\n@@ -63,11 +50,13 @@ def create_projection(parent_filename):\n )\n except Exception as invalid_projection_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_projection_filename.args[FIRST_ARGUMENT]}),\n+ jsonify({MESSAGE_RESULT: invalid_projection_filename.args[\n+ FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_CONFLICT,\n )\n \n try:\n+ parent_filename = request.json[PARENT_FILENAME_NAME]\n request_validator.filename_validator(parent_filename)\n except Exception as invalid_filename:\n return (\n@@ -85,37 +74,46 @@ def create_projection(parent_filename):\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n- database_url_input = collection_database_url(\n+ database_url_input = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n parent_filename,\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- database_url_output = collection_database_url(\n+ database_url_output = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n request.json[PROJECTION_FILENAME_NAME],\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- spark_manager = SparkManager(database_url_input, database_url_output)\n-\n- projection_fields = request.json[FIELDS_NAME]\n-\n- projection_fields.append(DOCUMENT_ID)\n-\n- spark_manager.projection(\n- parent_filename, request.json[PROJECTION_FILENAME_NAME], projection_fields\n- )\n+ thread_pool.submit(projection_async_processing, database_url_input,\n+ database_url_output, request.json[FIELDS_NAME],\n+ parent_filename, request.json[PROJECTION_FILENAME_NAME])\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({\n+ MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[PROJECTION_FILENAME_NAME] +\n+ MICROSERVICE_URI_GET_PARAMS}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n+def projection_async_processing(database_url_input, database_url_output,\n+ projection_fields, parent_filename,\n+ output_filename):\n+ spark_manager = SparkManager(database_url_input, database_url_output)\n+\n+ spark_manager.projection(\n+ parent_filename, output_filename,\n+ projection_fields)\n+\n+\n if __name__ == \"__main__\":\n app.run(\n- host=os.environ[PROJECTION_HOST_IP], port=int(os.environ[PROJECTION_HOST_PORT])\n+ host=os.environ[PROJECTION_HOST_IP],\n+ port=int(os.environ[PROJECTION_HOST_PORT])\n )\ndiff --git a/microservices/tsne_image/server.py b/microservices/tsne_image/server.py\n--- a/microservices/tsne_image/server.py\n+++ b/microservices/tsne_image/server.py\n@@ -1,6 +1,7 @@\n from flask import jsonify, request, Flask, send_file\n import os\n from tsne import TsneGenerator, MongoOperations, TsneRequestValidator\n+from concurrent.futures import ThreadPoolExecutor\n \n HTTP_STATUS_CODE_SUCESS = 200\n HTTP_STATUS_CODE_SUCESS_CREATED = 201\n@@ -18,46 +19,31 @@\n DATABASE_PORT = \"DATABASE_PORT\"\n DATABASE_NAME = \"DATABASE_NAME\"\n DATABASE_REPLICA_SET = \"DATABASE_REPLICA_SET\"\n-FULL_DATABASE_URL = (\n- os.environ[DATABASE_URL] + \"/?replicaSet=\" + os.environ[DATABASE_REPLICA_SET]\n-)\n-\n-GET = \"GET\"\n-POST = \"POST\"\n-DELETE = \"DELETE\"\n \n MESSAGE_RESULT = \"result\"\n-TSNE_FILENAME_NAME = \"tsne_filename\"\n-LABEL_NAME = \"label_name\"\n+PARENT_FILENAME_NAME = \"input_filename\"\n+TSNE_FILENAME_NAME = \"output_filename\"\n+LABEL_NAME = \"label\"\n \n-MESSAGE_CREATED_FILE = \"created_file\"\n MESSAGE_DELETED_FILE = \"deleted_file\"\n MESSAGE_NOT_FOUND = \"not_found_file\"\n \n FIRST_ARGUMENT = 0\n \n-app = Flask(__name__)\n+MICROSERVICE_URI_GET = \"/api/learningOrchestra/v1/explore/tsne/\"\n \n+thread_pool = ThreadPoolExecutor()\n \n-def collection_database_url(\n- database_url, database_name, database_filename, database_replica_set\n-):\n- return (\n- database_url\n- + \"/\"\n- + database_name\n- + \".\"\n- + database_filename\n- + \"?replicaSet=\"\n- + database_replica_set\n- + \"&authSource=admin\"\n- )\n+app = Flask(__name__)\n \n \n-@app.route(\"/images/<parent_filename>\", methods=[POST])\n-def create_tsne(parent_filename):\n+@app.route(\"/images\", methods=[\"POST\"])\n+def create_tsne():\n database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n+ os.environ[DATABASE_URL],\n+ os.environ[DATABASE_REPLICA_SET],\n+ os.environ[DATABASE_PORT],\n+ os.environ[DATABASE_NAME]\n )\n request_validator = TsneRequestValidator(database)\n \n@@ -67,12 +53,14 @@ def create_tsne(parent_filename):\n )\n except Exception as invalid_tsne_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_CONFLICT,\n )\n \n try:\n- request_validator.parent_filename_validator(parent_filename)\n+ request_validator.parent_filename_validator(\n+ request.json[PARENT_FILENAME_NAME])\n except Exception as invalid_filename:\n return (\n jsonify({MESSAGE_RESULT: invalid_filename.args[FIRST_ARGUMENT]}),\n@@ -81,7 +69,7 @@ def create_tsne(parent_filename):\n \n try:\n request_validator.filename_label_validator(\n- parent_filename, request.json[LABEL_NAME]\n+ request.json[PARENT_FILENAME_NAME], request.json[LABEL_NAME]\n )\n except Exception as invalid_label:\n return (\n@@ -89,43 +77,50 @@ def create_tsne(parent_filename):\n HTTP_STATUS_CODE_NOT_ACCEPTABLE,\n )\n \n- database_url_input = collection_database_url(\n+ database_url_input = MongoOperations.collection_database_url(\n os.environ[DATABASE_URL],\n os.environ[DATABASE_NAME],\n- parent_filename,\n+ request.json[PARENT_FILENAME_NAME],\n os.environ[DATABASE_REPLICA_SET],\n )\n \n- tsne_generator = TsneGenerator(database_url_input)\n-\n- tsne_generator.create_image(\n- parent_filename, request.json[LABEL_NAME], request.json[TSNE_FILENAME_NAME]\n- )\n+ thread_pool.submit(tsne_async_processing,\n+ database_url_input,\n+ request.json[LABEL_NAME],\n+ request.json[TSNE_FILENAME_NAME])\n \n return (\n- jsonify({MESSAGE_RESULT: MESSAGE_CREATED_FILE}),\n+ jsonify({MESSAGE_RESULT:\n+ MICROSERVICE_URI_GET +\n+ request.json[TSNE_FILENAME_NAME]}),\n HTTP_STATUS_CODE_SUCESS_CREATED,\n )\n \n \n-@app.route(\"/images\", methods=[GET])\n+def tsne_async_processing(database_url_input, label_name,\n+ tsne_filename):\n+ tsne_generator = TsneGenerator(database_url_input)\n+\n+ tsne_generator.create_image(\n+ label_name,\n+ tsne_filename\n+ )\n+\n+\n+@app.route(\"/images\", methods=[\"GET\"])\n def get_images():\n images = os.listdir(os.environ[IMAGES_PATH])\n return jsonify({MESSAGE_RESULT: images}), HTTP_STATUS_CODE_SUCESS\n \n \n-@app.route(\"/images/<filename>\", methods=[GET])\n+@app.route(\"/images/<filename>\", methods=[\"GET\"])\n def get_image(filename):\n- database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n- )\n- request_validator = TsneRequestValidator(database)\n-\n try:\n- request_validator.no_tsne_filename_existence_validator(filename)\n+ TsneRequestValidator.tsne_filename_inexistence_validator(filename)\n except Exception as invalid_tsne_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_FOUND,\n )\n \n@@ -134,26 +129,25 @@ def get_image(filename):\n return send_file(image_path, mimetype=\"image/png\")\n \n \n-@app.route(\"/images/<filename>\", methods=[DELETE])\n+@app.route(\"/images/<filename>\", methods=[\"DELETE\"])\n def delete_image(filename):\n- database = MongoOperations(\n- FULL_DATABASE_URL, os.environ[DATABASE_PORT], os.environ[DATABASE_NAME]\n- )\n- request_validator = TsneRequestValidator(database)\n-\n try:\n- request_validator.no_tsne_filename_existence_validator(filename)\n+ TsneRequestValidator.tsne_filename_inexistence_validator(filename)\n except Exception as invalid_tsne_filename:\n return (\n- jsonify({MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n+ jsonify(\n+ {MESSAGE_RESULT: invalid_tsne_filename.args[FIRST_ARGUMENT]}),\n HTTP_STATUS_CODE_NOT_FOUND,\n )\n \n image_path = os.environ[IMAGES_PATH] + \"/\" + filename + IMAGE_FORMAT\n- os.remove(image_path)\n \n- return jsonify({MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n+ thread_pool.submit(os.remove, image_path)\n+\n+ return jsonify(\n+ {MESSAGE_RESULT: MESSAGE_DELETED_FILE}), HTTP_STATUS_CODE_SUCESS\n \n \n if __name__ == \"__main__\":\n- app.run(host=os.environ[TSNE_HOST_IP], port=int(os.environ[TSNE_HOST_PORT]))\n+ app.run(host=os.environ[TSNE_HOST_IP],\n+ port=int(os.environ[TSNE_HOST_PORT]))\ndiff --git a/microservices/tsne_image/tsne.py b/microservices/tsne_image/tsne.py\n--- a/microservices/tsne_image/tsne.py\n+++ b/microservices/tsne_image/tsne.py\n@@ -23,23 +23,23 @@ class TsneGenerator:\n def __init__(self, database_url_input):\n self.spark_session = (\n SparkSession.builder.appName(\"tsne\")\n- .config(\"spark.mongodb.input.uri\", database_url_input)\n- .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n- .config(\"spark.driver.host\", os.environ[TSNE_HOST_NAME])\n- .config(\n+ .config(\"spark.mongodb.input.uri\", database_url_input)\n+ .config(\"spark.driver.port\", os.environ[SPARK_DRIVER_PORT])\n+ .config(\"spark.driver.host\", os.environ[TSNE_HOST_NAME])\n+ .config(\n \"spark.jars.packages\",\n \"org.mongodb.spark:mongo-spark\" + \"-connector_2.11:2.4.2\",\n )\n- .master(\n+ .master(\n \"spark://\"\n + os.environ[SPARKMASTER_HOST]\n + \":\"\n + str(os.environ[SPARKMASTER_PORT])\n )\n- .getOrCreate()\n+ .getOrCreate()\n )\n \n- def create_image(self, filename, label_name, tsne_filename):\n+ def create_image(self, label_name, tsne_filename):\n dataframe = self.file_processor()\n dataframe = dataframe.dropna()\n string_fields = self.fields_from_dataframe(dataframe, is_string=True)\n@@ -55,11 +55,13 @@ def create_image(self, filename, label_name, tsne_filename):\n treated_array = np.array(encoded_dataframe)\n embedded_array = TSNE().fit_transform(treated_array)\n embedded_array = pandas.DataFrame(embedded_array)\n- image_path = os.environ[IMAGES_PATH] + \"/\" + tsne_filename + IMAGE_FORMAT\n+ image_path = os.environ[\n+ IMAGES_PATH] + \"/\" + tsne_filename + IMAGE_FORMAT\n \n if label_name is not None:\n embedded_array[label_name] = encoded_dataframe[label_name]\n- sns_plot = sns.scatterplot(x=0, y=1, data=embedded_array, hue=label_name)\n+ sns_plot = sns.scatterplot(x=0, y=1, data=embedded_array,\n+ hue=label_name)\n sns_plot.get_figure().savefig(image_path)\n else:\n sns_plot = sns.scatterplot(\n@@ -69,6 +71,8 @@ def create_image(self, filename, label_name, tsne_filename):\n )\n sns_plot.get_figure().savefig(image_path)\n \n+ self.spark_session.stop()\n+\n def file_processor(self):\n file = self.spark_session.read.format(self.MONGO_SPARK_SOURCE).load()\n \n@@ -84,6 +88,7 @@ def file_processor(self):\n \"time_created\",\n \"url\",\n \"parent_filename\",\n+ \"type\"\n ]\n processed_file = file_without_metadata.drop(*metadata_fields)\n \n@@ -107,8 +112,9 @@ def fields_from_dataframe(dataframe, is_string):\n \n \n class MongoOperations:\n- def __init__(self, database_url, database_port, database_name):\n- self.mongo_client = MongoClient(database_url, int(database_port))\n+ def __init__(self, database_url, replica_set, database_port, database_name):\n+ self.mongo_client = MongoClient(\n+ database_url + '/?replicaSet=' + replica_set, int(database_port))\n self.database = self.mongo_client[database_name]\n \n def find_one(self, filename, query):\n@@ -118,6 +124,21 @@ def find_one(self, filename, query):\n def get_filenames(self):\n return self.database.list_collection_names()\n \n+ @staticmethod\n+ def collection_database_url(\n+ database_url, database_name, database_filename, database_replica_set\n+ ):\n+ return (\n+ database_url\n+ + \"/\"\n+ + database_name\n+ + \".\"\n+ + database_filename\n+ + \"?replicaSet=\"\n+ + database_replica_set\n+ + \"&authSource=admin\"\n+ )\n+\n \n class TsneRequestValidator:\n MESSAGE_INVALID_FILENAME = \"invalid_filename\"\n@@ -134,18 +155,20 @@ def parent_filename_validator(self, filename):\n if filename not in filenames:\n raise Exception(self.MESSAGE_INVALID_FILENAME)\n \n- def tsne_filename_existence_validator(self, tsne_filename):\n+ @staticmethod\n+ def tsne_filename_existence_validator(tsne_filename):\n images = os.listdir(os.environ[IMAGES_PATH])\n image_name = tsne_filename + IMAGE_FORMAT\n if image_name in images:\n- raise Exception(self.MESSAGE_DUPLICATE_FILE)\n+ raise Exception(TsneRequestValidator.MESSAGE_DUPLICATE_FILE)\n \n- def no_tsne_filename_existence_validator(self, tsne_filename):\n+ @staticmethod\n+ def tsne_filename_inexistence_validator(tsne_filename):\n images = os.listdir(os.environ[IMAGES_PATH])\n image_name = tsne_filename + IMAGE_FORMAT\n \n if image_name not in images:\n- raise Exception(self.MESSAGE_NOT_FOUND)\n+ raise Exception(TsneRequestValidator.MESSAGE_NOT_FOUND)\n \n def filename_label_validator(self, filename, label):\n if label is None:\n@@ -153,7 +176,8 @@ def filename_label_validator(self, filename, label):\n \n filename_metadata_query = {\"filename\": filename}\n \n- filename_metadata = self.database.find_one(filename, filename_metadata_query)\n+ filename_metadata = self.database.find_one(filename,\n+ filename_metadata_query)\n \n if label not in filename_metadata[\"fields\"]:\n raise Exception(self.MESSAGE_INVALID_LABEL)\n", "test_patch": "", "problem_statement": "", "hints_text": "", "created_at": "2020-11-04T14:28:51Z"}
|