1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
|
"""Copyright 2020 Deutsche Telekom.
Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.
"""
from enum import Enum, unique
from logging import Logger, getLogger
from types import TracebackType
from typing import Any, Dict, Generator, Optional, Type
from google.protobuf import json_format
from proto.BluePrintProcessing_pb2 import ExecutionServiceInput, ExecutionServiceOutput
from .client import Client
@unique
class WorkflowMode(Enum):
"""Workflow mode enumerator.
Workflow can be executed in two modes: synchronously and asynchronously.
This enumerator stores valid values to set the mode: SYNC for synchronously mode and ASYNC for asynchronously.
"""
SYNC = "sync"
ASYNC = "async"
class WorkflowExecution:
"""Wokflow execution class.
Describes workflow to call. Set blueprint name and version and workflow name to execute.
Workflow inputs are optional, by default set to empty directory.
Workflow mode is also optional. It is set by default to call workflow synchronously.
"""
def __init__(
self,
blueprint_name: str,
blueprint_version: str,
workflow_name: str,
workflow_inputs: Dict[str, Any] = None,
workflow_mode: WorkflowMode = WorkflowMode.SYNC,
) -> None:
"""Initialize workflow execution.
Get all needed information to execute workflow.
Args:
blueprint_name (str): Blueprint name to execute workflow from.
blueprint_version (str): Blueprint version.
workflow_name (str): Name of the workflow to execute
workflow_inputs (Dict[str, Any], optional): Key-value workflow inputs. Defaults to None.
workflow_mode (WorkflowMode, optional): Workflow execution mode. It can be run synchronously or
asynchronously. Defaults to WorkflowMode.SYNC.
"""
self.blueprint_name: str = blueprint_name
self.blueprint_version: str = blueprint_version
self.workflow_name: str = workflow_name
if workflow_inputs is None:
workflow_inputs = {}
self.workflow_inputs: Dict[str, Any] = workflow_inputs
self.workflow_mode: WorkflowMode = workflow_mode
@property
def message(self) -> ExecutionServiceInput:
"""Workflow execution protobuf message.
This message is going to be sent to gRPC server to execute workflow.
Returns:
ExecutionServiceInput: Properly filled protobuf message.
"""
execution_msg: ExecutionServiceInput = ExecutionServiceInput()
execution_msg.actionIdentifiers.mode = self.workflow_mode.value
execution_msg.actionIdentifiers.blueprintName = self.blueprint_name
execution_msg.actionIdentifiers.blueprintVersion = self.blueprint_version
execution_msg.actionIdentifiers.actionName = self.workflow_name
execution_msg.payload.update({f"{self.workflow_name}-request": self.workflow_inputs})
return execution_msg
class WorkflowExecutionResult:
"""Result of workflow execution.
Store both workflow data and the result returns by server.
"""
def __init__(self, workflow_execution: WorkflowExecution, execution_output: ExecutionServiceOutput) -> None:
"""Initialize workflow execution result object.
Stores workflow execution data and execution result.
Args:
workflow_execution (WorkflowExecution): WorkflowExecution object which was used to call request.
execution_output (ExecutionServiceOutput): gRPC server response.
"""
self.workflow_execution: WorkflowExecution = workflow_execution
self.execution_output: ExecutionServiceOutput = execution_output
@property
def blueprint_name(self) -> str:
"""Name of blueprint used to call workflow.
This value is taken from server response not request (should be the same).
Returns:
str: Blueprint name
"""
return self.execution_output.actionIdentifiers.blueprintName
@property
def blueprint_version(self) -> str:
"""Blueprint version.
This value is taken from server response not request (should be the same).
Returns:
str: Blueprint version
"""
return self.execution_output.actionIdentifiers.blueprintVersion
@property
def workflow_name(self) -> str:
"""Workflow name.
This value is taken from server response not request (should be the same).
Returns:
str: Workflow name
"""
return self.execution_output.actionIdentifiers.actionName
@property
def has_error(self) -> bool:
"""Returns bool if request returns error or not.
Returns:
bool: True if response has status code different than 200
"""
return self.execution_output.status.code != 200
@property
def error_message(self) -> str:
"""Error message.
This property is available only if response has error. Otherwise AttributeError will be raised.
Raises:
AttributeError: Response has 200 response code and hasn't error message.
Returns:
str: Error message returned by server
"""
if self.has_error:
return self.execution_output.status.errorMessage
raise AttributeError("Execution does not finish with error")
@property
def payload(self) -> dict:
"""Response payload.
Payload retured by the server is migrated to Python dict.
Returns:
dict: Response's payload.
"""
return json_format.MessageToDict(self.execution_output.payload)
class ResourceResolution:
"""Resource resolution class.
Helper class to connect to blueprintprocessor's gRPC server, send request to execute workflow and parse responses.
Blueprint with workflow must be deployed before workflow request.
It's possible to create both secre or unsecure connection (without SSL/TLS).
"""
def __init__(
self,
*,
server_address: str = "127.0.0.1",
server_port: int = "9111",
use_ssl: bool = False,
root_certificates: bytes = None,
private_key: bytes = None,
certificate_chain: bytes = None,
# Authentication header configuration
use_header_auth: bool = False,
header_auth_token: str = None,
) -> None:
"""Resource resolution object initialization.
Args:
server_address (str, optional): gRPC server address. Defaults to "127.0.0.1".
server_port (int, optional): gRPC server address port. Defaults to "9111".
use_ssl (bool, optional): Boolean flag to determine if secure channel should be created or not.
Defaults to False.
root_certificates (bytes, optional): The PEM-encoded root certificates. None if it shouldn't be used.
Defaults to None.
private_key (bytes, optional): The PEM-encoded private key as a byte string, or None if no private key
should be used. Defaults to None.
certificate_chain (bytes, optional): The PEM-encoded certificate chain as a byte string to use or or None if
no certificate chain should be used. Defaults to None.
use_header_auth (bool, optional): Boolean flag to determine if authorization headed shoud be added for
every call or not. Defaults to False.
header_auth_token (str, optional): Authorization token value. Defaults to None.
"""
# Logger
self.logger: Logger = getLogger(__name__)
# Client settings
self.client_server_address: str = server_address
self.client_server_port: str = server_port
self.client_use_ssl: bool = use_ssl
self.client_root_certificates: bytes = root_certificates
self.client_private_key: bytes = private_key
self.client_certificate_chain: bytes = certificate_chain
self.client_use_header_auth: bool = use_header_auth
self.client_header_auth_token: str = header_auth_token
self.client: Client = None
def __enter__(self) -> "ResourceResolution":
"""Enter ResourceResolution instance context.
Client connection is created.
"""
self.client = Client(
server_address=f"{self.client_server_address}:{self.client_server_port}",
use_ssl=self.client_use_ssl,
root_certificates=self.client_root_certificates,
private_key=self.client_private_key,
certificate_chain=self.client_certificate_chain,
use_header_auth=self.client_use_header_auth,
header_auth_token=self.client_header_auth_token,
)
return self
def __exit__(
self,
unused_exc_type: Optional[Type[BaseException]],
unused_exc_value: Optional[BaseException],
unused_traceback: Optional[TracebackType],
) -> None:
"""Exit ResourceResolution instance context.
Client connection is closed.
"""
self.client.close()
def execute_workflows(self, *workflows: WorkflowExecution) -> Generator[WorkflowExecutionResult, None, None]:
"""Execute provided workflows.
Workflows are going to be execured using one gRPC API call. Depends of implementation that may has
some consequences. In some cases if any request fails all requests after that won't be called.
Responses and zipped with workflows and WorkflowExecutionResult object is initialized and yielded.
Raises:
AttributeError: Raises if client object is not created. It occurs only if you not uses context manager.
Then user have to create client instance for ResourceResolution object by himself calling:
```
resource_resoulution.client = Client(
server_address=f"{resource_resoulution.client_server_address}:{resource_resoulution.client_server_port}",
use_ssl=resource_resoulution.client_use_ssl,
root_certificates=resource_resoulution.client_root_certificates,
private_key=resource_resoulution.client_private_key,
certificate_chain=resource_resoulution.client_certificate_chain,
use_header_auth=resource_resoulution.client_use_header_auth,
header_auth_token=resource_resoulution.client_header_auth_token,
)
```
Remeber also to close client connection.
Returns:
Generator[WorkflowExecutionResult, None, None]: WorkflowExecutionResult object
with both WorkflowExection object and server response for it's request.
"""
self.logger.debug("Execute workflows")
if not self.client:
raise AttributeError("gRPC client not connected")
for response, workflow in zip(self.client.process((workflow.message for workflow in workflows)), workflows):
yield WorkflowExecutionResult(workflow, response)
|