rflow 0.0.5
Sign up to get free protection for your applications and to get access to all the features.
- checksums.yaml +7 -0
- data/.gitignore +6 -0
- data/.rvmrc +1 -0
- data/Gemfile +5 -0
- data/NOTES +187 -0
- data/README +0 -0
- data/Rakefile +16 -0
- data/bin/rflow +215 -0
- data/example/basic_config.rb +49 -0
- data/example/basic_extensions.rb +142 -0
- data/example/http_config.rb +21 -0
- data/example/http_extensions.rb +262 -0
- data/lib/rflow.rb +440 -0
- data/lib/rflow/component.rb +192 -0
- data/lib/rflow/component/port.rb +150 -0
- data/lib/rflow/components.rb +10 -0
- data/lib/rflow/components/raw.rb +26 -0
- data/lib/rflow/components/raw/extensions.rb +18 -0
- data/lib/rflow/configuration.rb +290 -0
- data/lib/rflow/configuration/component.rb +27 -0
- data/lib/rflow/configuration/connection.rb +98 -0
- data/lib/rflow/configuration/migrations/20010101000001_create_settings.rb +14 -0
- data/lib/rflow/configuration/migrations/20010101000002_create_components.rb +19 -0
- data/lib/rflow/configuration/migrations/20010101000003_create_ports.rb +24 -0
- data/lib/rflow/configuration/migrations/20010101000004_create_connections.rb +27 -0
- data/lib/rflow/configuration/port.rb +30 -0
- data/lib/rflow/configuration/ruby_dsl.rb +183 -0
- data/lib/rflow/configuration/setting.rb +67 -0
- data/lib/rflow/configuration/uuid_keyed.rb +18 -0
- data/lib/rflow/connection.rb +59 -0
- data/lib/rflow/connections.rb +2 -0
- data/lib/rflow/connections/zmq_connection.rb +101 -0
- data/lib/rflow/message.rb +191 -0
- data/lib/rflow/port.rb +4 -0
- data/lib/rflow/util.rb +19 -0
- data/lib/rflow/version.rb +3 -0
- data/rflow.gemspec +42 -0
- data/schema/message.avsc +36 -0
- data/schema/raw.avsc +9 -0
- data/spec/fixtures/config_ints.rb +61 -0
- data/spec/fixtures/extensions_ints.rb +141 -0
- data/spec/rflow_configuration_spec.rb +73 -0
- data/spec/rflow_message_data_raw.rb +26 -0
- data/spec/rflow_message_data_spec.rb +60 -0
- data/spec/rflow_message_spec.rb +182 -0
- data/spec/rflow_spec.rb +100 -0
- data/spec/schema_spec.rb +28 -0
- data/spec/spec_helper.rb +37 -0
- data/temp.rb +295 -0
- metadata +270 -0
data/spec/rflow_spec.rb
ADDED
@@ -0,0 +1,100 @@
|
|
1
|
+
require 'spec_helper.rb'
|
2
|
+
|
3
|
+
require 'rflow'
|
4
|
+
|
5
|
+
describe RFlow do
|
6
|
+
before(:each) do
|
7
|
+
@fixture_directory_path = File.join(File.dirname(__FILE__), 'fixtures')
|
8
|
+
end
|
9
|
+
|
10
|
+
|
11
|
+
describe 'logger' do
|
12
|
+
it "should initialize correctly" do
|
13
|
+
log_file_path = File.join(@temp_directory_path, 'logfile')
|
14
|
+
RFlow.initialize_logger log_file_path
|
15
|
+
|
16
|
+
File.exist?(log_file_path).should_not be_nil
|
17
|
+
|
18
|
+
RFlow.logger.error "TESTTESTTEST"
|
19
|
+
File.read(log_file_path).should match(/TESTTESTTEST/)
|
20
|
+
|
21
|
+
RFlow.close_log_file
|
22
|
+
end
|
23
|
+
|
24
|
+
it "should reopen correctly" do
|
25
|
+
log_file_path = File.join(@temp_directory_path, 'logfile')
|
26
|
+
moved_path = log_file_path + '.old'
|
27
|
+
|
28
|
+
RFlow.initialize_logger log_file_path
|
29
|
+
File.exist?(log_file_path).should be_true
|
30
|
+
File.exist?(moved_path).should be_false
|
31
|
+
|
32
|
+
File.rename log_file_path, moved_path
|
33
|
+
|
34
|
+
RFlow.reopen_log_file
|
35
|
+
|
36
|
+
RFlow.logger.error "TESTTESTTEST"
|
37
|
+
File.read(log_file_path).should match(/TESTTESTTEST/)
|
38
|
+
File.read(moved_path).should_not match(/TESTTESTTEST/)
|
39
|
+
|
40
|
+
RFlow.close_log_file
|
41
|
+
end
|
42
|
+
|
43
|
+
it "should toggle log level" do
|
44
|
+
end
|
45
|
+
end
|
46
|
+
|
47
|
+
describe '.run' do
|
48
|
+
before(:each) do
|
49
|
+
@run_directory_path = File.join(@temp_directory_path, 'run')
|
50
|
+
@log_directory_path = File.join(@temp_directory_path, 'log')
|
51
|
+
Dir.mkdir @run_directory_path
|
52
|
+
Dir.mkdir @log_directory_path
|
53
|
+
end
|
54
|
+
|
55
|
+
it "should startup and run correctly with non-trivial workflow" do
|
56
|
+
config_file_path = File.join(@fixture_directory_path, 'config_ints.rb')
|
57
|
+
extensions_path = File.join(@fixture_directory_path, 'extensions_ints.rb')
|
58
|
+
config_database_path = File.join(@temp_directory_path, 'config.sqlite')
|
59
|
+
|
60
|
+
# Load the new database with the fixtured config file
|
61
|
+
RFlow::Configuration::initialize_database(config_database_path, config_file_path)
|
62
|
+
File.exist?(config_database_path).should be_true
|
63
|
+
|
64
|
+
# Load the fixtured extensions
|
65
|
+
load extensions_path
|
66
|
+
|
67
|
+
# Startup RFlow in its own thread
|
68
|
+
rflow_thread = Thread.new do
|
69
|
+
RFlow.run config_database_path, false
|
70
|
+
end
|
71
|
+
|
72
|
+
# TODO: figure out a way to get rid of this sleep, as there
|
73
|
+
# should be a better way
|
74
|
+
sleep(5)
|
75
|
+
|
76
|
+
all_file_path = File.join(@temp_directory_path, 'out')
|
77
|
+
all2_file_path = File.join(@temp_directory_path, 'out2')
|
78
|
+
even_file_path = File.join(@temp_directory_path, 'out_even')
|
79
|
+
odd_file_path = File.join(@temp_directory_path, 'out_odd')
|
80
|
+
even_odd_file_path = File.join(@temp_directory_path, 'out_even_odd')
|
81
|
+
even_odd2_file_path = File.join(@temp_directory_path, 'out_even_odd2')
|
82
|
+
|
83
|
+
File.exist?(all_file_path).should be_true
|
84
|
+
File.exist?(all2_file_path).should be_true
|
85
|
+
File.exist?(even_file_path).should be_true
|
86
|
+
File.exist?(odd_file_path).should be_true
|
87
|
+
File.exist?(even_odd_file_path).should be_true
|
88
|
+
File.exist?(even_odd2_file_path).should be_true
|
89
|
+
|
90
|
+
File.readlines(all_file_path).map(&:to_i).should == [20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
|
91
|
+
File.readlines(all2_file_path).map(&:to_i).should == [20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
|
92
|
+
File.readlines(even_file_path).map(&:to_i).should == [20, 22, 24, 26, 28, 30]
|
93
|
+
File.readlines(odd_file_path).map(&:to_i).should == [21, 23, 25, 27, 29]
|
94
|
+
File.readlines(even_odd_file_path).map(&:to_i).should == [20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
|
95
|
+
File.readlines(even_odd2_file_path).map(&:to_i).should == [20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30]
|
96
|
+
end
|
97
|
+
end
|
98
|
+
|
99
|
+
|
100
|
+
end
|
data/spec/schema_spec.rb
ADDED
@@ -0,0 +1,28 @@
|
|
1
|
+
require 'spec_helper.rb'
|
2
|
+
|
3
|
+
describe 'RFlow::Message::Data::Raw Avro Schema' do
|
4
|
+
before(:each) do
|
5
|
+
@schema_string = RFlow::Configuration.available_data_types['RFlow::Message::Data::Raw']['avro']
|
6
|
+
end
|
7
|
+
|
8
|
+
it "should encode and decode an object" do
|
9
|
+
raw = {
|
10
|
+
'raw' => Array.new(256) { rand(256) }.pack('c*')
|
11
|
+
}
|
12
|
+
|
13
|
+
expect {encode_avro(@schema_string, raw)}.to_not raise_error
|
14
|
+
avro_encoded_raw = encode_avro(@schema_string, raw)
|
15
|
+
|
16
|
+
expect {decode_avro(@schema_string, avro_encoded_raw)}.to_not raise_error
|
17
|
+
decoded_raw = decode_avro(@schema_string, avro_encoded_raw)
|
18
|
+
|
19
|
+
decoded_raw.should == raw
|
20
|
+
|
21
|
+
p decoded_raw['raw'].encoding
|
22
|
+
p raw['raw'].encoding
|
23
|
+
|
24
|
+
decoded_raw['raw'].should == raw['raw']
|
25
|
+
|
26
|
+
end
|
27
|
+
|
28
|
+
end
|
data/spec/spec_helper.rb
ADDED
@@ -0,0 +1,37 @@
|
|
1
|
+
require File.expand_path(File.join(File.dirname(__FILE__), '..', 'lib', 'rflow'))
|
2
|
+
|
3
|
+
require 'fileutils'
|
4
|
+
require 'log4r'
|
5
|
+
|
6
|
+
RSpec.configure do |config|
|
7
|
+
config.before(:all) do
|
8
|
+
RFlow.logger = Log4r::Logger.new 'test'
|
9
|
+
RFlow.logger.add Log4r::StdoutOutputter.new('test_stdout', :formatter => RFlow::LOG_PATTERN_FORMATTER)
|
10
|
+
end
|
11
|
+
|
12
|
+
config.before(:each) do
|
13
|
+
@temp_directory_path = File.expand_path(File.join(File.dirname(__FILE__), 'tmp'))
|
14
|
+
Dir.mkdir @temp_directory_path
|
15
|
+
end
|
16
|
+
|
17
|
+
config.after(:each) do
|
18
|
+
FileUtils.rm_rf @temp_directory_path
|
19
|
+
end
|
20
|
+
end
|
21
|
+
|
22
|
+
|
23
|
+
def decode_avro(schema_string, serialized_object)
|
24
|
+
schema = Avro::Schema.parse(schema_string)
|
25
|
+
serialized_object.force_encoding 'BINARY'
|
26
|
+
sio = StringIO.new(serialized_object)
|
27
|
+
Avro::IO::DatumReader.new(schema, schema).read Avro::IO::BinaryDecoder.new(sio)
|
28
|
+
end
|
29
|
+
|
30
|
+
def encode_avro(schema_string, object)
|
31
|
+
encoded_string = ''
|
32
|
+
encoded_string.force_encoding 'BINARY'
|
33
|
+
schema = Avro::Schema.parse(schema_string)
|
34
|
+
sio = StringIO.new(encoded_string)
|
35
|
+
Avro::IO::DatumWriter.new(schema).write object, Avro::IO::BinaryEncoder.new(sio)
|
36
|
+
encoded_string
|
37
|
+
end
|
data/temp.rb
ADDED
@@ -0,0 +1,295 @@
|
|
1
|
+
module RFlow
|
2
|
+
def self.run(config)
|
3
|
+
# Take in the config file
|
4
|
+
# Set a module-level config
|
5
|
+
# Set module-level attributes (logger)
|
6
|
+
# Create manager
|
7
|
+
# Start manager with parsed config elements
|
8
|
+
end
|
9
|
+
|
10
|
+
class Manager
|
11
|
+
|
12
|
+
def initialize(config)
|
13
|
+
end
|
14
|
+
# Find each component
|
15
|
+
# Instantiate (process management)
|
16
|
+
end
|
17
|
+
end
|
18
|
+
|
19
|
+
class SchemaRegistry
|
20
|
+
# maps data type names to schemas based on schema type
|
21
|
+
find_by_data_type_name
|
22
|
+
end
|
23
|
+
|
24
|
+
class MessageDataRegistry
|
25
|
+
def find(data_type_name)
|
26
|
+
# returns a data type class if registered, nil otherwise
|
27
|
+
end
|
28
|
+
end
|
29
|
+
|
30
|
+
class Message::Data
|
31
|
+
# contains the schema + data information
|
32
|
+
# subclasses can add extra functionality, otherwise will just have
|
33
|
+
# acces to standard messagedata stuffs (i.e. standard avro data types)
|
34
|
+
# delegates a lot to standard Avro types
|
35
|
+
|
36
|
+
# how does this get access to the registry at the class level?
|
37
|
+
class << self
|
38
|
+
attr_accessor :class_registry
|
39
|
+
attr_accessor :schema_registry
|
40
|
+
end
|
41
|
+
|
42
|
+
# Pointer to encapsulating message
|
43
|
+
attr_accessor :message
|
44
|
+
|
45
|
+
def initialize(data_type_name, serialized_data=nil, schema_name=nil, schema_type=nil, schema=nil, message=nil)
|
46
|
+
# schema_name ||= 'org.rflow.Messages.GenericStringMap'
|
47
|
+
# schema_type ||= 'avro'
|
48
|
+
# schema ||= 'default avro schema'
|
49
|
+
|
50
|
+
merge_options
|
51
|
+
|
52
|
+
# TODO: think about schema resolution and conflicts between passed
|
53
|
+
# data and schema registry
|
54
|
+
# Lookup schema based on data type name
|
55
|
+
registered_schema_name, registered_schema, registered_schema_type = self.class.schema_registry.find(data_type_name)
|
56
|
+
if registered_schema.nil? && schema
|
57
|
+
# If you were given a schema and didn't get one from the
|
58
|
+
# registry register the schema?
|
59
|
+
self.class.schema_registry.register(data_type_name, schema_name, schema_type, schema)
|
60
|
+
else
|
61
|
+
|
62
|
+
end
|
63
|
+
|
64
|
+
end
|
65
|
+
|
66
|
+
def self.create(data_type_name, data=nil, schema_name=nil, schema_type=nil, schema=nil)
|
67
|
+
# look for object in registry by data_type_name
|
68
|
+
# if object found, call new on that object
|
69
|
+
# otherwise, call new on the default object
|
70
|
+
message_class = self.class.data_class_registry.find(data_type_name)
|
71
|
+
if message_class.nil?
|
72
|
+
MessageData.new(data_type_name, data, schema_name, schema_type, schema)
|
73
|
+
else
|
74
|
+
message_class.create(data_type_name, data, schema_name, schema_type, schema)
|
75
|
+
end
|
76
|
+
end
|
77
|
+
end
|
78
|
+
|
79
|
+
module HTTPResponse
|
80
|
+
end
|
81
|
+
|
82
|
+
Message.new.extend(HTTPResponse)
|
83
|
+
|
84
|
+
class HTTPRequest < RFlow::Message::Data
|
85
|
+
# used to add methods, defaults, and more to data object, if required
|
86
|
+
|
87
|
+
# Put this in the registry
|
88
|
+
AVRO_SCHEMA_NAME = 'org.rflow.http_request'
|
89
|
+
DATA_TYPE_NAME = "HTTPRequest"
|
90
|
+
|
91
|
+
# All subclasses must have the same initialize signature. They need
|
92
|
+
# to figure out what to do when they get the extra parameters that
|
93
|
+
# might conflict with expectations. Subclasses are usually meant to
|
94
|
+
# enable extra functionality on a given data type, so as long as it
|
95
|
+
# operates properly, it might not care (duck typing)
|
96
|
+
def initialize(data_type_name, data, schema_name, schema_type, schema)
|
97
|
+
super(DATA_TYPE_NAME, data, AVRO_SCHEMA_NAME)
|
98
|
+
# do nice stuff with data here
|
99
|
+
end
|
100
|
+
|
101
|
+
def self.create(data_type_name, data, schema_name, schema_type, schema)
|
102
|
+
# figure out if you are being called with incompatible arguments,
|
103
|
+
# i.e. schema stuff
|
104
|
+
end
|
105
|
+
|
106
|
+
end
|
107
|
+
|
108
|
+
class Message
|
109
|
+
# contains all definitions about what to do for a message
|
110
|
+
# has a default Avro schema for a data type
|
111
|
+
|
112
|
+
class << self
|
113
|
+
attr_accessor :data_class_registry
|
114
|
+
end
|
115
|
+
|
116
|
+
|
117
|
+
# Should load all the data stuff, perhaps to top level method on object
|
118
|
+
|
119
|
+
attr_accessor :data_type_name, :provenance, :origination_context, :data_type_schema, :data
|
120
|
+
|
121
|
+
def initialize(data_type_name, provenance=nil, origination_context=nil, data_type_schema=nil, data=nil)
|
122
|
+
if data
|
123
|
+
# Potentially register this data_type_name to the schema
|
124
|
+
else
|
125
|
+
# Lookup MessageData type in the MessageDataRegistry
|
126
|
+
# if found and a class, create a specific MessageData object
|
127
|
+
# extend it with the module
|
128
|
+
# else, create generic MessageData object which will use
|
129
|
+
# the schema registry, under the hood
|
130
|
+
# if found and a module, extend object with found module
|
131
|
+
|
132
|
+
message_data_class = self.class.data_class_registry.find(data_type_name)
|
133
|
+
if message_data_class && message_data_class.class.is_a? Class
|
134
|
+
message_data = message_data_class.new
|
135
|
+
else
|
136
|
+
message_data = Message::Data.new
|
137
|
+
message_data.extend message_data_class if message_data_class.is_a? Module
|
138
|
+
end
|
139
|
+
end
|
140
|
+
end
|
141
|
+
|
142
|
+
end
|
143
|
+
|
144
|
+
class Port
|
145
|
+
def read_message
|
146
|
+
parts = read_all_parts
|
147
|
+
parts.assemble
|
148
|
+
data_type_name = read_message_part
|
149
|
+
provenance = read_message_part
|
150
|
+
origination_context = read_message_part
|
151
|
+
data_type_schema = read_message_part
|
152
|
+
data = read_message_part
|
153
|
+
|
154
|
+
message = Message.new(data_type_name, provenance, origination_context, data_type_schema, data)
|
155
|
+
|
156
|
+
message
|
157
|
+
end
|
158
|
+
end
|
159
|
+
|
160
|
+
class PortCollection
|
161
|
+
end
|
162
|
+
|
163
|
+
class Logger
|
164
|
+
end
|
165
|
+
|
166
|
+
class Component
|
167
|
+
def self.input_port(port_def)
|
168
|
+
@@input_ports ||= PortCollection.new
|
169
|
+
if port_def.is_a? Array
|
170
|
+
port_name = port_def.first.to_sym
|
171
|
+
port_incidence = :array
|
172
|
+
else
|
173
|
+
port_name = port_def
|
174
|
+
port_incidence = :single
|
175
|
+
end
|
176
|
+
@@input_ports[port_name] = InputPort.new port_name, port_incidence
|
177
|
+
end
|
178
|
+
|
179
|
+
def self.output_port
|
180
|
+
# same as input port with different stuffs
|
181
|
+
end
|
182
|
+
|
183
|
+
STATES = [:initialized, :started, :configured, :running, :stopping, :stopped]
|
184
|
+
attr_accessor :state
|
185
|
+
attr_accessor :input_ports
|
186
|
+
attr_accessor :output_ports
|
187
|
+
|
188
|
+
attr_accessor :uuid
|
189
|
+
attr_accessor :name
|
190
|
+
|
191
|
+
CONFIG_DEFAULTS = {
|
192
|
+
:logger,
|
193
|
+
:working_directory_path,
|
194
|
+
}
|
195
|
+
|
196
|
+
def initialize(config, run_directory)
|
197
|
+
# configure component
|
198
|
+
config = {
|
199
|
+
}
|
200
|
+
|
201
|
+
# TODO: where is the management bus listener configured/started
|
202
|
+
end
|
203
|
+
|
204
|
+
def run
|
205
|
+
input_ports.ready do |port|
|
206
|
+
message = port.read_message
|
207
|
+
process_input(port, message)
|
208
|
+
# read from the port and think about things
|
209
|
+
out.send('stuff')
|
210
|
+
another_out.send('more stuff')
|
211
|
+
end
|
212
|
+
# listen to
|
213
|
+
end
|
214
|
+
|
215
|
+
def process_message(input_port, message)
|
216
|
+
|
217
|
+
end
|
218
|
+
|
219
|
+
def receive_message(port)
|
220
|
+
port.receive
|
221
|
+
end
|
222
|
+
|
223
|
+
def send_message(port, message)
|
224
|
+
port.send(message)
|
225
|
+
end
|
226
|
+
|
227
|
+
end
|
228
|
+
|
229
|
+
class HTTPServer < RFlow::Component
|
230
|
+
input_port :responses
|
231
|
+
output_port :requests
|
232
|
+
|
233
|
+
input_types "HTTP::Response"
|
234
|
+
output_types "HTTP::Request"
|
235
|
+
|
236
|
+
|
237
|
+
end
|
238
|
+
|
239
|
+
class PassThrough < RFlow::Component
|
240
|
+
input_port [:in]
|
241
|
+
input_port :another_in
|
242
|
+
output_port :out
|
243
|
+
output_port :another_out
|
244
|
+
|
245
|
+
output_types
|
246
|
+
|
247
|
+
def initialize(config, run_directory)
|
248
|
+
# This will initialize the ports
|
249
|
+
super
|
250
|
+
# Do stuff to initialize component. Don't assume singleton
|
251
|
+
end
|
252
|
+
|
253
|
+
|
254
|
+
def process_message(input_port, data)
|
255
|
+
out.send(message)
|
256
|
+
another_out.send(message)
|
257
|
+
|
258
|
+
|
259
|
+
end
|
260
|
+
|
261
|
+
def process_data(input_port
|
262
|
+
|
263
|
+
end
|
264
|
+
|
265
|
+
|
266
|
+
class Transform < RFlow::Component
|
267
|
+
|
268
|
+
end
|
269
|
+
|
270
|
+
# Plugins:
|
271
|
+
|
272
|
+
# MessageData subclass: rflow-data-http_request
|
273
|
+
# lib/rflow-data-http_request.rb
|
274
|
+
require 'rflow'
|
275
|
+
require 'lib/data_name'
|
276
|
+
RFlow.available_data_types << data_name_object
|
277
|
+
|
278
|
+
|
279
|
+
# Component: rflow-component-http_server
|
280
|
+
# lib/rflow-component-http_server
|
281
|
+
require 'rflow'
|
282
|
+
require 'lib/component_name'
|
283
|
+
RFlow.available_components << component_class
|
284
|
+
|
285
|
+
|
286
|
+
|
287
|
+
|
288
|
+
# lib/component_name.rb ->
|
289
|
+
# data_type_name => schema + registration: just register in the application
|
290
|
+
|
291
|
+
|
292
|
+
|
293
|
+
# Server -> (HttpRequest -> Translate -> HTTPResponse) -> Server
|
294
|
+
|
295
|
+
|