This repository was archived by the owner on Mar 9, 2022. It is now read-only.
forked from snowplow/snowplow-python-tracker
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathemitters.py
More file actions
444 lines (383 loc) · 16 KB
/
emitters.py
File metadata and controls
444 lines (383 loc) · 16 KB
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
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
"""
emitters.py
Copyright (c) 2013-2014 Snowplow Analytics Ltd. All rights reserved.
This program is licensed to you under the Apache License Version 2.0,
and you may not use this file except in compliance with the Apache License
Version 2.0. You may obtain a copy of the Apache License Version 2.0 at
http://www.apache.org/licenses/LICENSE-2.0.
Unless required by applicable law or agreed to in writing,
software distributed under the Apache License Version 2.0 is distributed on
an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either
express or implied. See the Apache License Version 2.0 for the specific
language governing permissions and limitations there under.
Authors: Anuj More, Alex Dean, Fred Blundun
Copyright: Copyright (c) 2013-2014 Snowplow Analytics Ltd
License: Apache License Version 2.0
"""
import json
import logging
import time
import threading
try:
# Python 2
from Queue import Queue
except ImportError:
# Python 3
from queue import Queue
from celery import Celery
import redis
import requests
from contracts import contract, new_contract
from snowplow_tracker.self_describing_json import SelfDescribingJson
logger = logging.getLogger(__name__)
logger.setLevel(logging.INFO)
DEFAULT_MAX_LENGTH = 10
PAYLOAD_DATA_SCHEMA = "iglu:com.snowplowanalytics.snowplow/payload_data/jsonschema/1-0-4"
new_contract("protocol", lambda x: x == "http" or x == "https")
new_contract("method", lambda x: x == "get" or x == "post")
new_contract("function", lambda x: hasattr(x, "__call__"))
new_contract("redis", lambda x: isinstance(x, (redis.Redis, redis.StrictRedis)))
try:
# Check whether a custom Celery configuration module named "snowplow_celery_config" exists
import snowplow_celery_config
app = Celery()
app.config_from_object(snowplow_celery_config)
except ImportError:
# Otherwise configure Celery with default settings
snowplow_celery_config = None
app = Celery("Snowplow", broker="redis://guest@localhost//")
class Emitter(object):
"""
Synchronously send Snowplow events to a Snowplow collector
Supports both GET and POST requests
"""
@contract
def __init__(self, endpoint, protocol="http", port=None, method="get", buffer_size=None, on_success=None, on_failure=None, byte_limit=None):
"""
:param endpoint: The collector URL. Don't include "http://" - this is done automatically.
:type endpoint: string
:param protocol: The protocol to use - http or https. Defaults to http.
:type protocol: protocol
:param port: The collector port to connect to
:type port: int | None
:param method: The HTTP request method
:type method: method
:param buffer_size: The maximum number of queued events before the buffer is flushed. Default is 10.
:type buffer_size: int | None
:param on_success: Callback executed after every HTTP request in a flush has status code 200
Gets passed the number of events flushed.
:type on_success: function | None
:param on_failure: Callback executed if at least one HTTP request in a flush has status code 200
Gets passed two arguments:
1) The number of events which were successfully sent
2) If method is "post": The unsent data in string form;
If method is "get": An array of dictionaries corresponding to the unsent events' payloads
:type on_failure: function | None
:param byte_limit: The size event list after reaching which queued events will be flushed
:type byte_limit: int | None
"""
self.endpoint = Emitter.as_collector_uri(endpoint, protocol, port, method)
self.method = method
if buffer_size is None:
if method == "post":
buffer_size = DEFAULT_MAX_LENGTH
else:
buffer_size = 1
self.buffer_size = buffer_size
self.buffer = []
self.byte_limit = byte_limit
self.bytes_queued = None if byte_limit is None else 0
self.on_success = on_success
self.on_failure = on_failure
self.lock = threading.RLock()
self.timer = None
logger.info("Emitter initialized with endpoint " + self.endpoint)
@staticmethod
@contract
def as_collector_uri(endpoint, protocol="http", port=None, method="get"):
"""
:param endpoint: The raw endpoint provided by the user
:type endpoint: string
:param protocol: The protocol to use - http or https
:type protocol: protocol
:param port: The collector port to connect to
:type port: int | None
:param method: Either `get` or `post` HTTP method
:type method: method
:rtype: string
"""
if method == "get":
path = "/i"
else:
path = "/com.snowplowanalytics.snowplow/tp2"
if port is None:
return protocol + "://" + endpoint + path
else:
return protocol + "://" + endpoint + ":" + str(port) + path
@contract
def input(self, payload):
"""
Adds an event to the buffer.
If the maximum size has been reached, flushes the buffer.
:param payload: The name-value pairs for the event
:type payload: dict(string:*)
"""
with self.lock:
if self.bytes_queued is not None:
self.bytes_queued += len(str(payload))
if self.method == "post":
self.buffer.append({key: str(payload[key]) for key in payload})
else:
self.buffer.append(payload)
if self.reached_limit():
self.flush()
def reached_limit(self):
"""
Checks if event-size or bytes limit are reached
:rtype: bool
"""
if self.byte_limit is None:
return len(self.buffer) >= self.buffer_size
else:
return self.bytes_queued >= self.byte_limit or len(self.buffer) >= self.buffer_size
def flush(self):
"""
Sends all events in the buffer to the collector.
"""
with self.lock:
self.send_events(self.buffer)
self.buffer = []
if self.bytes_queued is not None:
self.bytes_queued = 0
@contract
def http_post(self, data):
"""
:param data: The array of JSONs to be sent
:type data: string
"""
logger.info("Sending POST request to %s..." % self.endpoint)
logger.debug("Payload: %s" % data)
r = requests.post(self.endpoint, data=data, headers={'content-type': 'application/json; charset=utf-8'})
getattr(logger, "info" if self.is_good_status_code(r.status_code) else "warn")("POST request finished with status code: " + str(r.status_code))
return r
@contract
def http_get(self, payload):
"""
:param payload: The event properties
:type payload: dict(string:*)
"""
logger.info("Sending GET request to %s..." % self.endpoint)
logger.debug("Payload: %s" % payload)
r = requests.get(self.endpoint, params=payload)
getattr(logger, "info" if self.is_good_status_code(r.status_code) else "warn")("GET request finished with status code: " + str(r.status_code))
return r
def sync_flush(self):
"""
Calls the flush method of the base Emitter class.
This is guaranteed to be blocking, not asynchronous.
"""
logger.debug("Starting synchronous flush...")
Emitter.flush(self)
logger.info("Finished synchrous flush")
@staticmethod
@contract
def is_good_status_code(status_code):
"""
:param status_code: HTTP status code
:type status_code: int
:rtype: bool
"""
return 200 <= status_code < 400
@contract
def send_events(self, evts):
"""
:param evts: Array of events to be sent
:type evts: list(dict(string:*))
"""
if len(evts) > 0:
logger.info("Attempting to send %s requests" % len(evts))
Emitter.attach_sent_timestamp(evts)
if self.method == 'post':
data = SelfDescribingJson(PAYLOAD_DATA_SCHEMA, evts).to_string()
post_succeeded = False
try:
status_code = self.http_post(data).status_code
post_succeeded = self.is_good_status_code(status_code)
except requests.RequestException as e:
logger.warn(e)
if post_succeeded:
if self.on_success is not None:
self.on_success(len(evts))
elif self.on_failure is not None:
self.on_failure(0, evts)
elif self.method == 'get':
success_count = 0
unsent_requests = []
for evt in evts:
get_succeeded = False
try:
status_code = self.http_get(evt).status_code
get_succeeded = self.is_good_status_code(status_code)
except requests.RequestException as e:
logger.warn(e)
if get_succeeded:
success_count += 1
else:
unsent_requests.append(evt)
if len(unsent_requests) == 0:
if self.on_success is not None:
self.on_success(success_count)
elif self.on_failure is not None:
self.on_failure(success_count, unsent_requests)
else:
logger.info("Skipping flush since buffer is empty")
@contract
def set_flush_timer(self, timeout, flush_now=False):
"""
Set an interval at which the buffer will be flushed
:param timeout: interval in seconds
:type timeout: int | float
:param flush_now: immediately flush buffer
:type flush_now: bool
"""
# Repeatable create new timer
if flush_now:
self.flush()
self.timer = threading.Timer(timeout, self.set_flush_timer, [timeout, True])
self.timer.daemon = True
self.timer.start()
def cancel_flush_timer(self):
"""
Abort automatic async flushing
"""
if self.timer is not None:
self.timer.cancel()
@staticmethod
def attach_sent_timestamp(events):
"""
Attach (by mutating in-place) current timestamp in milliseconds
as `stm` param
:param events: Array of events to be sent
:type events: list(dict(string:*))
:rtype: None
"""
def update(e):
e.update({'stm': str(int(time.time()) * 1000)})
[update(event) for event in events]
class AsyncEmitter(Emitter):
"""
Uses threads to send HTTP requests asynchronously
"""
@contract
def __init__(
self,
endpoint,
protocol="http",
port=None,
method="get",
buffer_size=None,
on_success=None,
on_failure=None,
thread_count=1,
byte_limit=None):
"""
:param endpoint: The collector URL. Don't include "http://" - this is done automatically.
:type endpoint: string
:param protocol: The protocol to use - http or https. Defaults to http.
:type protocol: protocol
:param port: The collector port to connect to
:type port: int | None
:param method: The HTTP request method
:type method: method
:param buffer_size: The maximum number of queued events before the buffer is flushed. Default is 10.
:type buffer_size: int | None
:param on_success: Callback executed after every HTTP request in a flush has status code 200
Gets passed the number of events flushed.
:type on_success: function | None
:param on_failure: Callback executed if at least one HTTP request in a flush has status code 200
Gets passed two arguments:
1) The number of events which were successfully sent
2) If method is "post": The unsent data in string form;
If method is "get": An array of dictionaries corresponding to the unsent events' payloads
:type on_failure: function | None
:param thread_count: Number of worker threads to use for HTTP requests
:type thread_count: int
:param byte_limit: The size event list after reaching which queued events will be flushed
:type byte_limit: int | None
"""
super(AsyncEmitter, self).__init__(endpoint, protocol, port, method, buffer_size, on_success, on_failure, byte_limit)
self.queue = Queue()
for i in range(thread_count):
t = threading.Thread(target=self.consume)
t.daemon = True
t.start()
def sync_flush(self):
while True:
self.flush()
self.queue.join()
if len(self.buffer) < 1:
break
def flush(self):
"""
Removes all dead threads, then creates a new thread which
executes the flush method of the base Emitter class
"""
with self.lock:
self.queue.put(self.buffer)
self.buffer = []
if self.bytes_queued is not None:
self.bytes_queued = 0
def consume(self):
while True:
evts = self.queue.get()
self.send_events(evts)
self.queue.task_done()
@app.task(bind=True, name='tasks.flush') # the self passed with bind can be used for on_fail/retrying
def flush_emitter(self, emitter):
try:
emitter.flush()
finally:
logger.info("Flush called on emitter")
class CeleryEmitter(Emitter):
"""
Uses a Celery worker to send HTTP requests asynchronously.
Works like the base Emitter class,
but on_success and on_failure callbacks cannot be set.
"""
def __init__(self, endpoint, protocol="http", port=None, method="get", buffer_size=None, byte_limit=None):
super(CeleryEmitter, self).__init__(endpoint, protocol, port, method, buffer_size, None, None, byte_limit)
def flush(self):
"""
Schedules a flush task
"""
flush_emitter.delay(self) # passes emitter (self - CeleryEmitter) to task
logger.info("Scheduled a Celery task to flush the event queue")
class RedisEmitter(object):
"""
Sends Snowplow events to a Redis database
"""
@contract
def __init__(self, rdb=None, key="snowplow"):
"""
:param rdb: Optional custom Redis database
:type rdb: redis | None
:param key: The Redis key for the list of events
:type key: string
"""
if rdb is None:
rdb = redis.StrictRedis()
self.rdb = rdb
self.key = key
@contract
def input(self, payload):
"""
:param payload: The event properties
:type payload: dict(string:*)
"""
logger.debug("Pushing event to Redis queue...")
self.rdb.rpush(self.key, json.dumps(payload))
logger.info("Finished sending event to Redis.")
def flush(self):
logger.warn("The RedisEmitter class does not need to be flushed")
def sync_flush(self):
self.flush()