-
Notifications
You must be signed in to change notification settings - Fork 7
/
parsing_lib.py
582 lines (512 loc) · 23 KB
/
parsing_lib.py
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
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
import datetime
import json
import os
import sys
import tempfile
import collections
import contextlib
import time
from pathlib import Path
from typing import Callable, Dict, Generator, List
import boto3
import requests
import requests.exceptions
import iso3166
try:
import common_lib
except ModuleNotFoundError:
import common.common_lib as common_lib
try:
import common.ingestion_logging as logging
except ModuleNotFoundError:
import ingestion_logging as logging
ENV_FIELD = "env"
SOURCE_URL_FIELD = "sourceUrl"
S3_BUCKET_FIELD = "s3Bucket"
S3_KEY_FIELD = "s3Key"
SOURCE_ID_FIELD = "sourceId"
UPLOAD_ID_FIELD = "uploadId"
DATE_FILTER_FIELD = "dateFilter"
DATE_RANGE_FIELD = "dateRange"
AUTH_FIELD = "auth"
DELTAS_FIELD = "deltas"
# Maximum exponential backoff times
MAX_WAIT_TIME = 600 # 5xx errors from data service
MAX_CONN_WAIT_TIME = 900 # connection errors from data service
# Expected date fields format.
DATE_FORMATS = ["%m/%d/%YZ", "%m/%d/%Y"]
# Number of cases to upload in batch.
# Increasing that number will speed-up the ingestion but will increase memory
# usage on the server-side and is known to cause OOMs so increase with caution.
CASES_BATCH_SIZE = 250
logger = logging.getLogger(__name__)
logger.setLevel("INFO")
try:
with (Path(__file__).parent / "geocoding_countries.json").open() as g:
GEOCODING_COUNTRIES = json.load(g)
COUNTRY_ISO2 = sorted(GEOCODING_COUNTRIES.keys())
except json.decoder.JSONDecodeError as e:
logger.exception(f"geocoding_countries.json JSONDecodeError: {e}")
logging.flushAll()
sys.exit(1)
s3_client = boto3.client("s3")
if os.environ.get("DOCKERIZED"):
s3_client = boto3.client(
"s3",
endpoint_url=os.environ.get("AWS_ENDPOINT", "https://localhost.localstack.cloud:4566"),
aws_access_key_id=os.environ.get("AWS_ACCESS_KEY_ID", "test"),
aws_secret_access_key=os.environ.get("AWS_SECRET_ACCESS_KEY", "test"),
region_name=os.environ.get("AWS_REGION", "eu-central-1")
)
def safe_int(x):
try:
return int(x)
except (ValueError, TypeError):
return None
def geocode_country(two_letter_iso_code):
two_letter_iso_code = two_letter_iso_code.upper()
try:
lat, lon, country = GEOCODING_COUNTRIES[two_letter_iso_code]
return {
"country": two_letter_iso_code,
"geoResolution": "Country",
"name": country,
"geometry": {"latitude": lat, "longitude": lon}
}
except KeyError:
logger.error(f"Code not found in country geocoding DB: {two_letter_iso_code}")
return None
def extract_event_fields(event: Dict):
logger.info(f"Extracting fields from event {event}")
if any(
field not in event
for field
in
[ENV_FIELD, SOURCE_URL_FIELD, SOURCE_ID_FIELD, S3_BUCKET_FIELD,
S3_KEY_FIELD]):
error_message = (
f"Required fields {ENV_FIELD}; {SOURCE_URL_FIELD}; {S3_BUCKET_FIELD}; "
f"{SOURCE_ID_FIELD}; {S3_KEY_FIELD} not found in input event json.")
e = ValueError(error_message)
common_lib.complete_with_error(e)
return event[
ENV_FIELD], event[
SOURCE_URL_FIELD], event[
SOURCE_ID_FIELD], event.get(
UPLOAD_ID_FIELD), event[
S3_BUCKET_FIELD], event[
S3_KEY_FIELD], event.get(
DATE_FILTER_FIELD, None), event.get(
DATE_RANGE_FIELD, None), event.get(
AUTH_FIELD, None), event.get(
DELTAS_FIELD, None)
def retrieve_raw_data_file(s3_bucket: str, s3_key: str, out_file):
try:
logger.info(f"Retrieving raw data from s3://{s3_bucket}/{s3_key}")
s3_client.download_fileobj(s3_bucket, s3_key, out_file)
out_file.seek(0)
except Exception as e:
common_lib.complete_with_error(e)
def retrieve_excluded_case_ids(source_id: str, date_filter: Dict, date_range: Dict, env: str,
headers=None, cookies=None):
if env == "locale2e":
return None
if date_range:
start_date = date_range["start"]
end_date = date_range["end"]
date_limits = f"&dateFrom={start_date}&dateTo={end_date}"
elif date_filter:
now = get_today()
delta = datetime.timedelta(days=date_filter["numDaysBeforeToday"])
cutoff_date = now - delta
start_date = datetime.datetime.strftime(cutoff_date, "%Y-%m-%d")
end_date = datetime.datetime.strftime(now, "%Y-%m-%d")
date_limits = f"&dateFrom={start_date}&dateTo={end_date}"
else:
now = get_today()
start_date = "2019-12-01"
end_date = datetime.datetime.strftime(now, "%Y-%m-%d")
date_limits = f"&dateFrom={start_date}&dateTo={end_date}"
excluded_case_ids_endpoint_url = (
f"{common_lib.get_source_api_url(env)}"
f"/excludedCaseIds?sourceId={source_id}{date_limits}")
res = requests.get(excluded_case_ids_endpoint_url, headers=headers, cookies=cookies)
if res and res.status_code == 200:
res_json = res.json()
logger.info("Excluded cases: Returning excluded cases.")
return res_json["cases"]
logger.info("Excluded cases: Returning None.")
return None
# This structure is needed for the partial name matching below.
countries_index = {c.name.upper(): c.alpha2 for c in iso3166.countries}
country_to_iso_fixes = {
"CZECH REPUBLIC": "CZ",
"UNITED STATES": "US",
"REUNION": "RE",
"DEMOCRATIC REPUBLIC OF THE CONGO": "CD",
"VIETNAM": "VN",
"SOUTH KOREA": "KR",
"REPUBLIC OF CONGO": "CG",
"COTE D'IVOIRE": "CI",
}
def iso3166_country_code(country_name: str) -> str:
"""
Given the name of a country, find the alpha2 code. Hopefully we find that the country name
is an exact match for the actual name of a country. If it is not, we try partial matches.
If there is neither an exact match nor EXACTLY ONE partial match, raise KeyError: it is a
bug in the parser if it generates locations with bad country names.
This is only used in cases where the parser supplies the location info: typically we would
expect the parser to give a location query which is later geocoded anyway.
"""
if len(country_name) == 2 and country_name.isalpha():
return country_name.upper()
if (ucase_name := country_name.upper()) in country_to_iso_fixes:
return country_to_iso_fixes[ucase_name]
country = iso3166.countries.get(ucase_name, None)
if country is not None:
return country.alpha2
# Didn't find an exact match, so try a partial match
matched_countries = [countries_index[k] for k in countries_index.keys() if ucase_name in k]
if len(matched_countries) == 1:
return matched_countries[0]
# Found 0 or many partial matches
raise KeyError
def prepare_cases(cases: Generator[Dict, None, None], upload_id: str, excluded_case_ids: list):
"""
Populates standard required fields for the G.h Case API.
TODO: Migrate source_id/source_url to this method.
"""
for case in cases:
case["caseReference"]["uploadIds"] = [upload_id]
if country := common_lib.deep_get(case, "location.country"):
case["location"]["country"] = iso3166_country_code(country)
for travel in common_lib.deep_get(case, "travelHistory.travel", default=[]):
if travel_country := common_lib.deep_get(travel, "location.country"):
travel["location"]["country"] = iso3166_country_code(travel_country)
if ((excluded_case_ids is None)
or ("sourceEntryId" not in case["caseReference"])
or (not case["caseReference"]["sourceEntryId"] in excluded_case_ids)):
yield remove_nested_none_and_empty(case)
def remove_nested_none_and_empty(d):
if not isinstance(d, (dict, list)):
return d
if isinstance(d, list):
return [v for v in (remove_nested_none_and_empty(v)
for v in d) if v is not None and v != ""]
return {k: v for k, v in ((k, remove_nested_none_and_empty(v))
for k, v in d.items()) if v is not None and v != ""}
def batch_of(cases: Generator[Dict, None, None], max_items: int) -> List[Dict]:
n = 0
batch = []
try:
while n < max_items:
batch.append(next(cases))
n += 1
return batch
except StopIteration:
return batch
def write_to_server(
cases: Generator[Dict, None, None],
env: str, source_id: str, upload_id: str, headers, cookies,
cases_batch_size: int):
"""Upserts the provided cases via the G.h Case API."""
source_api_url = common_lib.get_source_api_url(env)
if env == "locale2e":
source_api_url = common_lib.get_source_api_url("local")
put_api_url = f"{source_api_url}/cases/batchUpsert"
upload_status_url = f"{source_api_url}/sources/{source_id}/uploads/{upload_id}"
logger.info(f"Prod URL: {put_api_url}")
counter = collections.defaultdict(int)
counter["batch_num"] = 0
start_time = time.time()
while True:
batch_num = counter["batch_num"]
counter["batch_num"] += 1
batch = batch_of(cases, cases_batch_size)
# End of batch.
if not batch:
break
# There are two exponential backoffs:
# * wait, total_wait keeps track of the backoff from 5xx errors
# * conn_wait, total_conn_wait tracks backoff for connection errors
total_wait = 0
total_conn_wait = 0
wait = 10 # initial wait time in seconds
conn_wait = 30
logger.info(f"Sending {len(batch)} cases, total so far: {counter['total']}")
# Exponential backoff in dev and prod, but not for local testing
while (
total_wait <= (MAX_WAIT_TIME if env in ["dev", "qa", "prod"] else 0)
and total_conn_wait < MAX_CONN_WAIT_TIME
):
try:
res = requests.post(put_api_url, json={"cases": batch},
headers=headers, cookies=cookies)
except requests.exceptions.ConnectionError:
logger.warning(f"Failed to connect to data service, waiting {conn_wait}s")
time.sleep(conn_wait)
total_conn_wait += conn_wait
conn_wait *= 2
continue
if res.status_code in [200, 207]: # 207 is used for validation error
break
if res.status_code == 500 and "401" in res.text:
logger.warning(f"Request failed, status={res.status_code}, "
f"response={res.text}, reauthenticating...")
headers = common_lib.obtain_api_credentials(s3_client)
continue
logger.warning(f"Request failed, status={res.status_code}, "
f"response={res.text}, retrying in {wait} seconds...")
time.sleep(wait)
total_wait += wait
wait *= 2
if total_conn_wait >= MAX_CONN_WAIT_TIME:
# data service has failed, raise alert
notifymsg = f"[!] *Failed to connect to data-{env}* during {source_id} ingestion"
if webhook_url := os.getenv("NOTIFY_WEBHOOK_URL"):
with contextlib.suppress(requests.exceptions.RequestException):
requests.post(webhook_url, json={"text": notifymsg})
common_lib.complete_with_error(
ConnectionError("Could not connect to data service"),
env,
common_lib.UploadError.INTERNAL_ERROR,
source_id, upload_id, headers, cookies,
count_created=counter["numCreated"],
count_updated=counter["numUpdated"],
count_error=counter["numError"]
)
return
if res and res.status_code in [200, 207]:
counter["total"] += len(batch)
now = time.time()
try:
cps = int(counter["total"] / (now - start_time))
except ZeroDivisionError:
cps = 0
logger.info(f"\tCurrent speed: {cps} cases/sec")
res_json = res.json()
counter["numCreated"] += res_json["numCreated"]
counter["numUpdated"] += res_json["numUpdated"]
if res.status_code == 207:
# 207 encompasses both geocoding and case schema validation errors.
# We can consider separating geocoding issues, but for now classifying it
# as a validation problem is pretty reasonable.
# The motivation for continuing past 207 errors is
# https://github.com/globaldothealth/list/issues/1849
# The errors from the backend tell us which cases failed and
# for what reason. Make it easier to diagnose by extracting the
# failing case and attaching it to the error message.
res_json = res.json()
if "errors" in res_json:
def add_input_to_error(error):
res = dict(error)
res['input'] = batch[error['index']]
return res
augmented_errors = [add_input_to_error(e) for e in res_json['errors']]
reported_error = dict(res_json)
reported_error["errors"] = augmented_errors
logger.warning(f"Validation error in batch {batch_num}: "
f"{json.dumps(reported_error)}")
counter["numError"] += len(res_json["errors"])
else:
logger.warning(f"Validation error in batch {batch_num}: {res.text}")
update_status = {
"status": "IN_PROGRESS",
"summary": {
"numCreated": counter["numCreated"],
"numUpdated": counter["numUpdated"],
"numError": counter["numError"]
}
}
with contextlib.suppress(requests.exceptions.RequestException):
requests.put(upload_status_url, json=update_status,
headers=headers, cookies=cookies)
continue
# Response can contain an 'error' field which describe each error that
# occurred, it will be contained in the res.text here below.
e = RuntimeError(
f"Error sending cases to server, status={res.status_code}, response={res.text}")
upload_error = common_lib.UploadError.DATA_UPLOAD_ERROR
common_lib.complete_with_error(
e, env, upload_error,
source_id, upload_id, headers, cookies,
count_created=counter["numCreated"],
count_updated=counter["numUpdated"],
count_error=counter["numError"]
)
return
logger.info(f"sent {counter['total']} cases in {time.time() - start_time} seconds")
return counter["numCreated"], counter["numUpdated"], counter["numError"]
def get_today() -> datetime.datetime:
"""Return today's datetime, just here for easier mocking."""
return datetime.datetime.today()
def get_case_date(date_string) -> datetime.datetime:
"""Return a datetime parsed from a case."""
case_date = ""
for fmt in (DATE_FORMATS):
try:
return datetime.datetime.strptime(
date_string,
fmt)
except ValueError:
pass
if not case_date:
raise ValueError(f"Date {date_string} from case could not be parsed.")
return case_date
def filter_cases_by_date(
case_data: Generator[Dict, None, None],
date_filter: Dict, date_range: Dict, env: str,
source_id: str, upload_id: str, api_creds, cookies):
"""
Filter cases according to the date_range or date_filter provided.
If a date_range is provided, returns only cases within the specified start
and end bounds (inclusive). Else if date_filter is provided, returns the
cases within that specification. Else, returns all cases.
Notice that if _both_ date_range and date_filter are provided, then date_range is used
and date_filter is ignored.
"""
if date_range:
logger.info(f"Filtering cases using date range {date_range}")
def case_is_within_range(case, start, end):
confirmed_event = [e for e in case["events"]
if e["name"] == "confirmed"][0]
case_date = get_case_date(confirmed_event["dateRange"]["start"])
return start <= case_date <= end
start = datetime.datetime.strptime(date_range["start"], "%Y-%m-%d")
end = datetime.datetime.strptime(date_range["end"], "%Y-%m-%d")
return (case for case in case_data if case_is_within_range(case, start, end))
elif date_filter:
logger.info(f"Filtering cases using date filter {date_filter}")
now = get_today()
delta = datetime.timedelta(days=date_filter["numDaysBeforeToday"])
cutoff_date = now - delta
op = date_filter["op"]
def case_is_within_range(case, cutoff_date, op):
confirmed_event = [e for e in case["events"]
if e["name"] == "confirmed"][0]
case_date = get_case_date(confirmed_event["dateRange"]["start"])
delta_days = (case_date - cutoff_date).days
if op == "EQ":
return delta_days == 0
elif op == "LT":
return delta_days < 0
elif op == "GT":
return delta_days > 0
else:
e = ValueError(f"Unsupported date filter operand: {op}")
common_lib.complete_with_error(
e, env, common_lib.UploadError.SOURCE_CONFIGURATION_ERROR,
source_id, upload_id, api_creds, cookies)
return (case for case in case_data if case_is_within_range(case, cutoff_date, op))
else:
return case_data
class ParserError(Exception):
pass
def run(
event: Dict,
parsing_function: Callable[[str, str, str], Generator[Dict, None, None]]):
"""
Encapsulates all of the work performed by a parsing Lambda.
Parameters
----------
event: dict, required
Input event JSON-as-dict.
This must contain `s3Bucket`, `s3Key`, and `sourceUrl` fields specifying
the details of the stored source content.
parsing_function: function, required
Python function that parses raw source data into G.h case data.
This function must accept (in order): a file containing raw source
data, a string representing the source UUID, and a string representing
the source URL. It must yield each case conforming to the G.h
case format as per https://data.covid-19.global.health/api-docs/.
For an example, see:
https://github.com/globaldothealth/list/blob/main/ingestion/functions/parsing/india/india.py#L57
Returns
------
JSON object containing the count of line list cases successfully written to
G.h servers in the format:
{"count_created": count_created, "count_updated": count_updated}
For more information on return types, see:
https://docs.aws.amazon.com/lambda/latest/dg/python-handler.html
"""
(env, source_url, source_id, upload_id, s3_bucket, s3_key, date_filter,
date_range, local_auth, deltas) = extract_event_fields(event)
logger.info(
f"Event fields extracted in parsing_lib.run...env: {env}, "
f"source_url: {source_url}, source_id: {source_id}, "
f"upload_id: {upload_id}, s3_bucket: {s3_bucket}, s3_key: {s3_key}, "
f"date_filter: {date_filter}, date_range: {date_range}, "
f"local_auth: {local_auth}")
api_creds = None
cookies = None
if local_auth and env in ["local", "locale2e"]:
cookies = common_lib.login(local_auth["email"])
else:
api_creds = common_lib.obtain_api_credentials(s3_client)
if not upload_id:
upload_id = common_lib.create_upload_record(
env, source_id, api_creds, cookies)
# grab the source object
base_url = common_lib.get_source_api_url(env)
source_info_url = f"{base_url}/sources/{source_id}"
source_info_request = requests.get(source_info_url, headers=api_creds, cookies=cookies)
# if that failed then just bail, we can't ingest the cases
if source_info_request.status_code > 299: # yes I'm ignoring redirects
common_lib.complete_with_error(
ParserError(f"Retrieving source info for source {source_id} "
f"yielded HTTP status {source_info_request.status_code}"),
env, common_lib.UploadError.INTERNAL_ERROR, source_id, upload_id,
api_creds, cookies)
try:
# retrieve source from s3 bucket
fd, local_data_file_name = tempfile.mkstemp()
local_data_file = os.fdopen(fd, "wb")
retrieve_raw_data_file(s3_bucket, s3_key, local_data_file)
logger.info(f"Raw file retrieved at {local_data_file_name}")
# construct parsing generator
logger.info("Running parsing function...")
case_data = parsing_function(
local_data_file_name, source_id,
source_url)
logger.info("Retrieving excluded case IDs...")
excluded_case_ids = retrieve_excluded_case_ids(source_id, date_filter, date_range, env,
headers=api_creds, cookies=cookies)
logger.info("Preparing cases...")
final_cases = prepare_cases(case_data, upload_id, excluded_case_ids)
logger.info("Writing to server...")
count_created, count_updated, count_error = write_to_server(
filter_cases_by_date(
final_cases,
date_filter,
date_range,
env, source_id, upload_id,
api_creds, cookies),
env, source_id, upload_id,
api_creds, cookies,
CASES_BATCH_SIZE)
for _ in range(5): # Maximum number of attempts to finalize upload
logger.info("Attempting to finalise upload...")
status, text = common_lib.finalize_upload(
env, source_id, upload_id, api_creds, cookies, count_created,
count_updated, count_error, deltas=deltas
)
if status == 200:
break
elif status == 500 and "401" in text:
logger.warning("Finalizing upload failed with 401, reauthenticating...")
api_creds = common_lib.obtain_api_credentials(s3_client)
continue
else:
raise RuntimeError(f"Error updating upload record, "
f"status={status}, response={text}")
logger.info(f"count_created={count_created}, count_updated={count_updated}")
return {"count_created": count_created, "count_updated": count_updated}
except Exception as e:
common_lib.complete_with_error(
e, env, common_lib.UploadError.INTERNAL_ERROR, source_id, upload_id,
api_creds, cookies)
finally:
local_data_file.close()
if os.path.exists(local_data_file_name):
os.remove(local_data_file_name)
logging.flushAll()