def test_ingest_article_zip_starter_(self, fake_boto_conn, fake_logger):
fake_boto_conn.return_value = FakeBotoConnection()
self.stater_ingest_article_zip.start(
settings=settings_mock,
run=run_example,
info=S3NotificationInfo.from_dict(
test_data.ingest_article_zip_data))
def start(self, ENV="dev", info=None, run=None):
# TODO : much of this is common to many starters and could probably be streamlined
# Specify run environment settings
settings = settingsLib.get_settings(ENV)
# Log
identity = "starter_%s" % int(random.random() * 1000)
log_file = "starter.log"
# logFile = None
logger = log.logger(log_file, settings.setLevel, identity)
filename = info.file_name
if filename is None:
logger.error("Did not get a filename")
return
# Simple connect
conn = boto.swf.layer1.Layer1(settings.aws_access_key_id,
settings.aws_secret_access_key)
# Start a workflow execution
workflow_id = "PublishPerfectArticle_%s" % filename.replace(
'/', '_') + str(int(random.random() * 1000))
workflow_name = "PublishPerfectArticle"
workflow_version = "1"
child_policy = None
execution_start_to_close_timeout = str(60 * 30)
workflow_input = S3NotificationInfo.to_dict(info)
workflow_input['run'] = run
workflow_input = json.dumps(workflow_input,
default=lambda ob: ob.__dict__)
try:
response = conn.start_workflow_execution(
settings.domain, workflow_id, workflow_name, workflow_version,
settings.default_task_list, child_policy,
execution_start_to_close_timeout, workflow_input)
logger.info('got response: \n%s' %
json.dumps(response, sort_keys=True, indent=4))
except boto.swf.exceptions.SWFWorkflowExecutionAlreadyStartedError:
# There is already a running workflow with that ID, cannot start another
message = 'SWFWorkflowExecutionAlreadyStartedError: There is already a running workflow with ID %s' % workflow_id
logger.info(message)
def do_activity(self, data=None):
"""
Do the work
"""
self.expanded_bucket_name = self.settings.publishing_buckets_prefix + self.settings.expanded_bucket
info = S3NotificationInfo.from_dict(data)
session = Session(self.settings)
version = session.get_value(self.get_workflowId(), 'version')
article_id = session.get_value(self.get_workflowId(), 'article_id')
article_version_id = article_id + '.' + version
run = session.get_value(self.get_workflowId(), 'run')
self.emit_monitor_event(self.settings, article_id, version, run, "Apply Version Number", "start",
"Starting applying version number to files for " + article_id)
try:
if self.logger:
self.logger.info('data: %s' % json.dumps(data, sort_keys=True, indent=4))
# Do not rename files if a version number is in the file_name
m = re.search(ur'-v([0-9]*?)[\.|-]', info.file_name)
if m is not None:
# Nothing to do
pass
elif m is None and version is not None:
expanded_folder_name = session.get_value(self.get_workflowId(), 'expanded_folder')
bucket_folder_name = expanded_folder_name.replace(os.sep, '/')
self.rename_article_s3_objects(bucket_folder_name, version)
self.emit_monitor_event(self.settings, article_id, version, run, "Apply Version Number", "end",
"Finished applying version number to article " + article_id +
" for version " + version + " run " + str(run))
except Exception as e:
self.logger.exception("Exception when applying version number to article")
self.emit_monitor_event(self.settings, article_id, version, run, "Convert JATS", "error",
"Error in applying version number to files for " + article_id +
" message:" + e.message)
return True
def start(self, settings, info=None, run=None):
# Log
logger = helper.get_starter_logger(
settings.setLevel, helper.get_starter_identity(self.const_name))
if hasattr(info, 'file_name') == False or info.file_name is None:
raise NullRequiredDataException(
"filename is Null / Did not get a filename.")
input = S3NotificationInfo.to_dict(info)
input['run'] = run
input['version_lookup_function'] = "article_highest_version"
input['force'] = True
workflow_id, \
workflow_name, \
workflow_version, \
child_policy, \
execution_start_to_close_timeout, \
workflow_input = helper.set_workflow_information(self.const_name, "1", None, input,
info.file_name.replace('/', '_'))
# Simple connect
conn = boto.swf.layer1.Layer1(settings.aws_access_key_id,
settings.aws_secret_access_key)
try:
response = conn.start_workflow_execution(
settings.domain, workflow_id, workflow_name, workflow_version,
settings.default_task_list, child_policy,
execution_start_to_close_timeout, workflow_input)
logger.info('got response: \n%s' %
json.dumps(response, sort_keys=True, indent=4))
except NullRequiredDataException as e:
logger.exception(e.message)
raise
except boto.swf.exceptions.SWFWorkflowExecutionAlreadyStartedError:
# There is already a running workflow with that ID, cannot start another
message = 'SWFWorkflowExecutionAlreadyStartedError: There is already a running workflow with ID %s' % workflow_id
logger.info(message)
def start(self, ENV="dev", info=None, run=None):
# TODO : much of this is common to many starters and could probably be streamlined
# Specify run environment settings
settings = settingsLib.get_settings(ENV)
# Log
identity = "starter_%s" % int(random.random() * 1000)
log_file = "starter.log"
# logFile = None
logger = log.logger(log_file, settings.setLevel, identity)
filename = info.file_name
if filename is None:
logger.error("Did not get a filename")
return
# Simple connect
conn = boto.swf.layer1.Layer1(settings.aws_access_key_id, settings.aws_secret_access_key)
# Start a workflow execution
workflow_id = "PublishPerfectArticle_%s" % filename.replace('/', '_') + str(int(random.random() * 1000))
workflow_name = "PublishPerfectArticle"
workflow_version = "1"
child_policy = None
execution_start_to_close_timeout = str(60 * 30)
workflow_input = S3NotificationInfo.to_dict(info)
workflow_input['run'] = run
workflow_input = json.dumps(workflow_input, default=lambda ob: ob.__dict__)
try:
response = conn.start_workflow_execution(settings.domain, workflow_id, workflow_name, workflow_version,
settings.default_task_list, child_policy,
execution_start_to_close_timeout, workflow_input)
logger.info('got response: \n%s' % json.dumps(response, sort_keys=True, indent=4))
except boto.swf.exceptions.SWFWorkflowExecutionAlreadyStartedError:
# There is already a running workflow with that ID, cannot start another
message = 'SWFWorkflowExecutionAlreadyStartedError: There is already a running workflow with ID %s' % workflow_id
logger.info(message)
def start(self, settings, run, info):
# Log
logger = helper.get_starter_logger(settings.setLevel, helper.get_starter_identity(self.const_name))
if hasattr(info, 'file_name') == False or info.file_name is None:
raise NullRequiredDataException("filename is Null. Did not get a filename.")
input = S3NotificationInfo.to_dict(info)
input['run'] = run
input['version_lookup_function'] = "article_next_version"
workflow_id, \
workflow_name, \
workflow_version, \
child_policy, \
execution_start_to_close_timeout, \
workflow_input = helper.set_workflow_information(self.const_name, "1", None, input,
info.file_name.replace('/', '_'),
start_to_close_timeout=str(60 * 60 * 5))
# Simple connect
conn = boto.swf.layer1.Layer1(settings.aws_access_key_id, settings.aws_secret_access_key)
try:
response = conn.start_workflow_execution(settings.domain, workflow_id, workflow_name, workflow_version,
settings.default_task_list, child_policy,
execution_start_to_close_timeout, workflow_input)
logger.info('got response: \n%s' % json.dumps(response, sort_keys=True, indent=4))
except NullRequiredDataException as e:
logger.exception(e.message)
raise
except boto.swf.exceptions.SWFWorkflowExecutionAlreadyStartedError:
# There is already a running workflow with that ID, cannot start another
message = 'SWFWorkflowExecutionAlreadyStartedError: ' \
'There is already a running workflow with ID %s' % workflow_id
logger.info(message)
def do_activity(self, data=None):
try:
info = S3NotificationInfo.from_dict(data)
filename = info.file_name[info.file_name.rfind('/')+1:]
session = Session(self.settings)
session.store_value(data['run'], 'filename_last_element', filename)
article_structure = ArticleInfo(filename)
if article_structure.article_id is None:
self.logger.error("Name '%s' did not match expected pattern for article id" % filename)
raise RuntimeError("article_structure.article_id is None. File pattern problem.")
version = self.get_version(self.settings, article_structure, data['version_lookup_function'])
session.store_value(data['run'], 'version', version)
article_id = article_structure.article_id
self.emit_monitor_event(self.settings, article_id, version, data['run'],
self.pretty_name, "start",
" ".join(("Version Lookup for article", article_id, "version:", version)))
self.set_monitor_property(self.settings, article_id, "article-id", article_id, "text")
self.set_monitor_property(self.settings, article_id, "publication-status", "publication in progress",
"text",
version=version)
self.emit_monitor_event(self.settings, article_structure.article_id, version, data['run'],
self.pretty_name, "end",
" ".join(("Finished Version Lookup for article", article_structure.article_id,
"version:", version)))
return activity.activity.ACTIVITY_SUCCESS
except Exception as e:
self.logger.exception("Exception when trying to Lookup Version. Error: " + str(e))
return activity.activity.ACTIVITY_PERMANENT_FAILURE
parser.add_option("-e",
"--env",
default="dev",
action="store",
type="string",
dest="env",
help="set the environment to run, either dev or live")
parser.add_option("-f",
"--filename",
default=None,
action="store",
type="string",
dest="filename",
help="specify the DOI id the article to process")
(options, args) = parser.parse_args()
if options.env:
ENV = options.env
if options.filename:
filename = options.filename
import settings as settingsLib
settings = settingsLib.get_settings(ENV)
o = starter_NewS3File()
info = S3NotificationInfo("S3Event", "", "xxawsxx-drop-bucket",
"elife-kitchen-sink.xml",
"3f53f5c808dd58973cd93a368be739b4", "1")
o.start(settings=settings, info=info)
def process_data_publishperfectarticle(workflow_name, workflow_data):
data = {
'info': S3NotificationInfo.from_dict(workflow_data),
'run': str(uuid.uuid4())
}
return data
def process_data_publishperfectarticle(workflow_name, workflow_data):
data = {'info': S3NotificationInfo.from_dict(workflow_data),
'run': str(uuid.uuid4())}
return data
def process_data_ingestarticlezip(workflow_name, workflow_data):
data = {'info': S3NotificationInfo.from_dict(workflow_data),
'run': str(uuid.uuid4())}
return data
def do_activity(self, data=None):
"""
Do the work
"""
run = data['run']
if self.logger:
self.logger.info('data: %s' % json.dumps(data, sort_keys=True, indent=4))
info = S3NotificationInfo.from_dict(data)
storage_context = StorageContext(self.settings)
session = Session(self.settings)
filename_last_element = session.get_value(run, 'filename_last_element')
# zip name contains version information for previously archived zip files
article_structure = ArticleInfo(filename_last_element)
article_id = article_structure.article_id
session.store_value(run, 'article_id', article_id)
session.store_value(run, 'file_name', info.file_name)
if self.logger:
self.logger.info("Expanding file %s" % info.file_name)
version = session.get_value(run, 'version')
status = article_structure.status
if status is None or (status != 'vor' and status != 'poa'):
self.logger.error("Name '%s' did not match expected pattern for status" %
filename_last_element)
return activity.activity.ACTIVITY_PERMANENT_FAILURE # status could not be determined, exit workflow.
article_version_id = article_id + '.' + version
session.store_value(run, 'article_version_id', article_version_id)
session.store_value(run, 'run', run)
session.store_value(run, 'status', status)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article", "start",
"Starting expansion of article " + article_id)
try:
# download zip to temp folder
tmp = self.get_tmp_dir()
local_zip_file = self.open_file_from_tmp_dir(filename_last_element, mode='wb')
storage_resource_origin = self.settings.storage_provider + "://" + info.bucket_name + "/" + info.file_name
storage_context.get_resource_to_file(storage_resource_origin, local_zip_file)
local_zip_file.close()
# extract zip contents
folder_name = path.join(article_version_id, run)
content_folder = path.join(tmp, folder_name)
makedirs(content_folder)
with ZipFile(path.join(tmp, filename_last_element)) as zf:
zf.extractall(content_folder)
upload_filenames = []
for f in listdir(content_folder):
if isfile(join(content_folder, f)) and f[0] != '.' and not f[0] == '_':
upload_filenames.append(f)
self.check_filenames(upload_filenames)
bucket_folder_name = article_version_id + '/' + run
for filename in upload_filenames:
source_path = path.join(content_folder, filename)
dest_path = bucket_folder_name + '/' + filename
storage_resource_dest = self.settings.storage_provider + "://" + self.settings.publishing_buckets_prefix + \
self.settings.expanded_bucket + "/" + dest_path
storage_context.set_resource_from_filename(storage_resource_dest, source_path)
self.clean_tmp_dir()
session.store_value(run, 'expanded_folder', bucket_folder_name)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article",
"end", "Finished expansion of article " + article_id +
" for version " + version + " run " + str(run) +
" into " + bucket_folder_name)
except Exception as e:
self.logger.exception("Exception when expanding article")
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article",
"error", "Error expanding article " + article_id +
" message:" + e.message)
return activity.activity.ACTIVITY_PERMANENT_FAILURE
return True
def do_activity(self, data=None):
"""
Do the work
"""
run = data['run']
if self.logger:
self.logger.info('data: %s' % json.dumps(data, sort_keys=True, indent=4))
info = S3NotificationInfo.from_dict(data)
storage_context = StorageContext(self.settings)
session = Session(self.settings)
filename_last_element = session.get_value(run, 'filename_last_element')
# zip name contains version information for previously archived zip files
article_structure = ArticleInfo(filename_last_element)
article_id = article_structure.article_id
session.store_value(run, 'article_id', article_id)
session.store_value(run, 'file_name', info.file_name)
if self.logger:
self.logger.info("Expanding file %s" % info.file_name)
version = session.get_value(run, 'version')
status = article_structure.status
if status is None or (status != 'vor' and status != 'poa'):
self.logger.error("Name '%s' did not match expected pattern for status" %
filename_last_element)
return activity.activity.ACTIVITY_PERMANENT_FAILURE # status could not be determined, exit workflow.
article_version_id = article_id + '.' + version
session.store_value(run, 'article_version_id', article_version_id)
session.store_value(run, 'run', run)
session.store_value(run, 'status', status)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article", "start",
"Starting expansion of article " + article_id)
try:
# download zip to temp folder
tmp = self.get_tmp_dir()
local_zip_file = self.open_file_from_tmp_dir(filename_last_element, mode='wb')
storage_resource_origin = self.settings.storage_provider + "://" + info.bucket_name + "/" + info.file_name
storage_context.get_resource_to_file(storage_resource_origin, local_zip_file)
local_zip_file.close()
# extract zip contents
folder_name = path.join(article_version_id, run)
content_folder = path.join(tmp, folder_name)
makedirs(content_folder)
with ZipFile(path.join(tmp, filename_last_element)) as zf:
zf.extractall(content_folder)
upload_filenames = []
for f in listdir(content_folder):
if isfile(join(content_folder, f)) and f[0] != '.' and not f[0] == '_':
upload_filenames.append(f)
bucket_folder_name = article_version_id + '/' + run
for filename in upload_filenames:
source_path = path.join(content_folder, filename)
dest_path = bucket_folder_name + '/' + filename
storage_resource_dest = self.settings.storage_provider + "://" + self.settings.publishing_buckets_prefix + \
self.settings.expanded_bucket + "/" + dest_path
storage_context.set_resource_from_filename(storage_resource_dest, source_path)
self.clean_tmp_dir()
session.store_value(run, 'expanded_folder', bucket_folder_name)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article",
"end", "Finished expansion of article " + article_id +
" for version " + version + " run " + str(run) +
" into " + bucket_folder_name)
except Exception as e:
self.logger.exception("Exception when expanding article")
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article",
"error", "Error expanding article " + article_id +
" message:" + e.message)
return activity.activity.ACTIVITY_PERMANENT_FAILURE
return True
def test_ingest_article_zip_starter_(self, fake_boto_conn, fake_logger):
fake_boto_conn.return_value = FakeBotoConnection()
self.stater_ingest_article_zip.start(settings=settings_mock, run=run_example,
info=S3NotificationInfo.from_dict(test_data.ingest_article_zip_data))
def work(ENV, flag):
# Specify run environment settings
settings = settings_lib.get_settings(ENV)
# Log
identity = "queue_worker_%s" % os.getpid()
log_file = "queue_worker.log"
# logFile = None
logger = log.logger(log_file, settings.setLevel, identity)
# Simple connect
conn = boto.sqs.connect_to_region(settings.sqs_region,
aws_access_key_id=settings.aws_access_key_id,
aws_secret_access_key=settings.aws_secret_access_key)
queue = conn.get_queue(settings.S3_monitor_queue)
queue.set_message_class(S3SQSMessage)
rules = load_rules()
application = newrelic.agent.application()
# Poll for an activity task indefinitely
if queue is not None:
while flag.green():
logger.info('reading message')
queue_message = queue.read(30)
# TODO : check for more-than-once delivery
# ( Dynamo conditional write? http://tinyurl.com/of3tmop )
if queue_message is None:
logger.info('no messages available')
else:
with newrelic.agent.BackgroundTask(application, name=queue_message.notification_type, group='queue_worker.py'):
logger.info('got message id: %s' % queue_message.id)
if queue_message.notification_type == 'S3Event':
info = S3NotificationInfo.from_S3SQSMessage(queue_message)
logger.info("S3NotificationInfo: %s", info.to_dict())
workflow_name = get_starter_name(rules, info)
if workflow_name is None:
logger.info("Could not handle file %s in bucket %s" % (info.file_name, info.bucket_name))
return False
# build message
message = {
'workflow_name': workflow_name,
'workflow_data': info.to_dict()
}
# send workflow initiation message
out_queue = conn.get_queue(settings.workflow_starter_queue)
m = Message()
m.set_body(json.dumps(message))
out_queue.write(m)
# cancel incoming message
logger.info("cancelling message")
queue.delete_message(queue_message)
logger.info("message cancelled")
else:
# TODO : log
pass
time.sleep(10)
logger.info("graceful shutdown")
else:
logger.error('error obtaining queue')
def work(ENV, flag):
# Specify run environment settings
settings = settings_lib.get_settings(ENV)
# Log
identity = "queue_worker_%s" % os.getpid()
log_file = "queue_worker.log"
# logFile = None
logger = log.logger(log_file, settings.setLevel, identity)
# Simple connect
conn = boto.sqs.connect_to_region(
settings.sqs_region,
aws_access_key_id=settings.aws_access_key_id,
aws_secret_access_key=settings.aws_secret_access_key)
queue = conn.get_queue(settings.S3_monitor_queue)
queue.set_message_class(S3SQSMessage)
rules = load_rules()
application = newrelic.agent.application()
# Poll for an activity task indefinitely
if queue is not None:
while flag.green():
logger.info('reading message')
queue_message = queue.read(30)
# TODO : check for more-than-once delivery
# ( Dynamo conditional write? http://tinyurl.com/of3tmop )
if queue_message is None:
logger.info('no messages available')
else:
with newrelic.agent.BackgroundTask(
application,
name=queue_message.notification_type,
group='queue_worker.py'):
logger.info('got message id: %s' % queue_message.id)
if queue_message.notification_type == 'S3Event':
info = S3NotificationInfo.from_S3SQSMessage(
queue_message)
logger.info("S3NotificationInfo: %s", info.to_dict())
workflow_name = get_starter_name(rules, info)
if workflow_name is None:
logger.info(
"Could not handle file %s in bucket %s" %
(info.file_name, info.bucket_name))
return False
# build message
message = {
'workflow_name': workflow_name,
'workflow_data': info.to_dict()
}
# send workflow initiation message
out_queue = conn.get_queue(
settings.workflow_starter_queue)
m = Message()
m.set_body(json.dumps(message))
out_queue.write(m)
# cancel incoming message
logger.info("cancelling message")
queue.delete_message(queue_message)
logger.info("message cancelled")
else:
# TODO : log
pass
time.sleep(10)
logger.info("graceful shutdown")
else:
logger.error('error obtaining queue')
def do_activity(self, data=None):
"""
Do the work
"""
if self.logger:
self.logger.info('data: %s' % json.dumps(data, sort_keys=True, indent=4))
info = S3NotificationInfo.from_dict(data)
# set up required connections
conn = S3Connection(self.settings.aws_access_key_id, self.settings.aws_secret_access_key)
source_bucket = conn.get_bucket(info.bucket_name)
dest_bucket = conn.get_bucket(self.settings.publishing_buckets_prefix + self.settings.expanded_bucket)
session = Session(self.settings)
article_id_match = re.match(ur'elife-(.*?)-', info.file_name)
article_id = article_id_match.group(1)
session.store_value(self.get_workflowId(), 'article_id', article_id)
if self.logger:
self.logger.info("Expanding file %s" % info.file_name)
# extract any doi, version and updated date information from the filename
version = None
# zip name contains version information for previously archived zip files
m = re.search(ur'-v([0-9]*?)[\.|-]', info.file_name)
if m is not None:
version = m.group(1)
if version is None:
version = self.get_next_version(article_id)
if version == '-1':
return False # version could not be determined, exit workflow. Can't emit event as no version.
sm = re.search(ur'.*?-.*?-(.*?)-', info.file_name)
if sm is not None:
status = sm.group(1)
if status is None:
return False # version could not be determined, exit workflow. Can't emit event as no version.
run = str(uuid.uuid4())
# store version for other activities in this workflow execution
session.store_value(self.get_workflowId(), 'version', version)
# TODO : extract and store updated date if supplied
article_version_id = article_id + '.' + version
session.store_value(self.get_workflowId(), 'article_version_id', article_version_id)
session.store_value(self.get_workflowId(), 'run', run)
session.store_value(self.get_workflowId(), 'status', status)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article", "start",
"Starting expansion of article " + article_id)
self.set_monitor_property(self.settings, article_id, "article_id", article_id, "text")
try:
# download zip to temp folder
tmp = self.get_tmp_dir()
key = Key(source_bucket)
key.key = info.file_name
local_zip_file = self.open_file_from_tmp_dir(info.file_name, mode='wb')
key.get_contents_to_file(local_zip_file)
local_zip_file.close()
bucket_folder_name = article_version_id + '/' + run
folder_name = path.join(article_version_id, run)
# extract zip contents
content_folder = path.join(tmp, folder_name)
makedirs(content_folder)
with ZipFile(path.join(tmp, info.file_name)) as zf:
zf.extractall(content_folder)
# TODO : rename files (versions!)
# TODO : edit xml and rename references
upload_filenames = []
for f in listdir(content_folder):
if isfile(join(content_folder, f)) and f[0] != '.' and not f[0] == '_':
upload_filenames.append(f)
for filename in upload_filenames:
source_path = path.join(content_folder, filename)
dest_path = bucket_folder_name + '/' + filename
k = Key(dest_bucket)
k.key = dest_path
k.set_contents_from_filename(source_path)
session.store_value(self.get_workflowId(), 'expanded_folder', bucket_folder_name)
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article", "end",
"Finished expansion of article " + article_id +
" for version " + version + " run " + str(run) + " into " + bucket_folder_name)
except Exception as e:
self.logger.exception("Exception when expanding article")
self.emit_monitor_event(self.settings, article_id, version, run, "Expand Article", "error",
"Error expanding article " + article_id + " message:" + e.message)
return False
return True
def process_data_ingestarticlezip(workflow_name, workflow_data):
data = {
'info': S3NotificationInfo.from_dict(workflow_data),
'run': str(uuid.uuid4())
}
return data
