class File_Processing:
def __init__(self, events_log, events_elastic, report_elastic,
analysis_elastic, meta_service):
self.meta_service = meta_service
self.events_log = events_log
self.events_elastic = events_elastic
self.storage = Storage()
self.config = Config()
self.status = Status()
self.hash_json = Hash_Json()
self.report_elastic = report_elastic
self.sdk_api_version = "Not available"
self.sdk_engine_version = "Not available"
self.analysis_json = Analysis_Json()
self.analysis_elastic = analysis_elastic
def add_event_log(self, message, event_data={}):
json_data = self.events_log.add_log(message, event_data)
self.events_elastic.add_event_log(json_data)
def base64request(self, endpoint, api_route, base64enc_file):
try:
url = endpoint + "/" + api_route
payload = json.dumps({"Base64": base64enc_file})
headers = {'Content-Type': 'application/json'}
return requests.request("POST",
url,
headers=headers,
data=payload,
timeout=int(self.config.request_timeout))
except Exception as e:
log_error(str(e))
raise ValueError(str(e))
def xmlreport_request(self, endpoint, fileID):
try:
url = endpoint + "/api/Analyse/xmlreport?fileId=" + fileID
payload = ""
headers = {'Content-Type': 'application/octet-stream'}
response = requests.request("GET",
url,
headers=headers,
data=payload,
timeout=int(
self.config.request_timeout))
return response.text
except Exception as e:
raise ValueError(str(e))
def rebuild(self, endpoint, base64enc_file):
return self.base64request(endpoint, "api/rebuild/base64",
base64enc_file)
def get_xmlreport(self, endpoint, fileId, dir):
log_info(message=f"getting XML Report for {fileId} at {endpoint}")
xmlreport = self.xmlreport_request(endpoint, fileId)
if not xmlreport:
raise ValueError('Failed to obtain the XML report')
try:
json_obj = xmltodict.parse(xmlreport)
file_extension = json_obj["gw:GWallInfo"]["gw:DocumentStatistics"][
"gw:DocumentSummary"]["gw:FileType"]
self.meta_service.set_rebuild_file_extension(dir, file_extension)
json_obj['original_hash'] = os.path.basename(dir)
json_save_file_pretty(json_obj, os.path.join(dir, "report.json"))
#self.report_elastic.add_report(json_obj)
analysis_obj = self.analysis_json.get_file_analysis(
os.path.basename(dir), json_obj)
json_save_file_pretty(analysis_obj,
os.path.join(dir, "analysis.json"))
self.analysis_elastic.add_analysis(analysis_obj)
return True
except Exception as error:
log_error(
message=f"Error in parsing xmlreport for {fileId} : {error}")
return False
# Save to HD3
def save_file(self, result, processed_path):
self.add_event_log('Saving to: ' + processed_path)
dirname = ntpath.dirname(processed_path)
basename = ntpath.basename(processed_path)
folder_create(dirname)
decoded = FileService.base64decode(result)
if decoded:
FileService.wrtie_binary_file(dirname, basename, decoded)
self.add_event_log('The decoded file has been saved')
return processed_path
else:
FileService.wrtie_file(
dirname, basename + ".html",
result) # todo: capture better this workflow
self.add_event_log('Decoding FAILED. The HTML file has been saved')
return processed_path + '.html' # todo: refactor this workflow and how this is calculated
@log_duration
def do_rebuild(self, endpoint, hash, source_path, dir):
log_info(
message=f"Starting rebuild for file {hash} on endpoint {endpoint}")
with Duration() as duration:
event_data = {
"endpoint": endpoint,
"hash": hash,
"source_path": source_path,
"dir": dir
} # todo: see if we can use a variable that holds the params data
self.add_event_log('Starting File rebuild', event_data)
self.meta_service.set_rebuild_server(dir, endpoint)
encodedFile = FileService.base64encode(source_path)
if not encodedFile:
message = f"Failed to encode the file: {hash}"
log_error(message=message)
self.add_event_log(message)
self.meta_service.set_error(dir, message)
return False
response = self.rebuild(endpoint, encodedFile)
result = response.text
if not result:
message = f"Failed to rebuild the file : {hash}"
log_error(message=message)
self.add_event_log(message)
self.meta_service.set_error(dir, message)
return False
try:
for path in self.meta_service.get_original_file_paths(dir):
#rebuild_file_path = path
if path.startswith(self.config.hd1_location):
rebuild_file_path = path.replace(
self.config.hd1_location, self.config.hd3_location)
else:
rebuild_file_path = os.path.join(
self.config.hd3_location, path)
folder_create(parent_folder(
rebuild_file_path)) # make sure parent folder exists
final_rebuild_file_path = self.save_file(
result, rebuild_file_path
) # returns actual file saved (which could be .html)
# todo: improve the performance of these update since each will trigger a save
file_size = os.path.getsize(
final_rebuild_file_path) # calculate rebuilt file fize
rebuild_hash = self.meta_service.file_hash(
final_rebuild_file_path
) # calculate hash of final_rebuild_file_path
self.meta_service.set_rebuild_file_size(dir, file_size)
self.meta_service.set_rebuild_file_path(
dir, final_rebuild_file_path
) # capture final_rebuild_file_path
self.meta_service.set_rebuild_hash(
dir, rebuild_hash) # capture it
if not FileService.base64decode(result):
message = f"Engine response could not be decoded"
log_error(message=message, data=f"{result}")
self.meta_service.set_error(dir, message)
return False
except Exception as error:
message = f"Error Saving file for {hash} : {error}"
log_error(message=message)
self.meta_service.set_xml_report_status(dir, "No Report")
self.meta_service.set_error(dir, message)
return False
headers = response.headers
fileIdKey = "X-Adaptation-File-Id"
# get XML report
if fileIdKey in headers:
if self.get_xmlreport(endpoint, headers[fileIdKey], dir):
self.add_event_log('The XML report has been saved')
self.meta_service.set_xml_report_status(dir, "Obtained")
else:
self.meta_service.set_xml_report_status(
dir, "No XML Report")
else:
self.meta_service.set_xml_report_status(
dir, "Failed to obtain")
message = f'No X-Adaptation-File-Id header found in the response for {hash}'
log_error(message)
self.add_event_log(message)
self.meta_service.set_error(dir, message)
return False
#raise ValueError("No X-Adaptation-File-Id header found in the response")
# todo: add when server side supports this
# SDKEngineVersionKey = "X-SDK-Engine-Version"
# SDKAPIVersionKey = "X-SDK-Api-Version"
#
# if SDKEngineVersionKey in headers:
# self.sdk_engine_version = headers[SDKEngineVersionKey]
# if SDKAPIVersionKey in headers:
# self.sdk_api_version = headers[SDKAPIVersionKey]
#
# self.meta_service.set_server_version(dir, "Engine:" + self.sdk_engine_version + " API:" + self.sdk_api_version )
log_info(
message=
f"rebuild ok for file {hash} on endpoint {endpoint} took {duration.seconds()} seconds"
)
return True
@log_duration
def processDirectory(self, endpoint, dir):
self.add_event_log("Processing Directory: " + dir)
hash = ntpath.basename(dir)
if len(hash) != 64:
self.add_event_log("Unexpected hash length")
#raise ValueError("Unexpected hash length")
return False
metadata_file_path = os.path.join(dir,
Metadata_Service.METADATA_FILE_NAME)
if not (FileService.file_exist(metadata_file_path)):
self.add_event_log("The metadate.json file does not exist")
#raise ValueError("The metadate.json file does not exist")
return False
if self.meta_service.is_completed_status(dir):
self.add_event_log("Metadata is in the COMPLETED state")
return False
self.add_event_log("Set metadata status IN_PROGRESS")
self.meta_service.set_status_inprogress(dir)
self.status.add_in_progress()
source_path = os.path.join(dir, "source")
if not (FileService.file_exist(source_path)):
self.add_event_log("File does not exist")
#raise ValueError("File does not exist")
return False
self.add_event_log("Sending to rebuild")
tik = datetime.now()
status = self.do_rebuild(endpoint, hash, source_path, dir)
# if status:
# self.meta_service.set_status(dir, FileStatus.COMPLETED)
# self.meta_service.set_error(dir, "none")
# else:
if not status:
self.meta_service.set_status(dir, FileStatus.FAILED)
self.hash_json.update_status(hash, FileStatus.FAILED)
tok = datetime.now()
delta = tok - tik
self.meta_service.set_rebuild_file_duration(dir, delta.total_seconds())
return status
class Loops(object):
continue_processing = False
processing_started = False
lock = asyncio.Lock()
def __init__(self):
self.use_es = False
self.config = Config()
self.status = Status()
self.storage = Storage()
self.hash_json = Hash_Json()
self.events = Events_Log(self.config.hd2_status_location)
self.events_elastic = Events_Log_Elastic()
self.hash = None
self.report_elastic = Report_Elastic()
self.analysis_elastic = Analysis_Elastic()
self.report_elastic.setup()
self.analysis_elastic.setup()
create_folder(self.storage.hd2_processed())
create_folder(self.storage.hd2_not_processed())
def IsProcessing(self):
return Loops.processing_started
def StopProcessing(self):
Loops.continue_processing = False
def HasBeenStopped(self):
return not Loops.continue_processing
def git_commit(self):
git_commit = 'Not available'
try:
git_commit = subprocess.check_output(['git', 'rev-parse', 'HEAD'
]).decode("utf-8").rstrip()
except Exception as e:
pass
return git_commit
def ProcessDirectoryWithEndpoint(self, itempath, file_hash,
endpoint_index):
if not os.path.isdir(itempath):
return False
log_info(
message=
f"Starting ProcessDirectoryWithEndpoint on endpoint # {endpoint_index} for file {file_hash}"
)
meta_service = Metadata_Service()
original_file_path = meta_service.get_original_file_paths(itempath)
events = Events_Log(itempath)
endpoint = "http://" + self.config.endpoints['Endpoints'][
endpoint_index]['IP'] + ":" + self.config.endpoints['Endpoints'][
endpoint_index]['Port']
events.add_log("Processing with: " + endpoint)
meta_service.set_f2f_plugin_version(itempath, API_VERSION)
meta_service.set_f2f_plugin_git_commit(itempath, self.git_commit())
try:
file_processing = File_Processing(events, self.events_elastic,
self.report_elastic,
self.analysis_elastic,
meta_service)
if not file_processing.processDirectory(endpoint, itempath):
events.add_log("CANNOT be processed")
return False
log_data = {
'file': original_file_path,
'status': FileStatus.COMPLETED,
'error': 'none',
'timestamp': datetime.now(),
}
log_info('ProcessDirectoryWithEndpoint', data=log_data)
meta_service.set_error(itempath, "none")
meta_service.set_status(itempath, FileStatus.COMPLETED)
self.hash_json.update_status(file_hash, FileStatus.COMPLETED)
events.add_log("Has been processed")
return True
except Exception as error:
log_data = {
'file': original_file_path,
'status': FileStatus.FAILED,
'error': str(error),
}
log_error(message='error in ProcessDirectoryWithEndpoint',
data=log_data)
meta_service.set_error(itempath, str(error))
meta_service.set_status(itempath, FileStatus.FAILED)
self.hash_json.update_status(file_hash, FileStatus.FAILED)
events.add_log("ERROR:" + str(error))
return False
def ProcessDirectory(self, thread_data):
(itempath, file_hash, process_index) = thread_data
endpoint_index = process_index % self.config.endpoints_count
if not Loops.continue_processing:
return False
tik = datetime.now()
process_result = self.ProcessDirectoryWithEndpoint(
itempath, file_hash, endpoint_index)
if process_result:
self.status.add_completed()
tok = datetime.now()
delta = tok - tik
meta_service = Metadata_Service()
meta_service.set_hd2_to_hd3_copy_time(itempath,
delta.total_seconds())
else:
self.status.add_failed()
return process_result
# note: removing retries from this method (it should not be handled like this
#for idx in range(self.config.endpoints_count):
# if self.ProcessDirectoryWithEndpoint(itempath, file_hash, endpoint_index):
# return
# # The Endpoint failed to process the file
# # Retry it with the next one
# endpoint_index = (endpoint_index + 1) % self.config.endpoints_count
def updateHashJson(self):
self.hash_json.reset()
meta_service = Metadata_Service()
for hash_folder in os.listdir(self.storage.hd2_data()):
metadata_folder = self.storage.hd2_data(hash_folder)
if not os.path.isdir(metadata_folder):
continue
metadata = meta_service.get_from_file(metadata_folder)
file_name = metadata.get_file_name()
original_hash = metadata.get_original_hash()
status = metadata.get_rebuild_status()
if status != FileStatus.COMPLETED:
self.hash_json.add_file(original_hash, file_name)
self.hash_json.save()
self.status.set_processing_counters(len(self.hash_json.data()))
return self.hash_json.data()
def moveProcessedFiles(self):
json_list = self.hash_json.data()
for key in json_list:
source_path = self.storage.hd2_data(key)
if (FileStatus.COMPLETED == json_list[key]["file_status"]):
destination_path = self.storage.hd2_processed(key)
if folder_exists(destination_path):
folder_delete_all(destination_path)
shutil.move(source_path, destination_path)
if (FileStatus.FAILED == json_list[key]["file_status"]):
meta_service = Metadata_Service()
meta_service.get_from_file(source_path)
metadata = meta_service.metadata
if ("Engine response could not be decoded" == metadata.get_error()) and \
metadata.get_original_file_extension() in ['.xml', '.json']:
destination_path = self.storage.hd2_not_processed(key)
if folder_exists(destination_path):
folder_delete_all(destination_path)
shutil.move(source_path, destination_path)
def LoopHashDirectoriesInternal(self, thread_count, do_single):
if folder_exists(self.storage.hd2_data()) is False:
log_message = "ERROR: rootdir does not exist: " + self.storage.hd2_data(
)
log_error(log_message)
return False
if not isinstance(thread_count, int):
raise TypeError("thread_count must be a integer")
if not isinstance(do_single, bool):
raise TypeError("thread_count must be a integer")
log_message = f"LoopHashDirectoriesInternal started with {thread_count} threads"
self.events.add_log(log_message)
log_info(log_message)
json_list = self.updateHashJson()
log_message = f"LoopHashDirectoriesInternal started with {thread_count} threads"
self.events.add_log(log_message)
log_info(log_message)
threads = list()
process_index = 0
log_info(
message=f'before Mapping thread_data for {len(json_list)} files')
thread_data = []
for key in json_list:
file_hash = key
itempath = self.storage.hd2_data(key)
if (FileStatus.COMPLETED == json_list[key]["file_status"]):
self.events.add_log(
f"The file processing has been already completed")
continue
if not os.path.exists(itempath):
self.events.add_log(
f"ERROR: Path \"{itempath}\" does not exist")
json_list[key]["file_status"] = FileStatus.FAILED
continue
process_index += 1
thread_data.append((
itempath,
file_hash,
process_index,
))
# # limit the number of parallel threads
#
# if process_index % int(thread_count) == 0: # todo: refactor this workflow to use multiprocess and queues
# # Clean up the threads
# for index, thread in enumerate(threads): # todo: since at the moment this will block allocating new threads until
# thread.join() # all have finishing execution
#
# process_index += 1
# log_info(message=f"in LoopHashDirectoriesInternal process_index={process_index} , thread #{process_index % int(thread_count) }")
# x = threading.Thread(target=self.ProcessDirectory, args=(itempath, file_hash, process_index,))
# threads.append(x)
# x.start()
#
# if do_single:
# break
#
# if not Loops.continue_processing:
# break
# for index, thread in enumerate(threads):
# thread.join()
log_info(
message=
f'after mapped thread_data, there are {len(thread_data)} mapped items'
)
#thread_data = thread_data[:500]
#log_info(message=f'to start with only processing {len(thread_data)} thread_data items')
pool = ThreadPool(thread_count)
results = pool.map(self.ProcessDirectory, thread_data)
pool.close()
pool.join()
self.moveProcessedFiles()
self.events.add_log("LoopHashDirectoriesInternal finished")
return True
async def LoopHashDirectoriesAsync(self, thread_count, do_single=False):
await Loops.lock.acquire()
try:
Loops.continue_processing = True
Loops.processing_started = True
self.status.set_started()
self.LoopHashDirectoriesInternal(thread_count, do_single)
finally:
Loops.processing_started = False
Loops.lock.release()
self.status.set_stopped()
self.hash_json.save()
@log_duration
def LoopHashDirectories(self, thread_count=None):
#Allow only a single loop to be run at a time
if self.IsProcessing():
log_error(
message=
"ERROR: Attempt to start processing while processing is in progress"
)
return False
self.status.StartStatusThread()
thread_count = thread_count or self.config.thread_count
log_info(message="in LoopHashDirectories, about to start main loop")
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
loop.run_until_complete(self.LoopHashDirectoriesAsync(thread_count))
log_info(message="in LoopHashDirectories, Loop completed")
self.status.StopStatusThread()
return True
@log_duration
def LoopHashDirectoriesSequential(self):
#Allow only a single loop to be run at a time
if self.IsProcessing():
log_error(
"ERROR: Attempt to start processing while processing is in progress"
)
return False
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
loop.run_until_complete(self.LoopHashDirectoriesAsync(1))
return True
@log_duration
def ProcessSingleFile(self):
if self.IsProcessing():
log_error(
"ERROR: Attempt to start processing while processing is in progress"
)
return False
loop = asyncio.new_event_loop()
asyncio.set_event_loop(loop)
loop.run_until_complete(self.LoopHashDirectoriesAsync(1, True))
return True
class test_Hash_Json(TestCase):
test_file = None
@classmethod
def setUpClass(cls) -> None:
cls.test_file = temp_file(contents='Static text so that we have a static hash')
cls.test_file_name = file_name(cls.test_file)
cls.test_file_hash = '500286533bf75d769e9180a19414d1c3502dd52093e7351a0a9b1385d8f8961c'
@classmethod
def tearDownClass(cls) -> None:
file_delete(cls.test_file)
def setUp(self) -> None:
self.hash_json = Hash_Json()
self.storage = self.hash_json.storage
def test___init__(self):
assert abspath(self.hash_json.folder()) == self.storage.hd2_status()
@patch("multiprocessing.queues.Queue.put_nowait")
def test_add_file(self, patch_log_error):
hash_data = self.hash_json.load()
if hash_data.get('self.test_file_hash'):
del hash_data[self.test_file_hash]
assert self.hash_json.add_file(self.test_file_hash, self.test_file_name) is True
self.hash_json.save()
assert hash_data.get(self.test_file_hash) == {'file_name': self.test_file_name, 'file_status': 'Initial'}
assert self.hash_json.add_file('AAAA' , self.test_file_name) is False
assert self.hash_json.add_file(self.test_file_hash , None ) is False
assert self.hash_json.add_file(None , None ) is False
assert patch_log_error.mock_calls == [call({'level': 'ERROR', 'message': 'in Hash_Json.add_file bad data provided', 'data': {'file_hash': 'AAAA' , 'file_name': self.test_file_name}, 'duration': 0, 'from_method': 'add_file', 'from_class': 'Hash_Json'}),
call({'level': 'ERROR', 'message': 'in Hash_Json.add_file bad data provided', 'data': {'file_hash': self.test_file_hash, 'file_name': None }, 'duration': 0, 'from_method': 'add_file', 'from_class': 'Hash_Json'}),
call({'level': 'ERROR', 'message': 'in Hash_Json.add_file bad data provided', 'data': {'file_hash': None , 'file_name': None }, 'duration': 0, 'from_method': 'add_file', 'from_class': 'Hash_Json'})]
def test_get_file_path(self):
file_path = abspath(self.hash_json.get_file_path())
assert file_exists(file_path)
assert file_path == path_combine(self.storage.hd2_status(), Hash_Json.HASH_FILE_NAME)
def test_load(self):
data = self.hash_json.load()
assert type(data) is dict
assert self.hash_json.data() == data
def test_data(self):
assert self.hash_json.data() == self.hash_json._hash_json_data
def test_is_hash(self):
test_file = temp_file(contents='aaaa')
file_hash = Metadata_Utils().file_hash(test_file) # create hash from file
text_hash = str_sha256('asd') # create hash from string
assert self.hash_json.is_hash(file_hash ) is True # confirm both are valid hashes
assert self.hash_json.is_hash(text_hash ) is True
assert self.hash_json.is_hash(None ) is False # testing all sorts of conner cases
assert self.hash_json.is_hash('' ) is False # empty strings
assert self.hash_json.is_hash('aaaa' ) is False # non hash string
assert self.hash_json.is_hash(file_hash + 'aaaa') is False # confirm only exact matches work
assert self.hash_json.is_hash(text_hash + 'aaaa') is False
assert self.hash_json.is_hash('aaa' + file_hash ) is False
assert self.hash_json.is_hash(text_hash + '\nb`') is False # confirm content in new lines is also not a match
assert self.hash_json.is_hash('a\n' + file_hash ) is False
file_delete(test_file)
def test_save(self):
target_file = temp_file() # temp file to save data
assert file_not_exists(target_file) # confirm it doesn't exist
with patch.object(Hash_Json, 'get_file_path', return_value=target_file): # patch get_file_path to return temp file path
assert self.hash_json.get_file_path() == target_file # confirm patch is in place
self.hash_json.save() # call write_to_file
assert file_exists(target_file) # confirm temp file now exists
assert self.hash_json.load() == self.hash_json.data() # confirm reloaded data is correct
assert json_load_file(target_file) == self.hash_json.data() # also confirm using direct json load of temp file
assert self.hash_json.get_file_path() != target_file # confirm pathc is not there (after 'with' ends)
file_delete(target_file) # delete temp file
def test_update_status(self):
temp_data_file = temp_file()
with patch.object(Hash_Json, 'get_file_path', return_value=temp_data_file):
self.hash_json.add_file(self.test_file_hash, self.test_file_name)
assert self.hash_json.data()[self.test_file_hash]['file_status'] == 'Initial'
self.hash_json.update_status(self.test_file_hash, 'BBBB')
self.hash_json.save()
assert self.hash_json.data()[self.test_file_hash]['file_status'] == 'BBBB'
assert json_load_file(temp_data_file)[self.test_file_hash]['file_status'] == 'BBBB'
pprint(self.hash_json.load())
def test_data_bug(self): # this test confirms the bug
hashes = self.hash_json.data()
for hash in self.hash_json.data():
if len(hash) == 64: # all keys in this object should be a hash
assert len(hash) == 64
assert type(hashes[hash]) == dict # with all items being a dictionary
assert list_set(hashes[hash]) == ['file_name', 'file_status']
else:
assert hash == "file_list" # but the old schema is still present
assert type(hashes[hash]) == list # with the data being a list
assert list_set(hashes[hash][0]) == ['file_name', 'file_status', 'hash', 'id']