def test_ratom_report_enron_027(isolated_cli_runner, enron_dataset_part027,
params, expected):
msg_id = 2390436
result = generate_report(params, enron_dataset_part027,
isolated_cli_runner, expected)
with db_session_from_cmd_out(result) as session:
# Verify total message count
assert session.query(Message).count() == 9297
# Get message contents from DB
msg = session.query(Message).filter_by(pff_identifier=msg_id).one()
headers, body = msg.headers, msg.body
if expected.with_messages:
# Access message directly and compare
archive_file = list(enron_dataset_part027.glob("*.pst"))[0]
with open_mail_archive(archive_file) as archive:
message = archive.get_message_by_id(msg_id)
assert cleanup_message_body(
*archive.get_message_body(message)) == body
assert archive.get_message_headers(message) == headers
else:
assert headers is None
assert body is None
def test_get_mbox_message_by_id(sample_mbox_file):
with open_mail_archive(sample_mbox_file) as archive:
assert archive.message_count == 113
for index, message in enumerate(archive.messages(), start=1):
msg = archive.get_message_by_id(index)
assert extract_message_from_archive(archive, index)
assert archive.format_message(msg) == archive.format_message(
message)
assert archive.get_message_headers(message)
def test_ratom_entities_enron_001(isolated_cli_runner, enron_dataset_part001,
params, expected):
msg_id = 2097572
# Run entity extraction job with message content flag on
result = extract_entities(params, enron_dataset_part001,
isolated_cli_runner, expected)
# Get message contents from DB
with db_session_from_cmd_out(result) as session:
msg = session.query(Message).filter_by(pff_identifier=msg_id).one()
headers, body = msg.headers, msg.body
# Access message directly and compare
archive_file = list(enron_dataset_part001.glob("*.pst"))[0]
with open_mail_archive(archive_file) as archive:
message = archive.get_message_by_id(msg_id)
assert cleanup_message_body(*archive.get_message_body(message)) == body
assert archive.get_message_headers(message) == headers
def test_apply_spacy_model(sample_pst_file, model_name, expected_entity_types):
# Extract a known (short) message
msg_id = 2112164
with open_mail_archive(sample_pst_file) as archive:
msg_body = archive.get_message_body(
archive.get_message_by_id(msg_id))[0]
# Sanity check
assert len(msg_body) == 564
# Pre-load our model to install any missing dependencies
assert get_cached_spacy_model(model_name)
# Apply our model pretending to be in a forked process
with patch(
"libratom.lib.entities.current_process") as mock_current_process:
mock_current_process.return_value.name = "NotMainProcess"
# pylint:disable=no-value-for-parameter
res, error = process_message(
# Must use dictionary form if function is called explicitly
{
"filepath": sample_pst_file,
"message_id": msg_id,
"date": datetime.datetime.utcnow(),
"body": msg_body,
"body_type": BodyType.PLAIN,
"spacy_model_name": model_name,
"attachments": None,
})
assert res and not error
# Check that the expected entity types were found
assert expected_entity_types.issubset(
set(entity[1] for entity in res["entities"]))
def get_file_info(path: Path) -> Tuple[Dict, Optional[str]]:
"""
For a given file path, returns the size, md5 and sha256 checksums
"""
path_str, name = str(path), path.name
res = {"path": path_str, "name": name}
try:
size = os.stat(path_str).st_size
md5 = hashlib.md5()
sha256 = hashlib.sha256()
# First we read the file one block at a time and update digests
with open(path_str, "rb") as f:
for block in iter(partial(f.read, 128), b""):
md5.update(block)
sha256.update(block)
md5, sha256 = md5.hexdigest(), sha256.hexdigest()
res.update({"size": size, "md5": md5, "sha256": sha256})
# Then we try to get a message count
try:
with open_mail_archive(path) as archive:
res["msg_count"] = archive.message_count
except Exception as exc:
res["error"] = str(exc)
except Exception as exc:
return res, str(exc)
return res, None
def get_messages(
files: Iterable[Path],
progress_callback: Callable,
with_content=True,
with_headers=False,
**kwargs,
) -> Generator[Dict, None, None]:
"""
Message generator to feed a pool of processes from a directory of PST files
"""
msg_count = 0
# Iterate over files
for file in files:
try:
with open_mail_archive(file) as archive:
# Iterate over messages
for message in archive.messages():
try:
# Keyword arguments for process_message()
res = {
"filepath":
archive.filepath,
"message_id":
getattr(message, "identifier", None),
"attachments":
archive.get_attachment_metadata(message),
}
try:
res["date"] = archive.get_message_date(message)
except Exception as exc:
res["date"] = None
logger.debug(
"Unable to extract date from message: {message_id} in file: {filepath}"
.format(**res))
logger.debug(exc, exc_info=True)
if with_content:
body, body_type = archive.get_message_body(message)
res["body"] = body
res["body_type"] = body_type
if with_headers:
res["headers"] = archive.get_message_headers(
message)
# Add any optional arguments
res.update(kwargs)
yield res
except Exception as exc:
# Log and move on to the next message
message_id = getattr(message, "identifier", None)
message_str = (f"message {message_id}"
if message_id else "a message")
logger.info(f"Skipping {message_str} from {file}")
logger.debug(exc, exc_info=True)
finally:
msg_count += 1
# Update progress every N messages
if not msg_count % RATOM_MSG_PROGRESS_STEP:
progress_callback(RATOM_MSG_PROGRESS_STEP)
except Exception as exc:
# Log and move on to the next file
logger.info(f"Skipping file {file}")
logger.debug(exc, exc_info=True)
# Update progress with remaining message count
progress_callback(msg_count % RATOM_MSG_PROGRESS_STEP)
