def to_python(cls, data):
name = data.get('name')
if not name:
raise InterfaceValidationError("No 'name' value")
version = data.get('version')
if not version:
raise InterfaceValidationError("No 'version' value")
integrations = data.get('integrations')
if integrations and not isinstance(integrations, list):
raise InterfaceValidationError("'integrations' must be a list")
packages = data.get('packages')
if packages and not isinstance(packages, list):
raise InterfaceValidationError("'packages' must be a list")
kwargs = {
'name': trim(name, 128),
'version': trim(version, 128),
'integrations': integrations,
'packages': packages,
}
return cls(**kwargs)
def merge_symbolicated_frame(new_frame, sfrm, platform=None):
if sfrm.get('function'):
raw_func = trim(sfrm['function'], 256)
func = trim(trim_function_name(sfrm['function'], platform), 256)
# if function and raw function match, we can get away without
# storing a raw function
if func == raw_func:
new_frame['function'] = raw_func
# otherwise we store both
else:
new_frame['raw_function'] = raw_func
new_frame['function'] = func
if sfrm.get('instruction_addr'):
new_frame['instruction_addr'] = sfrm['instruction_addr']
if sfrm.get('symbol'):
new_frame['symbol'] = sfrm['symbol']
if sfrm.get('abs_path'):
new_frame['abs_path'] = sfrm['abs_path']
new_frame['filename'] = posixpath.basename(sfrm['abs_path'])
if sfrm.get('filename'):
new_frame['filename'] = sfrm['filename']
if sfrm.get('lineno'):
new_frame['lineno'] = sfrm['lineno']
if sfrm.get('colno'):
new_frame['colno'] = sfrm['colno']
if sfrm.get('package'):
new_frame['package'] = sfrm['package']
if sfrm.get('trust'):
new_frame['trust'] = sfrm['trust']
if sfrm.get('status'):
frame_meta = new_frame.setdefault('data', {})
frame_meta['symbolicator_status'] = sfrm['status']
def to_python(cls, data):
data = data.copy()
extra_data = data.pop('data', data)
if not isinstance(extra_data, dict):
extra_data = {}
try:
name = trim(data.pop('name'), 64)
except KeyError:
raise InterfaceValidationError("Missing or invalid value for 'name'")
try:
version = trim(data.pop('version'), 64)
except KeyError:
raise InterfaceValidationError("Missing or invalid value for 'version'")
build = trim(data.pop('build', None), 64)
kwargs = {
'name': name,
'version': version,
'build': build,
'data': trim_dict(data),
}
kwargs['data'] = trim_dict(data)
return cls(**kwargs)
def to_python(cls, data):
formatted = stringify(data.get('formatted'))
message = stringify(data.get('message'))
if formatted is None and message is None:
raise InterfaceValidationError("No message present")
params = data.get('params')
if isinstance(params, (list, tuple)):
params = tuple(p for p in params)
elif isinstance(params, dict):
params = {k: v for k, v in six.iteritems(params)}
else:
params = ()
if formatted is None and params:
try:
if '%' in message:
formatted = message % params
elif '{}' in message and isinstance(params, tuple):
formatted = message.format(*params)
# NB: Named newstyle arguments were never supported
except Exception:
pass
if formatted is None or message == formatted:
formatted = message
message = None
return cls(
formatted=trim(formatted, settings.SENTRY_MAX_MESSAGE_LENGTH),
message=trim(message, settings.SENTRY_MAX_MESSAGE_LENGTH),
params=trim(params, 1024),
)
def normalize_crumb(cls, crumb):
ty = crumb.get('type') or 'default'
ts = parse_new_timestamp(crumb.get('timestamp'))
if ts is None:
raise InterfaceValidationError('Unable to determine timestamp '
'for crumb')
rv = {
'type': ty,
'timestamp': to_timestamp(ts),
}
level = crumb.get('level')
if level not in (None, 'info'):
rv['level'] = level
msg = crumb.get('message')
if msg is not None:
rv['message'] = trim(unicode(msg), 4096)
category = crumb.get('category')
if category is not None:
rv['category'] = trim(unicode(category), 256)
event_id = crumb.get('event_id')
if event_id is not None:
rv['event_id'] = event_id
if 'data' in crumb:
rv['data'] = trim(crumb['data'], 4096)
return rv
def normalize_crumb(cls, crumb):
ty = crumb.get('type') or 'default'
ts = parse_timestamp(crumb.get('timestamp'))
if ts is None:
raise InterfaceValidationError('Unable to determine timestamp '
'for crumb')
rv = {
'type': ty,
'timestamp': to_timestamp(ts),
}
level = crumb.get('level')
if level not in (None, 'info'):
rv['level'] = level
msg = crumb.get('message')
if msg is not None:
rv['message'] = trim(six.text_type(msg), 4096)
category = crumb.get('category')
if category is not None:
rv['category'] = trim(six.text_type(category), 256)
event_id = crumb.get('event_id')
if event_id is not None:
rv['event_id'] = event_id
if crumb.get('data'):
for key, value in six.iteritems(crumb['data']):
if not isinstance(value, six.string_types):
crumb['data'][key] = json.dumps(value)
rv['data'] = trim(crumb['data'], 4096)
return rv
def to_python(cls, data, rust_renormalized=RUST_RENORMALIZED_DEFAULT):
if rust_renormalized:
for key in (
'abs_path',
'filename',
'context_line',
'lineno',
'pre_context',
'post_context',
):
data.setdefault(key, None)
return cls(**data)
is_valid, errors = validate_and_default_interface(data, cls.path)
if not is_valid:
raise InterfaceValidationError("Invalid template")
kwargs = {
'abs_path': trim(data.get('abs_path', None), 256),
'filename': trim(data.get('filename', None), 256),
'context_line': trim(data.get('context_line', None), 256),
'lineno': int(data['lineno']) if data.get('lineno', None) is not None else None,
# TODO(dcramer): trim pre/post_context
'pre_context': data.get('pre_context'),
'post_context': data.get('post_context'),
}
return cls(**kwargs)
def to_python(cls, data, has_system_frames=None):
if not (data.get('type') or data.get('value')):
raise InterfaceValidationError("No 'type' or 'value' present")
if data.get('stacktrace') and data['stacktrace'].get('frames'):
stacktrace = Stacktrace.to_python(
data['stacktrace'],
has_system_frames=has_system_frames,
)
else:
stacktrace = None
type = data.get('type')
value = data.get('value')
if not type and ':' in value.split(' ', 1)[0]:
type, value = value.split(':', 1)
# in case of TypeError: foo (no space)
value = value.strip()
if value is not None and not isinstance(value, basestring):
value = json.dumps(value)
value = trim(value, 4096)
kwargs = {
'type': trim(type, 128),
'value': value,
'module': trim(data.get('module'), 128),
'stacktrace': stacktrace,
}
return cls(**kwargs)
def to_python(cls, data, rust_renormalized=RUST_RENORMALIZED_DEFAULT):
if rust_renormalized:
for key in (
'name',
'version',
'integrations',
'packages',
):
data.setdefault(key, None)
return cls(**data)
name = data.get('name')
version = data.get('version')
integrations = data.get('integrations')
if integrations and not isinstance(integrations, list):
raise InterfaceValidationError("'integrations' must be a list")
packages = data.get('packages')
if packages and not isinstance(packages, list):
raise InterfaceValidationError("'packages' must be a list")
kwargs = {
'name': trim(name, 128),
'version': trim(version, 128),
'integrations': integrations,
'packages': packages,
}
return cls(**kwargs)
def to_python(cls, data):
data = data.copy()
extra_data = data.pop("data", data)
if not isinstance(extra_data, dict):
extra_data = {}
ident = trim(data.pop("id", None), 128)
if ident:
ident = unicode(ident)
try:
email = trim(validate_email(data.pop("email", None), False), 128)
except ValueError:
raise InterfaceValidationError("Invalid value for 'email'")
username = trim(data.pop("username", None), 128)
if username:
username = unicode(username)
try:
ip_address = validate_ip(data.pop("ip_address", None), False)
except ValueError:
raise InterfaceValidationError("Invalid value for 'ip_address'")
# TODO(dcramer): patch in fix to deal w/ old data but not allow new
# if not (ident or email or username or ip_address):
# raise ValueError('No identifying value')
kwargs = {"id": ident, "email": email, "username": username, "ip_address": ip_address}
kwargs["data"] = trim_dict(extra_data)
return cls(**kwargs)
def to_python(cls, data):
abs_path = data.get('abs_path')
filename = data.get('filename')
if not abs_path:
abs_path = filename
if not filename:
filename = abs_path
if abs_path and is_url(abs_path):
urlparts = urlparse(abs_path)
if urlparts.path:
filename = urlparts.path
assert filename or data.get('function') or data.get('module')
context_locals = data.get('vars') or {}
if isinstance(context_locals, (list, tuple)):
context_locals = dict(enumerate(context_locals))
elif not isinstance(context_locals, dict):
context_locals = {}
context_locals = trim_dict(context_locals)
# extra data is used purely by internal systems,
# so we dont trim it
extra_data = data.get('data') or {}
if isinstance(extra_data, (list, tuple)):
extra_data = dict(enumerate(extra_data))
kwargs = {
'abs_path': trim(abs_path, 256),
'filename': trim(filename, 256),
'module': trim(data.get('module'), 256),
'function': trim(data.get('function'), 256),
'in_app': validate_bool(data.get('in_app'), False),
'context_line': trim(data.get('context_line'), 256),
# TODO(dcramer): trim pre/post_context
'pre_context': data.get('pre_context'),
'post_context': data.get('post_context'),
'vars': context_locals,
'data': extra_data,
'errors': data.get('errors'),
}
if data.get('lineno') is not None:
lineno = int(data['lineno'])
if lineno < 0:
lineno = None
kwargs['lineno'] = lineno
else:
kwargs['lineno'] = None
if data.get('colno') is not None:
kwargs['colno'] = int(data['colno'])
else:
kwargs['colno'] = None
return cls(**kwargs)
def to_python(cls, data):
if not data.get('url'):
raise InterfaceValidationError("No value for 'url'")
kwargs = {}
if data.get('method'):
method = data['method'].upper()
if method not in HTTP_METHODS:
raise InterfaceValidationError("Invalid value for 'method'")
kwargs['method'] = method
else:
kwargs['method'] = None
scheme, netloc, path, query_bit, fragment_bit = urlsplit(data['url'])
query_string = data.get('query_string') or query_bit
if query_string:
# if querystring was a dict, convert it to a string
if isinstance(query_string, dict):
query_string = urlencode([(to_bytes(k), to_bytes(v))
for k, v in query_string.items()])
else:
query_string = query_string
if query_string[0] == '?':
# remove '?' prefix
query_string = query_string[1:]
kwargs['query_string'] = trim(query_string, 4096)
else:
kwargs['query_string'] = ''
fragment = data.get('fragment') or fragment_bit
cookies = data.get('cookies')
# if cookies were [also] included in headers we
# strip them out
headers = data.get('headers')
if headers:
headers, cookie_header = format_headers(headers)
if not cookies and cookie_header:
cookies = cookie_header
else:
headers = ()
body = data.get('data')
if isinstance(body, dict):
body = json.dumps(body)
if body:
body = trim(body, settings.SENTRY_MAX_HTTP_BODY_SIZE)
kwargs['cookies'] = trim_pairs(format_cookies(cookies))
kwargs['env'] = trim_dict(data.get('env') or {})
kwargs['headers'] = trim_pairs(headers)
kwargs['data'] = body
kwargs['url'] = urlunsplit((scheme, netloc, path, '', ''))
kwargs['fragment'] = trim(fragment, 1024)
return cls(**kwargs)
def to_python(cls, data):
assert data.get('query')
kwargs = {
'query': trim(data['query'], 1024),
'engine': trim(data.get('engine'), 128),
}
return cls(**kwargs)
def get_metadata(self):
exception = self.data['sentry.interfaces.Exception']['values'][0]
# in some situations clients are submitting non-string data for these
return {
'type': trim(exception.get('type', 'Error'), 128),
'value': trim(exception.get('value', ''), 1024),
}
def to_python(cls, data, slim_frames=True, rust_renormalized=RUST_RENORMALIZED_DEFAULT):
if not rust_renormalized:
is_valid, errors = validate_and_default_interface(data, cls.path)
if not is_valid:
raise InterfaceValidationError("Invalid exception")
if not (data.get('type') or data.get('value')):
raise InterfaceValidationError("No 'type' or 'value' present")
if get_path(data, 'stacktrace', 'frames', filter=True):
stacktrace = Stacktrace.to_python(
data['stacktrace'],
slim_frames=slim_frames,
rust_renormalized=rust_renormalized
)
else:
stacktrace = None
if get_path(data, 'raw_stacktrace', 'frames', filter=True):
raw_stacktrace = Stacktrace.to_python(
data['raw_stacktrace'], slim_frames=slim_frames, raw=True,
rust_renormalized=rust_renormalized
)
else:
raw_stacktrace = None
type = data.get('type')
value = data.get('value')
if not rust_renormalized:
if isinstance(value, six.string_types):
if type is None:
m = _type_value_re.match(value)
if m:
type = m.group(1)
value = m.group(2).strip()
elif value is not None:
value = json.dumps(value)
value = trim(value, 4096)
if data.get('mechanism'):
mechanism = Mechanism.to_python(data['mechanism'],
rust_renormalized=rust_renormalized)
else:
mechanism = None
kwargs = {
'type': trim(type, 128),
'value': value,
'module': trim(data.get('module'), 128),
'mechanism': mechanism,
'stacktrace': stacktrace,
'thread_id': trim(data.get('thread_id'), 40),
'raw_stacktrace': raw_stacktrace,
}
return cls(**kwargs)
def to_python(cls, data):
if not data.get('query'):
raise InterfaceValidationError("No 'query' value")
kwargs = {
'query': trim(data['query'], 1024),
'engine': trim(data.get('engine'), 128),
}
return cls(**kwargs)
def to_python(cls, data):
abs_path = data.get("abs_path")
filename = data.get("filename")
if not abs_path:
abs_path = filename
if not filename:
filename = abs_path
if abs_path and is_url(abs_path):
urlparts = urlparse(abs_path)
if urlparts.path:
filename = urlparts.path
assert filename or data.get("function") or data.get("module")
context_locals = data.get("vars") or {}
if isinstance(context_locals, (list, tuple)):
context_locals = dict(enumerate(context_locals))
elif not isinstance(context_locals, dict):
context_locals = {}
context_locals = trim_dict(context_locals)
# extra data is used purely by internal systems,
# so we dont trim it
extra_data = data.get("data") or {}
if isinstance(extra_data, (list, tuple)):
extra_data = dict(enumerate(extra_data))
kwargs = {
"abs_path": trim(abs_path, 256),
"filename": trim(filename, 256),
"module": trim(data.get("module"), 256),
"function": trim(data.get("function"), 256),
"in_app": validate_bool(data.get("in_app"), False),
"context_line": trim(data.get("context_line"), 256),
# TODO(dcramer): trim pre/post_context
"pre_context": data.get("pre_context"),
"post_context": data.get("post_context"),
"vars": context_locals,
"data": extra_data,
}
if data.get("lineno") is not None:
kwargs["lineno"] = int(data["lineno"])
else:
kwargs["lineno"] = None
if data.get("colno") is not None:
kwargs["colno"] = int(data["colno"])
else:
kwargs["colno"] = None
return cls(**kwargs)
def to_python(cls, data):
name = data.get("name")
if not name:
raise InterfaceValidationError("No 'name' value")
version = data.get("version")
if not version:
raise InterfaceValidationError("No 'version' value")
kwargs = {"name": trim(name, 128), "version": trim(version, 128)}
return cls(**kwargs)
def validate_query(payload):
rv = {}
for key in 'params', 'duration', 'classifier':
value = payload.get(key)
if value is not None:
rv[key] = trim(value, 1024)
if 'query' not in payload:
raise InterfaceValidationError("Query not provided for 'query' "
"breadcrumb.")
rv['query'] = trim(payload['query'], 4096)
return rv
def to_python(cls, data):
data = data.copy()
kwargs = {
'id': trim(data.pop('id', None), 128),
'email': trim(data.pop('email', None), 128),
'username': trim(data.pop('username', None), 128),
'ip_address': validate_ip(data.pop('ip_address', None), False),
}
kwargs['data'] = trim_dict(data.pop('data', data))
return cls(**kwargs)
def to_python(cls, data, has_system_frames=None, slim_frames=True):
if not (data.get('type') or data.get('value')):
raise InterfaceValidationError("No 'type' or 'value' present")
if data.get('stacktrace') and data['stacktrace'].get('frames'):
stacktrace = Stacktrace.to_python(
data['stacktrace'],
has_system_frames=has_system_frames,
slim_frames=slim_frames,
)
else:
stacktrace = None
if data.get('raw_stacktrace') and data['raw_stacktrace'].get('frames'):
raw_stacktrace = Stacktrace.to_python(
data['raw_stacktrace'],
has_system_frames=has_system_frames,
slim_frames=slim_frames,
raw=True
)
else:
raw_stacktrace = None
type = data.get('type')
value = data.get('value')
if not type and ':' in value.split(' ', 1)[0]:
type, value = value.split(':', 1)
# in case of TypeError: foo (no space)
value = value.strip()
if value is not None and not isinstance(value, six.string_types):
value = json.dumps(value)
value = trim(value, 4096)
mechanism = data.get('mechanism')
if mechanism is not None:
if not isinstance(mechanism, dict):
raise InterfaceValidationError('Bad value for mechanism')
mechanism = trim(data.get('mechanism'), 4096)
mechanism.setdefault('type', 'generic')
kwargs = {
'type': trim(type, 128),
'value': value,
'module': trim(data.get('module'), 128),
'mechanism': mechanism,
'stacktrace': stacktrace,
'thread_id': trim(data.get('thread_id'), 40),
'raw_stacktrace': raw_stacktrace,
}
return cls(**kwargs)
def to_python(cls, data):
assert data.get('message')
kwargs = {
'message': trim(data['message'], 2048)
}
if data.get('params'):
kwargs['params'] = trim(data['params'], 1024)
else:
kwargs['params'] = ()
return cls(**kwargs)
def to_python(cls, data):
data = data.copy()
ident = data.pop('id', None)
if ident is not None:
ident = trim(six.text_type(ident), 128)
try:
email = trim(validate_email(data.pop('email', None), False), MAX_EMAIL_FIELD_LENGTH)
except ValueError:
raise InterfaceValidationError("Invalid value for 'email'")
username = data.pop('username', None)
if username is not None:
username = trim(six.text_type(username), 128)
name = data.pop('name', None)
if name is not None:
name = trim(six.text_type(name), 128)
try:
ip_address = validate_ip(data.pop('ip_address', None), False)
except ValueError:
raise InterfaceValidationError("Invalid value for 'ip_address'")
geo = data.pop('geo', None)
if not geo and ip_address:
geo = Geo.from_ip_address(ip_address)
elif geo:
geo = Geo.to_python(geo)
extra_data = data.pop('data', None)
if not isinstance(extra_data, dict):
extra_data = {}
extra_data.update(data)
# TODO(dcramer): patch in fix to deal w/ old data but not allow new
# if not (ident or email or username or ip_address):
# raise ValueError('No identifying value')
kwargs = {
'id': ident,
'email': email,
'username': username,
'ip_address': ip_address,
'name': name,
'geo': geo,
'data': trim_dict(extra_data)
}
return cls(**kwargs)
def to_python(cls, data):
if not data.get('message'):
raise InterfaceValidationError("No 'message' present")
kwargs = {
'message': trim(data['message'], 2048)
}
if data.get('params'):
kwargs['params'] = trim(data['params'], 1024)
else:
kwargs['params'] = ()
return cls(**kwargs)
def to_python(cls, data):
if not data.get('message'):
raise InterfaceValidationError("No 'message' present")
# TODO(dcramer): some day we should stop people from sending arbitrary
# crap to the server
if not isinstance(data['message'], six.string_types):
data['message'] = json.dumps(data['message'])
kwargs = {
'message': trim(data['message'], settings.SENTRY_MAX_MESSAGE_LENGTH),
'formatted': data.get('formatted'),
}
if data.get('params'):
kwargs['params'] = trim(data['params'], 1024)
else:
kwargs['params'] = ()
if kwargs['formatted']:
if not isinstance(kwargs['formatted'], six.string_types):
data['formatted'] = json.dumps(data['formatted'])
# support python-esque formatting (e.g. %s)
elif '%' in kwargs['message'] and kwargs['params']:
if isinstance(kwargs['params'], list):
kwargs['params'] = tuple(kwargs['params'])
try:
kwargs['formatted'] = trim(
kwargs['message'] % kwargs['params'],
settings.SENTRY_MAX_MESSAGE_LENGTH,
)
except Exception:
pass
# support very basic placeholder formatters (non-typed)
elif '{}' in kwargs['message'] and kwargs['params']:
try:
kwargs['formatted'] = trim(
kwargs['message'].format(kwargs['params']),
settings.SENTRY_MAX_MESSAGE_LENGTH,
)
except Exception:
pass
# don't wastefully store formatted message twice
if kwargs['formatted'] == kwargs['message']:
kwargs['formatted'] = None
return cls(**kwargs)
def to_python(cls, data):
data = upgrade_legacy_mechanism(data)
is_valid, errors = validate_and_default_interface(data, cls.path)
if not is_valid:
raise InterfaceValidationError("Invalid mechanism")
if not data.get('type'):
raise InterfaceValidationError("No 'type' present")
meta = data.get('meta', {})
mach_exception = meta.get('mach_exception')
if mach_exception is not None:
mach_exception = prune_empty_keys({
'exception': mach_exception['exception'],
'code': mach_exception['code'],
'subcode': mach_exception['subcode'],
'name': mach_exception.get('name'),
})
signal = meta.get('signal')
if signal is not None:
signal = prune_empty_keys({
'number': signal['number'],
'code': signal.get('code'),
'name': signal.get('name'),
'code_name': signal.get('code_name'),
})
errno = meta.get('errno')
if errno is not None:
errno = prune_empty_keys({
'number': errno['number'],
'name': errno.get('name'),
})
kwargs = {
'type': trim(data['type'], 128),
'description': trim(data.get('description'), 1024),
'help_link': trim(data.get('help_link'), 1024),
'handled': data.get('handled'),
'data': trim(data.get('data'), 4096),
'meta': {
'errno': errno,
'mach_exception': mach_exception,
'signal': signal,
},
}
return cls(**kwargs)
def to_python(cls, data):
assert data.get("filename")
assert data.get("context_line")
assert data.get("lineno")
kwargs = {
"abs_path": trim(data.get("abs_path", None), 256),
"filename": trim(data["filename"], 256),
"context_line": trim(data.get("context_line", None), 256),
"lineno": int(data["lineno"]),
# TODO(dcramer): trim pre/post_context
"pre_context": data.get("pre_context"),
"post_context": data.get("post_context"),
}
return cls(**kwargs)
def to_python(cls, data):
is_valid, errors = validate_and_default_interface(data, cls.path)
if not is_valid:
raise InterfaceValidationError("Invalid template")
kwargs = {
'abs_path': trim(data.get('abs_path', None), 256),
'filename': trim(data.get('filename', None), 256),
'context_line': trim(data.get('context_line', None), 256),
'lineno': int(data['lineno']) if data.get('lineno', None) is not None else None,
# TODO(dcramer): trim pre/post_context
'pre_context': data.get('pre_context'),
'post_context': data.get('post_context'),
}
return cls(**kwargs)
def to_python(cls, data):
threads = []
for thread in data.get('values') or ():
threads.append({
'stacktrace': get_stacktrace(thread.get('stacktrace')),
'raw_stacktrace': get_stacktrace(thread.get('raw_stacktrace'),
raw=True),
'id': trim(thread.get('id'), 40),
'crashed': bool(thread.get('crashed')),
'current': bool(thread.get('current')),
'name': trim(thread.get('name'), 200),
})
return cls(values=threads)
def to_python(cls, data):
assert data.get('filename')
assert data.get('context_line')
assert data.get('lineno')
kwargs = {
'abs_path': trim(data.get('abs_path', None), 256),
'filename': trim(data['filename'], 256),
'context_line': trim(data.get('context_line', None), 256),
'lineno': int(data['lineno']),
# TODO(dcramer): trim pre/post_context
'pre_context': data.get('pre_context'),
'post_context': data.get('post_context'),
}
return cls(**kwargs)
def to_python(cls, data):
assert data.get('type') or data.get('value')
if data.get('stacktrace'):
stacktrace = Stacktrace.to_python(data['stacktrace'])
else:
stacktrace = None
kwargs = {
'type': trim(data.get('type'), 128),
'value': trim(data.get('value'), 256),
'module': trim(data.get('module'), 128),
'stacktrace': stacktrace,
}
return cls(**kwargs)
def __init__(self, alias, data):
self.alias = alias
ctx_data = {}
for key, value in trim(data).iteritems():
if value not in EMPTY_VALUES:
ctx_data[force_text(key)] = force_text(value)
self.data = ctx_data
def to_python(cls, data):
kwargs = {k: trim(data.get(k, None), 1024) for k in REPORT_KEYS}
if kwargs['effective_directive'] not in DIRECTIVES:
raise InterfaceValidationError(
"Invalid value for 'effective-directive'")
# Some reports from Chrome report blocked-uri as just 'about'.
# In this case, this is not actionable and is just noisy.
# Observed in Chrome 45 and 46.
if kwargs['blocked_uri'] == 'about':
raise InterfaceValidationError("Invalid value for 'blocked-uri'")
# Here, we want to block reports that are coming from browser extensions
# and other sources that are meaningless
if kwargs['source_file'] is not None:
if kwargs['source_file'].startswith(DISALLOWED_SOURCES):
raise InterfaceValidationError(
"Invalid value for 'source-file'")
# Anything resulting from an "inline" whatever violation is either sent
# as 'self', or left off. In the case if it missing, we want to noramalize.
if not kwargs['blocked_uri']:
kwargs['blocked_uri'] = 'self'
return cls(**kwargs)
def from_raw(cls, raw):
# Firefox doesn't send effective-directive, so parse it from
# violated-directive but prefer effective-directive when present
#
# refs: https://bugzil.la/1192684#c8
try:
report = raw['csp-report']
report['effective-directive'] = report.get(
'effective-directive',
report['violated-directive'].split(None, 1)[0])
except (KeyError, IndexError):
pass
# Validate the raw data against the input schema (raises on failure)
schema = INPUT_SCHEMAS[cls.path]
jsonschema.validate(raw, schema)
# For CSP, the values we want are nested under the 'csp-report' key.
raw = raw['csp-report']
# Trim values and convert keys to use underscores
kwargs = {
k.replace('-', '_'): trim(v, 1024)
for k, v in six.iteritems(raw)
}
return cls.to_python(kwargs)
def validate_event(payload):
rv = {}
for key in 'type', 'target', 'classifier':
value = payload.get(key)
if value is not None:
rv[key] = trim(value, 1024)
return rv
def validate_error(payload):
rv = {}
for key in 'type', 'message', 'event_id':
value = payload.get(key)
if value is not None:
rv[key] = trim(value, 1024)
return rv
def search_message(self):
"""
The internal search_message attribute is only used for search purposes.
It adds a bunch of data from the metadata and the culprit.
"""
data = self.data
culprit = self.culprit
event_metadata = self.get_event_metadata()
if event_metadata is None:
event_metadata = eventtypes.get(
self.get_event_type())().get_metadata(self.data)
message = ""
if data.get("logentry"):
message += data["logentry"].get(
"formatted") or data["logentry"].get("message") or ""
if event_metadata:
for value in event_metadata.values():
value_u = force_text(value, errors="replace")
if value_u not in message:
message = f"{message} {value_u}"
if culprit and culprit not in message:
culprit_u = force_text(culprit, errors="replace")
message = f"{message} {culprit_u}"
return trim(message.strip(), settings.SENTRY_MAX_MESSAGE_LENGTH)
def to_python(cls, data):
data = data.copy()
extra_data = data.pop('data', data)
if not isinstance(extra_data, dict):
extra_data = {}
kwargs = {
'id': trim(data.pop('id', None), 128),
'email': trim(data.pop('email', None), 128),
'username': trim(data.pop('username', None), 128),
'ip_address': validate_ip(data.pop('ip_address', None), False),
}
kwargs['data'] = trim_dict(extra_data)
return cls(**kwargs)
def get_search_message(self, event_metadata=None, culprit=None):
"""This generates the internal event.message attribute which is used
for search purposes. It adds a bunch of data from the metadata and
the culprit.
"""
if event_metadata is None:
event_metadata = self.get_event_type().get_metadata(self._data)
if culprit is None:
culprit = self.get_culprit()
data = self._data
message = ""
if data.get("logentry"):
message += data["logentry"].get(
"formatted") or data["logentry"].get("message") or ""
if event_metadata:
for value in six.itervalues(event_metadata):
value_u = force_text(value, errors="replace")
if value_u not in message:
message = u"{} {}".format(message, value_u)
if culprit and culprit not in message:
culprit_u = force_text(culprit, errors="replace")
message = u"{} {}".format(message, culprit_u)
return trim(message.strip(), settings.SENTRY_MAX_MESSAGE_LENGTH)
def to_python(cls, data):
threads = []
for thread in data.get('values') or ():
threads.append(
{
'stacktrace': get_stacktrace(thread.get('stacktrace')),
'raw_stacktrace': get_stacktrace(thread.get('raw_stacktrace'), raw=True),
'id': trim(thread.get('id'), 40),
'crashed': bool(thread.get('crashed')),
'current': bool(thread.get('current')),
'name': trim(thread.get('name'), 200),
}
)
return cls(values=threads)
def _merge_frame(new_frame, symbolicated):
if symbolicated.get("function"):
raw_func = trim(symbolicated["function"], 256)
func = trim(trim_function_name(symbolicated["function"], "native"),
256)
# if function and raw function match, we can get away without
# storing a raw function
if func == raw_func:
new_frame["function"] = raw_func
# otherwise we store both
else:
new_frame["raw_function"] = raw_func
new_frame["function"] = func
if symbolicated.get("instruction_addr"):
new_frame["instruction_addr"] = symbolicated["instruction_addr"]
if symbolicated.get("symbol"):
new_frame["symbol"] = symbolicated["symbol"]
if symbolicated.get("abs_path"):
new_frame["abs_path"] = symbolicated["abs_path"]
new_frame["filename"] = posixpath.basename(symbolicated["abs_path"])
if symbolicated.get("filename"):
new_frame["filename"] = symbolicated["filename"]
if symbolicated.get("lineno"):
new_frame["lineno"] = symbolicated["lineno"]
if symbolicated.get("colno"):
new_frame["colno"] = symbolicated["colno"]
if symbolicated.get("package"):
new_frame["package"] = symbolicated["package"]
if symbolicated.get("trust"):
new_frame["trust"] = symbolicated["trust"]
if symbolicated.get("pre_context"):
new_frame["pre_context"] = symbolicated["pre_context"]
if symbolicated.get("context_line") is not None:
new_frame["context_line"] = symbolicated["context_line"]
if symbolicated.get("post_context"):
new_frame["post_context"] = symbolicated["post_context"]
addr_mode = symbolicated.get("addr_mode")
if addr_mode is None:
new_frame.pop("addr_mode", None)
else:
new_frame["addr_mode"] = addr_mode
if symbolicated.get("status"):
frame_meta = new_frame.setdefault("data", {})
frame_meta["symbolicator_status"] = symbolicated["status"]
def get_metadata(self):
exception = get_path(self.data, 'exception', 'values', -1)
# in some situations clients are submitting non-string data for these
rv = {
'type': trim(get_path(exception, 'type', default='Error'), 128),
'value': trim(get_path(exception, 'value', default=''), 1024),
}
# Attach crash location
stacktrace = exception.get('stacktrace')
if stacktrace:
fn = get_crash_file(stacktrace)
if fn is not None:
rv['filename'] = fn
return rv
def get_metadata(self):
exception = self.data['sentry.interfaces.Exception']['values'][-1]
# in some situations clients are submitting non-string data for these
rv = {
'type': trim(exception.get('type', 'Error'), 128),
'value': trim(exception.get('value', ''), 1024),
}
# Attach crash location
stacktrace = exception.get('stacktrace')
if stacktrace:
fn = get_crash_file(stacktrace)
if fn is not None:
rv['filename'] = fn
return rv
def normalize_crumb(cls, crumb):
ty = crumb.get('type') or 'default'
ts = parse_timestamp(crumb.get('timestamp'))
if ts is None:
raise InterfaceValidationError('Unable to determine timestamp '
'for crumb')
rv = {
'type': ty,
'timestamp': to_timestamp(ts),
}
level = crumb.get('level')
if level not in (None, 'info'):
rv['level'] = level
msg = crumb.get('message')
if msg is not None:
rv['message'] = trim(six.text_type(msg), 4096)
category = crumb.get('category')
if category is not None:
rv['category'] = trim(six.text_type(category), 256)
event_id = crumb.get('event_id')
if event_id is not None:
rv['event_id'] = event_id
if crumb.get('data'):
try:
for key, value in six.iteritems(crumb['data']):
if not isinstance(value, six.string_types):
crumb['data'][key] = json.dumps(value)
except AttributeError:
# TODO(dcramer): we dont want to discard the the rest of the
# crumb, but it'd be nice if we could record an error
# raise InterfaceValidationError(
# 'The ``data`` on breadcrumbs must be a mapping (received {})'.format(
# type(crumb['data']),
# )
# )
pass
else:
rv['data'] = trim(crumb['data'], 4096)
return rv
def to_python(cls, data):
if not data.get('message'):
raise InterfaceValidationError("No 'message' present")
# TODO(dcramer): some day we should stop people from sending arbitrary
# crap to the server
if not isinstance(data['message'], basestring):
data['message'] = json.dumps(data['message'])
kwargs = {
'message': trim(data['message'], settings.SENTRY_MAX_MESSAGE_LENGTH),
'formatted': None,
}
if data.get('params'):
kwargs['params'] = trim(data['params'], 1024)
else:
kwargs['params'] = ()
if kwargs['formatted']:
if not isinstance(kwargs['formatted'], basestring):
data['formatted'] = json.dumps(data['formatted'])
# support python-esque formatting (e.g. %s)
elif '%' in kwargs['message'] and kwargs['params']:
if isinstance(kwargs['params'], list):
kwargs['params'] = tuple(kwargs['params'])
try:
kwargs['formatted'] = trim(
kwargs['message'] % kwargs['params'],
settings.SENTRY_MAX_MESSAGE_LENGTH,
)
except Exception:
pass
# support very basic placeholder formatters (non-typed)
elif '{}' in kwargs['message'] and kwargs['params']:
try:
kwargs['formatted'] = trim(
kwargs['message'].format(kwargs['params']),
settings.SENTRY_MAX_MESSAGE_LENGTH,
)
except Exception:
pass
return cls(**kwargs)
def normalize_crumb(cls, crumb):
ty = crumb.get('type') or 'default'
level = crumb.get('level')
if not isinstance(level, six.string_types) or \
(level not in LOG_LEVELS_MAP and level != 'critical'):
level = 'info'
ts = parse_timestamp(crumb.get('timestamp'))
if ts is None:
raise InterfaceValidationError(
'Unable to determine timestamp for crumb')
ts = to_timestamp(ts)
msg = crumb.get('message')
if msg is not None:
msg = trim(six.text_type(msg), 4096)
category = crumb.get('category')
if category is not None:
category = trim(six.text_type(category), 256)
event_id = crumb.get('event_id')
data = crumb.get('data')
if not isinstance(data, dict):
# TODO(dcramer): we dont want to discard the the rest of the
# crumb, but it'd be nice if we could record an error
# raise InterfaceValidationError(
# 'The ``data`` on breadcrumbs must be a mapping (received {})'.format(
# type(crumb['data']),
# )
# )
data = None
else:
data = trim(data, 4096)
return {
'type': ty,
'level': level,
'timestamp': ts,
'message': msg,
'category': category,
'event_id': event_id,
'data': data
}
def normalize_crumb(cls, crumb):
ty = crumb.get('type') or 'default'
level = crumb.get('level') or 'info'
ts = parse_timestamp(crumb.get('timestamp'))
if ts is None:
raise InterfaceValidationError('Unable to determine timestamp for crumb')
ts = to_timestamp(ts)
msg = crumb.get('message')
if msg is not None:
msg = trim(six.text_type(msg), 4096)
category = crumb.get('category')
if category is not None:
category = trim(six.text_type(category), 256)
event_id = crumb.get('event_id')
data = crumb.get('data')
if data:
try:
for key, value in six.iteritems(data):
if not isinstance(value, six.string_types):
data[key] = json.dumps(value)
except AttributeError:
# TODO(dcramer): we dont want to discard the the rest of the
# crumb, but it'd be nice if we could record an error
# raise InterfaceValidationError(
# 'The ``data`` on breadcrumbs must be a mapping (received {})'.format(
# type(crumb['data']),
# )
# )
data = None
else:
data = trim(data, 4096)
return {
'type': ty,
'level': level,
'timestamp': ts,
'message': msg,
'category': category,
'event_id': event_id,
'data': data
}
def to_python(cls, data):
kwargs = {
'abs_path':
trim(data.get('abs_path', None), 256),
'filename':
trim(data.get('filename', None), 256),
'context_line':
trim(data.get('context_line', None), 256),
'lineno':
int(data['lineno'])
if data.get('lineno', None) is not None else None,
# TODO(dcramer): trim pre/post_context
'pre_context':
data.get('pre_context'),
'post_context':
data.get('post_context'),
}
return cls(**kwargs)
def from_raw(cls, raw):
# Validate the raw data against the input schema (raises on failure)
schema = INPUT_SCHEMAS[cls.path]
jsonschema.validate(raw, schema)
# Trim values and convert keys to use underscores
kwargs = {k.replace('-', '_'): trim(v, 1024) for k, v in six.iteritems(raw)}
return cls.to_python(kwargs)
def to_python(cls, data):
if not data.get('filename'):
raise InterfaceValidationError("Missing 'filename'")
if not data.get('context_line'):
raise InterfaceValidationError("Missing 'context_line'")
if not data.get('lineno'):
raise InterfaceValidationError("Missing 'lineno'")
kwargs = {
'abs_path': trim(data.get('abs_path', None), 256),
'filename': trim(data['filename'], 256),
'context_line': trim(data.get('context_line', None), 256),
'lineno': int(data['lineno']),
# TODO(dcramer): trim pre/post_context
'pre_context': data.get('pre_context'),
'post_context': data.get('post_context'),
}
return cls(**kwargs)
def __init__(self, alias, data):
self.alias = alias
ctx_data = {}
for key, value in six.iteritems(trim(data)):
# we use simple checks here, rathern than ' in set()' to avoid
# issues with maps/lists
if value is not None and value != '':
ctx_data[force_text(key)] = value
self.data = ctx_data
def to_python(cls, data):
kwargs = {k: trim(data.get(k, None), 1024) for k in REPORT_KEYS}
# Anything resulting from an "inline" whatever violation is either sent
# as 'self', or left off. In the case if it missing, we want to noramalize.
if not kwargs['blocked_uri']:
kwargs['blocked_uri'] = 'self'
return cls(**kwargs)
def to_python(cls, data):
is_valid, errors = validate_and_default_interface(data, cls.path)
if not is_valid:
raise InterfaceValidationError("Invalid device")
data = data.copy()
extra_data = data.pop('data', data)
name = trim(data.pop('name'), 64)
version = trim(data.pop('version'), 64)
build = trim(data.pop('build', None), 64)
kwargs = {
'name': name,
'version': version,
'build': build,
'data': trim_dict(extra_data),
}
return cls(**kwargs)
def to_python(cls, data):
name = data.get('name')
if not name:
raise InterfaceValidationError("No 'name' value")
version = data.get('version')
if not version:
raise InterfaceValidationError("No 'version' value")
integrations = data.get('integrations')
if integrations and not isinstance(integrations, list):
raise InterfaceValidationError("'integrations' must be a list")
kwargs = {
'name': trim(name, 128),
'version': trim(version, 128),
'integrations': integrations,
}
return cls(**kwargs)
def to_python(cls, data):
data = data.copy()
extra_data = data.pop('data', data)
if not isinstance(extra_data, dict):
extra_data = {}
ident = trim(data.pop('id', None), 128)
if ident:
ident = six.text_type(ident)
try:
email = trim(validate_email(data.pop('email', None), False),
MAX_EMAIL_FIELD_LENGTH)
except ValueError:
raise InterfaceValidationError("Invalid value for 'email'")
username = trim(data.pop('username', None), 128)
if username:
username = six.text_type(username)
name = trim(data.pop('name', None), 128)
if name:
name = six.text_type(name)
try:
ip_address = validate_ip(data.pop('ip_address', None), False)
except ValueError:
raise InterfaceValidationError("Invalid value for 'ip_address'")
# TODO(dcramer): patch in fix to deal w/ old data but not allow new
# if not (ident or email or username or ip_address):
# raise ValueError('No identifying value')
kwargs = {
'id': ident,
'email': email,
'username': username,
'ip_address': ip_address,
'name': name,
}
kwargs['data'] = trim_dict(extra_data)
return cls(**kwargs)
def validate_rpc(payload):
rv = {}
for key in 'endpoint', 'params', 'classifier':
value = payload.get(key)
if value is not None:
rv[key] = trim(value, 1024)
if not rv.get('endpoint'):
raise InterfaceValidationError("No endpoint provided for "
"'rpc' breadcrumb.")
return rv
def validate_query(payload):
rv = {}
for key in 'query', 'params', 'duration', 'classifier':
value = payload.get(key)
if value is not None:
rv[key] = trim(value, 1024)
if 'query' not in rv:
raise InterfaceValidationError("Query not provided for 'query' "
"breadcrumb.")
return rv
def to_python(cls, data, rust_renormalized=RUST_RENORMALIZED_DEFAULT):
if rust_renormalized:
for key in (
'message',
'formatted',
'params',
):
data.setdefault(key, None)
return cls(**data)
formatted = stringify(data.get('formatted'))
message = stringify(data.get('message'))
if formatted is None and message is None:
raise InterfaceValidationError("No message present")
params = data.get('params')
if isinstance(params, (list, tuple)):
params = tuple(p for p in params)
elif isinstance(params, dict):
params = {k: v for k, v in six.iteritems(params)}
else:
params = ()
if formatted is None and params:
try:
if '%' in message:
formatted = message % params
elif '{}' in message and isinstance(params, tuple):
formatted = message.format(*params)
# NB: Named newstyle arguments were never supported
except Exception:
pass
if formatted is None or message == formatted:
formatted = message
message = None
return cls(
formatted=trim(formatted, settings.SENTRY_MAX_MESSAGE_LENGTH),
message=trim(message, settings.SENTRY_MAX_MESSAGE_LENGTH),
params=trim(params, 1024),
)
def to_python(cls, data, has_system_frames=None, slim_frames=True):
if not (data.get('type') or data.get('value')):
raise InterfaceValidationError("No 'type' or 'value' present")
if data.get('stacktrace') and data['stacktrace'].get('frames'):
stacktrace = Stacktrace.to_python(
data['stacktrace'],
has_system_frames=has_system_frames,
slim_frames=slim_frames,
)
else:
stacktrace = None
if data.get('raw_stacktrace') and data['raw_stacktrace'].get('frames'):
raw_stacktrace = Stacktrace.to_python(
data['raw_stacktrace'],
has_system_frames=has_system_frames,
slim_frames=slim_frames,
)
else:
raw_stacktrace = None
type = data.get('type')
value = data.get('value')
if not type and ':' in value.split(' ', 1)[0]:
type, value = value.split(':', 1)
# in case of TypeError: foo (no space)
value = value.strip()
if value is not None and not isinstance(value, basestring):
value = json.dumps(value)
value = trim(value, 4096)
kwargs = {
'type': trim(type, 128),
'value': value,
'module': trim(data.get('module'), 128),
'stacktrace': stacktrace,
'raw_stacktrace': raw_stacktrace,
}
return cls(**kwargs)
def to_python(cls, data):
threads = []
for thread in data.get("values") or ():
if thread is None:
# XXX(markus): We should handle this in the UI and other
# consumers of this interface
continue
threads.append(
{
"stacktrace": get_stacktrace(thread.get("stacktrace")),
"raw_stacktrace": get_stacktrace(thread.get("raw_stacktrace"), raw=True),
"id": trim(thread.get("id"), 40),
"crashed": bool(thread.get("crashed")),
"current": bool(thread.get("current")),
"name": trim(thread.get("name"), 200),
}
)
return cls(values=threads)