def __init__(self, object, format='TEXT', _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(144, _hsid, 0, _hflags)
self.type = 144
self['t'] = 144
self['object'] = object
self['format'] = format
def specific_config_gen(IC,args):
IC.base_dir = args['base_dir']
IC.annotate_dir = args['annotate_dir']
IC.groundtruth_dir = args['groundtruth_dir']
IC.correspondence = tb.tabarray(SVfile = args['frame_correspondence'])
IC.size = args['size']
IC.prefix = prefix = args.get('image_extension','.jpg')
IC.current_frame_path = None
csvs = [x for x in os.listdir(IC.annotate_dir) if x.endswith('.csv')]
csvs.sort()
Xs = [tb.tabarray(SVfile = os.path.join(IC.annotate_dir,csv)) for csv in csvs]
cns = [csv.split('.')[0] for csv in csvs]
cns = [[cn]*len(X) for (cn,X) in zip(cns,Xs)]
Xs = [X.addcols(cn,names=['clip_num']) for (cn,X) in zip(cns,Xs)]
csvs = [x for x in os.listdir(IC.groundtruth_dir) if x.endswith('.csv')]
csvs.sort()
Gs = []
fields = ['clip_num','Frame'] + xfields + yfields
for ind,csv in enumerate(csvs):
try:
g = tb.tabarray(SVfile = os.path.join(IC.groundtruth_dir,csv))
except:
x = Xs[ind].addcols([-1]*len(Xs[ind]),names=['Correctness'])
else:
g = g.addcols([csv.split('.')[0]]*len(g),names = ['clip_num'])
g = g[fields + ['Confidence']]
g.renamecol('Confidence','Correctness')
x = Xs[ind].join(g,keycols=fields)
Gs.append(x)
X = tb.tab_rowstack(Gs)
X.sort(order=['clip_num','Frame'])
Y = IC.correspondence
F = tb.fast.recarrayisin(Y[['clip_num','Frame']],X[['clip_num','Frame']])
Y = Y[F]
X = X.join(Y,keycols=['clip_num','Frame'])
params = []
for t in X:
print(t)
cn = t['clip_num']
fr = t['Frame']
box = get_darpa_box(t)
bb = box.pop('box')
xc,yc = bb.center
center = correct_center((xc,yc),IC.size,(1920,1080))
bb_new = bbox.BoundingBox(center = center,width = IC.size[0], height = IC.size[1])
p = SON([('size',IC.size),
('bounding_box',SON([('xfields',list(bb_new.xs)),('yfields',list(bb_new.ys))])),
('original_bounding_box',SON([('xfields',list(bb.xs)),('yfields',list(bb.ys))])),
('clip_num',cn),
('Frame',int(t['Original'])),
('base_dir',IC.base_dir),
('correctness',int(t['Correctness']))])
p.update(box)
p['GuessObjectType'] = p['ObjectType']
p['ObjectType'] = p['ObjectType'] if t['Correctness'] == 1 else ''
params.append(SON([('image',p)]))
return params
def __init__(self, text, format='TEXT', _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(140, _hsid, 0, _hflags)
self.type = 140
self['t'] = 140
self['text'] = text
self['format'] = format
def __init__(self, text, format, completions, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(141, _hsid, 0, _hflags)
self.type = 141
self['t'] = 141
self['text'] = text
self['format'] = format
self['completions'] = completions
def __init__(self, object, format, obj_found=False, doc=None, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(143, _hsid, 0, _hflags)
self.type = 143
self['t'] = 143
self['object'] = object
self['format'] = format
self['obj_found'] = obj_found
self['doc'] = doc
def __init__(self, object, format, obj_found=False, source=None, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(145, _hsid, 0, _hflags)
self.type = 145
self['t'] = 145
self['object'] = object
self['format'] = format
self['obj_found'] = obj_found
self['source'] = source
def to_mongo(self, use_db_field=True, fields=None):
"""
Return as SON data ready for use with MongoDB.
"""
if not fields:
fields = []
data = SON()
data['_id'] = None
data['_cls'] = self._class_name
# only root fields ['test1.a', 'test2'] => ['test1', 'test2']
root_fields = {f.split('.')[0] for f in fields}
for field_name in self:
if root_fields and field_name not in root_fields:
continue
value = self._data.get(field_name, None)
field = self._fields.get(field_name)
if field is None and self._dynamic:
field = self._dynamic_fields.get(field_name)
if value is not None:
f_inputs = field.to_mongo.__code__.co_varnames
ex_vars = {}
if fields and 'fields' in f_inputs:
key = '%s.' % field_name
embedded_fields = [
i.replace(key, '') for i in fields
if i.startswith(key)]
ex_vars['fields'] = embedded_fields
if 'use_db_field' in f_inputs:
ex_vars['use_db_field'] = use_db_field
value = field.to_mongo(value, **ex_vars)
# Handle self generating fields
if value is None and field._auto_gen:
value = field.generate()
self._data[field_name] = value
if (value is not None) or (field.null):
if use_db_field:
data[field.db_field] = value
else:
data[field.name] = value
# Only add _cls if allow_inheritance is True
if not self._meta.get('allow_inheritance'):
data.pop('_cls')
return data
def __init__(self, stderr, stack=None, etype=None, value=None, syntax=None, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(10, _hsid, 0, _hflags)
self.type = 10
self['t'] = 10
self['stderr'] = stderr
self['stack'] = stack
self['etype'] = etype
self['value'] = value
self['syntax'] = syntax
def command(self, command, value=1, read_preference=None,
callback=None, check=True, allowable_errors=[], connection=None, **kwargs):
"""Issue a MongoDB command.
Send command `command` to the database and return the
response. If `command` is an instance of :class:`basestring`
then the command {`command`: `value`} will be sent. Otherwise,
`command` must be an instance of :class:`dict` and will be
sent as is.
Any additional keyword arguments will be added to the final
command document before it is sent.
For example, a command like ``{buildinfo: 1}`` can be sent
using:
>>> db.command("buildinfo")
For a command where the value matters, like ``{collstats:
collection_name}`` we can do:
>>> db.command("collstats", collection_name)
For commands that take additional arguments we can use
kwargs. So ``{filemd5: object_id, root: file_root}`` becomes:
>>> db.command("filemd5", object_id, root=file_root)
:Parameters:
- `command`: document representing the command to be issued,
or the name of the command (for simple commands only).
.. note:: the order of keys in the `command` document is
significant (the "verb" must come first), so commands
which require multiple keys (e.g. `findandmodify`)
should use an instance of :class:`~bson.son.SON` or
a string and kwargs instead of a Python `dict`.
- `value` (optional): value to use for the command verb when
`command` is passed as a string
- `**kwargs` (optional): additional keyword arguments will
be added to the command document before it is sent
"""
if isinstance(command, basestring):
command = SON([(command, value)])
command.update(kwargs)
if read_preference is None:
read_preference = self._read_preference
Client(self, '$cmd').find_one(command, is_command=True,
connection=connection, read_preference=read_preference, callback=callback)
def near(col, params):
if params is None or params.count(':') != 2:
raise TypeError("$near requires three arguments. Use like /%s/$near=-73.10:42.18:0.5/ to return all records within a 0.5-mile radius of %s" % (col, col))
params = params.split(":")
params[0] = float(params[0])
params[1] = float(params[1])
params[2] = float(params[2]) / 69.0
near_dict = {"$near": [params[0], params[1]]}
dist_dict = {"$maxDistance": params[2]}
q = SON(near_dict)
q.update(dist_dict)
return q
def __init__(self, source, cid=0, echo_stdin=True, displayhook='LAST', assignhook='NONE', print_ast=False, except_msg=False, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(120, _hsid, 0, _hflags)
self.type = 120
self['t'] = 120
self['source'] = source
self['cid'] = cid
self['echo_stdin'] = echo_stdin
self['displayhook'] = displayhook
self['assignhook'] = assignhook
self['print_ast'] = print_ast
self['except_msg'] = except_msg
def generate_splits(task_config,hash,colname):
base_query = SON([('__hash__',hash)])
ntrain = task_config['ntrain']
ntest = task_config['ntest']
ntrain_pos = task_config.get('ntrain_pos')
ntest_pos = task_config.get('ntest_pos')
N = task_config.get('N',10)
query = task_config['query']
base_query.update(reach_in('config',task_config.get('universe',SON([]))))
cquery = reach_in('config',query)
print('q',cquery)
print('u',base_query)
return traintest.generate_split2(DB_NAME,colname,cquery,N,ntrain,ntest,ntrain_pos=ntrain_pos,ntest_pos = ntest_pos,universe=base_query,use_negate = True)
def put_in_performance(split_results,image_config_gen,m,model_hash,image_hash,perf_coll,task,ext_hash):
model_results = SON([])
for stat in STATS:
if stat in split_results[0] and split_results[0][stat] != None:
model_results[stat] = sp.array([split_result[stat] for split_result in split_results]).mean()
out_record = SON([('model',m['config']['model']),
('model_hash',model_hash),
('model_filename',m['filename']),
('images',son_escape(image_config_gen['images'])),
('image_hash',image_hash),
('task',son_escape(task)),
('__hash__',ext_hash)
])
out_record.update(model_results)
perf_coll.insert(out_record)
def put_in_split_result(res,image_config_gen,m,task,ext_hash,split_id,splitres_fs):
out_record = SON([('model',m['config']['model']),
('images',son_escape(image_config_gen['images'])),
('task',son_escape(task)),
('split_id',split_id),
])
split_result = SON([])
for stat in STATS:
if stat in res and res[stat] != None:
split_result[stat] = res[stat]
filename = get_filename(out_record)
out_record['filename'] = filename
out_record['__hash__'] = ext_hash
out_record.update(split_result)
print('dump out split result...')
out_data = cPickle.dumps(SON([('split_result',res)]))
splitres_fs.put(out_data,**out_record)
def to_mongo(self):
from bson import SON
from mongoengine import Document
from mongoengine.base.common import ALLOW_INHERITANCE
"""Return as SON data ready for use with MongoDB.
"""
data = SON()
data["_id"] = None
data['_cls'] = self._class_name
for field_name in self._fields_ordered:
value = self._data.get(field_name, None)
field = self._fields.get(field_name)
if field is None and self._dynamic:
field = self._dynamic_fields.get(field_name)
if value is not None:
value = field.to_mongo(value)
# Handle self generating fields
if value is None and field._auto_gen:
value = field.generate()
self._data[field_name] = value
if value is not None:
data[field.db_field] = value
# If "_id" has not been set, then try and set it
if isinstance(self, Document):
if data["_id"] is None:
data["_id"] = self._data.get("id", None)
if data['_id'] is None:
data.pop('_id')
# Only add _cls if allow_inheritance is True
if (not hasattr(self, '_meta') or
not self._meta.get('allow_inheritance', ALLOW_INHERITANCE)):
data.pop('_cls')
return data
def MRsimple(collection, FunMap, FunReduce=None, query={}, out={"replace": 'mr_tmp'}, finalize=None,
scope={}, sort=None, jsMode=False, verbose=1):
""" simplified generic Map Reduce
see: http://docs.mongodb.org/manual/reference/method/db.collection.mapReduce/
returns (MR response object, results collection or results list if out={"inline":1})
Reduce function defaults to one that increments value count
optimize by sorting on emit fields
see: http://edgystuff.tumblr.com/post/7624019777/optimizing-map-reduce-with-mongodb
docs.mongodb.org/manual/reference/method/db.collection.mapReduce/#db.collection.mapReduce
sort i.e: sort= { "_id":1 }
jsMode should be False if we expect more than 500K dictinct ids
"""
if len(out.viewkeys()) > 1:
command = MRCommand_(out)
out = SON([(command, out[command]), ('db', out.get('db')),
('nonAtomic', out.get('nonAtomic', False))])
#nonAtomic not allowed on replace
FunMap = Code(FunMap, {})
if FunReduce is None:
FunReduce = u"""function (key, values) {var total = 0; for (var i = 0;
i < values.length; i++) { total += values[i]; } return total;}
"""
FunReduce = Code(FunReduce, {})
if verbose > 2:
print "Start MRsimple collection = %s"\
"query = %s\nMap=\n%s\nReduce=\n%s\nFinalize=%s\nscope=%s sort=%s" \
% tuple(map(str, (out, query, FunMap, FunReduce, finalize, scope, sort)))
if sort:
sort = SON(sort)
r = collection.map_reduce(FunMap, FunReduce, out=out, query=query,
finalize=finalize, scope=scope, sort=sort, full_response=True)
if verbose > 1:
print "End MRsimple collection=%s, query=%s\nresulsts=\n %s"\
% (collection.name, str(query), str(r))
if 'db' in out.viewkeys():
#@note: can be dict or SON, either way it has property viewkeys
results = collection.database.connection[r['result']['db']][r['result']['collection']]
else:
results = r['results'] if out.keys()[0] == 'inline' else collection.database[r['result']]
#@note: results is a list if inline else a collection
return r, results
def get_last_error(self, db, **options):
command = SON([("getlasterror", 1)])
db = "%s.$cmd" % db.split('.', 1)[0]
command.update(options)
query = Query(collection=db, query=command)
reply = yield self.send_QUERY(query)
assert len(reply.documents) == 1
document = reply.documents[0].decode()
err = document.get("err", None)
code = document.get("code", None)
if err is not None:
if code == 11000:
raise DuplicateKeyError(err, code=code)
else:
raise OperationFailure(err, code=code)
defer.returnValue(document)
def write_outcerts(func, configs, incertdicts, outcertpaths, db):
if incertdicts:
old_param_names = dict_union([op["param_names"] for op in incertdicts])
else:
old_param_names = SON([])
new_param_names = uniqify(ListUnion([x.keys() for x in configs]))
for (outcertpath, outroot) in zip(outcertpaths, func.outroots):
param_names = old_param_names.copy()
param_names[outroot] = new_param_names
remove_incorrect(db, outroot, func.outconfig_string, func.outrun_hash)
createCertificateDict(
outcertpath,
{
"run_hash": func.outrun_hash,
"db": func.dbname,
"out_args": func.out_args,
"root": outroot,
"config_hash": func.outconfig_string,
"param_names": param_names,
},
)
def to_dict_document(self, serializer=None):
data = SON()
data["_id"] = None
# data['_cls'] = self._class_name
for field_name in self:
value = self._data.get(field_name, None)
field = self._fields.get(field_name)
if field is None and self._dynamic:
field = self._dynamic_fields.get(field_name)
if value and isinstance(field, (ResourceField, ResourceIntField, DynamicResourceField)):
pass
# value = value.to_dict(fields=field.fields)
if value is not None:
value = field.to_dict(value, serializer=serializer)
# Handle self generating fields
if value is None and field._auto_gen:
value = field.generate()
self._data[field_name] = value
if value is not None:
data[field.db_field] = value
# If "_id" has not been set, then try and set it
Document = _import_class("Document")
if isinstance(self, Document):
if data["_id"] is None:
data["_id"] = self._data.get("id", None)
if data['_id'] is None:
data.pop('_id')
# Only add _cls if allow_inheritance is True
# if (not hasattr(self, '_meta') or
# not self._meta.get('allow_inheritance', ALLOW_INHERITANCE)):
# data.pop('_cls')
return data
def get_op_gen(op, oplist):
if op.get("outcertpaths") is None:
func = op["func"]
params = op.get("params")
inroots = func.inroots
outroots = func.outroots
if func.action_name == "inject":
args = op["params"]
out_args = SON([(outroot, params) for outroot in outroots])
else:
params = op.get("params", SON([]))
parents = []
for ir in inroots:
try:
parent = [op0 for op0 in oplist if ir in op0["func"].outroots][0]
except IndexError:
raise IndexError, "No parent found for at least one collection in " + repr(op0["func"].outroots)
else:
parents.append(parent)
for parent in parents:
get_op_gen(parent, oplist)
in_args = [parent["out_args"] for parent in parents]
op["incertpaths"] = [
get_cert_path(func.dbname, inroot, get_config_string(in_arg))
for (inroot, in_arg) in zip(inroots, in_args)
]
out_args = dict_union(in_args)
out_args.update(params)
op["out_args"] = out_args
op["outcertpaths"] = [
get_cert_path(func.dbname, outroot, get_config_string(out_args)) for outroot in func.outroots
]
def oid_date_range_filter(dt_from=None, dt_upto=None, field_name='_id'):
"""
constructs a range query usefull to query an ObjectId field by date
:Parameters:
- dt_from (datetime or tuple): starting date_time if tuple a datetime is constucted from tuple
- dt_upto (datetime or tuple): end date_time if tuple a datetime is constucted from tuple
- field_name: (str): optional default to '_id' field to query or None if None returns range only else returns full query
:Returns:
- range query (due to objectId structure $gt includes dt_from) while $lt dt_upto (not included)
"""
def dt(dt_or_tuple):
if isinstance(dt_or_tuple, datetime):
return dt_or_tuple
elif isinstance(dt_or_tuple, tuple):
return datetime(*dt_or_tuple)
else:
raise TypeError('dt must be a date or tuple')
q = SON()
if dt_from is not None:
q.update(SON([('$gte', ObjectId.from_datetime(dt(dt_from)))]))
if dt_upto is not None:
q.update(SON([('$lte', ObjectId.from_datetime(dt(dt_upto)))]))
return q if field_name is None else SON([(field_name, q)])
def train_test_loop(outfile,extract_creates,task_config,feature_config_path,hash):
feature_config = get_config(feature_config_path)
base_query = SON([('__config_hash__',hash)])
image_params = SON([('image',feature_config['image'])])
models_params = feature_config['models']
ntrain = task_config['ntrain']
ntest = task_config['ntest']
ntrain_pos = task_config.get('ntrain_pos')
N = task_config.get('N',10)
query = task_config['query']
base_query.update(reach_in('config',task_config.get('universe',SON([]))))
print('\n')
print('BASE',base_query)
print('\n')
conn = pm.Connection(document_class=SON)
db = conn['v1']
fs = gridfs.GridFS(db, collection = 'model_performance')
cquery = reach_in('config',query)
for m in models_params:
base_query_copy = base_query.copy()
base_query_copy.update(reach_in('config.model',m))
splitdata, results = train_test(cquery,'v1','features',ntrain,ntest,ntrain_pos=ntrain_pos,N=N,universe=base_query_copy)
splitpickle = cPickle.dumps(splitdata)
data = SON([('feature_config_path',feature_config_path),
('model',m),
('task',son_escape(task_config)),
('image__aggregate__',son_escape(feature_config['image']))])
filename = get_filename(data)
data.update(results)
data['filename'] = filename
fs.put(splitpickle,**data)
createCertificateDict(outfile,{'task_config':task_config,'feature_config':feature_config,'feature_config_path':feature_config_path})
class BaseDocument(object):
# TODO simplify how `_changed_fields` is used.
# Currently, handling of `_changed_fields` seems unnecessarily convoluted:
# 1. `BaseDocument` defines `_changed_fields` in its `__slots__`, yet it's
# not setting it to `[]` (or any other value) in `__init__`.
# 2. `EmbeddedDocument` sets `_changed_fields` to `[]` it its overloaded
# `__init__`.
# 3. `Document` does NOT set `_changed_fields` upon initialization. The
# field is primarily set via `_from_son` or `_clear_changed_fields`,
# though there are also other methods that manipulate it.
# 4. The codebase is littered with `hasattr` calls for `_changed_fields`.
__slots__ = (
"_changed_fields",
"_initialised",
"_created",
"_data",
"_dynamic_fields",
"_auto_id_field",
"_db_field_map",
"__weakref__",
)
_dynamic = False
_dynamic_lock = True
STRICT = False
def __init__(self, *args, **values):
"""
Initialise a document or an embedded document.
:param values: A dictionary of keys and values for the document.
It may contain additional reserved keywords, e.g. "__auto_convert".
:param __auto_convert: If True, supplied values will be converted
to Python-type values via each field's `to_python` method.
:param __only_fields: A set of fields that have been loaded for
this document. Empty if all fields have been loaded.
:param _created: Indicates whether this is a brand new document
or whether it's already been persisted before. Defaults to true.
"""
self._initialised = False
self._created = True
if args:
raise TypeError(
"Instantiating a document with positional arguments is not "
"supported. Please use `field_name=value` keyword arguments."
)
__auto_convert = values.pop("__auto_convert", True)
__only_fields = set(values.pop("__only_fields", values))
_created = values.pop("_created", True)
signals.pre_init.send(self.__class__, document=self, values=values)
# Check if there are undefined fields supplied to the constructor,
# if so raise an Exception.
if not self._dynamic and (self._meta.get("strict", True) or _created):
_undefined_fields = set(values.keys()) - set(
list(self._fields.keys()) + ["id", "pk", "_cls", "_text_score"]
)
if _undefined_fields:
msg = ('The fields "{0}" do not exist on the document "{1}"').format(
_undefined_fields, self._class_name
)
raise FieldDoesNotExist(msg)
if self.STRICT and not self._dynamic:
self._data = StrictDict.create(allowed_keys=self._fields_ordered)()
else:
self._data = {}
self._dynamic_fields = SON()
# Assign default values to the instance.
# We set default values only for fields loaded from DB. See
# https://github.com/mongoengine/mongoengine/issues/399 for more info.
for key, field in iteritems(self._fields):
if self._db_field_map.get(key, key) in __only_fields:
continue
value = getattr(self, key, None)
setattr(self, key, value)
if "_cls" not in values:
self._cls = self._class_name
# Set passed values after initialisation
if self._dynamic:
dynamic_data = {}
for key, value in iteritems(values):
if key in self._fields or key == "_id":
setattr(self, key, value)
else:
dynamic_data[key] = value
else:
FileField = _import_class("FileField")
for key, value in iteritems(values):
key = self._reverse_db_field_map.get(key, key)
if key in self._fields or key in ("id", "pk", "_cls"):
if __auto_convert and value is not None:
field = self._fields.get(key)
if field and not isinstance(field, FileField):
value = field.to_python(value)
setattr(self, key, value)
else:
self._data[key] = value
# Set any get_<field>_display methods
self.__set_field_display()
if self._dynamic:
self._dynamic_lock = False
for key, value in iteritems(dynamic_data):
setattr(self, key, value)
# Flag initialised
self._initialised = True
self._created = _created
signals.post_init.send(self.__class__, document=self)
def __delattr__(self, *args, **kwargs):
"""Handle deletions of fields"""
field_name = args[0]
if field_name in self._fields:
default = self._fields[field_name].default
if callable(default):
default = default()
setattr(self, field_name, default)
else:
super(BaseDocument, self).__delattr__(*args, **kwargs)
def __setattr__(self, name, value):
# Handle dynamic data only if an initialised dynamic document
if self._dynamic and not self._dynamic_lock:
if not hasattr(self, name) and not name.startswith("_"):
DynamicField = _import_class("DynamicField")
field = DynamicField(db_field=name, null=True)
field.name = name
self._dynamic_fields[name] = field
self._fields_ordered += (name,)
if not name.startswith("_"):
value = self.__expand_dynamic_values(name, value)
# Handle marking data as changed
if name in self._dynamic_fields:
self._data[name] = value
if hasattr(self, "_changed_fields"):
self._mark_as_changed(name)
try:
self__created = self._created
except AttributeError:
self__created = True
if (
self._is_document
and not self__created
and name in self._meta.get("shard_key", tuple())
and self._data.get(name) != value
):
msg = "Shard Keys are immutable. Tried to update %s" % name
raise OperationError(msg)
try:
self__initialised = self._initialised
except AttributeError:
self__initialised = False
# Check if the user has created a new instance of a class
if (
self._is_document
and self__initialised
and self__created
and name == self._meta.get("id_field")
):
super(BaseDocument, self).__setattr__("_created", False)
super(BaseDocument, self).__setattr__(name, value)
def __getstate__(self):
data = {}
for k in (
"_changed_fields",
"_initialised",
"_created",
"_dynamic_fields",
"_fields_ordered",
):
if hasattr(self, k):
data[k] = getattr(self, k)
data["_data"] = self.to_mongo()
return data
def __setstate__(self, data):
if isinstance(data["_data"], SON):
data["_data"] = self.__class__._from_son(data["_data"])._data
for k in (
"_changed_fields",
"_initialised",
"_created",
"_data",
"_dynamic_fields",
):
if k in data:
setattr(self, k, data[k])
if "_fields_ordered" in data:
if self._dynamic:
setattr(self, "_fields_ordered", data["_fields_ordered"])
else:
_super_fields_ordered = type(self)._fields_ordered
setattr(self, "_fields_ordered", _super_fields_ordered)
dynamic_fields = data.get("_dynamic_fields") or SON()
for k in dynamic_fields.keys():
setattr(self, k, data["_data"].get(k))
def __iter__(self):
return iter(self._fields_ordered)
def __getitem__(self, name):
"""Dictionary-style field access, return a field's value if present.
"""
try:
if name in self._fields_ordered:
return getattr(self, name)
except AttributeError:
pass
raise KeyError(name)
def __setitem__(self, name, value):
"""Dictionary-style field access, set a field's value.
"""
# Ensure that the field exists before settings its value
if not self._dynamic and name not in self._fields:
raise KeyError(name)
return setattr(self, name, value)
def __contains__(self, name):
try:
val = getattr(self, name)
return val is not None
except AttributeError:
return False
def __len__(self):
return len(self._data)
def __repr__(self):
try:
u = self.__str__()
except (UnicodeEncodeError, UnicodeDecodeError):
u = "[Bad Unicode data]"
repr_type = str if u is None else type(u)
return repr_type("<%s: %s>" % (self.__class__.__name__, u))
def __str__(self):
# TODO this could be simpler?
if hasattr(self, "__unicode__"):
if six.PY3:
return self.__unicode__()
else:
return six.text_type(self).encode("utf-8")
return six.text_type("%s object" % self.__class__.__name__)
def __eq__(self, other):
if (
isinstance(other, self.__class__)
and hasattr(other, "id")
and other.id is not None
):
return self.id == other.id
if isinstance(other, DBRef):
return (
self._get_collection_name() == other.collection and self.id == other.id
)
if self.id is None:
return self is other
return False
def __ne__(self, other):
return not self.__eq__(other)
def clean(self):
"""
Hook for doing document level data cleaning before validation is run.
Any ValidationError raised by this method will not be associated with
a particular field; it will have a special-case association with the
field defined by NON_FIELD_ERRORS.
"""
pass
def get_text_score(self):
"""
Get text score from text query
"""
if "_text_score" not in self._data:
raise InvalidDocumentError(
"This document is not originally built from a text query"
)
return self._data["_text_score"]
def to_mongo(self, use_db_field=True, fields=None):
"""
Return as SON data ready for use with MongoDB.
"""
fields = fields or []
data = SON()
data["_id"] = None
data["_cls"] = self._class_name
# only root fields ['test1.a', 'test2'] => ['test1', 'test2']
root_fields = {f.split(".")[0] for f in fields}
for field_name in self:
if root_fields and field_name not in root_fields:
continue
value = self._data.get(field_name, None)
field = self._fields.get(field_name)
if field is None and self._dynamic:
field = self._dynamic_fields.get(field_name)
if value is not None:
f_inputs = field.to_mongo.__code__.co_varnames
ex_vars = {}
if fields and "fields" in f_inputs:
key = "%s." % field_name
embedded_fields = [
i.replace(key, "") for i in fields if i.startswith(key)
]
ex_vars["fields"] = embedded_fields
if "use_db_field" in f_inputs:
ex_vars["use_db_field"] = use_db_field
value = field.to_mongo(value, **ex_vars)
# Handle self generating fields
if value is None and field._auto_gen:
value = field.generate()
self._data[field_name] = value
if (value is not None) or (field.null):
if use_db_field:
data[field.db_field] = value
else:
data[field.name] = value
# Only add _cls if allow_inheritance is True
if not self._meta.get("allow_inheritance"):
data.pop("_cls")
return data
def validate(self, clean=True):
"""Ensure that all fields' values are valid and that required fields
are present.
Raises :class:`ValidationError` if any of the fields' values are found
to be invalid.
"""
# Ensure that each field is matched to a valid value
errors = {}
if clean:
try:
self.clean()
except ValidationError as error:
errors[NON_FIELD_ERRORS] = error
# Get a list of tuples of field names and their current values
fields = [
(
self._fields.get(name, self._dynamic_fields.get(name)),
self._data.get(name),
)
for name in self._fields_ordered
]
EmbeddedDocumentField = _import_class("EmbeddedDocumentField")
GenericEmbeddedDocumentField = _import_class("GenericEmbeddedDocumentField")
for field, value in fields:
if value is not None:
try:
if isinstance(
field, (EmbeddedDocumentField, GenericEmbeddedDocumentField)
):
field._validate(value, clean=clean)
else:
field._validate(value)
except ValidationError as error:
errors[field.name] = error.errors or error
except (ValueError, AttributeError, AssertionError) as error:
errors[field.name] = error
elif field.required and not getattr(field, "_auto_gen", False):
errors[field.name] = ValidationError(
"Field is required", field_name=field.name
)
if errors:
pk = "None"
if hasattr(self, "pk"):
pk = self.pk
elif self._instance and hasattr(self._instance, "pk"):
pk = self._instance.pk
message = "ValidationError (%s:%s) " % (self._class_name, pk)
raise ValidationError(message, errors=errors)
def to_json(self, *args, **kwargs):
"""Convert this document to JSON.
:param use_db_field: Serialize field names as they appear in
MongoDB (as opposed to attribute names on this document).
Defaults to True.
"""
use_db_field = kwargs.pop("use_db_field", True)
return json_util.dumps(self.to_mongo(use_db_field), *args, **kwargs)
@classmethod
def from_json(cls, json_data, created=False):
"""Converts json data to a Document instance
:param str json_data: The json data to load into the Document
:param bool created: Boolean defining whether to consider the newly
instantiated document as brand new or as persisted already:
* If True, consider the document as brand new, no matter what data
it's loaded with (i.e. even if an ID is loaded).
* If False and an ID is NOT provided, consider the document as
brand new.
* If False and an ID is provided, assume that the object has
already been persisted (this has an impact on the subsequent
call to .save()).
* Defaults to ``False``.
"""
# TODO should `created` default to False? If the object already exists
# in the DB, you would likely retrieve it from MongoDB itself through
# a query, not load it from JSON data.
return cls._from_son(json_util.loads(json_data), created=created)
def __expand_dynamic_values(self, name, value):
"""Expand any dynamic values to their correct types / values."""
if not isinstance(value, (dict, list, tuple)):
return value
# If the value is a dict with '_cls' in it, turn it into a document
is_dict = isinstance(value, dict)
if is_dict and "_cls" in value:
cls = get_document(value["_cls"])
return cls(**value)
if is_dict:
value = {k: self.__expand_dynamic_values(k, v) for k, v in value.items()}
else:
value = [self.__expand_dynamic_values(name, v) for v in value]
# Convert lists / values so we can watch for any changes on them
EmbeddedDocumentListField = _import_class("EmbeddedDocumentListField")
if isinstance(value, (list, tuple)) and not isinstance(value, BaseList):
if issubclass(type(self), EmbeddedDocumentListField):
value = EmbeddedDocumentList(value, self, name)
else:
value = BaseList(value, self, name)
elif isinstance(value, dict) and not isinstance(value, BaseDict):
value = BaseDict(value, self, name)
return value
def _mark_as_changed(self, key):
"""Mark a key as explicitly changed by the user."""
if not key:
return
if not hasattr(self, "_changed_fields"):
return
if "." in key:
key, rest = key.split(".", 1)
key = self._db_field_map.get(key, key)
key = "%s.%s" % (key, rest)
else:
key = self._db_field_map.get(key, key)
if key not in self._changed_fields:
levels, idx = key.split("."), 1
while idx <= len(levels):
if ".".join(levels[:idx]) in self._changed_fields:
break
idx += 1
else:
self._changed_fields.append(key)
# remove lower level changed fields
level = ".".join(levels[:idx]) + "."
remove = self._changed_fields.remove
for field in self._changed_fields[:]:
if field.startswith(level):
remove(field)
def _clear_changed_fields(self):
"""Using _get_changed_fields iterate and remove any fields that
are marked as changed.
"""
for changed in self._get_changed_fields():
parts = changed.split(".")
data = self
for part in parts:
if isinstance(data, list):
try:
data = data[int(part)]
except IndexError:
data = None
elif isinstance(data, dict):
data = data.get(part, None)
else:
data = getattr(data, part, None)
if not isinstance(data, LazyReference) and hasattr(
data, "_changed_fields"
):
if getattr(data, "_is_document", False):
continue
data._changed_fields = []
self._changed_fields = []
def _nestable_types_changed_fields(self, changed_fields, base_key, data):
"""Inspect nested data for changed fields
:param changed_fields: Previously collected changed fields
:param base_key: The base key that must be used to prepend changes to this data
:param data: data to inspect for changes
"""
# Loop list / dict fields as they contain documents
# Determine the iterator to use
if not hasattr(data, "items"):
iterator = enumerate(data)
else:
iterator = iteritems(data)
for index_or_key, value in iterator:
item_key = "%s%s." % (base_key, index_or_key)
# don't check anything lower if this key is already marked
# as changed.
if item_key[:-1] in changed_fields:
continue
if hasattr(value, "_get_changed_fields"):
changed = value._get_changed_fields()
changed_fields += ["%s%s" % (item_key, k) for k in changed if k]
elif isinstance(value, (list, tuple, dict)):
self._nestable_types_changed_fields(changed_fields, item_key, value)
def _get_changed_fields(self):
"""Return a list of all fields that have explicitly been changed.
"""
EmbeddedDocument = _import_class("EmbeddedDocument")
ReferenceField = _import_class("ReferenceField")
GenericReferenceField = _import_class("GenericReferenceField")
SortedListField = _import_class("SortedListField")
changed_fields = []
changed_fields += getattr(self, "_changed_fields", [])
for field_name in self._fields_ordered:
db_field_name = self._db_field_map.get(field_name, field_name)
key = "%s." % db_field_name
data = self._data.get(field_name, None)
field = self._fields.get(field_name)
if db_field_name in changed_fields:
# Whole field already marked as changed, no need to go further
continue
if isinstance(field, ReferenceField): # Don't follow referenced documents
continue
if isinstance(data, EmbeddedDocument):
# Find all embedded fields that have been changed
changed = data._get_changed_fields()
changed_fields += ["%s%s" % (key, k) for k in changed if k]
elif isinstance(data, (list, tuple, dict)):
if hasattr(field, "field") and isinstance(
field.field, (ReferenceField, GenericReferenceField)
):
continue
elif isinstance(field, SortedListField) and field._ordering:
# if ordering is affected whole list is changed
if any(field._ordering in d._changed_fields for d in data):
changed_fields.append(db_field_name)
continue
self._nestable_types_changed_fields(changed_fields, key, data)
return changed_fields
def _delta(self):
"""Returns the delta (set, unset) of the changes for a document.
Gets any values that have been explicitly changed.
"""
# Handles cases where not loaded from_son but has _id
doc = self.to_mongo()
set_fields = self._get_changed_fields()
unset_data = {}
if hasattr(self, "_changed_fields"):
set_data = {}
# Fetch each set item from its path
for path in set_fields:
parts = path.split(".")
d = doc
new_path = []
for p in parts:
if isinstance(d, (ObjectId, DBRef)):
# Don't dig in the references
break
elif isinstance(d, list) and p.isdigit():
# An item of a list (identified by its index) is updated
d = d[int(p)]
elif hasattr(d, "get"):
# dict-like (dict, embedded document)
d = d.get(p)
new_path.append(p)
path = ".".join(new_path)
set_data[path] = d
else:
set_data = doc
if "_id" in set_data:
del set_data["_id"]
# Determine if any changed items were actually unset.
for path, value in set_data.items():
if value or isinstance(
value, (numbers.Number, bool)
): # Account for 0 and True that are truthy
continue
parts = path.split(".")
if self._dynamic and len(parts) and parts[0] in self._dynamic_fields:
del set_data[path]
unset_data[path] = 1
continue
# If we've set a value that ain't the default value don't unset it.
default = None
if path in self._fields:
default = self._fields[path].default
else: # Perform a full lookup for lists / embedded lookups
d = self
db_field_name = parts.pop()
for p in parts:
if isinstance(d, list) and p.isdigit():
d = d[int(p)]
elif hasattr(d, "__getattribute__") and not isinstance(d, dict):
real_path = d._reverse_db_field_map.get(p, p)
d = getattr(d, real_path)
else:
d = d.get(p)
if hasattr(d, "_fields"):
field_name = d._reverse_db_field_map.get(
db_field_name, db_field_name
)
if field_name in d._fields:
default = d._fields.get(field_name).default
else:
default = None
if default is not None:
default = default() if callable(default) else default
if value != default:
continue
del set_data[path]
unset_data[path] = 1
return set_data, unset_data
@classmethod
def _get_collection_name(cls):
"""Return the collection name for this class. None for abstract
class.
"""
return cls._meta.get("collection", None)
@classmethod
def _from_son(cls, son, _auto_dereference=True, only_fields=None, created=False):
"""Create an instance of a Document (subclass) from a PyMongo SON."""
if not only_fields:
only_fields = []
if son and not isinstance(son, dict):
raise ValueError(
"The source SON object needs to be of type 'dict' but a '%s' was found"
% type(son)
)
# Get the class name from the document, falling back to the given
# class if unavailable
class_name = son.get("_cls", cls._class_name)
# Convert SON to a data dict, making sure each key is a string and
# corresponds to the right db field.
data = {}
for key, value in iteritems(son):
key = str(key)
key = cls._db_field_map.get(key, key)
data[key] = value
# Return correct subclass for document type
if class_name != cls._class_name:
cls = get_document(class_name)
errors_dict = {}
fields = cls._fields
if not _auto_dereference:
fields = copy.deepcopy(fields)
for field_name, field in iteritems(fields):
field._auto_dereference = _auto_dereference
if field.db_field in data:
value = data[field.db_field]
try:
data[field_name] = (
value if value is None else field.to_python(value)
)
if field_name != field.db_field:
del data[field.db_field]
except (AttributeError, ValueError) as e:
errors_dict[field_name] = e
if errors_dict:
errors = "\n".join(
["Field '%s' - %s" % (k, v) for k, v in errors_dict.items()]
)
msg = "Invalid data to create a `%s` instance.\n%s" % (
cls._class_name,
errors,
)
raise InvalidDocumentError(msg)
# In STRICT documents, remove any keys that aren't in cls._fields
if cls.STRICT:
data = {k: v for k, v in iteritems(data) if k in cls._fields}
obj = cls(
__auto_convert=False, _created=created, __only_fields=only_fields, **data
)
obj._changed_fields = []
if not _auto_dereference:
obj._fields = fields
return obj
@classmethod
def _build_index_specs(cls, meta_indexes):
"""Generate and merge the full index specs."""
geo_indices = cls._geo_indices()
unique_indices = cls._unique_with_indexes()
index_specs = [cls._build_index_spec(spec) for spec in meta_indexes]
def merge_index_specs(index_specs, indices):
"""Helper method for merging index specs."""
if not indices:
return index_specs
# Create a map of index fields to index spec. We're converting
# the fields from a list to a tuple so that it's hashable.
spec_fields = {tuple(index["fields"]): index for index in index_specs}
# For each new index, if there's an existing index with the same
# fields list, update the existing spec with all data from the
# new spec.
for new_index in indices:
candidate = spec_fields.get(tuple(new_index["fields"]))
if candidate is None:
index_specs.append(new_index)
else:
candidate.update(new_index)
return index_specs
# Merge geo indexes and unique_with indexes into the meta index specs.
index_specs = merge_index_specs(index_specs, geo_indices)
index_specs = merge_index_specs(index_specs, unique_indices)
return index_specs
@classmethod
def _build_index_spec(cls, spec):
"""Build a PyMongo index spec from a MongoEngine index spec."""
if isinstance(spec, six.string_types):
spec = {"fields": [spec]}
elif isinstance(spec, (list, tuple)):
spec = {"fields": list(spec)}
elif isinstance(spec, dict):
spec = dict(spec)
index_list = []
direction = None
# Check to see if we need to include _cls
allow_inheritance = cls._meta.get("allow_inheritance")
include_cls = (
allow_inheritance
and not spec.get("sparse", False)
and spec.get("cls", True)
and "_cls" not in spec["fields"]
)
# 733: don't include cls if index_cls is False unless there is an explicit cls with the index
include_cls = include_cls and (
spec.get("cls", False) or cls._meta.get("index_cls", True)
)
if "cls" in spec:
spec.pop("cls")
for key in spec["fields"]:
# If inherited spec continue
if isinstance(key, (list, tuple)):
continue
# ASCENDING from +
# DESCENDING from -
# TEXT from $
# HASHED from #
# GEOSPHERE from (
# GEOHAYSTACK from )
# GEO2D from *
direction = pymongo.ASCENDING
if key.startswith("-"):
direction = pymongo.DESCENDING
elif key.startswith("$"):
direction = pymongo.TEXT
elif key.startswith("#"):
direction = pymongo.HASHED
elif key.startswith("("):
direction = pymongo.GEOSPHERE
elif key.startswith(")"):
direction = pymongo.GEOHAYSTACK
elif key.startswith("*"):
direction = pymongo.GEO2D
if key.startswith(("+", "-", "*", "$", "#", "(", ")")):
key = key[1:]
# Use real field name, do it manually because we need field
# objects for the next part (list field checking)
parts = key.split(".")
if parts in (["pk"], ["id"], ["_id"]):
key = "_id"
else:
fields = cls._lookup_field(parts)
parts = []
for field in fields:
try:
if field != "_id":
field = field.db_field
except AttributeError:
pass
parts.append(field)
key = ".".join(parts)
index_list.append((key, direction))
# Don't add cls to a geo index
if include_cls and direction not in (
pymongo.GEO2D,
pymongo.GEOHAYSTACK,
pymongo.GEOSPHERE,
):
index_list.insert(0, ("_cls", 1))
if index_list:
spec["fields"] = index_list
return spec
@classmethod
def _unique_with_indexes(cls, namespace=""):
"""Find unique indexes in the document schema and return them."""
unique_indexes = []
for field_name, field in cls._fields.items():
sparse = field.sparse
# Generate a list of indexes needed by uniqueness constraints
if field.unique:
unique_fields = [field.db_field]
# Add any unique_with fields to the back of the index spec
if field.unique_with:
if isinstance(field.unique_with, six.string_types):
field.unique_with = [field.unique_with]
# Convert unique_with field names to real field names
unique_with = []
for other_name in field.unique_with:
parts = other_name.split(".")
# Lookup real name
parts = cls._lookup_field(parts)
name_parts = [part.db_field for part in parts]
unique_with.append(".".join(name_parts))
# Unique field should be required
parts[-1].required = True
sparse = not sparse and parts[-1].name not in cls.__dict__
unique_fields += unique_with
# Add the new index to the list
fields = [
("%s%s" % (namespace, f), pymongo.ASCENDING) for f in unique_fields
]
index = {"fields": fields, "unique": True, "sparse": sparse}
unique_indexes.append(index)
if field.__class__.__name__ in {
"EmbeddedDocumentListField",
"ListField",
"SortedListField",
}:
field = field.field
# Grab any embedded document field unique indexes
if (
field.__class__.__name__ == "EmbeddedDocumentField"
and field.document_type != cls
):
field_namespace = "%s." % field_name
doc_cls = field.document_type
unique_indexes += doc_cls._unique_with_indexes(field_namespace)
return unique_indexes
@classmethod
def _geo_indices(cls, inspected=None, parent_field=None):
inspected = inspected or []
geo_indices = []
inspected.append(cls)
geo_field_type_names = (
"EmbeddedDocumentField",
"GeoPointField",
"PointField",
"LineStringField",
"PolygonField",
)
geo_field_types = tuple(
[_import_class(field) for field in geo_field_type_names]
)
for field in cls._fields.values():
if not isinstance(field, geo_field_types):
continue
if hasattr(field, "document_type"):
field_cls = field.document_type
if field_cls in inspected:
continue
if hasattr(field_cls, "_geo_indices"):
geo_indices += field_cls._geo_indices(
inspected, parent_field=field.db_field
)
elif field._geo_index:
field_name = field.db_field
if parent_field:
field_name = "%s.%s" % (parent_field, field_name)
geo_indices.append({"fields": [(field_name, field._geo_index)]})
return geo_indices
@classmethod
def _lookup_field(cls, parts):
"""Given the path to a given field, return a list containing
the Field object associated with that field and all of its parent
Field objects.
Args:
parts (str, list, or tuple) - path to the field. Should be a
string for simple fields existing on this document or a list
of strings for a field that exists deeper in embedded documents.
Returns:
A list of Field instances for fields that were found or
strings for sub-fields that weren't.
Example:
>>> user._lookup_field('name')
[<mongoengine.fields.StringField at 0x1119bff50>]
>>> user._lookup_field('roles')
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>]
>>> user._lookup_field(['roles', 'role'])
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>,
<mongoengine.fields.StringField at 0x1119ec050>]
>>> user._lookup_field('doesnt_exist')
raises LookUpError
>>> user._lookup_field(['roles', 'doesnt_exist'])
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>,
'doesnt_exist']
"""
# TODO this method is WAY too complicated. Simplify it.
# TODO don't think returning a string for embedded non-existent fields is desired
ListField = _import_class("ListField")
DynamicField = _import_class("DynamicField")
if not isinstance(parts, (list, tuple)):
parts = [parts]
fields = []
field = None
for field_name in parts:
# Handle ListField indexing:
if field_name.isdigit() and isinstance(field, ListField):
fields.append(field_name)
continue
# Look up first field from the document
if field is None:
if field_name == "pk":
# Deal with "primary key" alias
field_name = cls._meta["id_field"]
if field_name in cls._fields:
field = cls._fields[field_name]
elif cls._dynamic:
field = DynamicField(db_field=field_name)
elif cls._meta.get("allow_inheritance") or cls._meta.get(
"abstract", False
):
# 744: in case the field is defined in a subclass
for subcls in cls.__subclasses__():
try:
field = subcls._lookup_field([field_name])[0]
except LookUpError:
continue
if field is not None:
break
else:
raise LookUpError('Cannot resolve field "%s"' % field_name)
else:
raise LookUpError('Cannot resolve field "%s"' % field_name)
else:
ReferenceField = _import_class("ReferenceField")
GenericReferenceField = _import_class("GenericReferenceField")
# If previous field was a reference, throw an error (we
# cannot look up fields that are on references).
if isinstance(field, (ReferenceField, GenericReferenceField)):
raise LookUpError(
"Cannot perform join in mongoDB: %s" % "__".join(parts)
)
# If the parent field has a "field" attribute which has a
# lookup_member method, call it to find the field
# corresponding to this iteration.
if hasattr(getattr(field, "field", None), "lookup_member"):
new_field = field.field.lookup_member(field_name)
# If the parent field is a DynamicField or if it's part of
# a DynamicDocument, mark current field as a DynamicField
# with db_name equal to the field name.
elif cls._dynamic and (
isinstance(field, DynamicField)
or getattr(getattr(field, "document_type", None), "_dynamic", None)
):
new_field = DynamicField(db_field=field_name)
# Else, try to use the parent field's lookup_member method
# to find the subfield.
elif hasattr(field, "lookup_member"):
new_field = field.lookup_member(field_name)
# Raise a LookUpError if all the other conditions failed.
else:
raise LookUpError(
"Cannot resolve subfield or operator {} "
"on the field {}".format(field_name, field.name)
)
# If current field still wasn't found and the parent field
# is a ComplexBaseField, add the name current field name and
# move on.
if not new_field and isinstance(field, ComplexBaseField):
fields.append(field_name)
continue
elif not new_field:
raise LookUpError('Cannot resolve field "%s"' % field_name)
field = new_field # update field to the new field type
fields.append(field)
return fields
@classmethod
def _translate_field_name(cls, field, sep="."):
"""Translate a field attribute name to a database field name.
"""
parts = field.split(sep)
parts = [f.db_field for f in cls._lookup_field(parts)]
return ".".join(parts)
def __set_field_display(self):
"""For each field that specifies choices, create a
get_<field>_display method.
"""
fields_with_choices = [(n, f) for n, f in self._fields.items() if f.choices]
for attr_name, field in fields_with_choices:
setattr(
self,
"get_%s_display" % attr_name,
partial(self.__get_field_display, field=field),
)
def __get_field_display(self, field):
"""Return the display value for a choice field"""
value = getattr(self, field.name)
if field.choices and isinstance(field.choices[0], (list, tuple)):
if value is None:
return None
sep = getattr(field, "display_sep", " ")
values = (
value
if field.__class__.__name__ in ("ListField", "SortedListField")
else [value]
)
return sep.join(
[
six.text_type(dict(field.choices).get(val, val))
for val in values or []
]
)
return value
def build_itemsets(distance, undefined=True, only_tourists=True):
hash_table = dict()
count = 0
places = create_place_dict()
f = open('data/encoded_itemsets.txt', mode='w')
g = open('data/translation_table.txt', mode='w')
o = open('data/decoded_itemsets.txt', mode='w')
q1 = {'touristLocal': 'tourist'}
q2 = {'$or': [{'touristLocal': 'tourist'}, {'touristLocal': 'undefined'}]}
if only_tourists:
query = q1
else:
query = q2
for tourist in ucol.find(query):
tweets = []
item_set = set()
places_set = set()
for sequence in scol.find({
'user_id': tourist['id']
}).sort('serial', pymongo.ASCENDING):
for id in sequence['sequence']:
for tweet in tcol.find({'id': id}):
tweets.append(tweet)
lon = tweet['coordinates']['coordinates'][0]
lat = tweet['coordinates']['coordinates'][1]
queryNear = {
'coordinates':
SON([('$near', {
"type": "Point",
"coordinates": [lon, lat]
}), ('$maxDistance', distance)])
}
close_pois = pcol.find(queryNear)
closest_poi = 'UNDEFINED'
if close_pois.count() > 0:
place_id = close_pois.next()['id']
code_place = places.get(place_id, -1)
if code_place == -1:
continue
closest_poi = code_place
else:
if not undefined:
continue
code = hash_table.get(closest_poi, 0)
if code == 0:
count += 1
hash_table[closest_poi] = count
code = count
item_set.add(code)
places_set.add(code_place)
if len(item_set) > 0:
f.write(" ".join(list(map(str, sorted(item_set)))))
f.write("\n")
if len(places_set) > 0:
o.write(" ".join(list(map(str, places_set))))
o.write("\n")
aux = []
for key, value in hash_table.items():
aux += [(key, value)]
sorted_aux = sorted(aux, key=lambda tup: tup[1])
for item in sorted_aux:
g.write(str(item[1]) + "\t" + item[0] + "\n")
return hash_table
def to_mongo(self, value):
if isinstance(value, dict):
return value
return SON([("type", self._type), ("coordinates", value)])
def __init__(self, *args, **values):
"""
Initialise a document or an embedded document.
:param dict values: A dictionary of keys and values for the document.
It may contain additional reserved keywords, e.g. "__auto_convert".
:param bool __auto_convert: If True, supplied values will be converted
to Python-type values via each field's `to_python` method.
:param set __only_fields: A set of fields that have been loaded for
this document. Empty if all fields have been loaded.
:param bool _created: Indicates whether this is a brand new document
or whether it's already been persisted before. Defaults to true.
"""
self._initialised = False
self._created = True
if args:
# Combine positional arguments with named arguments.
# We only want named arguments.
field = iter(self._fields_ordered)
# If its an automatic id field then skip to the first defined field
if getattr(self, '_auto_id_field', False):
next(field)
for value in args:
name = next(field)
if name in values:
raise TypeError(
'Multiple values for keyword argument "%s"' % name)
values[name] = value
__auto_convert = values.pop('__auto_convert', True)
__only_fields = set(values.pop('__only_fields', values))
_created = values.pop('_created', True)
signals.pre_init.send(self.__class__, document=self, values=values)
# Check if there are undefined fields supplied to the constructor,
# if so raise an Exception.
if not self._dynamic and (self._meta.get('strict', True) or _created):
_undefined_fields = set(
values.keys()) - set(self._fields.keys() +
['id', 'pk', '_cls', '_text_score'])
if _undefined_fields:
msg = ('The fields "{0}" do not exist on the document "{1}"'
).format(_undefined_fields, self._class_name)
raise FieldDoesNotExist(msg)
if self.STRICT and not self._dynamic:
self._data = StrictDict.create(allowed_keys=self._fields_ordered)()
else:
self._data = {}
self._dynamic_fields = SON()
# Assign default values to the instance.
# We set default values only for fields loaded from DB. See
# https://github.com/mongoengine/mongoengine/issues/399 for more info.
for key, field in iteritems(self._fields):
if self._db_field_map.get(key, key) in __only_fields:
continue
value = getattr(self, key, None)
setattr(self, key, value)
if '_cls' not in values:
self._cls = self._class_name
# Set passed values after initialisation
if self._dynamic:
dynamic_data = {}
for key, value in iteritems(values):
if key in self._fields or key == '_id':
setattr(self, key, value)
else:
dynamic_data[key] = value
else:
FileField = _import_class('FileField')
for key, value in iteritems(values):
key = self._reverse_db_field_map.get(key, key)
if key in self._fields or key in ('id', 'pk', '_cls'):
if __auto_convert and value is not None:
field = self._fields.get(key)
if field and not isinstance(field, FileField):
value = field.to_python(value)
setattr(self, key, value)
else:
self._data[key] = value
# Set any get_<field>_display methods
self.__set_field_display()
if self._dynamic:
self._dynamic_lock = False
for key, value in iteritems(dynamic_data):
setattr(self, key, value)
# Flag initialised
self._initialised = True
self._created = _created
signals.post_init.send(self.__class__, document=self)
def __init__(self, *args, **values):
"""
Initialise a document or embedded document
:param __auto_convert: Try and will cast python objects to Object types
:param values: A dictionary of values for the document
"""
self._initialised = False
self._created = True
if args:
# Combine positional arguments with named arguments.
# We only want named arguments.
field = iter(self._fields_ordered)
# If its an automatic id field then skip to the first defined field
if getattr(self, '_auto_id_field', False):
next(field)
for value in args:
name = next(field)
if name in values:
raise TypeError(
'Multiple values for keyword argument "%s"' % name)
values[name] = value
__auto_convert = values.pop('__auto_convert', True)
# 399: set default values only to fields loaded from DB
__only_fields = set(values.pop('__only_fields', values))
_created = values.pop('_created', True)
signals.pre_init.send(self.__class__, document=self, values=values)
# Check if there are undefined fields supplied to the constructor,
# if so raise an Exception.
if not self._dynamic and (self._meta.get('strict', True) or _created):
_undefined_fields = set(
values.keys()) - set(self._fields.keys() +
['id', 'pk', '_cls', '_text_score'])
if _undefined_fields:
msg = ('The fields "{0}" do not exist on the document "{1}"'
).format(_undefined_fields, self._class_name)
raise FieldDoesNotExist(msg)
if self.STRICT and not self._dynamic:
self._data = StrictDict.create(allowed_keys=self._fields_ordered)()
else:
self._data = {}
self._dynamic_fields = SON()
# Assign default values to instance
for key, field in self._fields.iteritems():
if self._db_field_map.get(key, key) in __only_fields:
continue
value = getattr(self, key, None)
setattr(self, key, value)
if '_cls' not in values:
self._cls = self._class_name
# Set passed values after initialisation
if self._dynamic:
dynamic_data = {}
for key, value in values.iteritems():
if key in self._fields or key == '_id':
setattr(self, key, value)
else:
dynamic_data[key] = value
else:
FileField = _import_class('FileField')
for key, value in values.iteritems():
key = self._reverse_db_field_map.get(key, key)
if key in self._fields or key in ('id', 'pk', '_cls'):
if __auto_convert and value is not None:
field = self._fields.get(key)
if field and not isinstance(field, FileField):
value = field.to_python(value)
setattr(self, key, value)
else:
self._data[key] = value
# Set any get_<field>_display methods
self.__set_field_display()
if self._dynamic:
self._dynamic_lock = False
for key, value in dynamic_data.iteritems():
setattr(self, key, value)
# Flag initialised
self._initialised = True
self._created = _created
signals.post_init.send(self.__class__, document=self)
def import_from_handle(cls, handle, filename, identifiers, languages, mimetype, source):
'''Import a file using a file-like object. The file is
uploaded to the s3 bucket specified in `dlx.Config`. The
metadata is stored in the database.
All paramaters are required.
Parameters
----------
handle : Any file-like object
filename : str
The destination filename. Files with common identifiers,
languages, and filename are considered versions of each
other
identifiers : list(dlx.file.Identifier)
languages : list(str)
The ISO 639-1 codes of the languages of the content.
Codes will be stored in uppercase.
mimetype : str
Must be a value recognized by s3, otherwise the upload
will fail
source : str
Name of the process that called the import for auditing
Returns
-------
If succesful, the md5 checksum as a hex string (also used as
the database record ID) of the imported file, otherwise
`False`
Raises
------
FileExists : The file is already in the system
FileExistsIdentifierConflict : The file is already in the
system but with different identifiers
FileExistsLanguageConflict : The file is already in the
system different languages
'''
###
if len(identifiers) == 0 or len(languages) == 0:
raise ValueError('Params `identifiers` and `languages` cannot be an empty list')
for idx in identifiers:
if not isinstance(idx, Identifier):
raise TypeError('Identifier must be of type `dlx.file.Identifier`')
for lang in languages:
lang = lang.upper()
if lang.lower() not in ISO6391.codes:
raise ValueError('Invalid ISO 639-1 language code')
###
hasher = hashlib.md5()
while True:
chunk = handle.read(8192)
if chunk:
hasher.update(chunk)
else:
break
size = handle.tell()
if size == 0:
raise Exception('File-like object "{}" has no content'.format(handle))
checksum = hasher.hexdigest()
File._check_file_exists(checksum, identifiers, languages)
handle.seek(0)
###
if S3.upload(handle, checksum, mimetype):
db_result = DB.files.insert_one(SON({
'_id': checksum,
'filename': filename,
'identifiers': [SON({'type': idx.type, 'value': idx.value}) for idx in identifiers],
'languages': languages,
'mimetype': mimetype,
'size': size,
'source': source,
'timestamp': datetime.now(timezone.utc),
'uri': '{}.s3.amazonaws.com/{}'.format(S3.bucket, checksum),
}))
if db_result.acknowledged:
return checksum
return False
def to_mongo(self, value):
if isinstance(value, dict):
return value
return SON([('type', self._type), ('coordinates', value)])
def greedy_optimization(outfile,task,image_certificate_file,initial_model,convolve_func,rep_limit, modifier_args,modifier):
conn = pm.Connection(document_class=bson.SON)
db = conn['v1']
opt_fs = gridfs.GridFS(db,'optimized_performance')
image_coll = db['raw_images.files']
image_fs = gridfs.GridFS(db,'raw_images')
image_certdict = cPickle.load(open(image_certificate_file))
print('using image certificate', image_certificate_file)
image_hash = image_certdict['run_hash']
image_args = image_certdict['out_args']
if convolve_func == v1f.v1like_filter_pyfft:
v1_pyfft.setup_pyfft()
filterbanks = []
perfs = []
model_configs = []
center_config = initial_model
i = 0
improving = True
while ((i < rep_limit) or rep_limit is None):
i += 1
print('Round', i)
next_configs = [m for m in get_consistent_deltas(center_config,modifier) if m not in model_configs]
if next_configs:
next_results = [get_performance(task,image_hash,image_fs,m,convolve_func) for m in next_configs]
next_perfs = [x[0] for x in next_results]
next_filterbanks = [x[1] for x in next_results]
next_perf_ac_max = np.array([x['test_accuracy'] for x in next_perfs]).max()
perf_ac_max = max([x['test_accuracy'] for x in perfs]) if perfs else 0
if next_perf_ac_max > perf_ac_max:
next_perf_ac_argmax = np.array([x['test_accuracy'] for x in next_perfs]).argmax()
center_config = next_configs[next_perf_ac_argmax]
print('\n\n')
print('new best performance is', next_perf_ac_max, 'from model', center_config)
print('\n\n')
perfs.extend(next_perfs)
model_configs.extend(next_configs)
filterbanks.extend(next_filterbanks)
else:
print('Breaking because no further optimization could be done. Best existing performance was', perf_ac_max, 'while best next performance was', next_perf_ac_max)
break
else:
print('Breaking because no next configs')
break
perfargmax = np.array([p['test_accuracy'] for p in perfs]).argmax()
best_model = model_configs[perfargmax]
best_performance = perfs[perfargmax]
out_record = SON([('initial_model',initial_model),
('task',son_escape(task)),
('images',son_escape(image_args)),
('images_hash',image_hash),
('modifier_args',son_escape(modifier_args)),
('modifier',modifier.__class__.__module__ + '.' + modifier.__class__.__name__)
])
filename = get_filename(out_record)
out_record['filename'] = filename
out_record.update(SON([('performances',perfs)]))
out_record.update(SON([('best_model',best_model)]))
out_record.update(SON([('best_performance',best_performance)]))
out_record.update(SON([('num_steps',len(model_configs))]))
out_record.update(SON([('models',model_configs)]))
outdata = cPickle.dumps(filterbanks)
opt_fs.put(outdata,**out_record)
if convolve_func == v1f.v1like_filter_pyfft:
v1_pyfft.cleanup_pyfft()
createCertificateDict(outfile,{'image_file':image_certificate_file})
def distinct_cmd(self, key, query=None):
cmd = SON([('distinct', self._mongo_collection.name)])
cmd.update({'key': key})
cmd.update({'query': key}) if query else None
return self.db_command(cmd)['values']
def set_fail_point(self, command_args):
cmd = SON([("configureFailPoint", "failCommand")])
cmd.update(command_args)
self.client.admin.command(cmd)
async def use_count_command(self):
response = await self.db.command(SON([("count", "img_url")]))
print(f'response:{pprint.pformat(response)}')
def userRegistration(collection):
dataset = []
id = ''
cursor = collection.aggregate([{
'$unwind': '$products'
}, {
'$match': {
'products.project.registration_date': {
'$gte': dateutil.parser.parse('2017-09-01T00:00:00.000Z')
}
}
}, {
'$project': {
'email': 1,
'products': 1,
'pay_total': 1,
'contacts': 1,
'__v': 1
}
}, {
'$sort':
SON([('products.project.registration_date', 1)])
}])
for user in cursor:
products = user['products']
project = products['project']
company = project.get('company', '')
last_utm_source = ''
last_utm_medium = ''
last_utm_campaign = ''
sector = ''
productName = 'product_one' if (products['_id']
== id) else 'product_two'
try:
utm = map(mapUtm, project["utm_sources"])
list_sources = ','.join(utm)
except AttributeError:
print('Utm is empty')
if ('last_mark' in project):
if ('utm' in project['last_mark']):
lastUtm = utmParse(project['last_mark']['utm'], True)
last_utm_source = lastUtm['utm_source'] if (
type(lastUtm) == dict and 'utm_source' in lastUtm) else ''
last_utm_medium = lastUtm['utm_medium'] if (
type(lastUtm) == dict and 'utm_medium' in lastUtm) else ''
last_utm_campaign = lastUtm['utm_campaign'] if (
type(lastUtm) == dict
and 'utm_campaign' in lastUtm) else ''
if ('sector' in project):
if ('name' in project['sector']):
sector = project['sector']['name']
dataset.append({
'registration_date': project['registration_date'],
'product': productName,
'pay_total': project['pay_total'],
'name': project['name'],
'phone': project['phone'],
'company': company,
'email': user['email'],
'sector': sector,
'last_click_source': last_utm_source,
'last_click_channel': last_utm_medium,
'last_click_campaign': last_utm_campaign,
'multi_channeel': list_sources
})
else:
return dataset
def query(_doc_cls=None, **kwargs):
"""Transform a query from Django-style format to Mongo format."""
mongo_query = {}
merge_query = defaultdict(list)
for key, value in sorted(kwargs.items()):
if key == '__raw__':
mongo_query.update(value)
continue
parts = key.rsplit('__')
indices = [(i, p) for i, p in enumerate(parts) if p.isdigit()]
parts = [part for part in parts if not part.isdigit()]
# Check for an operator and transform to mongo-style if there is
op = None
if len(parts) > 1 and parts[-1] in MATCH_OPERATORS:
op = parts.pop()
# Allow to escape operator-like field name by __
if len(parts) > 1 and parts[-1] == '':
parts.pop()
negate = False
if len(parts) > 1 and parts[-1] == 'not':
parts.pop()
negate = True
if _doc_cls:
# Switch field names to proper names [set in Field(name='foo')]
try:
fields = _doc_cls._lookup_field(parts)
except Exception as e:
raise InvalidQueryError(e)
parts = []
CachedReferenceField = _import_class('CachedReferenceField')
GenericReferenceField = _import_class('GenericReferenceField')
cleaned_fields = []
for field in fields:
append_field = True
if isinstance(field, six.string_types):
parts.append(field)
append_field = False
# is last and CachedReferenceField
elif isinstance(field,
CachedReferenceField) and fields[-1] == field:
parts.append('%s._id' % field.db_field)
else:
parts.append(field.db_field)
if append_field:
cleaned_fields.append(field)
# Convert value to proper value
field = cleaned_fields[-1]
singular_ops = [None, 'ne', 'gt', 'gte', 'lt', 'lte', 'not']
singular_ops += STRING_OPERATORS
if op in singular_ops:
if isinstance(field, six.string_types):
if (op in STRING_OPERATORS
and isinstance(value, six.string_types)):
StringField = _import_class('StringField')
value = StringField.prepare_query_value(op, value)
else:
value = field
else:
value = field.prepare_query_value(op, value)
if isinstance(field, CachedReferenceField) and value:
value = value['_id']
elif op in ('in', 'nin', 'all',
'near') and not isinstance(value, dict):
# Raise an error if the in/nin/all/near param is not iterable.
value = _prepare_query_for_iterable(field, op, value)
# If we're querying a GenericReferenceField, we need to alter the
# key depending on the value:
# * If the value is a DBRef, the key should be "field_name._ref".
# * If the value is an ObjectId, the key should be "field_name._ref.$id".
if isinstance(field, GenericReferenceField):
if isinstance(value, DBRef):
parts[-1] += '._ref'
elif isinstance(value, ObjectId):
parts[-1] += '._ref.$id'
# if op and op not in COMPARISON_OPERATORS:
if op:
if op in GEO_OPERATORS:
value = _geo_operator(field, op, value)
elif op in ('match', 'elemMatch'):
ListField = _import_class('ListField')
EmbeddedDocumentField = _import_class('EmbeddedDocumentField')
if (isinstance(value, dict) and isinstance(field, ListField)
and isinstance(field.field, EmbeddedDocumentField)):
value = query(field.field.document_type, **value)
else:
value = field.prepare_query_value(op, value)
value = {'$elemMatch': value}
elif op in CUSTOM_OPERATORS:
NotImplementedError('Custom method "%s" has not '
'been implemented' % op)
elif op not in STRING_OPERATORS:
value = {'$' + op: value}
if negate:
value = {'$not': value}
for i, part in indices:
parts.insert(i, part)
key = '.'.join(parts)
if op is None or key not in mongo_query:
mongo_query[key] = value
elif key in mongo_query:
if isinstance(mongo_query[key], dict):
mongo_query[key].update(value)
# $max/minDistance needs to come last - convert to SON
value_dict = mongo_query[key]
if ('$maxDistance' in value_dict or '$minDistance' in value_dict) and \
('$near' in value_dict or '$nearSphere' in value_dict):
value_son = SON()
for k, v in value_dict.iteritems():
if k == '$maxDistance' or k == '$minDistance':
continue
value_son[k] = v
# Required for MongoDB >= 2.6, may fail when combining
# PyMongo 3+ and MongoDB < 2.6
near_embedded = False
for near_op in ('$near', '$nearSphere'):
if isinstance(value_dict.get(near_op), dict) and (
IS_PYMONGO_3
or get_connection().max_wire_version > 1):
value_son[near_op] = SON(value_son[near_op])
if '$maxDistance' in value_dict:
value_son[near_op][
'$maxDistance'] = value_dict[
'$maxDistance']
if '$minDistance' in value_dict:
value_son[near_op][
'$minDistance'] = value_dict[
'$minDistance']
near_embedded = True
if not near_embedded:
if '$maxDistance' in value_dict:
value_son['$maxDistance'] = value_dict[
'$maxDistance']
if '$minDistance' in value_dict:
value_son['$minDistance'] = value_dict[
'$minDistance']
mongo_query[key] = value_son
else:
# Store for manually merging later
merge_query[key].append(value)
# The queryset has been filter in such a way we must manually merge
for k, v in merge_query.items():
merge_query[k].append(mongo_query[k])
del mongo_query[k]
if isinstance(v, list):
value = [{k: val} for val in v]
if '$and' in mongo_query.keys():
mongo_query['$and'].extend(value)
else:
mongo_query['$and'] = value
return mongo_query
def from_json(cls, json_str, created=False, *args, **kwargs):
# Proposition: add a private method like from_json that allows
# dictionaries to be used as inputs to avoid having to use
# loads(dumps(data)) all the time
"""
Decode from human-readable json.
Parameters:
json_str: JSON string that should be passed to the serialized
created: a parameter that is passed to cls._from_son.
*args, **kwargs: Any additional arguments that is passed to
json.loads.
"""
from .fields import FollowReferenceField
kwargs.setdefault("object_hook", generate_object_hook(cls))
dct = json.loads(json_str, *args, **kwargs)
for name, fld in cls._fields.items():
if any([
getattr(fld, "exclude_from_json", None),
getattr(fld, "exclude_json", None)
]):
dct.pop(name, None)
from_son_result = cls._from_son(SON(dct), created=created)
atLeastOneReference = False
for fldname, fld in cls._fields.items():
if isinstance(fld, db.ListField):
target = fld.field
if not isinstance(target, db.ReferenceField) or \
isinstance(target, FollowReferenceField):
continue
atLeastOneReference = True
values = dct.get(fldname)
setattr(from_son_result, fldname, [])
for value in values:
valueDoc = value.as_doc()
if 'id' not in valueDoc['$id']:
valueDoc['$id']['id'] = str(ObjectId())
getattr(from_son_result, fldname).append(
target.document_type_obj.from_json(
json.dumps(value.as_doc()['$id'])))
elif isinstance(fld, db.DictField):
target = fld.field
if not isinstance(target, db.ReferenceField) or \
isinstance(target, FollowReferenceField):
continue
atLeastOneReference = True
values = dct.get(fldname)
setattr(from_son_result, fldname, {})
for k, value in values.items():
valueDoc = value.as_doc()
if 'id' not in valueDoc['$id']:
valueDoc['$id']['id'] = str(ObjectId())
getattr(from_son_result, fldname)[k] = \
target.document_type_obj.from_json(
json.dumps(valueDoc['$id']))
else:
target = fld
if not isinstance(target, db.ReferenceField) or \
isinstance(target, FollowReferenceField):
continue
atLeastOneReference = True
value = dct.get(fldname)
try:
valueDoc = value.as_doc()
# If there is no ID in the JSON (aka the JSON was not saved
# from mongoengine but rather created manually), create an
# ObjectId on the fly
if 'id' not in valueDoc['$id']:
valueDoc['$id']['id'] = ObjectId()
valueDoc['$id']['id'] = str(valueDoc['$id']['id'])
setattr(
from_son_result, fldname,
target.document_type_obj.from_json(
json.dumps(valueDoc['$id'])))
except TypeError:
setattr(
from_son_result, fldname,
normalize_reference(getattr(value, "id", value),
target))
# All fields have been changed, because the document was loaded from a
# JSON file. However, mongoengine does not detect it automatically. In
# order for all fields to be saved, we set the _changed_fields variable
# manually
from_son_result._changed_fields = list(cls._fields.keys())
if atLeastOneReference:
# If the document contains at least one reference, override the
# save() method to save the referenced documents at the same time
# as the master document. Otherwise, the referenced documents would
# not be saved and the document would not be valid anymore after a
# save and load from the database.
def save(self, *args, **kwargs):
for fldname, fld in cls._fields.items():
if isinstance(fld,
(db.ReferenceField, FollowReferenceField)):
getattr(self, fldname).save(*args, **kwargs)
elif isinstance(fld, db.fields.ComplexBaseField):
isReferences = isinstance(
fld.field,
(db.ReferenceField, FollowReferenceField))
if isinstance(fld, db.DictField) and isReferences:
field = getattr(self, fldname)
for key, value in field.items():
field[key].save(*args, **kwargs)
elif isinstance(fld, db.ListField) and isReferences:
field = getattr(self, fldname)
for valueIndex in range(len(field)):
field[valueIndex].save(*args, **kwargs)
super(self.__class__, self).save(*args, **kwargs)
from_son_result.save = types.MethodType(save, from_son_result)
return from_son_result
def select():
# 查询 文档中的第一条数据
print(site.find_one())
# 查询集合中所有数据
for s in site.find():
print(s)
# where
output(site.find_one({'_id': ObjectId('5cb73399efdf11190e096721')}))
# 嵌入doc
cursor = inventory.find(
{"size": SON([("h", 14), ("w", 21), ("uom", "cm")])})
inventory.find({"size.uom": "in"})
# and
output(site.find_one({
'name': 'RUNOOB',
'alexa': '100003',
}))
query = {"name": {"$lt": "H"}}
mydoc = site.find(query)
for x in mydoc:
print(x)
# 字段限制 or select xx .limit(1).skip(2) => limit(2,1)
query = {"$or": [{'alexa': '101'}, {'alexa': '12345'}]}
output(site.find(query, {'_id': 0}).limit(1).skip(2))
# order by 1 为升序排列，而 -1 是用于降序排列。
for x in site.find().sort([('alexa', pymongo.DESCENDING),
('name', pymongo.ASCENDING)]):
print(x)
# in
query = {'name': {'$in': ['Facebook', 'Taobao']}}
for x in site.find(query):
print(x)
# array
# 严格 有 顺序一样
inventory.find({"tags": ["red", "blank"]})
# 包含这两个元素 不考虑数组中的顺序或其他元素
inventory.find({"tags": {"$all": ["red", "blank"]}})
# 包含
inventory.find({"tags": "red"})
# 数组位置2
cursor = db.inventory.find({"dim_cm.1": {"$gt": 25}})
# 长度
cursor = db.inventory.find({"tags": {"$size": 3}})
# 数组嵌入
cursor = db.inventory.find({'instock.0.qty': {"$lte": 20}})
cursor = db.inventory.find({"instock.qty": {"$gt": 10, "$lte": 20}})
# 聚合
query = [
# {'$match': {'a': 1}},
{
'$group': {
'_id': "$name",
'a': {
'$sum': 1
},
'b': {
'$max': '$alexa'
},
}
},
{
'$sort': {
'a': pymongo.ASCENDING
}
},
{
'$sort': {
'b': pymongo.DESCENDING
}
},
{
'$limit': 20
},
{
'$match': {
'a': 1
}
},
]
# select _id, sum(*) as a, max(alexa) as b from xx group by name as _id having a=1 order by a asce, b desc
for x in site.aggregate(query):
print(x)
"$match": {
"when": {
"$gte": "$from",
"$lte": "$to"
},
"workflow.state": "$state"
}
}, {
"$group": {
"_id": "$type",
"count": {
"$sum": 1
}
}
}, {
"$sort": SON([("count", -1)])
}]
}
}
}
aggregate_types_discipline = {
'item_title': 'Observation Aggregations by discipline and types',
'url': '{}/aggregate/discipline'.format(BASE_URL),
'datasource': {
'source': RESOURCE_COLLECTION,
'aggregation': {
'pipeline': [{
"$unwind": "$type"
}, {
"$match": {
# costs.append(neighbor['properties']['price'])
# median_costs[doc['properties']['listing_id']]= stats(costs)
for doc in collection.find():
i = i + 1
print i
costs = []
for neighbor in collection.find({
"properties.bedrooms":
doc['properties']['bedrooms'],
"properties.bathrooms":
doc['properties']['bathrooms'],
"geometry":
SON([("$near", {
"$geometry":
SON([("type", "Point"),
("coordinates", doc['geometry']['coordinates'])])
})])
}).limit(30):
costs.append(neighbor['properties']['price'])
median_costs[doc['properties']['listing_id']] = stats(
costs, doc['properties']['price'])
t2 = time()
print("Run Time:", t2 - t1)
neighborhood_values = pd.DataFrame.from_dict(median_costs, orient='index')
neighborhood_values.columns = [
'0_per_30', '10_per_30', '20_per_30', '30_per_30', '40_per_30',
'50_per_30', '60_per_30', '70_per_30', '80_per_30', '90_per_30',
'100_per_30', 'median_30', 'mean_30'
def __init__(self, oscontrol, domain, operation, async):
SON.__init__(self)
self['$type'] = operation
self['_domain'] = domain
self['_asyncrequest'] = async
class BaseDocument(object):
__slots__ = ('_changed_fields', '_initialised', '_created', '_data',
'_dynamic_fields', '_auto_id_field', '_db_field_map',
'__weakref__')
_dynamic = False
_dynamic_lock = True
STRICT = False
def __init__(self, *args, **values):
"""
Initialise a document or embedded document
:param __auto_convert: Try and will cast python objects to Object types
:param values: A dictionary of values for the document
"""
self._initialised = False
self._created = True
if args:
# Combine positional arguments with named arguments.
# We only want named arguments.
field = iter(self._fields_ordered)
# If its an automatic id field then skip to the first defined field
if getattr(self, '_auto_id_field', False):
next(field)
for value in args:
name = next(field)
if name in values:
raise TypeError(
'Multiple values for keyword argument "%s"' % name)
values[name] = value
__auto_convert = values.pop('__auto_convert', True)
# 399: set default values only to fields loaded from DB
__only_fields = set(values.pop('__only_fields', values))
_created = values.pop('_created', True)
signals.pre_init.send(self.__class__, document=self, values=values)
# Check if there are undefined fields supplied to the constructor,
# if so raise an Exception.
if not self._dynamic and (self._meta.get('strict', True) or _created):
_undefined_fields = set(
values.keys()) - set(self._fields.keys() +
['id', 'pk', '_cls', '_text_score'])
if _undefined_fields:
msg = ('The fields "{0}" do not exist on the document "{1}"'
).format(_undefined_fields, self._class_name)
raise FieldDoesNotExist(msg)
if self.STRICT and not self._dynamic:
self._data = StrictDict.create(allowed_keys=self._fields_ordered)()
else:
self._data = {}
self._dynamic_fields = SON()
# Assign default values to instance
for key, field in self._fields.iteritems():
if self._db_field_map.get(key, key) in __only_fields:
continue
value = getattr(self, key, None)
setattr(self, key, value)
if '_cls' not in values:
self._cls = self._class_name
# Set passed values after initialisation
if self._dynamic:
dynamic_data = {}
for key, value in values.iteritems():
if key in self._fields or key == '_id':
setattr(self, key, value)
else:
dynamic_data[key] = value
else:
FileField = _import_class('FileField')
for key, value in values.iteritems():
key = self._reverse_db_field_map.get(key, key)
if key in self._fields or key in ('id', 'pk', '_cls'):
if __auto_convert and value is not None:
field = self._fields.get(key)
if field and not isinstance(field, FileField):
value = field.to_python(value)
setattr(self, key, value)
else:
self._data[key] = value
# Set any get_<field>_display methods
self.__set_field_display()
if self._dynamic:
self._dynamic_lock = False
for key, value in dynamic_data.iteritems():
setattr(self, key, value)
# Flag initialised
self._initialised = True
self._created = _created
signals.post_init.send(self.__class__, document=self)
def __delattr__(self, *args, **kwargs):
"""Handle deletions of fields"""
field_name = args[0]
if field_name in self._fields:
default = self._fields[field_name].default
if callable(default):
default = default()
setattr(self, field_name, default)
else:
super(BaseDocument, self).__delattr__(*args, **kwargs)
def __setattr__(self, name, value):
# Handle dynamic data only if an initialised dynamic document
if self._dynamic and not self._dynamic_lock:
if not hasattr(self, name) and not name.startswith('_'):
DynamicField = _import_class('DynamicField')
field = DynamicField(db_field=name, null=True)
field.name = name
self._dynamic_fields[name] = field
self._fields_ordered += (name, )
if not name.startswith('_'):
value = self.__expand_dynamic_values(name, value)
# Handle marking data as changed
if name in self._dynamic_fields:
self._data[name] = value
if hasattr(self, '_changed_fields'):
self._mark_as_changed(name)
try:
self__created = self._created
except AttributeError:
self__created = True
if (self._is_document and not self__created
and name in self._meta.get('shard_key', tuple())
and self._data.get(name) != value):
msg = 'Shard Keys are immutable. Tried to update %s' % name
raise OperationError(msg)
try:
self__initialised = self._initialised
except AttributeError:
self__initialised = False
# Check if the user has created a new instance of a class
if (self._is_document and self__initialised and self__created
and name == self._meta.get('id_field')):
super(BaseDocument, self).__setattr__('_created', False)
super(BaseDocument, self).__setattr__(name, value)
def __getstate__(self):
data = {}
for k in ('_changed_fields', '_initialised', '_created',
'_dynamic_fields', '_fields_ordered'):
if hasattr(self, k):
data[k] = getattr(self, k)
data['_data'] = self.to_mongo()
return data
def __setstate__(self, data):
if isinstance(data['_data'], SON):
data['_data'] = self.__class__._from_son(data['_data'])._data
for k in ('_changed_fields', '_initialised', '_created', '_data',
'_dynamic_fields'):
if k in data:
setattr(self, k, data[k])
if '_fields_ordered' in data:
if self._dynamic:
setattr(self, '_fields_ordered', data['_fields_ordered'])
else:
_super_fields_ordered = type(self)._fields_ordered
setattr(self, '_fields_ordered', _super_fields_ordered)
dynamic_fields = data.get('_dynamic_fields') or SON()
for k in dynamic_fields.keys():
setattr(self, k, data['_data'].get(k))
def __iter__(self):
return iter(self._fields_ordered)
def __getitem__(self, name):
"""Dictionary-style field access, return a field's value if present.
"""
try:
if name in self._fields_ordered:
return getattr(self, name)
except AttributeError:
pass
raise KeyError(name)
def __setitem__(self, name, value):
"""Dictionary-style field access, set a field's value.
"""
# Ensure that the field exists before settings its value
if not self._dynamic and name not in self._fields:
raise KeyError(name)
return setattr(self, name, value)
def __contains__(self, name):
try:
val = getattr(self, name)
return val is not None
except AttributeError:
return False
def __len__(self):
return len(self._data)
def __repr__(self):
try:
u = self.__str__()
except (UnicodeEncodeError, UnicodeDecodeError):
u = '[Bad Unicode data]'
repr_type = str if u is None else type(u)
return repr_type('<%s: %s>' % (self.__class__.__name__, u))
def __str__(self):
# TODO this could be simpler?
if hasattr(self, '__unicode__'):
if six.PY3:
return self.__unicode__()
else:
return six.text_type(self).encode('utf-8')
return six.text_type('%s object' % self.__class__.__name__)
def __eq__(self, other):
if isinstance(other, self.__class__) and hasattr(
other, 'id') and other.id is not None:
return self.id == other.id
if isinstance(other, DBRef):
return self._get_collection_name(
) == other.collection and self.id == other.id
if self.id is None:
return self is other
return False
def __ne__(self, other):
return not self.__eq__(other)
def clean(self):
"""
Hook for doing document level data cleaning before validation is run.
Any ValidationError raised by this method will not be associated with
a particular field; it will have a special-case association with the
field defined by NON_FIELD_ERRORS.
"""
pass
def get_text_score(self):
"""
Get text score from text query
"""
if '_text_score' not in self._data:
raise InvalidDocumentError(
'This document is not originally built from a text query')
return self._data['_text_score']
def to_mongo(self, use_db_field=True, fields=None):
"""
Return as SON data ready for use with MongoDB.
"""
if not fields:
fields = []
data = SON()
data['_id'] = None
data['_cls'] = self._class_name
# only root fields ['test1.a', 'test2'] => ['test1', 'test2']
root_fields = {f.split('.')[0] for f in fields}
for field_name in self:
if root_fields and field_name not in root_fields:
continue
value = self._data.get(field_name, None)
field = self._fields.get(field_name)
if field is None and self._dynamic:
field = self._dynamic_fields.get(field_name)
if value is not None:
f_inputs = field.to_mongo.__code__.co_varnames
ex_vars = {}
if fields and 'fields' in f_inputs:
key = '%s.' % field_name
embedded_fields = [
i.replace(key, '') for i in fields if i.startswith(key)
]
ex_vars['fields'] = embedded_fields
if 'use_db_field' in f_inputs:
ex_vars['use_db_field'] = use_db_field
value = field.to_mongo(value, **ex_vars)
# Handle self generating fields
if value is None and field._auto_gen:
value = field.generate()
self._data[field_name] = value
if (value is not None) or (field.null):
if use_db_field:
data[field.db_field] = value
else:
data[field.name] = value
# Only add _cls if allow_inheritance is True
if not self._meta.get('allow_inheritance'):
data.pop('_cls')
return data
def validate(self, clean=True):
"""Ensure that all fields' values are valid and that required fields
are present.
"""
# Ensure that each field is matched to a valid value
errors = {}
if clean:
try:
self.clean()
except ValidationError as error:
errors[NON_FIELD_ERRORS] = error
# Get a list of tuples of field names and their current values
fields = [(self._fields.get(name, self._dynamic_fields.get(name)),
self._data.get(name)) for name in self._fields_ordered]
EmbeddedDocumentField = _import_class('EmbeddedDocumentField')
GenericEmbeddedDocumentField = _import_class(
'GenericEmbeddedDocumentField')
for field, value in fields:
if value is not None:
try:
if isinstance(
field,
(EmbeddedDocumentField, GenericEmbeddedDocumentField)):
field._validate(value, clean=clean)
else:
field._validate(value)
except ValidationError as error:
errors[field.name] = error.errors or error
except (ValueError, AttributeError, AssertionError) as error:
errors[field.name] = error
elif field.required and not getattr(field, '_auto_gen', False):
errors[field.name] = ValidationError('Field is required',
field_name=field.name)
if errors:
pk = 'None'
if hasattr(self, 'pk'):
pk = self.pk
elif self._instance and hasattr(self._instance, 'pk'):
pk = self._instance.pk
message = 'ValidationError (%s:%s) ' % (self._class_name, pk)
raise ValidationError(message, errors=errors)
def to_json(self, *args, **kwargs):
"""Convert this document to JSON.
:param use_db_field: Serialize field names as they appear in
MongoDB (as opposed to attribute names on this document).
Defaults to True.
"""
use_db_field = kwargs.pop('use_db_field', True)
return json_util.dumps(self.to_mongo(use_db_field), *args, **kwargs)
@classmethod
def from_json(cls, json_data, created=False):
"""Converts json data to a Document instance
:param json_data: The json data to load into the Document
:param created: If True, the document will be considered as a brand new document
If False and an id is provided, it will consider that the data being
loaded corresponds to what's already in the database (This has an impact of subsequent call to .save())
If False and no id is provided, it will consider the data as a new document
(default ``False``)
"""
return cls._from_son(json_util.loads(json_data), created=created)
def __expand_dynamic_values(self, name, value):
"""Expand any dynamic values to their correct types / values."""
if not isinstance(value, (dict, list, tuple)):
return value
# If the value is a dict with '_cls' in it, turn it into a document
is_dict = isinstance(value, dict)
if is_dict and '_cls' in value:
cls = get_document(value['_cls'])
return cls(**value)
if is_dict:
value = {
k: self.__expand_dynamic_values(k, v)
for k, v in value.items()
}
else:
value = [self.__expand_dynamic_values(name, v) for v in value]
# Convert lists / values so we can watch for any changes on them
EmbeddedDocumentListField = _import_class('EmbeddedDocumentListField')
if (isinstance(value, (list, tuple))
and not isinstance(value, BaseList)):
if issubclass(type(self), EmbeddedDocumentListField):
value = EmbeddedDocumentList(value, self, name)
else:
value = BaseList(value, self, name)
elif isinstance(value, dict) and not isinstance(value, BaseDict):
value = BaseDict(value, self, name)
return value
def _mark_as_changed(self, key):
"""Mark a key as explicitly changed by the user."""
if not key:
return
if not hasattr(self, '_changed_fields'):
return
if '.' in key:
key, rest = key.split('.', 1)
key = self._db_field_map.get(key, key)
key = '%s.%s' % (key, rest)
else:
key = self._db_field_map.get(key, key)
if key not in self._changed_fields:
levels, idx = key.split('.'), 1
while idx <= len(levels):
if '.'.join(levels[:idx]) in self._changed_fields:
break
idx += 1
else:
self._changed_fields.append(key)
# remove lower level changed fields
level = '.'.join(levels[:idx]) + '.'
remove = self._changed_fields.remove
for field in self._changed_fields[:]:
if field.startswith(level):
remove(field)
def _clear_changed_fields(self):
"""Using _get_changed_fields iterate and remove any fields that
are marked as changed.
"""
for changed in self._get_changed_fields():
parts = changed.split('.')
data = self
for part in parts:
if isinstance(data, list):
try:
data = data[int(part)]
except IndexError:
data = None
elif isinstance(data, dict):
data = data.get(part, None)
else:
data = getattr(data, part, None)
if not isinstance(data, LazyReference) and hasattr(
data, '_changed_fields'):
if getattr(data, '_is_document', False):
continue
data._changed_fields = []
self._changed_fields = []
def _nestable_types_changed_fields(self, changed_fields, base_key, data):
"""Inspect nested data for changed fields
:param changed_fields: Previously collected changed fields
:param base_key: The base key that must be used to prepend changes to this data
:param data: data to inspect for changes
"""
# Loop list / dict fields as they contain documents
# Determine the iterator to use
if not hasattr(data, 'items'):
iterator = enumerate(data)
else:
iterator = data.iteritems()
for index_or_key, value in iterator:
item_key = '%s%s.' % (base_key, index_or_key)
# don't check anything lower if this key is already marked
# as changed.
if item_key[:-1] in changed_fields:
continue
if hasattr(value, '_get_changed_fields'):
changed = value._get_changed_fields()
changed_fields += [
'%s%s' % (item_key, k) for k in changed if k
]
elif isinstance(value, (list, tuple, dict)):
self._nestable_types_changed_fields(changed_fields, item_key,
value)
def _get_changed_fields(self):
"""Return a list of all fields that have explicitly been changed.
"""
EmbeddedDocument = _import_class('EmbeddedDocument')
ReferenceField = _import_class('ReferenceField')
GenericReferenceField = _import_class('GenericReferenceField')
SortedListField = _import_class('SortedListField')
changed_fields = []
changed_fields += getattr(self, '_changed_fields', [])
for field_name in self._fields_ordered:
db_field_name = self._db_field_map.get(field_name, field_name)
key = '%s.' % db_field_name
data = self._data.get(field_name, None)
field = self._fields.get(field_name)
if db_field_name in changed_fields:
# Whole field already marked as changed, no need to go further
continue
if isinstance(field,
ReferenceField): # Don't follow referenced documents
continue
if isinstance(data, EmbeddedDocument):
# Find all embedded fields that have been changed
changed = data._get_changed_fields()
changed_fields += ['%s%s' % (key, k) for k in changed if k]
elif isinstance(data, (list, tuple, dict)):
if (hasattr(field, 'field') and isinstance(
field.field, (ReferenceField, GenericReferenceField))):
continue
elif isinstance(field, SortedListField) and field._ordering:
# if ordering is affected whole list is changed
if any(field._ordering in d._changed_fields for d in data):
changed_fields.append(db_field_name)
continue
self._nestable_types_changed_fields(changed_fields, key, data)
return changed_fields
def _delta(self):
"""Returns the delta (set, unset) of the changes for a document.
Gets any values that have been explicitly changed.
"""
# Handles cases where not loaded from_son but has _id
doc = self.to_mongo()
set_fields = self._get_changed_fields()
unset_data = {}
if hasattr(self, '_changed_fields'):
set_data = {}
# Fetch each set item from its path
for path in set_fields:
parts = path.split('.')
d = doc
new_path = []
for p in parts:
if isinstance(d, (ObjectId, DBRef)):
# Don't dig in the references
break
elif isinstance(d, list) and p.isdigit():
# An item of a list (identified by its index) is updated
d = d[int(p)]
elif hasattr(d, 'get'):
# dict-like (dict, embedded document)
d = d.get(p)
new_path.append(p)
path = '.'.join(new_path)
set_data[path] = d
else:
set_data = doc
if '_id' in set_data:
del set_data['_id']
# Determine if any changed items were actually unset.
for path, value in set_data.items():
if value or isinstance(
value, (numbers.Number,
bool)): # Account for 0 and True that are truthy
continue
parts = path.split('.')
if (self._dynamic and len(parts)
and parts[0] in self._dynamic_fields):
del set_data[path]
unset_data[path] = 1
continue
# If we've set a value that ain't the default value don't unset it.
default = None
if path in self._fields:
default = self._fields[path].default
else: # Perform a full lookup for lists / embedded lookups
d = self
db_field_name = parts.pop()
for p in parts:
if isinstance(d, list) and p.isdigit():
d = d[int(p)]
elif (hasattr(d, '__getattribute__')
and not isinstance(d, dict)):
real_path = d._reverse_db_field_map.get(p, p)
d = getattr(d, real_path)
else:
d = d.get(p)
if hasattr(d, '_fields'):
field_name = d._reverse_db_field_map.get(
db_field_name, db_field_name)
if field_name in d._fields:
default = d._fields.get(field_name).default
else:
default = None
if default is not None:
default = default() if callable(default) else default
if value != default:
continue
del set_data[path]
unset_data[path] = 1
return set_data, unset_data
@classmethod
def _get_collection_name(cls):
"""Return the collection name for this class. None for abstract
class.
"""
return cls._meta.get('collection', None)
@classmethod
def _from_son(cls,
son,
_auto_dereference=True,
only_fields=None,
created=False):
"""Create an instance of a Document (subclass) from a PyMongo
SON.
"""
if not only_fields:
only_fields = []
if son and not isinstance(son, dict):
raise ValueError(
"The source SON object needs to be of type 'dict'")
# Get the class name from the document, falling back to the given
# class if unavailable
class_name = son.get('_cls', cls._class_name)
# Convert SON to a data dict, making sure each key is a string and
# corresponds to the right db field.
data = {}
for key, value in son.iteritems():
key = str(key)
key = cls._db_field_map.get(key, key)
data[key] = value
# Return correct subclass for document type
if class_name != cls._class_name:
cls = get_document(class_name)
changed_fields = []
errors_dict = {}
fields = cls._fields
if not _auto_dereference:
fields = copy.deepcopy(fields)
for field_name, field in fields.iteritems():
field._auto_dereference = _auto_dereference
if field.db_field in data:
value = data[field.db_field]
try:
data[field_name] = (value if value is None else
field.to_python(value))
if field_name != field.db_field:
del data[field.db_field]
except (AttributeError, ValueError) as e:
errors_dict[field_name] = e
if errors_dict:
errors = '\n'.join(
['%s - %s' % (k, v) for k, v in errors_dict.items()])
msg = ('Invalid data to create a `%s` instance.\n%s' %
(cls._class_name, errors))
raise InvalidDocumentError(msg)
# In STRICT documents, remove any keys that aren't in cls._fields
if cls.STRICT:
data = {k: v for k, v in data.iteritems() if k in cls._fields}
obj = cls(__auto_convert=False,
_created=created,
__only_fields=only_fields,
**data)
obj._changed_fields = changed_fields
if not _auto_dereference:
obj._fields = fields
return obj
@classmethod
def _build_index_specs(cls, meta_indexes):
"""Generate and merge the full index specs."""
geo_indices = cls._geo_indices()
unique_indices = cls._unique_with_indexes()
index_specs = [cls._build_index_spec(spec) for spec in meta_indexes]
def merge_index_specs(index_specs, indices):
"""Helper method for merging index specs."""
if not indices:
return index_specs
# Create a map of index fields to index spec. We're converting
# the fields from a list to a tuple so that it's hashable.
spec_fields = {
tuple(index['fields']): index
for index in index_specs
}
# For each new index, if there's an existing index with the same
# fields list, update the existing spec with all data from the
# new spec.
for new_index in indices:
candidate = spec_fields.get(tuple(new_index['fields']))
if candidate is None:
index_specs.append(new_index)
else:
candidate.update(new_index)
return index_specs
# Merge geo indexes and unique_with indexes into the meta index specs.
index_specs = merge_index_specs(index_specs, geo_indices)
index_specs = merge_index_specs(index_specs, unique_indices)
return index_specs
@classmethod
def _build_index_spec(cls, spec):
"""Build a PyMongo index spec from a MongoEngine index spec."""
if isinstance(spec, six.string_types):
spec = {'fields': [spec]}
elif isinstance(spec, (list, tuple)):
spec = {'fields': list(spec)}
elif isinstance(spec, dict):
spec = dict(spec)
index_list = []
direction = None
# Check to see if we need to include _cls
allow_inheritance = cls._meta.get('allow_inheritance')
include_cls = (allow_inheritance and not spec.get('sparse', False)
and spec.get('cls', True)
and '_cls' not in spec['fields'])
# 733: don't include cls if index_cls is False unless there is an explicit cls with the index
include_cls = include_cls and (spec.get('cls', False)
or cls._meta.get('index_cls', True))
if 'cls' in spec:
spec.pop('cls')
for key in spec['fields']:
# If inherited spec continue
if isinstance(key, (list, tuple)):
continue
# ASCENDING from +
# DESCENDING from -
# TEXT from $
# HASHED from #
# GEOSPHERE from (
# GEOHAYSTACK from )
# GEO2D from *
direction = pymongo.ASCENDING
if key.startswith('-'):
direction = pymongo.DESCENDING
elif key.startswith('$'):
direction = pymongo.TEXT
elif key.startswith('#'):
direction = pymongo.HASHED
elif key.startswith('('):
direction = pymongo.GEOSPHERE
elif key.startswith(')'):
direction = pymongo.GEOHAYSTACK
elif key.startswith('*'):
direction = pymongo.GEO2D
if key.startswith(('+', '-', '*', '$', '#', '(', ')')):
key = key[1:]
# Use real field name, do it manually because we need field
# objects for the next part (list field checking)
parts = key.split('.')
if parts in (['pk'], ['id'], ['_id']):
key = '_id'
else:
fields = cls._lookup_field(parts)
parts = []
for field in fields:
try:
if field != '_id':
field = field.db_field
except AttributeError:
pass
parts.append(field)
key = '.'.join(parts)
index_list.append((key, direction))
# Don't add cls to a geo index
if include_cls and direction not in (pymongo.GEO2D,
pymongo.GEOHAYSTACK,
pymongo.GEOSPHERE):
index_list.insert(0, ('_cls', 1))
if index_list:
spec['fields'] = index_list
return spec
@classmethod
def _unique_with_indexes(cls, namespace=''):
"""Find unique indexes in the document schema and return them."""
unique_indexes = []
for field_name, field in cls._fields.items():
sparse = field.sparse
# Generate a list of indexes needed by uniqueness constraints
if field.unique:
unique_fields = [field.db_field]
# Add any unique_with fields to the back of the index spec
if field.unique_with:
if isinstance(field.unique_with, six.string_types):
field.unique_with = [field.unique_with]
# Convert unique_with field names to real field names
unique_with = []
for other_name in field.unique_with:
parts = other_name.split('.')
# Lookup real name
parts = cls._lookup_field(parts)
name_parts = [part.db_field for part in parts]
unique_with.append('.'.join(name_parts))
# Unique field should be required
parts[-1].required = True
sparse = (not sparse
and parts[-1].name not in cls.__dict__)
unique_fields += unique_with
# Add the new index to the list
fields = [('%s%s' % (namespace, f), pymongo.ASCENDING)
for f in unique_fields]
index = {'fields': fields, 'unique': True, 'sparse': sparse}
unique_indexes.append(index)
if field.__class__.__name__ == 'ListField':
field = field.field
# Grab any embedded document field unique indexes
if (field.__class__.__name__ == 'EmbeddedDocumentField'
and field.document_type != cls):
field_namespace = '%s.' % field_name
doc_cls = field.document_type
unique_indexes += doc_cls._unique_with_indexes(field_namespace)
return unique_indexes
@classmethod
def _geo_indices(cls, inspected=None, parent_field=None):
inspected = inspected or []
geo_indices = []
inspected.append(cls)
geo_field_type_names = ('EmbeddedDocumentField', 'GeoPointField',
'PointField', 'LineStringField',
'PolygonField')
geo_field_types = tuple(
[_import_class(field) for field in geo_field_type_names])
for field in cls._fields.values():
if not isinstance(field, geo_field_types):
continue
if hasattr(field, 'document_type'):
field_cls = field.document_type
if field_cls in inspected:
continue
if hasattr(field_cls, '_geo_indices'):
geo_indices += field_cls._geo_indices(
inspected, parent_field=field.db_field)
elif field._geo_index:
field_name = field.db_field
if parent_field:
field_name = '%s.%s' % (parent_field, field_name)
geo_indices.append(
{'fields': [(field_name, field._geo_index)]})
return geo_indices
@classmethod
def _lookup_field(cls, parts):
"""Given the path to a given field, return a list containing
the Field object associated with that field and all of its parent
Field objects.
Args:
parts (str, list, or tuple) - path to the field. Should be a
string for simple fields existing on this document or a list
of strings for a field that exists deeper in embedded documents.
Returns:
A list of Field instances for fields that were found or
strings for sub-fields that weren't.
Example:
>>> user._lookup_field('name')
[<mongoengine.fields.StringField at 0x1119bff50>]
>>> user._lookup_field('roles')
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>]
>>> user._lookup_field(['roles', 'role'])
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>,
<mongoengine.fields.StringField at 0x1119ec050>]
>>> user._lookup_field('doesnt_exist')
raises LookUpError
>>> user._lookup_field(['roles', 'doesnt_exist'])
[<mongoengine.fields.EmbeddedDocumentListField at 0x1119ec250>,
'doesnt_exist']
"""
# TODO this method is WAY too complicated. Simplify it.
# TODO don't think returning a string for embedded non-existent fields is desired
ListField = _import_class('ListField')
DynamicField = _import_class('DynamicField')
if not isinstance(parts, (list, tuple)):
parts = [parts]
fields = []
field = None
for field_name in parts:
# Handle ListField indexing:
if field_name.isdigit() and isinstance(field, ListField):
fields.append(field_name)
continue
# Look up first field from the document
if field is None:
if field_name == 'pk':
# Deal with "primary key" alias
field_name = cls._meta['id_field']
if field_name in cls._fields:
field = cls._fields[field_name]
elif cls._dynamic:
field = DynamicField(db_field=field_name)
elif cls._meta.get('allow_inheritance') or cls._meta.get(
'abstract', False):
# 744: in case the field is defined in a subclass
for subcls in cls.__subclasses__():
try:
field = subcls._lookup_field([field_name])[0]
except LookUpError:
continue
if field is not None:
break
else:
raise LookUpError('Cannot resolve field "%s"' %
field_name)
else:
raise LookUpError('Cannot resolve field "%s"' % field_name)
else:
ReferenceField = _import_class('ReferenceField')
GenericReferenceField = _import_class('GenericReferenceField')
# If previous field was a reference, throw an error (we
# cannot look up fields that are on references).
if isinstance(field, (ReferenceField, GenericReferenceField)):
raise LookUpError('Cannot perform join in mongoDB: %s' %
'__'.join(parts))
# If the parent field has a "field" attribute which has a
# lookup_member method, call it to find the field
# corresponding to this iteration.
if hasattr(getattr(field, 'field', None), 'lookup_member'):
new_field = field.field.lookup_member(field_name)
# If the parent field is a DynamicField or if it's part of
# a DynamicDocument, mark current field as a DynamicField
# with db_name equal to the field name.
elif cls._dynamic and (isinstance(field, DynamicField)
or getattr(
getattr(field, 'document_type',
None), '_dynamic', None)):
new_field = DynamicField(db_field=field_name)
# Else, try to use the parent field's lookup_member method
# to find the subfield.
elif hasattr(field, 'lookup_member'):
new_field = field.lookup_member(field_name)
# Raise a LookUpError if all the other conditions failed.
else:
raise LookUpError('Cannot resolve subfield or operator {} '
'on the field {}'.format(
field_name, field.name))
# If current field still wasn't found and the parent field
# is a ComplexBaseField, add the name current field name and
# move on.
if not new_field and isinstance(field, ComplexBaseField):
fields.append(field_name)
continue
elif not new_field:
raise LookUpError('Cannot resolve field "%s"' % field_name)
field = new_field # update field to the new field type
fields.append(field)
return fields
@classmethod
def _translate_field_name(cls, field, sep='.'):
"""Translate a field attribute name to a database field name.
"""
parts = field.split(sep)
parts = [f.db_field for f in cls._lookup_field(parts)]
return '.'.join(parts)
def __set_field_display(self):
"""For each field that specifies choices, create a
get_<field>_display method.
"""
fields_with_choices = [(n, f) for n, f in self._fields.items()
if f.choices]
for attr_name, field in fields_with_choices:
setattr(self, 'get_%s_display' % attr_name,
partial(self.__get_field_display, field=field))
def __get_field_display(self, field):
"""Return the display value for a choice field"""
value = getattr(self, field.name)
if field.choices and isinstance(field.choices[0], (list, tuple)):
if value is None:
return None
sep = getattr(field, 'display_sep', ' ')
values = value if field.__class__.__name__ in (
'ListField', 'SortedListField') else [value]
return sep.join([
six.text_type(dict(field.choices).get(val, val))
for val in values or []
])
return value
def evaluate(cls, config, ctrl):
time_limit = time.time() + 60 * 60 # 1hr from now
rval = SON(dbn_train_fn_version=1)
ctrl.info('starting dbn_train_fn')
kv = config.items()
kv.sort()
for k, v in kv:
ctrl.info('key=%s\t%s' % (k, str(v)))
rng = numpy.random.RandomState(config['iseed'])
s_rng = RandomStreams(int(rng.randint(2**30)))
dataset, train_Xy, valid_Xy, test_Xy = preprocess_data(config, ctrl)
# allocate learning function parameters
s_inputs_all = tensor.fmatrix('inputs')
s_labels_all = tensor.ivector('labels')
s_idx = tensor.lscalar('batch_idx')
s_batchsize = tensor.lscalar('batch_size')
s_low = s_idx * s_batchsize
s_high = s_low + s_batchsize
s_inputs = s_inputs_all[s_low:s_high]
s_labels = s_labels_all[s_low:s_high]
s_lr = tensor.scalar('lr')
s_features = s_inputs # s_features will be modified in the model-building loop
weights = []
vbiases = []
hbiases = []
n_inputs_i = valid_Xy[0].get_value(borrow=True).shape[1]
rval['cd_reports'] = []
try:
layer_config = config['next_layer']
# allocate model parameters
while layer_config:
i = len(rval['cd_reports'])
n_hid_i = layer_config['n_hid']
if layer_config['W_init_dist'] == 'uniform':
W = rng.uniform(low=-1, high=1,
size=(n_hid_i,
n_inputs_i)).T.astype('float32')
elif layer_config['W_init_dist'] == 'normal':
W = rng.randn(n_hid_i, n_inputs_i).T.astype('float32')
else:
raise ValueError('W_init_dist',
layer_config['W_init_dist'])
if layer_config['W_init_algo'] == 'old':
#N.B. the weights are transposed so that as the number of hidden units changes,
# the first hidden units are always the same vectors.
# this makes it easier to isolate the effect of random initialization
# from the other hyper-parameters under review
W *= layer_config[
'W_init_algo_old_multiplier'] / numpy.sqrt(n_inputs_i)
elif layer_config['W_init_algo'] == 'Xavier':
W *= numpy.sqrt(6.0 / (n_inputs_i + n_hid_i))
else:
raise ValueError(layer_config['W_init_algo'])
layer_idx = len(rval['cd_reports'])
weights.append(theano.shared(W, 'W_%i' % layer_idx))
hbiases.append(
theano.shared(numpy.zeros(n_hid_i, dtype='float32'),
'h_%i' % layer_idx))
vbiases.append(
theano.shared(numpy.zeros(n_inputs_i, dtype='float32'),
'v_%i' % layer_idx))
del W
# allocate RBM training function for this layer
# this version re-calculates the training set every time
# TODO: cache the training set for each layer
# TODO: consider sparsity?
# TODO: consider momentum?
if layer_config['cd_epochs']:
cd_report = train_rbm(
s_rng,
s_idx,
s_batchsize,
s_features,
W=weights[-1],
vbias=vbiases[-1],
hbias=hbiases[-1],
n_in=n_inputs_i,
n_hid=n_hid_i,
batchsize=layer_config['cd_batchsize'],
sample_v0s=layer_config['cd_sample_v0s'],
cdlr=layer_config['cd_lr'] /
float(layer_config['cd_batchsize']),
n_epochs=layer_config['cd_epochs'],
n_batches_per_epoch=dataset.descr['n_train'] //
layer_config['cd_batchsize'],
lr_anneal_start=layer_config['cd_lr_anneal_start'],
givens={
s_inputs_all:
tensor.as_tensor_variable(train_Xy[0])
},
time_limit=time_limit)
else:
cd_report = None
rval['cd_reports'].append(cd_report)
# update s_features to point to top layer
s_features = tensor.nnet.sigmoid(
tensor.dot(s_features, weights[-1]) + hbiases[-1])
n_inputs_i = n_hid_i
layer_config = layer_config.get('next_layer', None)
except (MemoryError, ):
rval['abort'] = 'MemoryError'
rval['status'] = 'ok'
rval['loss'] = 1.0
rval['best_epoch_valid'] = 0.0
return rval
# allocate model
logreg = LogisticRegression.new(s_features,
n_in=n_inputs_i,
n_out=dataset.descr['n_classes'])
traincost = logreg.nll(s_labels).mean()
def ssq(X):
return (X**2).sum()
traincost = traincost + config['l2_penalty'] * (
sum([ssq(w_i) for w_i in weights]) + ssq(logreg.w))
# params = weights+hbiases+vbiases+logreg.params
# vbiases are not involved in the supervised network
params = weights + hbiases + logreg.params
train_logreg_fn = theano.function(
[s_idx, s_lr], [logreg.nll(s_labels).mean()],
updates=sgd_updates(params=params,
grads=tensor.grad(traincost, params),
stepsizes=[s_lr] * len(params)),
givens={
s_batchsize: config['batchsize'],
s_inputs_all: tensor.as_tensor_variable(train_Xy[0]),
s_labels_all: train_Xy[1]
})
valid_logreg_fn = theano.function(
[s_idx],
logreg.errors(s_labels).mean(),
givens={
s_batchsize: config['batchsize'],
s_inputs_all: tensor.as_tensor_variable(valid_Xy[0]),
s_labels_all: valid_Xy[1]
})
test_logreg_fn = theano.function(
[s_idx],
logreg.errors(s_labels).mean(),
givens={
s_batchsize: config['batchsize'],
s_inputs_all: tensor.as_tensor_variable(test_Xy[0]),
s_labels_all: test_Xy[1]
})
rval['best_epoch'] = -1
rval['best_epoch_valid'] = -1
rval['best_epoch_train'] = -1
rval['best_epoch_test'] = -1
rval['status'] = 'ok'
valid_rate = -1
test_rate = -1
train_rate = -1
n_train_batches = dataset.descr['n_train'] // config['batchsize']
n_valid_batches = dataset.descr['n_valid'] // config['batchsize']
n_test_batches = dataset.descr['n_test'] // config['batchsize']
n_iters = 0
for epoch in xrange(config['sup_max_epochs']):
e_lr = config['lr']
e_lr *= min(1, config['lr_anneal_start'] /
float(n_iters + 1)) #anneal learning rate
valid_rate = float(1 - numpy.mean(
[valid_logreg_fn(i) for i in range(n_valid_batches)]))
valid_rate_std_thresh = 0.5 * numpy.sqrt(
valid_rate * (1 - valid_rate) /
(n_valid_batches * config['batchsize']))
if valid_rate > (rval['best_epoch_valid'] + valid_rate_std_thresh):
rval['best_epoch'] = epoch
rval['best_epoch_test'] = test_rate
rval['best_epoch_valid'] = valid_rate
rval['best_epoch_train'] = train_rate
best_params = copy.deepcopy(params)
logger.info(
'Epoch=%i best epoch %i valid %f test %f best_epoch_train %f prev_train %f'
% (epoch, rval['best_epoch'], rval['best_epoch_valid'],
rval['best_epoch_test'], rval['best_epoch_train'],
train_rate))
#ctrl.info('Epoch %i train nll: %f'%(epoch, train_rate))
ctrl.checkpoint(rval)
if epoch > config[
'sup_min_epochs'] and epoch > 2 * rval['best_epoch']:
break
if time.time() > time_limit:
break
train_rate = float(
numpy.mean([
train_logreg_fn(i, e_lr) for i in range(n_train_batches)
]))
if not numpy.isfinite(train_rate):
do_test = False
rval['status'] = 'fail'
rval['status_info'] = 'train_rate %f' % train_rate
break
++n_iters
do_test = 1
if do_test and rval['status'] == 'ok':
# copy best params back into place
for p, bp in zip(params, best_params):
p.set_value(bp.get_value())
rval['best_epoch_test'] = 1 - float(
numpy.mean([test_logreg_fn(i) for i in range(n_test_batches)]))
rval['loss'] = 1.0 - rval['best_epoch_valid']
ctrl.info('rval: %s' % str(rval))
return rval
'versioning': False,
'resource_methods': ['GET', 'POST', 'DELETE'],
'item_methods': ['GET', 'PATCH', 'DELETE'],
'mongo_indexes': {'attr': ([('key', 1), ('id', 1)], {'background': True}),
'icao': ([('icao', 1)], {'background': True}),
'etag': ([('_etag', 1)], {'background': True, 'unique': True}),
'content': ([('descr', 'text'), ('label', 'text')], {'background': True}),
'rit_version': ([('rit_version', 1)], {'background': True})
},
'allow_unknown': True,
'schema': _schema
}
from bson import SON, ObjectId
agg_count_keys = {
'url': 'e5x/choices/keys/count',
'item_title': 'E5X choices count',
'pagination': False,
'datasource': {
'source': RESOURCE_COLLECTION,
'aggregation': {
'pipeline': [
{"$group": {"_id": {"key": "$key"}, "count": {"$sum": 1}}},
{"$sort": SON([("count", -1), ("_id", -1)])}
]
}
}
}
def __init__(self, *args, **values):
"""
Initialise a document or an embedded document.
:param values: A dictionary of keys and values for the document.
It may contain additional reserved keywords, e.g. "__auto_convert".
:param __auto_convert: If True, supplied values will be converted
to Python-type values via each field's `to_python` method.
:param _created: Indicates whether this is a brand new document
or whether it's already been persisted before. Defaults to true.
"""
self._initialised = False
self._created = True
if args:
raise TypeError(
"Instantiating a document with positional arguments is not "
"supported. Please use `field_name=value` keyword arguments.")
__auto_convert = values.pop("__auto_convert", True)
_created = values.pop("_created", True)
signals.pre_init.send(self.__class__, document=self, values=values)
# Check if there are undefined fields supplied to the constructor,
# if so raise an Exception.
if not self._dynamic and (self._meta.get("strict", True) or _created):
_undefined_fields = set(values.keys()) - set(
list(self._fields.keys()) +
["id", "pk", "_cls", "_text_score"])
if _undefined_fields:
msg = f'The fields "{_undefined_fields}" do not exist on the document "{self._class_name}"'
raise FieldDoesNotExist(msg)
if self.STRICT and not self._dynamic:
self._data = StrictDict.create(allowed_keys=self._fields_ordered)()
else:
self._data = {}
self._dynamic_fields = SON()
# Assign default values for fields
# not set in the constructor
for field_name in self._fields:
if field_name in values:
continue
value = getattr(self, field_name, None)
setattr(self, field_name, value)
if "_cls" not in values:
self._cls = self._class_name
# Set actual values
dynamic_data = {}
FileField = _import_class("FileField")
for key, value in values.items():
field = self._fields.get(key)
if field or key in ("id", "pk", "_cls"):
if __auto_convert and value is not None:
if field and not isinstance(field, FileField):
value = field.to_python(value)
setattr(self, key, value)
else:
if self._dynamic:
dynamic_data[key] = value
else:
# For strict Document
self._data[key] = value
# Set any get_<field>_display methods
self.__set_field_display()
if self._dynamic:
self._dynamic_lock = False
for key, value in dynamic_data.items():
setattr(self, key, value)
# Flag initialised
self._initialised = True
self._created = _created
signals.post_init.send(self.__class__, document=self)
def get_list_distinct_approval_year_and_period(args):
ret = {}
collection = common.get_collection('TMPER_AprPeriod').aggregate([{
"$lookup": {
'from': common.get_collection_name_with_schema('SYS_VW_ValueList'),
'localField': 'apr_period',
'foreignField': 'value',
'as': 'aprPeriod'
}
}, {
"$unwind": {
'path': '$aprPeriod',
"preserveNullAndEmptyArrays": True
}
}, {
"$match": {
"$and": [{
'aprPeriod.list_name': "LApprovalPeriod"
}, {
'aprPeriod.language': quicky.language.get_language()
}]
}
}, {
"$project": {
"apr_period_a": {
"$ifNull": ["$aprPeriod.caption", ""]
},
"apr_year_a": {
'$toString': "$apr_year"
},
"apr_period": {
'$toString': "$apr_period"
},
"apr_year": 1
}
}, {
"$project": {
"caption": {
'$concat': ["$apr_year_a", " / ", "$apr_period_a"]
},
"value": {
'$concat': ["$apr_year_a", "__", "$apr_period"]
},
"apr_year": 1,
"apr_period": {
'$toInt': "$apr_period"
},
}
}, {
"$sort":
SON([("apr_year", -1), ("apr_period", -1)])
}])
ret = list(collection)
ret1 = []
if (args['data'] != None
and args['data'].has_key('apr_period' and 'apr_year')):
for x in ret:
if (x['apr_period'] == args['data']['apr_period']
and x['apr_year'] == args['data']['apr_year']):
continue
else:
collection1 = common.get_collection(
'TMPER_AprPeriodEmpOut').aggregate([
{
"$match": {
"$and": [{
'apr_year': x['apr_year']
}, {
'apr_period': x['apr_period']
}]
}
},
])
if (len(list(collection1)) == 0):
ret1.append(x)
item_list = [e for e in ret if e not in ret1]
return item_list
def query(_doc_cls=None, **kwargs):
"""Transform a query from Django-style format to Mongo format."""
mongo_query = {}
merge_query = defaultdict(list)
for key, value in sorted(kwargs.items()):
if key == "__raw__":
mongo_query.update(value)
continue
parts = key.rsplit("__")
indices = [(i, p) for i, p in enumerate(parts) if p.isdigit()]
parts = [part for part in parts if not part.isdigit()]
# Check for an operator and transform to mongo-style if there is
op = None
if len(parts) > 1 and parts[-1] in MATCH_OPERATORS:
op = parts.pop()
# Allow to escape operator-like field name by __
if len(parts) > 1 and parts[-1] == "":
parts.pop()
negate = False
if len(parts) > 1 and parts[-1] == "not":
parts.pop()
negate = True
if _doc_cls:
# Switch field names to proper names [set in Field(name='foo')]
try:
fields = _doc_cls._lookup_field(parts)
except Exception as e:
raise InvalidQueryError(e)
parts = []
CachedReferenceField = _import_class("CachedReferenceField")
GenericReferenceField = _import_class("GenericReferenceField")
cleaned_fields = []
for field in fields:
append_field = True
if isinstance(field, six.string_types):
parts.append(field)
append_field = False
# is last and CachedReferenceField
elif isinstance(field,
CachedReferenceField) and fields[-1] == field:
parts.append("%s._id" % field.db_field)
else:
parts.append(field.db_field)
if append_field:
cleaned_fields.append(field)
# Convert value to proper value
field = cleaned_fields[-1]
singular_ops = [None, "ne", "gt", "gte", "lt", "lte", "not"]
singular_ops += STRING_OPERATORS
if op in singular_ops:
value = field.prepare_query_value(op, value)
if isinstance(field, CachedReferenceField) and value:
value = value["_id"]
elif op in ("in", "nin", "all",
"near") and not isinstance(value, dict):
# Raise an error if the in/nin/all/near param is not iterable.
value = _prepare_query_for_iterable(field, op, value)
# If we're querying a GenericReferenceField, we need to alter the
# key depending on the value:
# * If the value is a DBRef, the key should be "field_name._ref".
# * If the value is an ObjectId, the key should be "field_name._ref.$id".
if isinstance(field, GenericReferenceField):
if isinstance(value, DBRef):
parts[-1] += "._ref"
elif isinstance(value, ObjectId):
parts[-1] += "._ref.$id"
# if op and op not in COMPARISON_OPERATORS:
if op:
if op in GEO_OPERATORS:
value = _geo_operator(field, op, value)
elif op in ("match", "elemMatch"):
ListField = _import_class("ListField")
EmbeddedDocumentField = _import_class("EmbeddedDocumentField")
if (isinstance(value, dict) and isinstance(field, ListField)
and isinstance(field.field, EmbeddedDocumentField)):
value = query(field.field.document_type, **value)
else:
value = field.prepare_query_value(op, value)
value = {"$elemMatch": value}
elif op in CUSTOM_OPERATORS:
NotImplementedError('Custom method "%s" has not '
"been implemented" % op)
elif op not in STRING_OPERATORS:
value = {"$" + op: value}
if negate:
value = {"$not": value}
for i, part in indices:
parts.insert(i, part)
key = ".".join(parts)
if key not in mongo_query:
mongo_query[key] = value
else:
if isinstance(mongo_query[key], dict) and isinstance(value, dict):
mongo_query[key].update(value)
# $max/minDistance needs to come last - convert to SON
value_dict = mongo_query[key]
if ("$maxDistance" in value_dict or "$minDistance"
in value_dict) and ("$near" in value_dict
or "$nearSphere" in value_dict):
value_son = SON()
for k, v in iteritems(value_dict):
if k == "$maxDistance" or k == "$minDistance":
continue
value_son[k] = v
# Required for MongoDB >= 2.6, may fail when combining
# PyMongo 3+ and MongoDB < 2.6
near_embedded = False
for near_op in ("$near", "$nearSphere"):
if isinstance(value_dict.get(near_op), dict):
value_son[near_op] = SON(value_son[near_op])
if "$maxDistance" in value_dict:
value_son[near_op][
"$maxDistance"] = value_dict[
"$maxDistance"]
if "$minDistance" in value_dict:
value_son[near_op][
"$minDistance"] = value_dict[
"$minDistance"]
near_embedded = True
if not near_embedded:
if "$maxDistance" in value_dict:
value_son["$maxDistance"] = value_dict[
"$maxDistance"]
if "$minDistance" in value_dict:
value_son["$minDistance"] = value_dict[
"$minDistance"]
mongo_query[key] = value_son
else:
# Store for manually merging later
merge_query[key].append(value)
# The queryset has been filter in such a way we must manually merge
for k, v in merge_query.items():
merge_query[k].append(mongo_query[k])
del mongo_query[k]
if isinstance(v, list):
value = [{k: val} for val in v]
if "$and" in mongo_query.keys():
mongo_query["$and"].extend(value)
else:
mongo_query["$and"] = value
return mongo_query
def build_itemsets_by_day(distance, undefined=True, only_tourists=True):
hash_table = dict()
count = 0
f = open('DatasetBuilding/input/is_userdays.txt', mode='w')
g = open('DatasetBuilding/translation_tables/is_userdays_tt', mode='w')
q1 = {'touristLocal': 'tourist'}
q2 = {'$or': [{'touristLocal': 'tourist'}, {'touristLocal': 'undefined'}]}
if only_tourists:
query = q1
else:
query = q2
for tourist in ucol.find(query):
tweets = []
item_set = set()
prevDate = None
for sequence in scol.find({
'user_id': tourist['id']
}).sort('serial', pymongo.ASCENDING):
for id in sequence['sequence']:
for tweet in tcol.find({'id': id}):
currentDate = datetime.datetime.strptime(
tweet['created_at'], '%a %b %d %H:%M:%S +0000 %Y')
if prevDate is None:
prevDate = currentDate
if (currentDate.date() - prevDate.date()).days > 0:
f.write(" ".join(list(map(str, sorted(item_set)))))
f.write("\n")
item_set = set()
#print(currentDate)
tweets.append(tweet)
lon = tweet['coordinates']['coordinates'][0]
lat = tweet['coordinates']['coordinates'][1]
queryNear = {
'coordinates':
SON([('$near', {
"type": "Point",
"coordinates": [lon, lat]
}), ('$maxDistance', distance)])
}
close_pois = pcol.find(queryNear).limit(1)
closest_poi = 'UNDEFINED'
if close_pois.count() > 0:
closest_poi = close_pois.next()['name']
else:
if not undefined:
continue
code = hash_table.get(closest_poi, 0)
if code == 0:
count += 1
hash_table[closest_poi] = count
code = count
item_set.add(code)
if len(item_set) > 0:
f.write(" ".join(list(map(str, sorted(item_set)))))
f.write("\n")
aux = []
for key, value in hash_table.items():
aux += [(key, value)]
sorted_aux = sorted(aux, key=lambda tup: tup[1])
for item in sorted_aux:
g.write(str(item[1]) + "\t" + item[0] + "\n")
agg_count_types = {
'url': 'organizations/types/count',
'item_title': 'Organiuzations Types Count',
'pagination': False,
'datasource': {
'source': RESOURCE_COLLECTION,
'aggregation': {
'pipeline': [{
"$group": {
"_id": "$type_id",
"count": {
"$sum": 1
}
}
}, {
"$sort": SON([("count", -1), ("_id", -1)])
}]
}
}
}
agg_parents = {
'item_title': 'Content Parents Aggregation',
'pagination': False,
'datasource': {
'source': RESOURCE_COLLECTION,
'aggregation': {
'pipeline': [
{
"$match": {
"id": "$start_id",
def default(obj, json_options=DEFAULT_JSON_OPTIONS):
# We preserve key order when rendering SON, DBRef, etc. as JSON by
# returning a SON for those types instead of a dict.
if isinstance(obj, ObjectId):
return {"$oid": str(obj)}
if isinstance(obj, DBRef):
return _json_convert(obj.as_doc(), json_options=json_options)
if isinstance(obj, datetime.datetime):
if (json_options.datetime_representation ==
DatetimeRepresentation.ISO8601):
if not obj.tzinfo:
obj = obj.replace(tzinfo=utc)
if obj >= EPOCH_AWARE:
off = obj.tzinfo.utcoffset(obj)
if (off.days, off.seconds, off.microseconds) == (0, 0, 0):
tz_string = 'Z'
else:
tz_string = obj.strftime('%z')
millis = int(obj.microsecond / 1000)
fracsecs = ".%03d" % (millis,) if millis else ""
return {"$date": "%s%s%s" % (
obj.strftime("%Y-%m-%dT%H:%M:%S"), fracsecs, tz_string)}
millis = bson._datetime_to_millis(obj)
if (json_options.datetime_representation ==
DatetimeRepresentation.LEGACY):
return {"$date": millis}
return {"$date": {"$numberLong": str(millis)}}
if json_options.strict_number_long and isinstance(obj, Int64):
return {"$numberLong": str(obj)}
if isinstance(obj, (RE_TYPE, Regex)):
flags = ""
if obj.flags & re.IGNORECASE:
flags += "i"
if obj.flags & re.LOCALE:
flags += "l"
if obj.flags & re.MULTILINE:
flags += "m"
if obj.flags & re.DOTALL:
flags += "s"
if obj.flags & re.UNICODE:
flags += "u"
if obj.flags & re.VERBOSE:
flags += "x"
if isinstance(obj.pattern, str):
pattern = obj.pattern
else:
pattern = obj.pattern.decode('utf-8')
if json_options.json_mode == JSONMode.LEGACY:
return SON([("$regex", pattern), ("$options", flags)])
return {'$regularExpression': SON([("pattern", pattern),
("options", flags)])}
if isinstance(obj, MinKey):
return {"$minKey": 1}
if isinstance(obj, MaxKey):
return {"$maxKey": 1}
if isinstance(obj, Timestamp):
return {"$timestamp": SON([("t", obj.time), ("i", obj.inc)])}
if isinstance(obj, Code):
if obj.scope is None:
return {'$code': str(obj)}
return SON([
('$code', str(obj)),
('$scope', _json_convert(obj.scope, json_options))])
if isinstance(obj, Binary):
return _encode_binary(obj, obj.subtype, json_options)
if isinstance(obj, bytes):
return _encode_binary(obj, 0, json_options)
if isinstance(obj, uuid.UUID):
if json_options.strict_uuid:
binval = Binary.from_uuid(
obj, uuid_representation=json_options.uuid_representation)
return _encode_binary(binval, binval.subtype, json_options)
else:
return {"$uuid": obj.hex}
if isinstance(obj, Decimal128):
return {"$numberDecimal": str(obj)}
if isinstance(obj, bool):
return obj
if (json_options.json_mode == JSONMode.CANONICAL and
isinstance(obj, int)):
if -2 ** 31 <= obj < 2 ** 31:
return {'$numberInt': str(obj)}
return {'$numberLong': str(obj)}
if json_options.json_mode != JSONMode.LEGACY and isinstance(obj, float):
if math.isnan(obj):
return {'$numberDouble': 'NaN'}
elif math.isinf(obj):
representation = 'Infinity' if obj > 0 else '-Infinity'
return {'$numberDouble': representation}
elif json_options.json_mode == JSONMode.CANONICAL:
# repr() will return the shortest string guaranteed to produce the
# original value, when float() is called on it.
return {'$numberDouble': str(repr(obj))}
raise TypeError("%r is not JSON serializable" % obj)
def test_update(self):
self.coll.insert([{}, {}])
bulk = self.coll.initialize_ordered_bulk_op()
# update() requires find() first.
self.assertRaises(AttributeError,
lambda: bulk.update({'$set': {
'x': 1
}}))
self.assertRaises(TypeError, bulk.find({}).update, 1)
self.assertRaises(ValueError, bulk.find({}).update, {})
# All fields must be $-operators.
self.assertRaises(ValueError, bulk.find({}).update, {'foo': 'bar'})
bulk.find({}).update({'$set': {'foo': 'bar'}})
result = bulk.execute()
self.assertEqualResponse(
{
'nMatched': 2,
'nModified': 2,
'nUpserted': 0,
'nInserted': 0,
'nRemoved': 0,
'upserted': [],
'writeErrors': [],
'writeConcernErrors': []
}, result)
self.assertEqual(self.coll.find({'foo': 'bar'}).count(), 2)
# All fields must be $-operators -- validated server-side.
bulk = self.coll.initialize_ordered_bulk_op()
updates = SON([('$set', {'x': 1}), ('y', 1)])
bulk.find({}).update(updates)
self.assertRaises(BulkWriteError, bulk.execute)
self.coll.remove()
self.coll.insert([{}, {}])
bulk = self.coll.initialize_unordered_bulk_op()
bulk.find({}).update({'$set': {'bim': 'baz'}})
result = bulk.execute()
self.assertEqualResponse(
{
'nMatched': 2,
'nModified': 2,
'nUpserted': 0,
'nInserted': 0,
'nRemoved': 0,
'upserted': [],
'writeErrors': [],
'writeConcernErrors': []
}, result)
self.assertEqual(self.coll.find({'bim': 'baz'}).count(), 2)
self.coll.insert({'x': 1})
bulk = self.coll.initialize_unordered_bulk_op()
bulk.find({'x': 1}).update({'$set': {'x': 42}})
result = bulk.execute()
self.assertEqualResponse(
{
'nMatched': 1,
'nModified': 1,
'nUpserted': 0,
'nInserted': 0,
'nRemoved': 0,
'upserted': [],
'writeErrors': [],
'writeConcernErrors': []
}, result)
self.assertEqual(1, self.coll.find({'x': 42}).count())
# Second time, x is already 42 so nModified is 0.
bulk = self.coll.initialize_unordered_bulk_op()
bulk.find({'x': 42}).update({'$set': {'x': 42}})
result = bulk.execute()
self.assertEqualResponse(
{
'nMatched': 1,
'nModified': 0,
'nUpserted': 0,
'nInserted': 0,
'nRemoved': 0,
'upserted': [],
'writeErrors': [],
'writeConcernErrors': []
}, result)
def add_sorting(self, order_by: list):
ordering = [(field, 1) for field in order_by]
sorting = {"$sort": SON(ordering)}
return sorting
def query(_doc_cls=None, _field_operation=False, **query):
"""Transform a query from Django-style format to Mongo format.
"""
mongo_query = {}
merge_query = defaultdict(list)
for key, value in sorted(query.items()):
if key == "__raw__":
mongo_query.update(value)
continue
parts = key.split('__')
indices = [(i, p) for i, p in enumerate(parts) if p.isdigit()]
parts = [part for part in parts if not part.isdigit()]
# Check for an operator and transform to mongo-style if there is
op = None
if parts[-1] in MATCH_OPERATORS:
op = parts.pop()
negate = False
if parts[-1] == 'not':
parts.pop()
negate = True
if _doc_cls:
# Switch field names to proper names [set in Field(name='foo')]
try:
fields = _doc_cls._lookup_field(parts)
except Exception as e:
raise InvalidQueryError(e)
parts = []
cleaned_fields = []
for field in fields:
append_field = True
if isinstance(field, str):
parts.append(field)
append_field = False
else:
parts.append(field.db_field)
if append_field:
cleaned_fields.append(field)
# Convert value to proper value
field = cleaned_fields[-1]
singular_ops = [None, 'ne', 'gt', 'gte', 'lt', 'lte', 'not']
singular_ops += STRING_OPERATORS
if op in singular_ops:
if isinstance(field, str):
if (op in STRING_OPERATORS and isinstance(value, str)):
StringField = _import_class('StringField')
value = StringField.prepare_query_value(op, value)
else:
value = field
else:
value = field.prepare_query_value(op, value)
elif op in ('in', 'nin', 'all',
'near') and not isinstance(value, dict):
# 'in', 'nin' and 'all' require a list of values
value = [field.prepare_query_value(op, v) for v in value]
# if op and op not in COMPARISON_OPERATORS:
if op:
if op in GEO_OPERATORS:
value = _geo_operator(field, op, value)
elif op in CUSTOM_OPERATORS:
if op == 'match':
value = {"$elemMatch": value}
else:
NotImplementedError("Custom method '%s' has not "
"been implemented" % op)
elif op not in STRING_OPERATORS:
value = {'$' + op: value}
if negate:
value = {'$not': value}
for i, part in indices:
parts.insert(i, part)
key = '.'.join(parts)
if op is None or key not in mongo_query:
mongo_query[key] = value
elif key in mongo_query:
if key in mongo_query and isinstance(mongo_query[key], dict):
mongo_query[key].update(value)
# $maxDistance needs to come last - convert to SON
if '$maxDistance' in mongo_query[key]:
value_dict = mongo_query[key]
value_son = SON()
for k, v in value_dict.items():
if k == '$maxDistance':
continue
value_son[k] = v
value_son['$maxDistance'] = value_dict['$maxDistance']
mongo_query[key] = value_son
else:
# Store for manually merging later
merge_query[key].append(value)
# The queryset has been filter in such a way we must manually merge
for k, v in list(merge_query.items()):
merge_query[k].append(mongo_query[k])
del mongo_query[k]
if isinstance(v, list):
value = [{k: val} for val in v]
if '$and' in list(mongo_query.keys()):
mongo_query['$and'].append(value)
else:
mongo_query['$and'] = value
return mongo_query
def count_cmd(self, select=None, take=None, skip=None):
cmd = SON([('aggregate', self._mongo_collection.name)])
cmd.update({'query': select}) if select else None
cmd.update({'limit': take}) if take else None
cmd.update({'skip': skip}) if skip else None
return self.db_command(cmd)['n']
]
result = db.my.bulk_write(requests)
bulk()
# inventory.insert_one(
# {"item": "canvas",
# "qty": 100,
# "tags": ["cotton"],
# "size": {"h": 28, "w": 35.5, "uom": "cm"}})
# inventory.insert_many([
# {"item": "journal",
# "qty": 25,
# "tags": ["blank", "red"],
# "size": {"h": 14, "w": 21, "uom": "cm"}},
# {"item": "mat",
# "qty": 85,
# "tags": ["gray"],
# "size": {"h": 27.9, "w": 35.5, "uom": "cm"}},
# {"item": "mousepad",
# "qty": 25,
# "tags": ["gel", "blue"],
# "size": {"h": 19, "w": 22.85, "uom": "cm"}}])
cursor = inventory.find({"size": SON([("h", 14), ("w", 21), ("uom", "cm")])})
inventory.find_one({"size.uom": "in"})
output(inventory.find({"size.uom": "cm"}))
def find_cmd(self, select=None, project=None, sort=None, take=None, skip=None, tailable=False, reduce_by=None):
cmd = SON([('find', self._mongo_collection.name)])
cmd.update({'filter': select}) if select else None
cmd.update({'projection': project}) if project else None
cmd.update({'sort': sort}) if sort else None
cmd.update({'skip': skip}) if skip else None
cmd.update({'limit': take}) if take else None
cmd.update({'tailable': tailable}) if tailable else None
cmd.update({'singleBatch': True})
cmd.update({'batchSize': 1000})
docs = self.db_command(cmd)['cursor']['firstBatch']
if hasattr(docs, '__iter__'):
return [d if not reduce_by else reduce_by(**d) for d in docs]
# for d in docs:
# yield d if not reduce_by else reduce_by().objectify(d)
else:
return docs if not reduce_by else reduce_by(**docs)
def __init__(self, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(99, _hsid, 0, _hflags)
self.type = 99
self['t'] = 99
model1 = SON([
#global
('color_space' , 'rgb'),
('conv_mode' , 'valid'),
# prepare images before processing
('preproc', SON([
# resize input images by keeping aspect ratio and fix the biggest edge
('max_edge' , 150),
# kernel size of the box low pass filter
('lsum_ksize' , 3),
# how to resize the image
('resize_method' , 'bicubic'),
])),
# - input local normalization
# local zero-mean, unit-magnitude
('normin', SON([
# kernel shape of the local normalization
('kshape' , (3,3)),
# magnitude threshold
# if the vector's length is below, it doesn't get resized
('threshold' , 1.0),
])),
# - linear filtering
('filter', SON([
('model_name','gridded_gabor'),
# kernel shape of the gabors
('kshape' , [43,43]),
# list of orientations
('norients' , 16),
# list of frequencies
('divfreqs' , [2, 3, 4, 6, 11, 18]),
# list of phases
('phases' , [0]),
])),
# - simple non-linear activation
('activ', SON([
# minimum output (clamp)
('minout' , 0),
# maximum output (clamp)
('maxout' , 1),
])),
# - output local normalization
('normout', SON([
# kernel shape of the local normalization
('kshape', (3,3)),
# magnitude threshold
# if the vector's length is below, it doesn't get resized
('threshold', 1.0),
])),
# - pooling
('pool' , SON([
# kernel size of the local sum (2d slice)
('lsum_ksize' , 17 ),
# fixed output shape (only the first 2 dimensions, y and x)
('outshape' , (30,30)),
])),
# -- featsel details what features you want to be included in the vector
('featsel' , SON([
# Include representation output ? True or False
('output' , True),
# Include grayscale values ? None or (height, width)
('input_gray' , None),
# Include color histograms ? None or nbins per color
('input_colorhists' , None),
# Include input norm histograms ? None or (division, nfeatures)
('normin_hists' , None),
# Include filter output histograms ? None or (division, nfeatures)
('filter_hists' , None),
# Include activation output histograms ? None or (division, nfeatures)
('activ_hists' , None),
# Include output norm histograms ? None or (division, nfeatures)
('normout_hists' , None),
# Include representation output histograms ? None or (division, nfeatures)
('pool_hists' , None),
]))
])
def __init__(self, bytes, _hsid=0, _hflags=0):
SON.__init__(self)
self.hdr = Hdr(2, _hsid, 0, _hflags)
self.type = 2
self['t'] = 2
self['bytes'] = bytes
base_model = SON([
#global
('color_space' , 'rgb'),
('conv_mode' , 'valid'),
# prepare images before processing
('preproc', SON([
# resize input images by keeping aspect ratio and fix the biggest edge
('max_edge' , None),
# kernel size of the box low pass filter
('lsum_ksize' , None),
# how to resize the image
('resize_method' , None),
('whiten', False)
])),
# - input local normalization
# local zero-mean, unit-magnitude
('normin', SON([
# kernel shape of the local normalization
('kshape' , (3,3)),
# magnitude threshold
# if the vector's length is below, it doesn't get resized
('threshold' , 1.0),
])),
# - linear filtering
('filter', SON([
('model_name','cairo_generated'),
('kshape' , [32,32]),
# list of orientations
('specs',[SON([
('width',64),
('height',64),
('object' , cairo_objects.SQUARE),
('pattern' , cairo_objects.SOLID_RED),
]),SON([
('width',64),
('height',64),
('object' , cairo_objects.SQUARE),
('pattern' , cairo_objects.SOLID_RED),
('sx',.5),
('sy',2)
])] , )
])),
# - simple non-linear activation
('activ', SON([
# minimum output (clamp)
# maximum output (clamp)
('maxout' , 1),
])),
# - output local normalization
('normout', SON([
# kernel shape of the local normalization
('kshape', (3,3)),
# magnitude threshold
# if the vector's length is below, it doesn't get resized
('threshold', 1.0),
])),
])
def command(sock, dbname, spec, slave_ok, is_mongos,
read_preference, codec_options, session, client, check=True,
allowable_errors=None, address=None,
check_keys=False, listeners=None, max_bson_size=None,
read_concern=None,
parse_write_concern_error=False,
collation=None, retryable_write=False):
"""Execute a command over the socket, or raise socket.error.
:Parameters:
- `sock`: a raw socket instance
- `dbname`: name of the database on which to run the command
- `spec`: a command document as a dict, SON, or mapping object
- `slave_ok`: whether to set the SlaveOkay wire protocol bit
- `is_mongos`: are we connected to a mongos?
- `read_preference`: a read preference
- `codec_options`: a CodecOptions instance
- `session`: optional ClientSession instance.
- `client`: optional MongoClient instance for updating $clusterTime.
- `check`: raise OperationFailure if there are errors
- `allowable_errors`: errors to ignore if `check` is True
- `address`: the (host, port) of `sock`
- `check_keys`: if True, check `spec` for invalid keys
- `listeners`: An instance of :class:`~pymongo.monitoring.EventListeners`
- `max_bson_size`: The maximum encoded bson size for this server
- `read_concern`: The read concern for this command.
- `parse_write_concern_error`: Whether to parse the ``writeConcernError``
field in the command response.
- `collation`: The collation for this command.
- `retryable_write`: True if this command is a retryable write.
"""
name = next(iter(spec))
ns = dbname + '.$cmd'
flags = 4 if slave_ok else 0
if (client or session) and not isinstance(spec, ORDERED_TYPES):
# Ensure command name remains in first place.
spec = SON(spec)
if session:
spec['lsid'] = session._use_lsid()
if retryable_write:
spec['txnNumber'] = session._transaction_id()
if client:
client._send_cluster_time(spec, session)
# Publish the original command document, perhaps with lsid and $clusterTime.
orig = spec
if is_mongos:
spec = message._maybe_add_read_preference(spec, read_preference)
if read_concern:
if read_concern.level:
spec['readConcern'] = read_concern.document
if (session and session.options.causal_consistency
and session.operation_time is not None):
spec.setdefault(
'readConcern', {})['afterClusterTime'] = session.operation_time
if collation is not None:
spec['collation'] = collation
publish = listeners is not None and listeners.enabled_for_commands
if publish:
start = datetime.datetime.now()
request_id, msg, size = message.query(flags, ns, 0, -1, spec,
None, codec_options, check_keys)
if (max_bson_size is not None
and size > max_bson_size + message._COMMAND_OVERHEAD):
message._raise_document_too_large(
name, size, max_bson_size + message._COMMAND_OVERHEAD)
if publish:
encoding_duration = datetime.datetime.now() - start
listeners.publish_command_start(orig, dbname, request_id, address)
start = datetime.datetime.now()
try:
sock.sendall(msg)
reply = receive_message(sock, request_id)
unpacked_docs = reply.unpack_response(codec_options=codec_options)
response_doc = unpacked_docs[0]
if client:
client._receive_cluster_time(response_doc, session)
if check:
helpers._check_command_response(
response_doc, None, allowable_errors,
parse_write_concern_error=parse_write_concern_error)
except Exception as exc:
if publish:
duration = (datetime.datetime.now() - start) + encoding_duration
if isinstance(exc, (NotMasterError, OperationFailure)):
failure = exc.details
else:
failure = message._convert_exception(exc)
listeners.publish_command_failure(
duration, failure, name, request_id, address)
raise
if publish:
duration = (datetime.datetime.now() - start) + encoding_duration
listeners.publish_command_success(
duration, response_doc, name, request_id, address)
return response_doc