This repository has been archived by the owner on Aug 4, 2020. It is now read-only.
/
columnfamilymap.py
287 lines (230 loc) · 11.9 KB
/
columnfamilymap.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
"""
Provides a means for mapping an existing class to a column family.
.. seealso:: :mod:`pycassa.types`
In addition to the default :class:`~pycassa.types.Column` classes,
you may also define your own types for the mapper. For example, the
IntString may be defined as:
.. code-block:: python
>>> class IntString(pycassa.Column):
... def pack(self, val):
... return str(val)
... def unpack(self, val):
... return int(val)
...
"""
from pycassa.types import Column
from pycassa.cassandra.ttypes import IndexExpression, IndexClause
__all__ = ['ColumnFamilyMap']
def create_instance(cls, **kwargs):
instance = cls()
instance.__dict__.update(kwargs)
return instance
class ColumnFamilyMap(object):
""" Maps an existing class to a column family. """
def __init__(self, cls, column_family, columns=None, raw_columns=False):
"""
Maps an existing class to a column family. Class fields become columns,
and instances of that class can be represented as rows in standard column
families or super columns in super column families.
Instances of `cls` are returned from :meth:`get()`, :meth:`multiget()`,
:meth:`get_range()` and :meth:`get_indexed_slices()`.
`column_family` is a :class:`~pycassa.columnfamily.ColumnFamily` to
tie with `cls`. This :class:`ColumnFamily` should almost always have
`autopack_names` and `autopack_values` set to ``False``.
If `raw_columns` is ``True``, all columns will be fetched into the
`raw_columns` field in requests.
"""
self.cls = cls
self.column_family = column_family
self.raw_columns = raw_columns
self.dict_class = self.column_family.dict_class
self.columns = self.dict_class()
for name, column in self.cls.__dict__.iteritems():
if not isinstance(column, Column):
continue
self.columns[name] = column
def combine_columns(self, columns):
combined_columns = self.dict_class()
if self.raw_columns:
combined_columns['raw_columns'] = self.dict_class()
for column, type in self.columns.iteritems():
combined_columns[column] = type.default
for column, value in columns.iteritems():
col_cls = self.columns.get(column, None)
if col_cls is not None:
combined_columns[column] = col_cls.unpack(value)
if self.raw_columns:
combined_columns['raw_columns'][column] = value
return combined_columns
def get(self, key, *args, **kwargs):
"""
Creates one or more instances of `cls` from the row with key `key`.
The fields that are retreived may be specified using `columns`, which
should be a list of column names.
If the column family is a super column family, a list of `cls`
instances will be returned, one for each super column. If
the `super_column` parameter is not supplied, then `columns`
specifies which super columns will be used to create instances
of `cls`. If the `super_column` parameter *is* supplied, only
one instance of `cls` will be returned; if `columns` is specified
in this case, only those attributes listed in `columns` will be fetched.
"""
if 'columns' not in kwargs and not self.column_family.super and not self.raw_columns:
kwargs['columns'] = self.columns.keys()
columns = self.column_family.get(key, *args, **kwargs)
if self.column_family.super:
if 'super_column' not in kwargs:
vals = self.dict_class()
for super_column, subcols in columns.iteritems():
combined = self.combine_columns(subcols)
vals[super_column] = create_instance(self.cls, key=key,
super_column=super_column, **combined)
return vals
combined = self.combine_columns(columns)
return create_instance(self.cls, key=key,
super_column=kwargs['super_column'],
**combined)
combined = self.combine_columns(columns)
return create_instance(self.cls, key=key, **combined)
def multiget(self, *args, **kwargs):
"""
Like :meth:`get()`, but a list of keys may be specified.
The result of multiget will be a dictionary where the keys
are the keys from the `keys` argument, minus any missing rows.
The value for each key in the dictionary will be the same as
if :meth:`get()` were called on that individual key.
"""
if 'columns' not in kwargs and not self.column_family.super and not self.raw_columns:
kwargs['columns'] = self.columns.keys()
kcmap = self.column_family.multiget(*args, **kwargs)
ret = self.dict_class()
for key, columns in kcmap.iteritems():
if self.column_family.super:
if 'super_column' not in kwargs:
vals = self.dict_class()
for super_column, subcols in columns.iteritems():
combined = self.combine_columns(subcols)
vals[super_column] = create_instance(self.cls, key=key, super_column=super_column, **combined)
ret[key] = vals
else:
combined = self.combine_columns(columns)
ret[key] = create_instance(self.cls, key=key, super_column=kwargs['super_column'], **combined)
else:
combined = self.combine_columns(columns)
ret[key] = create_instance(self.cls, key=key, **combined)
return ret
def get_count(self, *args, **kwargs):
"""
Count the number of columns for a key.
.. deprecated:: 0.6.0
Use :meth:`pycassa.columnfamily.ColumnFamily.get()` instead.
"""
return self.column_family.get_count(*args, **kwargs)
def get_range(self, *args, **kwargs):
"""
Get an iterator over instances in a specified key range.
Like :meth:`multiget()`, whether a single instance or multiple
instances are returned per-row when the column family is a super
column family depends on what parameters are passed.
For an explanation of how :meth:`get_range` works and a description
of the parameters, see :meth:`pycassa.columnfamily.ColumnFamily.get_range()`.
Example usage with a standard column family:
.. code-block:: python
>>> pool = pycassa.ConnectionPool('Keyspace1')
>>> usercf = pycassa.ColumnFamily(pool, 'Users')
>>> cfmap = pycassa.ColumnFamilyMap(MyClass, usercf)
>>> users = cfmap.get_range(row_count=2, columns=['name', 'age'])
>>> for key, user in users:
... print user.name, user.age
Miles Davis 84
Winston Smith 42
"""
if 'columns' not in kwargs and not self.column_family.super and not self.raw_columns:
kwargs['columns'] = self.columns.keys()
for key, columns in self.column_family.get_range(*args, **kwargs):
if self.column_family.super:
if 'super_column' not in kwargs:
vals = self.dict_class()
for super_column, subcols in columns.iteritems():
combined = self.combine_columns(subcols)
vals[super_column] = create_instance(self.cls, key=key, super_column=super_column, **combined)
yield vals
else:
combined = self.combine_columns(columns)
yield create_instance(self.cls, key=key, super_column=kwargs['super_column'], **combined)
else:
combined = self.combine_columns(columns)
yield create_instance(self.cls, key=key, **combined)
def get_indexed_slices(self, instance=None, *args, **kwargs):
"""
Fetches a list of instances that satisfy an index clause. Similar
to :meth:`get_range()`, but uses an index clause instead of a key range.
If `instance` is supplied, its values will be used for each
:class:`IndexExpression` where the name matches one of the instance's
attribute names. This makes packing the values in the :class:`IndexExpresssion`
simpler when possible.
See :meth:`pycassa.columnfamily.ColumnFamily.get_indexed_slices()` for
an explanation of the parameters.
"""
if 'columns' not in kwargs and not self.column_family.super and not self.raw_columns:
kwargs['columns'] = self.columns.keys()
# Autopack the index clause's values
if instance is not None:
new_exprs = []
for expr in kwargs['index_clause'].expressions:
new_expr = IndexExpression(expr.column_name, expr.op,
value=self.columns[expr.column_name].pack(instance.__dict__[expr.column_name]))
new_exprs.append(new_expr)
old_clause = kwargs['index_clause']
new_clause = IndexClause(new_exprs, old_clause.start_key, old_clause.count)
kwargs['index_clause'] = new_clause
keyslice_map = self.column_family.get_indexed_slices(*args, **kwargs)
ret = self.dict_class()
for key, columns in keyslice_map:
if self.column_family.super:
if 'super_column' not in kwargs:
vals = self.dict_class()
for super_column, subcols in columns.iteritems():
combined = self.combine_columns(subcols)
vals[super_column] = create_instance(self.cls, key=key, super_column=super_column, **combined)
ret[key] = vals
else:
combined = self.combine_columns(columns)
ret[key] = create_instance(self.cls, key=key, super_column=kwargs['super_column'], **combined)
else:
combined = self.combine_columns(columns)
ret[key] = create_instance(self.cls, key=key, **combined)
return ret
def insert(self, instance, columns=None, write_consistency_level=None):
"""
Insert or update stored instances.
`instance` should be an instance of `cls` to store.
The `columns` parameter allows to you specify which attributes of
`instance` should be inserted or updated. If left as ``None``, all
attributes will be inserted.
"""
insert_dict = {}
if columns is None:
columns = self.columns.keys()
for column in columns:
if instance.__dict__.has_key(column) and instance.__dict__[column] is not None:
insert_dict[column] = self.columns[column].pack(instance.__dict__[column])
if self.column_family.super:
insert_dict = {instance.super_column: insert_dict}
return self.column_family.insert(instance.key, insert_dict,
write_consistency_level=write_consistency_level)
def remove(self, instance, columns=None, write_consistency_level=None):
"""
Removes a stored instance.
The `columns` parameter is a list of columns that should be removed.
If this is left as the default value of ``None``, the entire stored
instance will be removed.
"""
if self.column_family.super:
return self.column_family.remove(instance.key,
super_column=instance.super_column,
columns=columns,
write_consistency_level=write_consistency_level)
else:
return self.column_family.remove(instance.key, columns,
write_consistency_level=write_consistency_level)