/usr/lib/python3/dist-packages/minidb.py is in python3-minidb 2.0.2-2.
This file is owned by root:root, with mode 0o644.
The actual contents of the file can be viewed below.
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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833 834 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 850 851 852 853 854 855 856 857 858 859 860 861 862 863 864 865 866 867 868 869 870 871 872 873 874 875 876 877 878 879 880 881 882 883 884 885 886 887 888 889 890 891 892 893 894 895 896 897 898 899 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923 924 925 926 927 928 929 930 931 932 | # -*- coding: utf-8 -*-
# _ _ _ _
# _ __ (_)_ _ (_)__| | |__
# | ' \| | ' \| / _` | '_ \
# |_|_|_|_|_||_|_\__,_|_.__/
# simple python object store
#
# Copyright 2009-2010, 2014-2017 Thomas Perl <thp.io>. All rights reserved.
#
# Permission to use, copy, modify, and/or distribute this software for any
# purpose with or without fee is hereby granted, provided that the above
# copyright notice and this permission notice appear in all copies.
#
# THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
# WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
# MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
# ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
# WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
# ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
# OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
#
"""A simple SQLite3-based store for Python objects"""
__author__ = 'Thomas Perl <m@thp.io>'
__version__ = '2.0.2'
__url__ = 'http://thp.io/2010/minidb/'
__license__ = 'ISC'
__all__ = [
# Main classes
'Store', 'Model', 'JSON',
# Exceptions
'UnknownClass',
# Utility functions
'columns', 'func', 'literal',
# Decorator for registering converters
'converter_for',
# Debugging utilities
'pprint', 'pformat',
]
DEBUG_OBJECT_CACHE = False
CONVERTERS = {}
import sqlite3
import threading
import inspect
import functools
import types
import collections
import weakref
import sys
import json
import datetime
import logging
logger = logging.getLogger(__name__)
class UnknownClass(TypeError):
...
def converter_for(type_):
def decorator(f):
CONVERTERS[type_] = f
return f
return decorator
def _get_all_slots(class_, include_private=False):
for clazz in reversed(inspect.getmro(class_)):
if hasattr(clazz, '__minidb_slots__'):
for name, type_ in clazz.__minidb_slots__.items():
if include_private or not name.startswith('_'):
yield (name, type_)
def _set_attribute(o, slot, cls, value):
if value is None and hasattr(o.__class__, '__minidb_defaults__'):
value = getattr(o.__class__.__minidb_defaults__, slot, None)
if isinstance(value, types.FunctionType):
# Late-binding of default lambda (taking o as argument)
value = value(o)
if value is not None and cls not in CONVERTERS:
value = cls(value)
setattr(o, slot, value)
class RowProxy(object):
def __init__(self, row, keys):
self._row = row
self._keys = keys
def __getitem__(self, key):
if isinstance(key, str):
try:
index = self._keys.index(key)
except ValueError as ve:
raise KeyError(key)
return self._row[index]
return self._row[key]
def __getattr__(self, attr):
if attr not in self._keys:
raise AttributeError(attr)
return self[attr]
def __repr__(self):
return repr(self._row)
def keys(self):
return self._keys
class Store(object):
PRIMARY_KEY = ('id', int)
MINIDB_ATTR = '_minidb'
def __init__(self, filename=':memory:', debug=False, smartupdate=False):
self.db = sqlite3.connect(filename, check_same_thread=False)
self.debug = debug
self.smartupdate = smartupdate
self.registered = {}
self.lock = threading.RLock()
def __enter__(self):
return self
def __exit__(self, exc_type, exc_value, traceback):
if (exc_type, exc_value, traceback) is (None, None, None):
self.commit()
self.close()
def _execute(self, sql, args=None):
if args is None:
if self.debug:
logger.debug('%s', sql)
return self.db.execute(sql)
else:
if self.debug:
logger.debug('%s %r', sql, args)
return self.db.execute(sql, args)
def _schema(self, class_):
if class_ not in self.registered.values():
raise UnknownClass('{} was never registered'.format(class_))
return (class_.__name__, list(_get_all_slots(class_)))
def commit(self):
with self.lock:
self.db.commit()
def close(self):
with self.lock:
self.db.isolation_level = None
self._execute('VACUUM')
self.db.close()
def _ensure_schema(self, table, slots):
with self.lock:
cur = self._execute('PRAGMA table_info(%s)' % table)
available = cur.fetchall()
def column(name, type_, primary=True):
if (name, type_) == self.PRIMARY_KEY and primary:
return 'INTEGER PRIMARY KEY'
elif type_ in (int, bool):
return 'INTEGER'
elif type_ in (float,):
return 'REAL'
elif type_ in (bytes,):
return 'BLOB'
else:
return 'TEXT'
if available:
available = [(row[1], row[2]) for row in available]
modify_slots = [(name, type_) for name, type_ in slots if name in (name for name, _ in available) and
(name, column(name, type_, False)) not in available]
for name, type_ in modify_slots:
raise TypeError('Column {} is {}, but expected {}'.format(name, next(dbtype for n, dbtype in
available if n == name),
column(name, type_)))
# TODO: What to do with extraneous columns?
missing_slots = [(name, type_) for name, type_ in slots if name not in (n for n, _ in available)]
for name, type_ in missing_slots:
self._execute('ALTER TABLE %s ADD COLUMN %s %s' % (table, name, column(name, type_)))
else:
self._execute('CREATE TABLE %s (%s)' % (table, ', '.join('{} {}'.format(name, column(name, type_))
for name, type_ in slots)))
def register(self, class_, upgrade=False):
if not issubclass(class_, Model):
raise TypeError('{} is not a subclass of minidb.Model'.format(class_.__name__))
if class_ in self.registered.values():
raise TypeError('{} is already registered'.format(class_.__name__))
elif class_.__name__ in self.registered and not upgrade:
raise TypeError('{} is already registered {}'.format(class_.__name__, self.registered[class_.__name__]))
with self.lock:
self.registered[class_.__name__] = class_
table, slots = self._schema(class_)
self._ensure_schema(table, slots)
return class_
def serialize(self, v, t):
if v is None:
return None
elif t in CONVERTERS:
return CONVERTERS[t](v, True)
elif isinstance(v, bool):
return int(v)
elif isinstance(v, (int, float, bytes)):
return v
return str(v)
def deserialize(self, v, t):
if v is None:
return None
elif t in CONVERTERS:
return CONVERTERS[t](v, False)
elif isinstance(v, t):
return v
return t(v)
def save_or_update(self, o):
if o.id is None:
o.id = self.save(o)
else:
self._update(o)
def delete_by_pk(self, o):
with self.lock:
table, slots = self._schema(o.__class__)
assert self.PRIMARY_KEY in slots
pk_name, pk_type = self.PRIMARY_KEY
pk = getattr(o, pk_name)
assert pk is not None
res = self._execute('DELETE FROM %s WHERE %s = ?' % (table, pk_name), [pk])
setattr(o, pk_name, None)
def _update(self, o):
with self.lock:
table, slots = self._schema(o.__class__)
# Update requires a primary key
assert self.PRIMARY_KEY in slots
pk_name, pk_type = self.PRIMARY_KEY
if self.smartupdate:
existing = dict(next(self.query(o.__class__, where=lambda c:
getattr(c, pk_name) == getattr(o, pk_name))))
else:
existing = {}
values = [(name, type_, getattr(o, name, None))
for name, type_ in slots if (name, type_) != self.PRIMARY_KEY and
(name not in existing or getattr(o, name, None) != existing[name])]
if self.smartupdate and self.debug:
for name, type_, to_value in values:
logger.debug('%s %s', '{}(id={})'.format(table, o.id),
'{}: {} -> {}'.format(name, existing[name], to_value))
if not values:
# No values have changed - nothing to update
return
def gen_keys():
for name, type_, value in values:
if value is not None:
yield '{name}=?'.format(name=name)
else:
yield '{name}=NULL'.format(name=name)
def gen_values():
for name, type_, value in values:
if value is not None:
yield self.serialize(value, type_)
yield getattr(o, pk_name)
self._execute('UPDATE %s SET %s WHERE %s = ?' % (table, ', '.join(gen_keys()), pk_name), list(gen_values()))
def save(self, o):
with self.lock:
table, slots = self._schema(o.__class__)
# Save all values except for the primary key
slots = [(name, type_) for name, type_ in slots if (name, type_) != self.PRIMARY_KEY]
values = [self.serialize(getattr(o, name), type_) for name, type_ in slots]
return self._execute('INSERT INTO %s (%s) VALUES (%s)' % (table, ', '.join(name for name, type_ in slots),
', '.join('?' * len(slots))), values).lastrowid
def delete_where(self, class_, where):
with self.lock:
table, slots = self._schema(class_)
if isinstance(where, types.FunctionType):
# Late-binding of where
where = where(class_.c)
ssql, args = where.tosql()
sql = 'DELETE FROM %s WHERE %s' % (table, ssql)
return self._execute(sql, args).rowcount
def query(self, class_, select=None, where=None, order_by=None, group_by=None, limit=None):
with self.lock:
table, slots = self._schema(class_)
attr_to_type = dict(slots)
sql = []
args = []
if select is None:
select = literal('*')
if isinstance(select, types.FunctionType):
# Late-binding of columns
select = select(class_.c)
# Select can always be a sequence
if not isinstance(select, Sequence):
select = Sequence([select])
# Look for RenameOperation operations in the SELECT sequence and
# remember the column types, so we can decode values properly later
for arg in select.args:
if isinstance(arg, Operation):
if isinstance(arg.a, RenameOperation):
if isinstance(arg.a.column, Column):
attr_to_type[arg.a.name] = arg.a.column.type_
ssql, sargs = select.tosql()
sql.append('SELECT %s FROM %s' % (ssql, table))
args.extend(sargs)
if where is not None:
if isinstance(where, types.FunctionType):
# Late-binding of columns
where = where(class_.c)
wsql, wargs = where.tosql()
sql.append('WHERE %s' % (wsql,))
args.extend(wargs)
if order_by is not None:
if isinstance(order_by, types.FunctionType):
# Late-binding of columns
order_by = order_by(class_.c)
osql, oargs = order_by.tosql()
sql.append('ORDER BY %s' % (osql,))
args.extend(oargs)
if group_by is not None:
if isinstance(group_by, types.FunctionType):
# Late-binding of columns
group_by = group_by(class_.c)
gsql, gargs = group_by.tosql()
sql.append('GROUP BY %s' % (gsql,))
args.extend(gargs)
if limit is not None:
sql.append('LIMIT ?')
args.append(limit)
sql = ' '.join(sql)
result = self._execute(sql, args)
columns = [d[0] for d in result.description]
def _decode(row, columns):
for name, value in zip(columns, row):
type_ = attr_to_type.get(name, None)
yield (self.deserialize(value, type_) if type_ is not None else value)
return (RowProxy(tuple(_decode(row, columns)), columns) for row in result)
def load(self, class_, *args, **kwargs):
with self.lock:
query = kwargs.get('__query__', None)
if '__query__' in kwargs:
del kwargs['__query__']
table, slots = self._schema(class_)
sql = 'SELECT %s FROM %s' % (', '.join(name for name, type_ in slots), table)
if query:
if isinstance(query, types.FunctionType):
# Late-binding of query
query = query(class_.c)
ssql, aargs = query.tosql()
sql += ' WHERE %s' % ssql
sql_args = aargs
elif kwargs:
sql += ' WHERE %s' % (' AND '.join('%s = ?' % k for k in kwargs))
sql_args = list(kwargs.values())
else:
sql_args = []
cur = self._execute(sql, sql_args)
def apply(row):
row = zip(slots, row)
kwargs = {name: self.deserialize(v, type_) for (name, type_), v in row if v is not None}
o = class_(*args, **kwargs)
setattr(o, self.MINIDB_ATTR, self)
return o
return (x for x in (apply(row) for row in cur) if x is not None)
def get(self, class_, *args, **kwargs):
it = self.load(class_, *args, **kwargs)
result = next(it, None)
try:
next(it)
except StopIteration:
return result
raise ValueError('More than one row returned')
class Operation(object):
def __init__(self, a, op=None, b=None, brackets=False):
self.a = a
self.op = op
self.b = b
self.brackets = brackets
def _get_class(self, a):
if isinstance(a, Column):
return a.class_
elif isinstance(a, RenameOperation):
return self._get_class(a.column)
elif isinstance(a, Function):
return a.args[0].class_
elif isinstance(a, Sequence):
return a.args[0].class_
raise ValueError('Cannot determine class for query')
def query(self, db, where=None, order_by=None, group_by=None, limit=None):
return self._get_class(self.a).query(db, self, where=where, order_by=order_by, group_by=group_by, limit=limit)
def __floordiv__(self, other):
if self.b is not None:
raise ValueError('Cannot sequence columns')
return Sequence([self, other])
def argtosql(self, arg):
if isinstance(arg, Operation):
return arg.tosql(self.brackets)
elif isinstance(arg, Column):
return (arg.name, [])
elif isinstance(arg, RenameOperation):
columnname, args = arg.column.tosql()
return ('%s AS %s' % (columnname, arg.name), args)
elif isinstance(arg, Function):
sqls = []
argss = []
for farg in arg.args:
sql, args = self.argtosql(farg)
sqls.append(sql)
argss.extend(args)
return ['%s(%s)' % (arg.name, ', '.join(sqls)), argss]
elif isinstance(arg, Sequence):
sqls = []
argss = []
for farg in arg.args:
sql, args = self.argtosql(farg)
sqls.append(sql)
argss.extend(args)
return ['%s' % ', '.join(sqls), argss]
elif isinstance(arg, Literal):
return [arg.name, []]
if type(arg) in CONVERTERS:
return ('?', [CONVERTERS[type(arg)](arg, True)])
return ('?', [arg])
def tosql(self, brackets=False):
sql = []
args = []
ssql, aargs = self.argtosql(self.a)
sql.append(ssql)
args.extend(aargs)
if self.op is not None:
sql.append(self.op)
if self.b is not None:
ssql, aargs = self.argtosql(self.b)
sql.append(ssql)
args.extend(aargs)
if brackets:
sql.insert(0, '(')
sql.append(')')
return (' '.join(sql), args)
def __and__(self, other):
return Operation(self, 'AND', other, True)
def __or__(self, other):
return Operation(self, 'OR', other, True)
def __repr__(self):
if self.b is None:
if self.op is None:
return '{self.a!r}'.format(self=self)
return '{self.a!r} {self.op}'.format(self=self)
return '{self.a!r} {self.op} {self.b!r}'.format(self=self)
class Sequence(object):
def __init__(self, args):
self.args = args
def __repr__(self):
return ', '.join(repr(arg) for arg in self.args)
def tosql(self):
return Operation(self).tosql()
def query(self, db, order_by=None, group_by=None, limit=None):
return Operation(self).query(db, order_by=order_by, group_by=group_by, limit=limit)
def __floordiv__(self, other):
self.args.append(other)
return self
def columns(*args):
"""columns(a, b, c) -> a // b // c
Query multiple columns, like the // column sequence operator.
"""
return Sequence(args)
class func(object):
max = staticmethod(lambda *args: Function('max', *args))
min = staticmethod(lambda *args: Function('min', *args))
sum = staticmethod(lambda *args: Function('sum', *args))
distinct = staticmethod(lambda *args: Function('distinct', *args))
random = staticmethod(lambda: Function('random'))
abs = staticmethod(lambda a: Function('abs', a))
length = staticmethod(lambda a: Function('length', a))
lower = staticmethod(lambda a: Function('lower', a))
upper = staticmethod(lambda a: Function('upper', a))
ltrim = staticmethod(lambda a: Function('ltrim', a))
rtrim = staticmethod(lambda a: Function('rtrim', a))
trim = staticmethod(lambda a: Function('trim', a))
count = staticmethod(lambda a: Function('count', a))
__call__ = lambda a, name: RenameOperation(a, name)
class OperatorMixin(object):
__lt__ = lambda a, b: Operation(a, '<', b)
__le__ = lambda a, b: Operation(a, '<=', b)
__eq__ = lambda a, b: Operation(a, '=', b) if b is not None else Operation(a, 'IS NULL')
__ne__ = lambda a, b: Operation(a, '!=', b) if b is not None else Operation(a, 'IS NOT NULL')
__gt__ = lambda a, b: Operation(a, '>', b)
__ge__ = lambda a, b: Operation(a, '>=', b)
__call__ = lambda a, name: RenameOperation(a, name)
tosql = lambda a: Operation(a).tosql()
query = lambda a, db, where=None, order_by=None, group_by=None, limit=None: Operation(a).query(db, where=where,
order_by=order_by,
group_by=group_by,
limit=limit)
__floordiv__ = lambda a, b: Sequence([a, b])
like = lambda a, b: Operation(a, 'LIKE', b)
avg = property(lambda a: Function('avg', a))
max = property(lambda a: Function('max', a))
min = property(lambda a: Function('min', a))
sum = property(lambda a: Function('sum', a))
distinct = property(lambda a: Function('distinct', a))
asc = property(lambda a: Operation(a, 'ASC'))
desc = property(lambda a: Operation(a, 'DESC'))
abs = property(lambda a: Function('abs', a))
length = property(lambda a: Function('length', a))
lower = property(lambda a: Function('lower', a))
upper = property(lambda a: Function('upper', a))
ltrim = property(lambda a: Function('ltrim', a))
rtrim = property(lambda a: Function('rtrim', a))
trim = property(lambda a: Function('trim', a))
count = property(lambda a: Function('count', a))
class RenameOperation(OperatorMixin):
def __init__(self, column, name):
self.column = column
self.name = name
def __repr__(self):
return '%r AS %s' % (self.column, self.name)
class Literal(OperatorMixin):
def __init__(self, name):
self.name = name
def __repr__(self):
return self.name
def literal(name):
"""Insert a literal as-is into a SQL query
>>> func.count(literal('*'))
count(*)
"""
return Literal(name)
class Function(OperatorMixin):
def __init__(self, name, *args):
self.name = name
self.args = args
def __repr__(self):
return '%s(%s)' % (self.name, ', '.join(repr(arg) for arg in self.args))
class Column(OperatorMixin):
def __init__(self, class_, name, type_):
self.class_ = class_
self.name = name
self.type_ = type_
def __repr__(self):
return '.'.join((self.class_.__name__, self.name))
class Columns(object):
def __init__(self, name, slots):
self._class = None
self._name = name
self._slots = slots
def __repr__(self):
return '<{} for {} ({})>'.format(self.__class__.__name__, self._name, ', '.join(self._slots))
def __getattr__(self, name):
d = {k: v for k, v in _get_all_slots(self._class, include_private=True)}
if name not in d:
raise AttributeError(name)
return Column(self._class, name, d[name])
def model_init(self, *args, **kwargs):
slots = list(_get_all_slots(self.__class__, include_private=True))
unmatched_kwargs = set(kwargs.keys()).difference(set(key for key, type_ in slots))
if unmatched_kwargs:
raise KeyError('Invalid keyword argument(s): %r' % unmatched_kwargs)
for key, type_ in slots:
_set_attribute(self, key, type_, kwargs.get(key, None))
# Call redirected constructor
if '__minidb_init__' in self.__class__.__dict__:
getattr(self, '__minidb_init__')(*args)
class MetaModel(type):
@classmethod
def __prepare__(metacls, name, bases):
return collections.OrderedDict()
def __new__(mcs, name, bases, d):
# Redirect __init__() to __minidb_init__()
if '__init__' in d:
d['__minidb_init__'] = d['__init__']
d['__init__'] = model_init
# Caching of live objects
d['__minidb_cache__'] = weakref.WeakValueDictionary()
slots = collections.OrderedDict((k, v) for k, v in d.items()
if k.lower() == k and
not k.startswith('__') and
not isinstance(v, types.FunctionType) and
not isinstance(v, property) and
not isinstance(v, staticmethod) and
not isinstance(v, classmethod))
keep = collections.OrderedDict((k, v) for k, v in d.items() if k not in slots)
keep['__minidb_slots__'] = slots
keep['__slots__'] = tuple(slots.keys())
if not bases:
# Add weakref slot to Model (for caching)
keep['__slots__'] += ('__weakref__',)
columns = Columns(name, slots)
keep['c'] = columns
result = type.__new__(mcs, name, bases, keep)
columns._class = result
return result
def pformat(result, color=False):
def incolor(color_id, s):
return '\033[9%dm%s\033[0m' % (color_id, s) if sys.stdout.isatty() and color else s
inred, ingreen, inyellow, inblue = (functools.partial(incolor, x) for x in range(1, 5))
rows = list(result)
if not rows:
return '(no rows)'
def colorvalue(formatted, value):
if value is None:
return inred(formatted)
if isinstance(value, bool):
return ingreen(formatted)
return formatted
s = []
keys = rows[0].keys()
lengths = tuple(max(x) for x in zip(*[[len(str(column)) for column in row] for row in [keys] + rows]))
s.append(' | '.join(inyellow('%-{}s'.format(length) % key) for key, length in zip(keys, lengths)))
s.append('-+-'.join('-' * length for length in lengths))
for row in rows:
s.append(' | '.join(colorvalue('%-{}s'.format(length) % col, col) for col, length in zip(row, lengths)))
s.append('({} row(s))'.format(len(rows)))
return ('\n'.join(s))
def pprint(result, color=False):
print(pformat(result, color))
class JSON(object):
...
@converter_for(JSON)
def convert_json(v, serialize):
return json.dumps(v) if serialize else json.loads(v)
@converter_for(datetime.datetime)
def convert_datetime_datetime(v, serialize):
"""
>>> convert_datetime_datetime(datetime.datetime(2014, 12, 13, 14, 15), True)
'2014-12-13T14:15:00'
>>> convert_datetime_datetime('2014-12-13T14:15:16', False)
datetime.datetime(2014, 12, 13, 14, 15, 16)
"""
if serialize:
return v.isoformat()
else:
isoformat, microseconds = (v.rsplit('.', 1) if '.' in v else (v, 0))
return (datetime.datetime.strptime(isoformat, '%Y-%m-%dT%H:%M:%S') +
datetime.timedelta(microseconds=int(microseconds)))
@converter_for(datetime.date)
def convert_datetime_date(v, serialize):
"""
>>> convert_datetime_date(datetime.date(2014, 12, 13), True)
'2014-12-13'
>>> convert_datetime_date('2014-12-13', False)
datetime.date(2014, 12, 13)
"""
if serialize:
return v.isoformat()
else:
return datetime.datetime.strptime(v, '%Y-%m-%d').date()
@converter_for(datetime.time)
def convert_datetime_time(v, serialize):
"""
>>> convert_datetime_time(datetime.time(14, 15, 16), True)
'14:15:16'
>>> convert_datetime_time('14:15:16', False)
datetime.time(14, 15, 16)
"""
if serialize:
return v.isoformat()
else:
isoformat, microseconds = (v.rsplit('.', 1) if '.' in v else (v, 0))
return (datetime.datetime.strptime(isoformat, '%H:%M:%S') +
datetime.timedelta(microseconds=int(microseconds))).time()
class Model(metaclass=MetaModel):
id = int
_minidb = Store
@classmethod
def _finalize(cls, id):
if DEBUG_OBJECT_CACHE:
logger.debug('Finalizing {} id={}'.format(cls.__name__, id))
def __repr__(self):
def get_attrs():
for key, type_ in _get_all_slots(self.__class__):
yield key, getattr(self, key, None)
attrs = ['{key}={value!r}'.format(key=key, value=value) for key, value in get_attrs()]
return '<%(cls)s(%(attrs)s)>' % {
'cls': self.__class__.__name__,
'attrs': ', '.join(attrs),
}
@classmethod
def __lookup_single(cls, o):
if o is None:
return None
cache = cls.__minidb_cache__
if o.id not in cache:
if DEBUG_OBJECT_CACHE:
logger.debug('Storing id={} in cache {}'.format(o.id, o))
weakref.finalize(o, cls._finalize, o.id)
cache[o.id] = o
else:
if DEBUG_OBJECT_CACHE:
logger.debug('Getting id={} from cache'.format(o.id))
return cache[o.id]
@classmethod
def __lookup_cache(cls, objects):
for o in objects:
yield cls.__lookup_single(o)
@classmethod
def load(cls, db, query=None, **kwargs):
if query is not None:
kwargs['__query__'] = query
if '__minidb_init__' in cls.__dict__:
@functools.wraps(cls.__minidb_init__)
def init_wrapper(*args):
return cls.__lookup_cache(db.load(cls, *args, **kwargs))
return init_wrapper
else:
return cls.__lookup_cache(db.load(cls, **kwargs))
@classmethod
def get(cls, db, query=None, **kwargs):
if query is not None:
kwargs['__query__'] = query
if '__minidb_init__' in cls.__dict__:
@functools.wraps(cls.__minidb_init__)
def init_wrapper(*args):
return cls.__lookup_single(db.get(cls, *args, **kwargs))
return init_wrapper
else:
return cls.__lookup_single(db.get(cls, **kwargs))
def save(self, db=None):
if getattr(self, Store.MINIDB_ATTR, None) is None:
if db is None:
raise ValueError('Needs a db object')
setattr(self, Store.MINIDB_ATTR, db)
getattr(self, Store.MINIDB_ATTR).save_or_update(self)
if DEBUG_OBJECT_CACHE:
logger.debug('Storing id={} in cache {}'.format(self.id, self))
weakref.finalize(self, self.__class__._finalize, self.id)
self.__class__.__minidb_cache__[self.id] = self
return self
def delete(self):
if getattr(self, Store.MINIDB_ATTR) is None:
raise ValueError('Needs a db object')
elif self.id is None:
raise KeyError('id is None (not stored in db?)')
# drop from cache
cache = self.__class__.__minidb_cache__
if self.id in cache:
if DEBUG_OBJECT_CACHE:
logger.debug('Dropping id={} from cache {}'.format(self.id, self))
del cache[self.id]
getattr(self, Store.MINIDB_ATTR).delete_by_pk(self)
@classmethod
def delete_where(cls, db, query):
return db.delete_where(cls, query)
@classmethod
def query(cls, db, select=None, where=None, order_by=None, group_by=None, limit=None):
return db.query(cls, select=select, where=where, order_by=order_by, group_by=group_by, limit=limit)
@classmethod
def pquery(cls, db, select=None, where=None, order_by=None, group_by=None, limit=None, color=True):
pprint(db.query(cls, select=select, where=where, order_by=order_by, group_by=group_by, limit=limit), color)
|