| 1 |
"""Base classes and tools for writing database Storage Managers. |
|---|
| 2 |
|
|---|
| 3 |
DATA TYPES |
|---|
| 4 |
========== |
|---|
| 5 |
Database Storage Manager modules are mostly adapters to support round-trip |
|---|
| 6 |
data coercion: |
|---|
| 7 |
|
|---|
| 8 |
Unit type -> [SQL repr ->] DB -> incoming Python value -> Unit type |
|---|
| 9 |
|
|---|
| 10 |
Since Dejavu relies on external database servers for its persistence, |
|---|
| 11 |
Python datatypes must be converted to column types in the DB. When writing |
|---|
| 12 |
a StorageManager, you should make sure that your type conversions can handle |
|---|
| 13 |
at least the following limitations: If possible, implement the type with no |
|---|
| 14 |
limits. Also, follow UnitProperty.hints['bytes'] where possible. A value |
|---|
| 15 |
of zero for hints['bytes'] implies no limit. If no value is given, try to |
|---|
| 16 |
assume no limit, although you may choose whatever default size you wish |
|---|
| 17 |
(255 is common for strings). |
|---|
| 18 |
|
|---|
| 19 |
ENCODING ISSUES |
|---|
| 20 |
=============== |
|---|
| 21 |
All SQL sent to the database must be strings, not unicode. You can set the |
|---|
| 22 |
encoding of the Adapters (I may add a more centralized encoding context in |
|---|
| 23 |
the future). We must use encoded strings so that we can mix encodings |
|---|
| 24 |
within the same string; for example, we might have a DB which understands |
|---|
| 25 |
utf8, but a pickle value which will be encoded in raw-unicode-escape inline |
|---|
| 26 |
with that. All values, therefore, must be coerced before we try to join |
|---|
| 27 |
them into an SQL statement string. |
|---|
| 28 |
|
|---|
| 29 |
""" |
|---|
| 30 |
|
|---|
| 31 |
import threading |
|---|
| 32 |
import warnings |
|---|
| 33 |
|
|---|
| 34 |
|
|---|
| 35 |
import geniusql |
|---|
| 36 |
from geniusql import logic, logicfuncs |
|---|
| 37 |
|
|---|
| 38 |
import dejavu |
|---|
| 39 |
from dejavu import storage, logflags, xray |
|---|
| 40 |
from dejavu.errors import StorageWarning, MappingError |
|---|
| 41 |
|
|---|
| 42 |
|
|---|
| 43 |
|
|---|
| 44 |
|
|---|
| 45 |
|
|---|
| 46 |
class StorageManagerDB(storage.StorageManager): |
|---|
| 47 |
"""StoreManager base class to save and retrieve Units using a DB.""" |
|---|
| 48 |
|
|---|
| 49 |
databaseclass = geniusql.Database |
|---|
| 50 |
|
|---|
| 51 |
def __init__(self, allOptions={}): |
|---|
| 52 |
storage.StorageManager.__init__(self, allOptions) |
|---|
| 53 |
self.reserve_lock = threading.Lock() |
|---|
| 54 |
|
|---|
| 55 |
|
|---|
| 56 |
def get_option(name): |
|---|
| 57 |
item = allOptions.get(name) |
|---|
| 58 |
if isinstance(item, basestring): |
|---|
| 59 |
item = xray.classes(item) |
|---|
| 60 |
return item |
|---|
| 61 |
|
|---|
| 62 |
dbclass = get_option('Database Class') |
|---|
| 63 |
if dbclass: |
|---|
| 64 |
self.databaseclass = dbclass |
|---|
| 65 |
|
|---|
| 66 |
allOptions = dict([(str(k), v) for k, v in allOptions.iteritems()]) |
|---|
| 67 |
|
|---|
| 68 |
self.db = self.databaseclass(**allOptions) |
|---|
| 69 |
self.schema = self.db.schema() |
|---|
| 70 |
|
|---|
| 71 |
if 'Prefix' in allOptions: |
|---|
| 72 |
self.schema.prefix = allOptions['Prefix'] |
|---|
| 73 |
|
|---|
| 74 |
def logger(msg): |
|---|
| 75 |
if self.logflags & logflags.SQL: |
|---|
| 76 |
self.log(logflags.SQL.message(msg)) |
|---|
| 77 |
self.db.log = logger |
|---|
| 78 |
|
|---|
| 79 |
def version(self): |
|---|
| 80 |
return self.db.version() |
|---|
| 81 |
|
|---|
| 82 |
def shutdown(self): |
|---|
| 83 |
self.db.connections.shutdown() |
|---|
| 84 |
|
|---|
| 85 |
def xrecall(self, cls, expr=None, order=None, limit=None, offset=None): |
|---|
| 86 |
"""Yield a sequence of Unit instances which satisfy the expression.""" |
|---|
| 87 |
if self.logflags & logflags.RECALL: |
|---|
| 88 |
self.log(logflags.RECALL.message(cls, expr)) |
|---|
| 89 |
|
|---|
| 90 |
clsname = cls.__name__ |
|---|
| 91 |
|
|---|
| 92 |
|
|---|
| 93 |
|
|---|
| 94 |
idnames = list(cls.identifiers) |
|---|
| 95 |
attrs = idnames + [x for x in cls.properties if x not in idnames] |
|---|
| 96 |
coercers = [getattr(cls, key).coerce for key in attrs] |
|---|
| 97 |
|
|---|
| 98 |
data = self.select((cls, attrs, expr), order=order, |
|---|
| 99 |
limit=limit, offset=offset) |
|---|
| 100 |
for row in data: |
|---|
| 101 |
unit = cls.__new__(cls) |
|---|
| 102 |
unit._zombie = True |
|---|
| 103 |
unit.__init__() |
|---|
| 104 |
|
|---|
| 105 |
for key, value, propcoerce in zip(attrs, row, coercers): |
|---|
| 106 |
try: |
|---|
| 107 |
if propcoerce: |
|---|
| 108 |
value = propcoerce(unit, value) |
|---|
| 109 |
unit._properties[key] = value |
|---|
| 110 |
except UnicodeDecodeError, x: |
|---|
| 111 |
x.reason += " [%r: %r]" % (key, value) |
|---|
| 112 |
raise |
|---|
| 113 |
except Exception, x: |
|---|
| 114 |
x.args += (key, value) |
|---|
| 115 |
raise |
|---|
| 116 |
|
|---|
| 117 |
|
|---|
| 118 |
|
|---|
| 119 |
if expr and data.statement.imperfect: |
|---|
| 120 |
if not expr(unit): |
|---|
| 121 |
continue |
|---|
| 122 |
|
|---|
| 123 |
unit.cleanse() |
|---|
| 124 |
yield unit |
|---|
| 125 |
|
|---|
| 126 |
def reserve(self, unit): |
|---|
| 127 |
"""Reserve a persistent slot for unit.""" |
|---|
| 128 |
self.reserve_lock.acquire() |
|---|
| 129 |
try: |
|---|
| 130 |
|
|---|
| 131 |
|
|---|
| 132 |
seqclass = unit.sequencer.__class__.__name__ |
|---|
| 133 |
seq_handler = getattr(self, "_seq_%s" % seqclass, None) |
|---|
| 134 |
if seq_handler: |
|---|
| 135 |
seq_handler(unit) |
|---|
| 136 |
else: |
|---|
| 137 |
self._manual_reserve(unit) |
|---|
| 138 |
unit.cleanse() |
|---|
| 139 |
finally: |
|---|
| 140 |
self.reserve_lock.release() |
|---|
| 141 |
|
|---|
| 142 |
|
|---|
| 143 |
|
|---|
| 144 |
if self.logflags & logflags.RESERVE: |
|---|
| 145 |
self.log(logflags.RESERVE.message(unit)) |
|---|
| 146 |
|
|---|
| 147 |
def _seq_UnitSequencerInteger(self, unit): |
|---|
| 148 |
"""Reserve a unit (using the table's autoincrement fields).""" |
|---|
| 149 |
cls = unit.__class__ |
|---|
| 150 |
|
|---|
| 151 |
|
|---|
| 152 |
newids = self.schema[cls.__name__].insert(**unit._properties) |
|---|
| 153 |
for k, v in newids.iteritems(): |
|---|
| 154 |
setattr(unit, k, v) |
|---|
| 155 |
|
|---|
| 156 |
def _manual_reserve(self, unit): |
|---|
| 157 |
"""Use when the DB cannot automatically generate an identifier. |
|---|
| 158 |
The identifiers will be supplied by UnitSequencer.assign(). |
|---|
| 159 |
""" |
|---|
| 160 |
cls = unit.__class__ |
|---|
| 161 |
t = self.schema[cls.__name__] |
|---|
| 162 |
if not unit.sequencer.valid_id(unit.identity()): |
|---|
| 163 |
|
|---|
| 164 |
data = list(self.db.select((t, cls.identifiers))) |
|---|
| 165 |
cls.sequencer.assign(unit, data) |
|---|
| 166 |
t.insert(**unit._properties) |
|---|
| 167 |
|
|---|
| 168 |
def save(self, unit, forceSave=False): |
|---|
| 169 |
"""Update storage from unit's data (if unit.dirty()).""" |
|---|
| 170 |
if self.logflags & logflags.SAVE: |
|---|
| 171 |
self.log(logflags.SAVE.message(unit, forceSave)) |
|---|
| 172 |
|
|---|
| 173 |
if forceSave or unit.dirty(): |
|---|
| 174 |
self.schema[unit.__class__.__name__].save(**unit._properties) |
|---|
| 175 |
unit.cleanse() |
|---|
| 176 |
|
|---|
| 177 |
def destroy(self, unit): |
|---|
| 178 |
"""Delete the unit.""" |
|---|
| 179 |
if self.logflags & logflags.DESTROY: |
|---|
| 180 |
self.log(logflags.DESTROY.message(unit)) |
|---|
| 181 |
|
|---|
| 182 |
table = self.schema[unit.__class__.__name__] |
|---|
| 183 |
table.delete(**unit._properties) |
|---|
| 184 |
|
|---|
| 185 |
|
|---|
| 186 |
|
|---|
| 187 |
|
|---|
| 188 |
def tablejoin(self, join): |
|---|
| 189 |
"""Return a geniusql Join tree for the given UnitJoin.""" |
|---|
| 190 |
t1, t2 = join.class1, join.class2 |
|---|
| 191 |
|
|---|
| 192 |
if isinstance(t1, dejavu.UnitJoin): |
|---|
| 193 |
wt1 = self.tablejoin(t1) |
|---|
| 194 |
else: |
|---|
| 195 |
wt1 = self.schema[t1.__name__] |
|---|
| 196 |
|
|---|
| 197 |
if isinstance(t2, dejavu.UnitJoin): |
|---|
| 198 |
wt2 = self.tablejoin(t2) |
|---|
| 199 |
else: |
|---|
| 200 |
wt2 = self.schema[t2.__name__] |
|---|
| 201 |
|
|---|
| 202 |
uj = geniusql.sqlwriters.Join(wt1, wt2, join.leftbiased) |
|---|
| 203 |
|
|---|
| 204 |
|
|---|
| 205 |
uj.path = join.path |
|---|
| 206 |
return uj |
|---|
| 207 |
|
|---|
| 208 |
def _geniusql_query(self, query): |
|---|
| 209 |
"""Return a Geniusql Query object for the given Dejavu Query.""" |
|---|
| 210 |
rel = query.relation |
|---|
| 211 |
if isinstance(rel, dejavu.UnitJoin): |
|---|
| 212 |
rel = self.tablejoin(rel) |
|---|
| 213 |
elif rel is None: |
|---|
| 214 |
|
|---|
| 215 |
rel = self.schema |
|---|
| 216 |
else: |
|---|
| 217 |
rel = self.schema[rel.__name__] |
|---|
| 218 |
return geniusql.sqlwriters.Query(rel, query.attributes, query.restriction) |
|---|
| 219 |
|
|---|
| 220 |
def select(self, query, order=None, distinct=False, limit=None, offset=None): |
|---|
| 221 |
"""Return a geniusql Dataset for the given Query object.""" |
|---|
| 222 |
if not isinstance(query, dejavu.Query): |
|---|
| 223 |
query = dejavu.Query(*query) |
|---|
| 224 |
|
|---|
| 225 |
return self.db.select(self._geniusql_query(query), |
|---|
| 226 |
order=order, distinct=distinct, |
|---|
| 227 |
limit=limit, offset=offset, strict=False) |
|---|
| 228 |
|
|---|
| 229 |
def insert_into(self, name, query, distinct=False): |
|---|
| 230 |
"""INSERT matching data INTO a new class and return the class.""" |
|---|
| 231 |
if not isinstance(query, dejavu.Query): |
|---|
| 232 |
query = dejavu.Query(*query) |
|---|
| 233 |
|
|---|
| 234 |
self.db.insert_into(name, self._geniusql_query(query), |
|---|
| 235 |
distinct=distinct) |
|---|
| 236 |
return Modeler(self.schema).make_class(name) |
|---|
| 237 |
|
|---|
| 238 |
def make_class(self, name): |
|---|
| 239 |
"""Return a (new) Unit class for the given storage name.""" |
|---|
| 240 |
return Modeler(self.schema).make_class(name) |
|---|
| 241 |
|
|---|
| 242 |
def xview(self, query, distinct=False): |
|---|
| 243 |
"""Yield value tuples for the given query.""" |
|---|
| 244 |
if not isinstance(query, dejavu.Query): |
|---|
| 245 |
query = dejavu.Query(*query) |
|---|
| 246 |
|
|---|
| 247 |
if self.logflags & logflags.VIEW: |
|---|
| 248 |
self.log(logflags.VIEW.message(query, distinct)) |
|---|
| 249 |
|
|---|
| 250 |
data = self.select(query, distinct=distinct) |
|---|
| 251 |
if data.statement.imperfect: |
|---|
| 252 |
|
|---|
| 253 |
|
|---|
| 254 |
|
|---|
| 255 |
clsname = self.__class__.__name__ |
|---|
| 256 |
warnings.warn("The requested query cannot produce perfect SQL " |
|---|
| 257 |
"with a %s datasource. It may take an absurdly " |
|---|
| 258 |
"long time to run, since each unit must be fully-" |
|---|
| 259 |
"formed. %s" % (clsname, query), StorageWarning) |
|---|
| 260 |
for row in storage.StorageManager.xview(self, query, distinct): |
|---|
| 261 |
yield row |
|---|
| 262 |
else: |
|---|
| 263 |
|
|---|
| 264 |
for row in data: |
|---|
| 265 |
yield tuple(row) |
|---|
| 266 |
|
|---|
| 267 |
def count(self, cls, expr=None): |
|---|
| 268 |
"""Number of Units of the given cls which match the given expr.""" |
|---|
| 269 |
if cls.identifiers: |
|---|
| 270 |
uniq = cls.identifiers |
|---|
| 271 |
else: |
|---|
| 272 |
uniq = cls._properties.keys() |
|---|
| 273 |
|
|---|
| 274 |
counter = lambda x: [logicfuncs.count(getattr(x, uniq[0]))] |
|---|
| 275 |
|
|---|
| 276 |
query = dejavu.Query(cls, counter, expr) |
|---|
| 277 |
|
|---|
| 278 |
if self.logflags & logflags.VIEW: |
|---|
| 279 |
self.log(logflags.VIEW.message(query, False)) |
|---|
| 280 |
|
|---|
| 281 |
data = self.select(query) |
|---|
| 282 |
if data.statement.imperfect: |
|---|
| 283 |
|
|---|
| 284 |
|
|---|
| 285 |
|
|---|
| 286 |
clsname = self.__class__.__name__ |
|---|
| 287 |
warnings.warn("The requested query cannot produce perfect SQL " |
|---|
| 288 |
"with a %s datasource. It may take an absurdly " |
|---|
| 289 |
"long time to run, since each unit must be fully-" |
|---|
| 290 |
"formed. %s" % (clsname, query), StorageWarning) |
|---|
| 291 |
return storage.StorageManager.count(self, cls, expr) |
|---|
| 292 |
else: |
|---|
| 293 |
return data.scalar() |
|---|
| 294 |
|
|---|
| 295 |
def xmultirecall(self, classes, expr=None, order=None, limit=None, offset=None): |
|---|
| 296 |
"""Yield Unit instance sets which satisfy the expression.""" |
|---|
| 297 |
if self.logflags & logflags.RECALL: |
|---|
| 298 |
self.log(logflags.RECALL.message(classes, expr)) |
|---|
| 299 |
|
|---|
| 300 |
|
|---|
| 301 |
allattrs = [] |
|---|
| 302 |
props = [] |
|---|
| 303 |
for cls in classes: |
|---|
| 304 |
t = self.schema[cls.__name__] |
|---|
| 305 |
attrs = [] |
|---|
| 306 |
for key in cls.properties: |
|---|
| 307 |
attrs.append(key) |
|---|
| 308 |
props.append((cls, key, getattr(cls, key).coerce)) |
|---|
| 309 |
allattrs.append(attrs) |
|---|
| 310 |
|
|---|
| 311 |
data = self.select((classes, allattrs, expr), order=order, |
|---|
| 312 |
limit=limit, offset=offset) |
|---|
| 313 |
for row in data: |
|---|
| 314 |
|
|---|
| 315 |
units = {} |
|---|
| 316 |
for i, (cls, key, propcoerce) in enumerate(props): |
|---|
| 317 |
if cls in units: |
|---|
| 318 |
unit = units[cls] |
|---|
| 319 |
else: |
|---|
| 320 |
unit = cls.__new__(cls) |
|---|
| 321 |
unit._zombie = True |
|---|
| 322 |
unit.__init__() |
|---|
| 323 |
units[cls] = unit |
|---|
| 324 |
|
|---|
| 325 |
value = row[i] |
|---|
| 326 |
try: |
|---|
| 327 |
if propcoerce: |
|---|
| 328 |
value = propcoerce(unit, value) |
|---|
| 329 |
unit._properties[key] = value |
|---|
| 330 |
except Exception, x: |
|---|
| 331 |
x.args += (cls, key) |
|---|
| 332 |
raise |
|---|
| 333 |
|
|---|
| 334 |
unitset = [] |
|---|
| 335 |
for cls in classes: |
|---|
| 336 |
unit = units[cls] |
|---|
| 337 |
unit.cleanse() |
|---|
| 338 |
unitset.append(unit) |
|---|
| 339 |
|
|---|
| 340 |
|
|---|
| 341 |
|
|---|
| 342 |
acceptable = True |
|---|
| 343 |
if expr and data.statement.imperfect: |
|---|
| 344 |
acceptable = expr(*unitset) |
|---|
| 345 |
if acceptable: |
|---|
| 346 |
yield unitset |
|---|
| 347 |
|
|---|
| 348 |
|
|---|
| 349 |
|
|---|
| 350 |
def create_database(self): |
|---|
| 351 |
if self.logflags & logflags.DDL: |
|---|
| 352 |
self.log(logflags.DDL.message("create database")) |
|---|
| 353 |
self.db.create() |
|---|
| 354 |
self.schema.create() |
|---|
| 355 |
|
|---|
| 356 |
def drop_database(self): |
|---|
| 357 |
if self.logflags & logflags.DDL: |
|---|
| 358 |
self.log(logflags.DDL.message("drop database")) |
|---|
| 359 |
self.schema.drop() |
|---|
| 360 |
self.db.drop() |
|---|
| 361 |
|
|---|
| 362 |
def _make_table(self, cls): |
|---|
| 363 |
"""Create and return a Table object for the given class.""" |
|---|
| 364 |
t = self.schema.table(cls.__name__) |
|---|
| 365 |
|
|---|
| 366 |
indices = cls.indices() |
|---|
| 367 |
fields = [] |
|---|
| 368 |
for key in cls.properties: |
|---|
| 369 |
t[key] = self._make_column(cls, key) |
|---|
| 370 |
if key in indices: |
|---|
| 371 |
t.add_index(key) |
|---|
| 372 |
|
|---|
| 373 |
|
|---|
| 374 |
for k, v in cls._associations.iteritems(): |
|---|
| 375 |
t.references[k] = (v.nearKey, v.farClass.__name__, v.farKey) |
|---|
| 376 |
|
|---|
| 377 |
return t |
|---|
| 378 |
|
|---|
| 379 |
def create_storage(self, cls): |
|---|
| 380 |
"""Create storage for the given class.""" |
|---|
| 381 |
if self.logflags & logflags.DDL: |
|---|
| 382 |
self.log(logflags.DDL.message("create storage %s" % cls)) |
|---|
| 383 |
|
|---|
| 384 |
self.schema[cls.__name__] = self._make_table(cls) |
|---|
| 385 |
|
|---|
| 386 |
def _make_column(self, cls, key): |
|---|
| 387 |
prop = getattr(cls, key) |
|---|
| 388 |
col = self.schema.column(prop.type, default=prop.default, hints=prop.hints) |
|---|
| 389 |
if key in cls.identifiers: |
|---|
| 390 |
col.key = True |
|---|
| 391 |
if isinstance(cls.sequencer, dejavu.UnitSequencerInteger): |
|---|
| 392 |
col.autoincrement = True |
|---|
| 393 |
col.initial = cls.sequencer.initial |
|---|
| 394 |
return col |
|---|
| 395 |
|
|---|
| 396 |
def has_storage(self, cls): |
|---|
| 397 |
return cls.__name__ in self.schema |
|---|
| 398 |
|
|---|
| 399 |
def drop_storage(self, cls): |
|---|
| 400 |
if self.logflags & logflags.DDL: |
|---|
| 401 |
self.log(logflags.DDL.message("drop storage %s" % cls)) |
|---|
| 402 |
del self.schema[cls.__name__] |
|---|
| 403 |
|
|---|
| 404 |
def rename_storage(self, oldname, newname): |
|---|
| 405 |
if self.logflags & logflags.DDL: |
|---|
| 406 |
self.log(logflags.DDL.message("rename storage from %s to %s" |
|---|
| 407 |
% (oldname, newname))) |
|---|
| 408 |
self.schema.rename(oldname, newname) |
|---|
| 409 |
|
|---|
| 410 |
def add_property(self, cls, name): |
|---|
| 411 |
if self.logflags & logflags.DDL: |
|---|
| 412 |
self.log(logflags.DDL.message("add property %s %s" % |
|---|
| 413 |
(cls, name))) |
|---|
| 414 |
if not self.has_property(cls, name): |
|---|
| 415 |
table = self.schema[cls.__name__] |
|---|
| 416 |
table[name] = self._make_column(cls, name) |
|---|
| 417 |
|
|---|
| 418 |
def has_property(self, cls, name): |
|---|
| 419 |
return name in self.schema[cls.__name__] |
|---|
| 420 |
|
|---|
| 421 |
def drop_property(self, cls, name): |
|---|
| 422 |
if self.logflags & logflags.DDL: |
|---|
| 423 |
self.log(logflags.DDL.message("drop property %s %s" % |
|---|
| 424 |
(cls, name))) |
|---|
| 425 |
if self.has_property(cls, name): |
|---|
| 426 |
del self.schema[cls.__name__][name] |
|---|
| 427 |
|
|---|
| 428 |
def rename_property(self, cls, oldname, newname): |
|---|
| 429 |
if self.logflags & logflags.DDL: |
|---|
| 430 |
self.log(logflags.DDL.message( |
|---|
| 431 |
"rename property %s from %s to %s" % |
|---|
| 432 |
(cls, oldname, newname))) |
|---|
| 433 |
t = self.schema[cls.__name__] |
|---|
| 434 |
|
|---|
| 435 |
|
|---|
| 436 |
|
|---|
| 437 |
|
|---|
| 438 |
|
|---|
| 439 |
|
|---|
| 440 |
tempcol = None |
|---|
| 441 |
try: |
|---|
| 442 |
t[oldname] |
|---|
| 443 |
except KeyError: |
|---|
| 444 |
c = [x for x in self.schema._get_columns(t.name) |
|---|
| 445 |
if x.name == self.schema._column_name(t.name, oldname)] |
|---|
| 446 |
if not c: |
|---|
| 447 |
raise KeyError("Rename failed. Old column %r not found in %r." |
|---|
| 448 |
% (oldname, t.name)) |
|---|
| 449 |
oldcol = c[0] |
|---|
| 450 |
|
|---|
| 451 |
dict.__setitem__(t, oldname, oldcol) |
|---|
| 452 |
|
|---|
| 453 |
t.rename(oldname, newname) |
|---|
| 454 |
|
|---|
| 455 |
def add_index(self, cls, name): |
|---|
| 456 |
self.schema[cls.__name__].add_index(name) |
|---|
| 457 |
|
|---|
| 458 |
def has_index(self, cls, name): |
|---|
| 459 |
return name in self.schema[cls.__name__].indices |
|---|
| 460 |
|
|---|
| 461 |
def drop_index(self, cls, name): |
|---|
| 462 |
del self.schema[cls.__name__].indices[name] |
|---|
| 463 |
|
|---|
| 464 |
auto_discover = True |
|---|
| 465 |
|
|---|
| 466 |
def map(self, classes, conflict_mode='error'): |
|---|
| 467 |
"""Map classes to internal storage. |
|---|
| 468 |
|
|---|
| 469 |
If self.auto_discover is True (the default), then Table/Column/Index |
|---|
| 470 |
objects will be formed by inspecting the underlying database using |
|---|
| 471 |
self.sync(). |
|---|
| 472 |
|
|---|
| 473 |
If auto_discover is False, then mock Table/Column/Index objects |
|---|
| 474 |
will be used instead; this provides a performance improvement |
|---|
| 475 |
in scenarios where the model maps perfectly to the database |
|---|
| 476 |
and changes to the database are not expected outside the model. |
|---|
| 477 |
|
|---|
| 478 |
conflict_mode: This argument determines what happens when there are |
|---|
| 479 |
discrepancies between the Dejavu model and the actual database. |
|---|
| 480 |
|
|---|
| 481 |
If 'error' (the default), MappingError is raised for the |
|---|
| 482 |
first issue and the sync process is aborted. |
|---|
| 483 |
|
|---|
| 484 |
If 'warn', then a warning is raised (instead of an error) |
|---|
| 485 |
for each issue, and the sync process is not aborted. This |
|---|
| 486 |
allows you to see all errors at once, without having to stop |
|---|
| 487 |
and fix each one and then execute the process again. |
|---|
| 488 |
|
|---|
| 489 |
If 'repair', then each issue will be resolved by changing |
|---|
| 490 |
the database to match the model. |
|---|
| 491 |
""" |
|---|
| 492 |
if self.auto_discover: |
|---|
| 493 |
self.sync(classes, conflict_mode) |
|---|
| 494 |
else: |
|---|
| 495 |
for cls in classes: |
|---|
| 496 |
if self.has_storage(cls): |
|---|
| 497 |
|
|---|
| 498 |
|
|---|
| 499 |
|
|---|
| 500 |
continue |
|---|
| 501 |
|
|---|
| 502 |
t = self._make_table(cls) |
|---|
| 503 |
|
|---|
| 504 |
|
|---|
| 505 |
dict.__setitem__(self.schema, cls.__name__, t) |
|---|
| 506 |
|
|---|
| 507 |
def sync(self, classes, conflict_mode='error'): |
|---|
| 508 |
"""Map classes to existing Table objects (found via discovery). |
|---|
| 509 |
|
|---|
| 510 |
conflict_mode: This argument determines what happens when there are |
|---|
| 511 |
discrepancies between the Dejavu model and the actual database. |
|---|
| 512 |
|
|---|
| 513 |
If 'error' (the default), MappingError is raised for the |
|---|
| 514 |
first issue and the sync process is aborted. |
|---|
| 515 |
|
|---|
| 516 |
If 'warn', then a warning is raised (instead of an error) |
|---|
| 517 |
for each issue, and the sync process is not aborted. This |
|---|
| 518 |
allows you to see all errors at once, without having to stop |
|---|
| 519 |
and fix each one and then execute the process again. |
|---|
| 520 |
|
|---|
| 521 |
If 'repair', then each issue will be resolved by changing |
|---|
| 522 |
the database to match the model. |
|---|
| 523 |
|
|---|
| 524 |
If 'ignore', then each issue will be silently ignored. |
|---|
| 525 |
""" |
|---|
| 526 |
for cls in classes: |
|---|
| 527 |
if cls.__name__ in self.schema: |
|---|
| 528 |
|
|---|
| 529 |
|
|---|
| 530 |
|
|---|
| 531 |
continue |
|---|
| 532 |
self._find_table(self.schema, cls, conflict_mode) |
|---|
| 533 |
|
|---|
| 534 |
def _find_table(self, schema, cls, conflict_mode='error'): |
|---|
| 535 |
|
|---|
| 536 |
|
|---|
| 537 |
def notify(msg): |
|---|
| 538 |
if conflict_mode == 'warn': |
|---|
| 539 |
warnings.warn(msg) |
|---|
| 540 |
elif conflict_mode == 'ignore': |
|---|
| 541 |
pass |
|---|
| 542 |
else: |
|---|
| 543 |
raise MappingError(msg) |
|---|
| 544 |
|
|---|
| 545 |
|
|---|
| 546 |
clsname = cls.__name__ |
|---|
| 547 |
tablename = schema.table_name(clsname) |
|---|
| 548 |
try: |
|---|
| 549 |
|
|---|
| 550 |
table = schema[tablename] |
|---|
| 551 |
schema.alias(table.name, clsname) |
|---|
| 552 |
except KeyError: |
|---|
| 553 |
|
|---|
| 554 |
try: |
|---|
| 555 |
table = schema.discover(tablename) |
|---|
| 556 |
schema.alias(table.name, clsname) |
|---|
| 557 |
except geniusql.errors.MappingError: |
|---|
| 558 |
msg = "%s: no such table %r." % (clsname, tablename) |
|---|
| 559 |
if conflict_mode == 'repair': |
|---|
| 560 |
self.create_storage(cls) |
|---|
| 561 |
table = schema[clsname] |
|---|
| 562 |
else: |
|---|
| 563 |
notify(msg) |
|---|
| 564 |
return |
|---|
| 565 |
|
|---|
| 566 |
|
|---|
| 567 |
dbcols = dict([(c.name, c) for c in table.itervalues()]) |
|---|
| 568 |
indices = cls.indices() |
|---|
| 569 |
for pkey in cls.properties: |
|---|
| 570 |
colname = schema._column_name(table.name, pkey) |
|---|
| 571 |
try: |
|---|
| 572 |
col = dbcols[colname] |
|---|
| 573 |
table.alias(colname, pkey) |
|---|
| 574 |
except KeyError, x: |
|---|
| 575 |
msg = "%s: no column found for %r." % (clsname, pkey) |
|---|
| 576 |
if conflict_mode == 'repair': |
|---|
| 577 |
self.add_property(cls, pkey) |
|---|
| 578 |
if pkey in cls.indices() and pkey not in table.indices: |
|---|
| 579 |
self.add_index(cls, pkey) |
|---|
| 580 |
col = table[pkey] |
|---|
| 581 |
else: |
|---|
| 582 |
notify(msg) |
|---|
| 583 |
continue |
|---|
| 584 |
|
|---|
| 585 |
|
|---|
| 586 |
|
|---|
| 587 |
|
|---|
| 588 |
|
|---|
| 589 |
if pkey in cls.identifiers and not col.key: |
|---|
| 590 |
msg = ("%s: %r is an identifier, but the " |
|---|
| 591 |
"column is not marked as a primary key." |
|---|
| 592 |
% (clsname, pkey)) |
|---|
| 593 |
if conflict_mode == 'repair': |
|---|
| 594 |
col.key = True |
|---|
| 595 |
table.set_primary() |
|---|
| 596 |
else: |
|---|
| 597 |
notify(msg) |
|---|
| 598 |
continue |
|---|
| 599 |
elif col.key and not pkey in cls.identifiers: |
|---|
| 600 |
msg = ("%s: %r is not an identifier, but the " |
|---|
| 601 |
"column is marked as a primary key." |
|---|
| 602 |
% (clsname, pkey)) |
|---|
| 603 |
if conflict_mode == 'repair': |
|---|
| 604 |
col.key = False |
|---|
| 605 |
|
|---|
| 606 |
|
|---|
| 607 |
table.set_primary() |
|---|
| 608 |
else: |
|---|
| 609 |
notify(msg) |
|---|
| 610 |
continue |
|---|
| 611 |
|
|---|
| 612 |
col.pytype = getattr(cls, pkey).type |
|---|
| 613 |
|
|---|
| 614 |
|
|---|
| 615 |
|
|---|
| 616 |
try: |
|---|
| 617 |
col.adapter = col.dbtype.default_adapter(col.pytype) |
|---|
| 618 |
except TypeError, x: |
|---|
| 619 |
x.args += ("%s.%s" % (table.name, col.name),) |
|---|
| 620 |
raise |
|---|
| 621 |
|
|---|
| 622 |
|
|---|
| 623 |
|
|---|
| 624 |
if pkey in indices: |
|---|
| 625 |
for ikey, idx in table.indices.items(): |
|---|
| 626 |
if idx.colname == colname: |
|---|
| 627 |
a = schema.table_name("i" + clsname + pkey) |
|---|
| 628 |
table.indices.alias(ikey, a) |
|---|
| 629 |
break |
|---|
| 630 |
else: |
|---|
| 631 |
msg = "%s: no index found for %r." % (clsname, pkey) |
|---|
| 632 |
if conflict_mode == 'repair': |
|---|
| 633 |
self.add_index(cls, pkey) |
|---|
| 634 |
else: |
|---|
| 635 |
notify(msg) |
|---|
| 636 |
continue |
|---|
| 637 |
else: |
|---|
| 638 |
if pkey in cls.identifiers and self.db.pks_must_be_indexed: |
|---|
| 639 |
pass |
|---|
| 640 |
else: |
|---|
| 641 |
for ikey, idx in table.indices.items(): |
|---|
| 642 |
if idx.colname == colname: |
|---|
| 643 |
msg = ("%s: index found for non-indexed %r." |
|---|
| 644 |
% (clsname, pkey)) |
|---|
| 645 |
if conflict_mode == 'repair': |
|---|
| 646 |
self.drop_index(cls, ikey) |
|---|
| 647 |
else: |
|---|
| 648 |
notify(msg) |
|---|
| 649 |
continue |
|---|
| 650 |
|
|---|
| 651 |
|
|---|
| 652 |
for k, v in cls._associations.iteritems(): |
|---|
| 653 |
table.references[k] = (v.nearKey, v.farClass.__name__, v.farKey) |
|---|
| 654 |
|
|---|
| 655 |
return table |
|---|
| 656 |
|
|---|
| 657 |
|
|---|
| 658 |
|
|---|
| 659 |
def start(self, isolation=None): |
|---|
| 660 |
"Start a transaction (not needed if db.connections.implicit_trans)." |
|---|
| 661 |
self.db.connections.start(isolation) |
|---|
| 662 |
|
|---|
| 663 |
def rollback(self): |
|---|
| 664 |
"""Roll back the current transaction.""" |
|---|
| 665 |
self.db.connections.rollback() |
|---|
| 666 |
|
|---|
| 667 |
def commit(self): |
|---|
| 668 |
"""Commit the current transaction.""" |
|---|
| 669 |
self.db.connections.commit() |
|---|
| 670 |
|
|---|
| 671 |
|
|---|
| 672 |
class Modeler(object): |
|---|
| 673 |
"""Tool to automatically form Unit classes or source from existing DB's.""" |
|---|
| 674 |
|
|---|
| 675 |
ignore = ['Unit', 'DeployedVersion', |
|---|
| 676 |
'UnitEngine', 'UnitEngineRule', 'UnitCollection', |
|---|
| 677 |
] |
|---|
| 678 |
|
|---|
| 679 |
def __init__(self, schema): |
|---|
| 680 |
self.schema = schema |
|---|
| 681 |
self.ignore = self.ignore[:] |
|---|
| 682 |
|
|---|
| 683 |
def all_classes(self): |
|---|
| 684 |
"""Return a list of new classes for all tables in the Database.""" |
|---|
| 685 |
ignore = dict.fromkeys([self.schema.table_name(x) for x in self.ignore] |
|---|
| 686 |
+ self.ignore).keys() |
|---|
| 687 |
|
|---|
| 688 |
self.schema.discover_all(ignore=ignore) |
|---|
| 689 |
|
|---|
| 690 |
classes = [] |
|---|
| 691 |
seen = {} |
|---|
| 692 |
for key, table in self.schema.items(): |
|---|
| 693 |
if key not in ignore and table.name not in seen: |
|---|
| 694 |
cls = self.make_class(key) |
|---|
| 695 |
classes.append(cls) |
|---|
| 696 |
seen[table.name] = None |
|---|
| 697 |
return classes |
|---|
| 698 |
|
|---|
| 699 |
def make_class(self, tablename, newclassname=None): |
|---|
| 700 |
"""Create a Unit class automatically from the named table.""" |
|---|
| 701 |
if tablename not in self.schema: |
|---|
| 702 |
self.schema.discover(tablename) |
|---|
| 703 |
table = self.schema[tablename] |
|---|
| 704 |
|
|---|
| 705 |
class AutoUnitClass(dejavu.Unit): |
|---|
| 706 |
sequencer = dejavu.UnitSequencer() |
|---|
| 707 |
identifiers = tuple([k for k in table if table[k].key]) |
|---|
| 708 |
|
|---|
| 709 |
if newclassname is None: |
|---|
| 710 |
newclassname = table.name |
|---|
| 711 |
|
|---|
| 712 |
for key, t in self.schema.iteritems(): |
|---|
| 713 |
if t.name == newclassname: |
|---|
| 714 |
newclassname = key |
|---|
| 715 |
break |
|---|
| 716 |
AutoUnitClass.__name__ = newclassname |
|---|
| 717 |
|
|---|
| 718 |
indices = [idx.colname for idx in table.indices.itervalues()] |
|---|
| 719 |
for cname, c in table.iteritems(): |
|---|
| 720 |
ptype = c.pytype |
|---|
| 721 |
if ptype == int and c.dbtype.bytes == 1: |
|---|
| 722 |
|
|---|
| 723 |
ptype = bool |
|---|
| 724 |
p = AutoUnitClass.set_property(cname, ptype) |
|---|
| 725 |
if c.autoincrement: |
|---|
| 726 |
AutoUnitClass.sequencer = dejavu.UnitSequencerInteger(int, c.initial) |
|---|
| 727 |
p.default = c.default |
|---|
| 728 |
|
|---|
| 729 |
p.hints = dict([(k, getattr(c.dbtype, k)) |
|---|
| 730 |
for k in ("bytes", "precision", "scale") |
|---|
| 731 |
if hasattr(c.dbtype, k)]) |
|---|
| 732 |
if p.hints: |
|---|
| 733 |
|
|---|
| 734 |
|
|---|
| 735 |
for k, v in p.hints.iteritems(): |
|---|
| 736 |
if v.__class__.__name__ == 'ComparableInfinity': |
|---|
| 737 |
p.hints[k] = 0 |
|---|
| 738 |
|
|---|
| 739 |
p.index = (cname in indices) |
|---|
| 740 |
|
|---|
| 741 |
|
|---|
| 742 |
if "ID" not in table: |
|---|
| 743 |
AutoUnitClass.properties.remove('ID') |
|---|
| 744 |
AutoUnitClass.ID = None |
|---|
| 745 |
|
|---|
| 746 |
return AutoUnitClass |
|---|
| 747 |
|
|---|
| 748 |
def all_source(self): |
|---|
| 749 |
"""Return a list of strings of Unit source code for all tables.""" |
|---|
| 750 |
ignore = dict.fromkeys([self.schema.table_name(x) for x in self.ignore] |
|---|
| 751 |
+ self.ignore).keys() |
|---|
| 752 |
|
|---|
| 753 |
self.schema.discover_all(ignore=ignore) |
|---|
| 754 |
|
|---|
| 755 |
allcode = [] |
|---|
| 756 |
seen = {} |
|---|
| 757 |
tables = self.schema.items() |
|---|
| 758 |
tables.sort() |
|---|
| 759 |
for key, table in tables: |
|---|
| 760 |
if key not in ignore and table.name not in seen: |
|---|
| 761 |
code = self.make_source(key) |
|---|
| 762 |
allcode.append(code) |
|---|
| 763 |
seen[table.name] = None |
|---|
| 764 |
return allcode |
|---|
| 765 |
|
|---|
| 766 |
def make_source(self, tablename, newclassname=None): |
|---|
| 767 |
"""Create source code for a Unit class from the named table.""" |
|---|
| 768 |
if tablename not in self.schema: |
|---|
| 769 |
self.schema.discover(tablename) |
|---|
| 770 |
table = self.schema[tablename] |
|---|
| 771 |
|
|---|
| 772 |
code = [] |
|---|
| 773 |
|
|---|
| 774 |
if newclassname is None: |
|---|
| 775 |
newclassname = table.name |
|---|
| 776 |
|
|---|
| 777 |
for key, t in self.schema.iteritems(): |
|---|
| 778 |
if t.name == newclassname: |
|---|
| 779 |
newclassname = key |
|---|
| 780 |
break |
|---|
| 781 |
|
|---|
| 782 |
newclassname = newclassname.replace(".", "_") |
|---|
| 783 |
code.append("class %s(Unit):" % newclassname) |
|---|
| 784 |
|
|---|
| 785 |
if table.description: |
|---|
| 786 |
import textwrap |
|---|
| 787 |
block = textwrap.fill(table.description, subsequent_indent=' ') |
|---|
| 788 |
code.append(' """%s"""' % block) |
|---|
| 789 |
|
|---|
| 790 |
sequencer = None |
|---|
| 791 |
indices = [idx.colname for idx in table.indices.itervalues()] |
|---|
| 792 |
|
|---|
| 793 |
|
|---|
| 794 |
columns = table.items() |
|---|
| 795 |
columns.sort() |
|---|
| 796 |
for cname, c in columns: |
|---|
| 797 |
prop, seq = self._make_column_source(cname, c, cname in indices) |
|---|
| 798 |
if not prop.startswith(" ID = UnitProperty(int"): |
|---|
| 799 |
code.append(prop) |
|---|
| 800 |
if seq: |
|---|
| 801 |
sequencer = seq |
|---|
| 802 |
|
|---|
| 803 |
|
|---|
| 804 |
if "ID" not in table: |
|---|
| 805 |
code.append(" # Remove the default 'ID' property.") |
|---|
| 806 |
code.append(" ID = None") |
|---|
| 807 |
|
|---|
| 808 |
pk = tuple([k for k in table if table[k].key]) |
|---|
| 809 |
if pk not in [("ID",), ("id",)]: |
|---|
| 810 |
code.append(" identifiers = %s" % repr(pk)) |
|---|
| 811 |
|
|---|
| 812 |
if sequencer: |
|---|
| 813 |
if sequencer != " sequencer = UnitSequencerInteger(int, 1)": |
|---|
| 814 |
code.append(sequencer) |
|---|
| 815 |
else: |
|---|
| 816 |
code.append(" sequencer = UnitSequencer()") |
|---|
| 817 |
|
|---|
| 818 |
if len(code) == 1: |
|---|
| 819 |
code.append(" pass") |
|---|
| 820 |
|
|---|
| 821 |
return "\n".join(code) |
|---|
| 822 |
|
|---|
| 823 |
def _make_column_source(self, colname, column, has_index): |
|---|
| 824 |
ptype = column.pytype |
|---|
| 825 |
if ptype == int and column.dbtype.bytes == 1: |
|---|
| 826 |
|
|---|
| 827 |
ptype = bool |
|---|
| 828 |
|
|---|
| 829 |
mod = ptype.__module__ |
|---|
| 830 |
if mod == '__builtin__': |
|---|
| 831 |
ptype = ptype.__name__ |
|---|
| 832 |
else: |
|---|
| 833 |
ptype = mod + "." + ptype.__name__ |
|---|
| 834 |
|
|---|
| 835 |
seq = None |
|---|
| 836 |
if column.autoincrement: |
|---|
| 837 |
seq = (" sequencer = UnitSequencerInteger(int, %r)" % |
|---|
| 838 |
column.initial) |
|---|
| 839 |
|
|---|
| 840 |
default = column.default |
|---|
| 841 |
if default is None: |
|---|
| 842 |
default = "" |
|---|
| 843 |
else: |
|---|
| 844 |
default = ", default=%r" % default |
|---|
| 845 |
|
|---|
| 846 |
index = "" |
|---|
| 847 |
if has_index: |
|---|
| 848 |
index = ", index=True" |
|---|
| 849 |
|
|---|
| 850 |
hints = dict([(k, getattr(column.dbtype, k)) |
|---|
| 851 |
for k in ("bytes", "precision", "scale") |
|---|
| 852 |
if hasattr(column.dbtype, k)]) |
|---|
| 853 |
if hints: |
|---|
| 854 |
|
|---|
| 855 |
|
|---|
| 856 |
for k, v in hints.iteritems(): |
|---|
| 857 |
if v.__class__.__name__ == 'ComparableInfinity': |
|---|
| 858 |
hints[k] = 0 |
|---|
| 859 |
hints = ", hints=%r" % hints |
|---|
| 860 |
else: |
|---|
| 861 |
hints = "" |
|---|
| 862 |
|
|---|
| 863 |
return (" %s = UnitProperty(%s%s%s%s)" % |
|---|
| 864 |
(colname, ptype, index, hints, default)), seq |
|---|
| 865 |
|
|---|