sqlitedict 2.1.0

写入

>>> from sqlitedict import SqliteDict
>>> db = SqliteDict("example.sqlite")
>>>
>>> db["1"] = {"name": "first item"}
>>> db["2"] = {"name": "second item"}
>>> db["3"] = {"name": "yet another item"}
>>>
>>> # Commit to save the objects.
>>> db.commit()
>>>
>>> db["4"] = {"name": "yet another item"}
>>> # Oops, forgot to commit here, that object will never be saved.
>>> # Always remember to commit, or enable autocommit with SqliteDict("example.sqlite", autocommit=True)
>>> # Autocommit is off by default for performance.
>>>
>>> db.close()

读取

>>> from sqlitedict import SqliteDict
>>> db = SqliteDict("example.sqlite")
>>>
>>> print("There are %d items in the database" % len(db))
There are 3 items in the database
>>>
>>> # Standard dict interface. items() values() keys() etc...
>>> for key, item in db.items():
...     print("%s=%s" % (key, item))
1={'name': 'first item'}
2={'name': 'second item'}
3={'name': 'yet another item'}
>>>
>>> db.close()

效率

默认情况下，sqlitedict的异常处理优先考虑详细性而不是效率。它从错误日志中提取和输出外层异常堆栈。如果您更倾向于效率，那么请用outer_stack=False初始化DB。

>>> from sqlitedict import SqliteDict
>>> db = SqliteDict("example.sqlite", outer_stack=False)  # True is the default
>>> db[1]
{'name': 'first item'}

上下文管理器

>>> from sqlitedict import SqliteDict
>>>
>>> # The database is automatically closed when leaving the with section.
>>> # Uncommitted objects are not saved on close. REMEMBER TO COMMIT!
>>>
>>> with SqliteDict("example.sqlite") as db:
...     print("There are %d items in the database" % len(db))
There are 3 items in the database

表

数据库文件可以存储多个表。如果没有指定表名，则使用默认表。

注意：写入是序列化的，拥有多个表不会提高性能。

>>> from sqlitedict import SqliteDict
>>>
>>> products = SqliteDict("example.sqlite", tablename="product", autocommit=True)
>>> manufacturers = SqliteDict("example.sqlite", tablename="manufacturer", autocommit=True)
>>>
>>> products["1"] = {"name": "first item",  "manufacturer_id": "1"}
>>> products["2"] = {"name": "second item", "manufacturer_id": "1"}
>>>
>>> manufacturers["1"] = {"manufacturer_name": "afactory", "location": "US"}
>>> manufacturers["2"] = {"manufacturer_name": "anotherfactory", "location": "UK"}
>>>
>>> tables = products.get_tablenames('example.sqlite')
>>> print(tables)
['unnamed', 'product', 'manufacturer']
>>>
>>> products.close()
>>> manufacturers.close()

如果您想知道，未命名的表来自之前的示例，其中我们没有指定表名。

序列化

键是字符串。值是任何可序列化对象。

默认情况下，内部使用Pickle来（反）序列化值。

可以使用自定义（反）序列化器，特别是用于JSON和压缩。

>>> # Use JSON instead of pickle
>>> import json
>>> with SqliteDict("example.sqlite", encode=json.dumps, decode=json.loads) as mydict:
...     pass
>>>
>>> # Apply zlib compression after pickling
>>> import zlib, pickle, sqlite3
>>>
>>> def my_encode(obj):
...     return sqlite3.Binary(zlib.compress(pickle.dumps(obj, pickle.HIGHEST_PROTOCOL)))
>>>
>>> def my_decode(obj):
...     return pickle.loads(zlib.decompress(bytes(obj)))
>>>
>>> with SqliteDict("example.sqlite", encode=my_encode, decode=my_decode) as mydict:
...     pass

还可以使用自定义（反）序列化器来允许非字符串键。

>>> # Use key encoding instead of default string keys only
>>> from sqlitedict import encode_key, decode_key
>>> with SqliteDict("example.sqlite", encode_key=encode_key, decode_key=decode_key) as mydict:
...     pass

更多

函数有良好的文档，直接在 sqlitedict.py 中的docstrings中查看，或调用 help(sqlitedict)。

注意：由于Python语义，sqlitedict 无法知道内存中基于SqliteDict的条目何时被修改。您需要显式地将修改后的对象分配回SqliteDict

>>> from sqlitedict import SqliteDict
>>> db = SqliteDict("example.sqlite")
>>> db["colors"] = {"red": (255, 0, 0)}
>>> db.commit()
>>>
>>> colors = db["colors"]
>>> colors["blue"] = (0, 0, 255) # sqlite DB not updated here!
>>> db["colors"] = colors  # now updated
>>>
>>> db.commit() # remember to commit (or set autocommit)
>>> db.close()

测试

安装

$ pip install pytest coverage pytest-coverage

执行所有测试

$ mkdir -p tests/db
$ pytest tests
$ python -m doctest README.rst

执行所有测试并包含覆盖率

$ pytest tests --cov=sqlitedict

评论，错误报告

sqlitedict 位于 github。您可以在那里提交问题或拉取请求。