技巧

使用 peewee 收集技巧。有很酷的技巧想分享吗？在 GitHub 上打开一个问题 或 联系我。

乐观锁

乐观锁适用于通常使用SELECT FOR UPDATE（或在 SQLite 中使用BEGIN IMMEDIATE）的情况。例如，你可以从数据库中获取一条用户记录，进行一些修改，然后保存修改后的用户记录。通常情况下，此场景要求我们在事务期间锁定用户记录，从选择记录的那一刻到保存更改的那一刻。

另一方面，在乐观锁中，我们不获取任何锁，而是依赖于要修改的行中的内部版本列。在读取时，我们查看当前行的版本，在保存时，我们确保仅在版本与最初读取的版本相同的情况下才执行更新。如果版本较高，则其他一些进程一定已经潜入并更改了行 - 保存修改后的版本可能会导致丢失重要更改。

在 Peewee 中实现乐观锁非常简单，这里有一个基本类，你可以将其用作起点

from peewee import *

class ConflictDetectedException(Exception): pass

class BaseVersionedModel(Model):
    version = IntegerField(default=1, index=True)

    def save_optimistic(self):
        if not self.id:
            # This is a new record, so the default logic is to perform an
            # INSERT. Ideally your model would also have a unique
            # constraint that made it impossible for two INSERTs to happen
            # at the same time.
            return self.save()

        # Update any data that has changed and bump the version counter.
        field_data = dict(self.__data__)
        current_version = field_data.pop('version', 1)
        self._populate_unsaved_relations(field_data)
        field_data = self._prune_fields(field_data, self.dirty_fields)
        if not field_data:
            raise ValueError('No changes have been made.')

        ModelClass = type(self)
        field_data['version'] = ModelClass.version + 1  # Atomic increment.

        query = ModelClass.update(**field_data).where(
            (ModelClass.version == current_version) &
            (ModelClass.id == self.id))
        if query.execute() == 0:
            # No rows were updated, indicating another process has saved
            # a new version. How you handle this situation is up to you,
            # but for simplicity I'm just raising an exception.
            raise ConflictDetectedException()
        else:
            # Increment local version to match what is now in the db.
            self.version += 1
            return True

以下是如何工作的示例。假设我们有以下模型定义。请注意，用户名上有一个唯一约束 - 这很重要，因为它提供了一种防止重复插入的方法。

class User(BaseVersionedModel):
    username = CharField(unique=True)
    favorite_animal = CharField()

示例

>>> u = User(username='charlie', favorite_animal='cat')
>>> u.save_optimistic()
True

>>> u.version
1

>>> u.save_optimistic()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "x.py", line 18, in save_optimistic
    raise ValueError('No changes have been made.')
ValueError: No changes have been made.

>>> u.favorite_animal = 'kitten'
>>> u.save_optimistic()
True

# Simulate a separate thread coming in and updating the model.
>>> u2 = User.get(User.username == 'charlie')
>>> u2.favorite_animal = 'macaw'
>>> u2.save_optimistic()
True

# Now, attempt to change and re-save the original instance:
>>> u.favorite_animal = 'little parrot'
>>> u.save_optimistic()
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "x.py", line 30, in save_optimistic
    raise ConflictDetectedException()
ConflictDetectedException: current version is out of sync

每组的顶部对象

这些示例描述了查询每组单个顶部项的几种方法。要全面讨论各种技术，请查看我的博客文章 使用 Peewee ORM 查询组中的顶部项。如果你对查询顶部N 个项的更一般问题感兴趣，请参阅下面的章节 每组的顶部 N 个对象。

在这些示例中，我们将使用用户和推文模型来查找每个用户及其最近的推文。

我在测试中发现的最有效的方法是使用 MAX() 聚合函数。

我们将在非关联子查询中执行聚合，因此我们可以确信此方法将具有高性能。其思想是，我们将按作者对帖子进行选择，其时间戳等于该用户观察到的最大时间戳。

# When referencing a table multiple times, we'll call Model.alias() to create
# a secondary reference to the table.
TweetAlias = Tweet.alias()

# Create a subquery that will calculate the maximum Tweet created_date for each
# user.
subquery = (TweetAlias
            .select(
                TweetAlias.user,
                fn.MAX(TweetAlias.created_date).alias('max_ts'))
            .group_by(TweetAlias.user)
            .alias('tweet_max_subquery'))

# Query for tweets and join using the subquery to match the tweet's user
# and created_date.
query = (Tweet
         .select(Tweet, User)
         .join(User)
         .switch(Tweet)
         .join(subquery, on=(
             (Tweet.created_date == subquery.c.max_ts) &
             (Tweet.user == subquery.c.user_id))))

SQLite 和 MySQL 稍微宽松一些，并允许按所选列的子集进行分组。这意味着我们可以取消子查询并简洁地表达它

query = (Tweet
         .select(Tweet, User)
         .join(User)
         .group_by(Tweet.user)
         .having(Tweet.created_date == fn.MAX(Tweet.created_date)))

每个组的 N 个首要对象

这些示例描述了以合理高效的方式查询每个组的顶部 N 个项目的几种方法。有关各种技术的详尽讨论，请查看我的博客文章 使用 Peewee ORM 查询每个组的顶部 N 个对象。

在这些示例中，我们将使用 User 和 Tweet 模型来查找每个用户及其最近的三条推文。

Postgres 横向连接

横向连接是一项简洁的 Postgres 功能，允许使用合理高效的相关子查询。它们通常被描述为 SQL for each 循环。

所需的 SQL 是

SELECT * FROM
  (SELECT id, username FROM user) AS uq
   LEFT JOIN LATERAL
  (SELECT message, created_date
   FROM tweet
   WHERE (user_id = uq.id)
   ORDER BY created_date DESC LIMIT 3)
  AS pq ON true

使用 peewee 实现这一点非常简单

subq = (Tweet
        .select(Tweet.message, Tweet.created_date)
        .where(Tweet.user == User.id)
        .order_by(Tweet.created_date.desc())
        .limit(3))

query = (User
         .select(User, subq.c.content, subq.c.created_date)
         .join(subq, JOIN.LEFT_LATERAL)
         .order_by(User.username, subq.c.created_date.desc()))

# We queried from the "perspective" of user, so the rows are User instances
# with the addition of a "content" and "created_date" attribute for each of
# the (up-to) 3 most-recent tweets for each user.
for row in query:
    print(row.username, row.content, row.created_date)

要从 Tweet 模型的“透视图”实现等效查询，我们可以改写为

# subq is the same as the above example.
subq = (Tweet
        .select(Tweet.message, Tweet.created_date)
        .where(Tweet.user == User.id)
        .order_by(Tweet.created_date.desc())
        .limit(3))

query = (Tweet
         .select(User.username, subq.c.content, subq.c.created_date)
         .from_(User)
         .join(subq, JOIN.LEFT_LATERAL)
         .order_by(User.username, subq.c.created_date.desc()))

# Each row is a "tweet" instance with an additional "username" attribute.
# This will print the (up-to) 3 most-recent tweets from each user.
for tweet in query:
    print(tweet.username, tweet.content, tweet.created_date)

窗口函数

窗口函数（受 peewee 支持）提供可扩展的高效性能。

所需的 SQL 是

SELECT subq.message, subq.username
FROM (
    SELECT
        t2.message,
        t3.username,
        RANK() OVER (
            PARTITION BY t2.user_id
            ORDER BY t2.created_date DESC
        ) AS rnk
    FROM tweet AS t2
    INNER JOIN user AS t3 ON (t2.user_id = t3.id)
) AS subq
WHERE (subq.rnk <= 3)

要使用 peewee 实现这一点，我们将把排名的推文包装在执行过滤的外查询中。

TweetAlias = Tweet.alias()

# The subquery will select the relevant data from the Tweet and
# User table, as well as ranking the tweets by user from newest
# to oldest.
subquery = (TweetAlias
            .select(
                TweetAlias.message,
                User.username,
                fn.RANK().over(
                    partition_by=[TweetAlias.user],
                    order_by=[TweetAlias.created_date.desc()]).alias('rnk'))
            .join(User, on=(TweetAlias.user == User.id))
            .alias('subq'))

# Since we can't filter on the rank, we are wrapping it in a query
# and performing the filtering in the outer query.
query = (Tweet
         .select(subquery.c.message, subquery.c.username)
         .from_(subquery)
         .where(subquery.c.rnk <= 3))

其他方法

如果您不使用 Postgres，那么不幸的是，您只剩下性能不太理想的选项。有关常用方法的更完整概述，请查看 这篇博客文章。下面我将总结这些方法和相应的 SQL。

使用 COUNT，我们可以获取所有推文，其中存在少于 N 条时间戳较新的推文

TweetAlias = Tweet.alias()

# Create a correlated subquery that calculates the number of
# tweets with a higher (newer) timestamp than the tweet we're
# looking at in the outer query.
subquery = (TweetAlias
            .select(fn.COUNT(TweetAlias.id))
            .where(
                (TweetAlias.created_date >= Tweet.created_date) &
                (TweetAlias.user == Tweet.user)))

# Wrap the subquery and filter on the count.
query = (Tweet
         .select(Tweet, User)
         .join(User)
         .where(subquery <= 3))

我们可以通过进行自连接并在 HAVING 子句中执行过滤来实现类似的结果

TweetAlias = Tweet.alias()

# Use a self-join and join predicates to count the number of
# newer tweets.
query = (Tweet
         .select(Tweet.id, Tweet.message, Tweet.user, User.username)
         .join(User)
         .switch(Tweet)
         .join(TweetAlias, on=(
             (TweetAlias.user == Tweet.user) &
             (TweetAlias.created_date >= Tweet.created_date)))
         .group_by(Tweet.id, Tweet.content, Tweet.user, User.username)
         .having(fn.COUNT(Tweet.id) <= 3))

最后一个示例在相关子查询中使用 LIMIT 子句。

TweetAlias = Tweet.alias()

# The subquery here will calculate, for the user who created the
# tweet in the outer loop, the three newest tweets. The expression
# will evaluate to `True` if the outer-loop tweet is in the set of
# tweets represented by the inner query.
query = (Tweet
         .select(Tweet, User)
         .join(User)
         .where(Tweet.id << (
             TweetAlias
             .select(TweetAlias.id)
             .where(TweetAlias.user == Tweet.user)
             .order_by(TweetAlias.created_date.desc())
             .limit(3))))

使用 SQLite 编写自定义函数

SQLite 非常容易扩展，可以使用 Python 编写的自定义函数，然后从 SQL 语句中调用这些函数。通过使用 SqliteExtDatabase 和 func() 装饰器，您可以非常轻松地定义自己的函数。

这是一个生成用户提供的密码的哈希版本的示例函数。我们还可以使用它来实现 login 功能，以匹配用户和密码。

from hashlib import sha1
from random import random
from playhouse.sqlite_ext import SqliteExtDatabase

db = SqliteExtDatabase('my-blog.db')

def get_hexdigest(salt, raw_password):
    data = salt + raw_password
    return sha1(data.encode('utf8')).hexdigest()

@db.func()
def make_password(raw_password):
    salt = get_hexdigest(str(random()), str(random()))[:5]
    hsh = get_hexdigest(salt, raw_password)
    return '%s$%s' % (salt, hsh)

@db.func()
def check_password(raw_password, enc_password):
    salt, hsh = enc_password.split('$', 1)
    return hsh == get_hexdigest(salt, raw_password)

以下是如何使用该函数添加新用户，存储哈希密码

query = User.insert(
    username='charlie',
    password=fn.make_password('testing')).execute()

如果我们从数据库中检索用户，则存储的密码已哈希并加盐

>>> user = User.get(User.username == 'charlie')
>>> print(user.password)
b76fa$88be1adcde66a1ac16054bc17c8a297523170949

要实现 login 类型功能，您可以编写类似这样的内容

def login(username, password):
    try:
        return (User
                .select()
                .where(
                    (User.username == username) &
                    (fn.check_password(password, User.password) == True))
                .get())
    except User.DoesNotExist:
        # Incorrect username and/or password.
        return False

日期计算

Peewee 支持的每个数据库都实现了它们自己的日期/时间算术函数和语义集。

本部分将提供一个简短的场景和示例代码，演示如何利用 Peewee 在 SQL 中执行动态日期操作。

场景：我们需要每隔 X 秒运行某些任务，并且任务间隔和任务本身都在数据库中定义。我们需要编写一些代码，告诉我们应该在给定时间运行哪些任务

class Schedule(Model):
    interval = IntegerField()  # Run this schedule every X seconds.


class Task(Model):
    schedule = ForeignKeyField(Schedule, backref='tasks')
    command = TextField()  # Run this command.
    last_run = DateTimeField()  # When was this run last?

我们的逻辑本质上归结为

# e.g., if the task was last run at 12:00:05, and the associated interval
# is 10 seconds, the next occurrence should be 12:00:15. So we check
# whether the current time (now) is 12:00:15 or later.
now >= task.last_run + schedule.interval

因此，我们可以编写以下代码

next_occurrence = something  # ??? how do we define this ???

# We can express the current time as a Python datetime value, or we could
# alternatively use the appropriate SQL function/name.
now = Value(datetime.datetime.now())  # Or SQL('current_timestamp'), e.g.

query = (Task
         .select(Task, Schedule)
         .join(Schedule)
         .where(now >= next_occurrence))

对于 Postgresql，我们将乘以一个静态的 1 秒间隔来动态计算偏移量

second = SQL("INTERVAL '1 second'")
next_occurrence = Task.last_run + (Schedule.interval * second)

对于 MySQL，我们可以直接引用计划的间隔

from peewee import NodeList  # Needed to construct sql entity.

interval = NodeList((SQL('INTERVAL'), Schedule.interval, SQL('SECOND')))
next_occurrence = fn.date_add(Task.last_run, interval)

对于 SQLite，事情有点棘手，因为 SQLite 没有专用的日期时间类型。因此，对于 SQLite，我们转换为 Unix 时间戳，添加计划秒数，然后转换回可比较的日期时间表示

next_ts = fn.strftime('%s', Task.last_run) + Schedule.interval
next_occurrence = fn.datetime(next_ts, 'unixepoch')