`
ryan.liu
  • 浏览: 136408 次
  • 性别: Icon_minigender_1
  • 来自: 深圳
社区版块
存档分类
最新评论

Flask中使用SqlAlchemy的较好方法

阅读更多

首先安装 Flask-SqlAlchem? v0.15 和 SqlAutoCode? v0.7

在配置文件settings中设置

 

SQLALCHEMY_DATABASE_URI = 'mysql://kw3:123456@localhost/kw3_article'
SQLALCHEMY_BINDS = {
    'game': 'mysql://kw3:123456@localhost/kw_game',
}
TABLE_PREFIX = 't_'

 

 将最后面的代码放到scripts下,并运行它。  如存为generate_models.py文件,python  scripts/generate_models.py

在项目入口文件中

 

#-*- coding: utf-8 -*-
from flask import Flask
from models.base.db import db

app = Flask(__name__)
db.init_app(app)

 你就可以在view中使用了,具体参考  http://packages.python.org/Flask-SQLAlchemy/index.html

 

 

自动产生models的脚本:

 

#-*- coding: utf-8 -*-
#使用SqlAutocode,根据数据库已有表,产生符合Flask-SqlAlchemy要求的models的定义

import os.path
from flask import Flask
from sqlautocode import config
from sqlautocode.declarative import *
from sqlautocode.formatter import _repr_coltype_as
from flask.ext.sqlalchemy import SQLAlchemy

singular = plural = lambda x: x
constants.COLUMN = 'db.' + constants.COLUMN
constants.HEADER_DECL = """from sqlalchemy import *
if __name__ == '__main__':
    from flask import Flask
    from flask.ext.sqlalchemy import SQLAlchemy

    app = Flask(__name__)
    app.config['SQLALCHEMY_DATABASE_URI'] = '%s'
    db = SQLAlchemy(app)
else:
    from .db import db


"""


def no_prefix_wrapper(f, prefix=None):
    def _name2label(name, schema=None):
        if schema:
            if name.startswith(schema+'.'):
                name = '.'.join(name.split('.')[1:])
        if prefix and name.startswith(prefix):
            name = name[ len(prefix):]
        label = str(''.join([s.capitalize() for s in
                   re.findall(r'([A-Z][a-z0-9]+|[a-z0-9]+|[A-Z0-9]+)', name)]))
        return label
    return _name2label


def column_repr(self):

    kwarg = []
    if self.key != self.name:
        kwarg.append( 'key')

    if hasattr(self, 'primary_key') and self.primary_key:
        self.primary_key = True
        kwarg.append( 'primary_key')

    if not self.nullable:
        kwarg.append( 'nullable')
    if self.onupdate:
        kwarg.append( 'onupdate')
    if self.default:
        kwarg.append( 'default')
    ks = ', '.join('%s=%r' % (k, getattr(self, k)) for k in kwarg)
    if self.server_default:
        ks += ', ' if ks else ''
        ks += ('default=%s' % self.server_default.arg.text)

    name = self.name

    if not hasattr(config, 'options') and self.config.options.generictypes:
        coltype = repr(self.type)
    elif type(self.type).__module__ == 'sqlalchemy.types':
        coltype = repr(self.type)
    else:
        # Try to 'cast' this column type to a cross-platform type
        # from sqlalchemy.types, dropping any database-specific type
        # arguments.
        for base in type(self.type).__mro__:
            if (base.__module__ == 'sqlalchemy.types' and
                base.__name__ in sqlalchemy.__all__):
                coltype = _repr_coltype_as(self.type, base)
                break
        # FIXME: if a dialect has a non-standard type that does not
        # derive from an ANSI type, there's no choice but to ignore
        # generic-types and output the exact type. However, import
        # headers have already been output and lack the required
        # dialect import.
        else:
            coltype = repr(self.type)

    data = {'name': self.name,
            'type': coltype,
            'constraints': ', '.join(["ForeignKey('%s')"%cn.target_fullname for cn in self.foreign_keys]),
            'args': ks and ks or '',
            }

    if data['constraints']:
        if data['constraints']: data['constraints'] = ', ' + data['constraints']
    if data['args']:
        if data['args']: data['args'] = ', ' + data['args']

    return constants.COLUMN % data


class FlaskModelFactory(ModelFactory):

    def __init__(self, name, conn):
        self.name = name
        argv = ['sqlautocode', conn, '-d', '-g', '-i']
        config.configure(argv)
        config.interactive, config.schema, config.example = None, None, False
        super(FlaskModelFactory, self).__init__(config)

    def _table_repr(self, table):
        s = "db.Table(u'%s',\n"%(table.name)
        for column in table.c:
            s += "    %s,\n"%column_repr(column)
        if self.name != "default":
            s +="    info={'bind_key': '%s'}\n"%self.name
        #if table.schema:
        #    s +="    schema='%s'\n"%table.schema
        s+=")"
        return s

    def create_model(self, table):
        #partially borrowed from Jorge Vargas' code
        #http://dpaste.org/V6YS/
        log.debug('Creating Model from table: %s'%table.name)

        model_name = self.find_new_name(singular(name2label(table.name)), self.used_model_names)
        self.used_model_names.append(model_name)
        is_many_to_many_table = self.is_many_to_many_table(table)
        table_name = self.find_new_name(table.name, self.used_table_names)
        self.used_table_names.append(table_name)

        mtl = self.model_table_lookup


        class Temporal(self.DeclarativeBase):
            __table__ = table

            @classmethod
            def _relation_repr(cls, rel):
                target = rel.argument
                if target and inspect.isfunction(target):
                    target = target()
                if isinstance(target, Mapper):
                    target = target.class_
                target = target.__name__
                primaryjoin=''
                lookup = mtl()
                if rel.primaryjoin is not None and hasattr(rel.primaryjoin, 'right'):
                    right_lookup = lookup.get(rel.primaryjoin.right.table.name, '%s.c'%rel.primaryjoin.right.table.name)
                    left_lookup = lookup.get(rel.primaryjoin.left.table.name, '%s.c'%rel.primaryjoin.left.table.name)

                    primaryjoin = ", primaryjoin='%s.%s==%s.%s'"%(left_lookup,
                                                                  rel.primaryjoin.left.name,
                                                                  right_lookup,
                                                                  rel.primaryjoin.right.name)
                elif hasattr(rel, '_as_string'):
                    primaryjoin=', primaryjoin="%s"'%rel._as_string

                secondary = ''
                secondaryjoin = ''
                if rel.secondary is not None:
                    secondary = ", secondary=%s"%rel.secondary.name
                    right_lookup = lookup.get(rel.secondaryjoin.right.table.name, '%s.c'%rel.secondaryjoin.right.table.name)
                    left_lookup = lookup.get(rel.secondaryjoin.left.table.name, '%s.c'%rel.secondaryjoin.left.table.name)
                    secondaryjoin = ", secondaryjoin='%s.%s==%s.%s'"%(left_lookup,
                                                                  rel.secondaryjoin.left.name,
                                                                  right_lookup,
                                                                  rel.secondaryjoin.right.name)
                backref=''
#                if rel.backref:
#                    backref=", backref='%s'"%rel.backref.key
                return "%s = relation('%s'%s%s%s%s)"%(rel.key, target, primaryjoin, secondary, secondaryjoin, backref)

            @classmethod
            def __repr__(cls):
                try:
                    mapper = None
                    try:
                        mapper = class_mapper(cls)
                    except exc.InvalidRequestError:
                        log.warn("A proper mapper could not be generated for the class %s, no relations will be created"%model_name)
                    s = ""
                    s += "class "+model_name+'(db.Model):\n'
                    if cls.__bind_key__ != "default":
                        s += "    __bind_key__ = '%s'\n"%cls.__bind_key__
                    if is_many_to_many_table:
                        s += "    __table__ = %s\n\n"%table_name
                    else:
                        s += "    __tablename__ = '%s'\n"%table_name
                        if hasattr(cls, '__table_args__'):
                            #if cls.__table_args__[0]:
                                #for fkc in cls.__table_args__[0]:
                                #    fkc.__class__.__repr__ = foreignkeyconstraint_repr
                                #    break
                            s+="    __table_args__ = %s\n"%cls.__table_args__
                        s+="\n"
                        for column in cls.__table__.c:
                            s += "    %s = %s\n"%(column.name, column_repr(column))
                    ess = s
                    # this is only required in SA 0.5
                    if mapper and RelationProperty:
                        for prop in mapper.iterate_properties:
                            if isinstance(prop, RelationshipProperty):
                                s+='    %s\n'%cls._relation_repr(prop)
                    return s

                except Exception, e:
                    log.error("Could not generate class for: %s"%cls.__name__)
                    from traceback import format_exc
                    log.error(format_exc())
                    return ''


        #hack the class to have the right classname
        Temporal.__name__ = model_name

        #set up some blank table args
        Temporal.__table_args__ = {}

        #add in the schema
        Temporal.__bind_key__ = self.name
        #if self.config.schema:
        #    Temporal.__table_args__[1]['schema'] = table.schema

        #trick sa's model registry to think the model is the correct name
        if model_name != 'Temporal':
            Temporal._decl_class_registry[model_name] = Temporal._decl_class_registry['Temporal']
            del Temporal._decl_class_registry['Temporal']

        #add in single relations
        fks = self.get_single_foreign_keys_by_column(table)
        for column, fk in fks.iteritems():
            related_table = fk.column.table
            if related_table not in self.tables:
                continue

            log.info('    Adding <primary> foreign key for:%s'%related_table.name)
            backref_name = plural(table_name)
            rel = relation(singular(name2label(related_table.name, related_table.schema)),
                           primaryjoin=column==fk.column)#, backref=backref_name)

            setattr(Temporal, related_table.name, _deferred_relationship(Temporal, rel))

        #add in the relations for the composites
        for constraint in table.constraints:
            if isinstance(constraint, ForeignKeyConstraint):
                if len(constraint.elements) >1:
                    related_table = constraint.elements[0].column.table
                    related_classname = singular(name2label(related_table.name, related_table.schema))

                    primary_join = "and_(%s)"%', '.join(["%s.%s==%s.%s"%(model_name,
                                                                        k.parent.name,
                                                                        related_classname,
                                                                        k.column.name)
                                                      for k in constraint.elements])
                    rel = relation(related_classname,
                                    primaryjoin=primary_join
                                    #foreign_keys=[k.parent for k in constraint.elements]
                               )

                    rel._as_string = primary_join
                    setattr(Temporal, related_table.name, rel) # _deferred_relationship(Temporal, rel))


        #add in many-to-many relations
        for join_table in self.get_related_many_to_many_tables(table.name):

            if join_table not in self.tables:
                continue
            primary_column = [c for c in join_table.columns if c.foreign_keys and list(c.foreign_keys)[0].column.table==table][0]

            for column in join_table.columns:
                if column.foreign_keys:
                    key = list(column.foreign_keys)[0]
                    if key.column.table is not table:
                        related_column = related_table = list(column.foreign_keys)[0].column
                        related_table = related_column.table
                        if related_table not in self.tables:
                            continue
                        log.info('    Adding <secondary> foreign key(%s) for:%s'%(key, related_table.name))
                        setattr(Temporal, plural(related_table.name), _deferred_relationship(Temporal,
                                 relation(singular(name2label(related_table.name,
                                                     related_table.schema)),
                                          secondary=join_table,
                                          primaryjoin=list(primary_column.foreign_keys)[0].column==primary_column,
                                          secondaryjoin=column==related_column
                                          )))
                        break;

        return Temporal


def gen_models_dir(app, models_dir):
    #找到并建立models文件夹和__init__.py文件
    if not models_dir:
        app_root = app.config.get('APPLICATION_ROOT', '')
        if not app_root:
            app_root = os.path.dirname( os.path.dirname( os.path.realpath(__file__) ) )
        models_dir = os.path.join(app_root, 'models')
    if not os.path.exists(models_dir):
        os.mkdir(models_dir)
    init_file = os.path.join(models_dir, '__init__.py')
    with open(init_file, 'wb') as fh:
        fh.write('#-*- coding: utf-8 -*-\n')
    return models_dir


def write_db_file(db_file):
    #建立数据库定义文件
    with open(db_file, 'wb') as fh:
        fh.write('#-*- coding: utf-8 -*-\n')
        fh.write('\n')
        fh.write('from flask.ext.sqlalchemy import SQLAlchemy\n')
        fh.write('\n\n')
        fh.write('db = SQLAlchemy()\n')


def write_schema_file(factory, schema_file, name='default'):
    #建立数据库定义文件
    with open(schema_file, 'wb') as fh:
        fh.write("#-*- coding: utf-8 -*-\n")
        fh.write('\n')
        fh.write( repr(factory) )
        fh.write('\n')
        fh.write("if __name__ == '__main__':\n")
        if name == 'default':
            fh.write("    db.create_all(bind=None)\n")
        else:
            fh.write("    db.create_all(bind=['%s'])\n" % name)


def generate_models(app, models_dir=None):
    db = SQLAlchemy(app)
    conns = {
        'default': app.config.get('SQLALCHEMY_DATABASE_URI') or {},
    }
    conns.update( app.config.get('SQLALCHEMY_BINDS') or {} )

    models_dir = gen_models_dir(app, models_dir)
    db_file = os.path.join(models_dir, 'db.py')
    if not os.path.exists(db_file):
        write_db_file(db_file)
    for name, conn in conns.items():
        if not conn:
            continue
        schema_file = os.path.join(models_dir, '%s.py' % name)
        if not os.path.exists(schema_file):
            factory = FlaskModelFactory(name, conn)
            write_schema_file(factory, schema_file, name)


if __name__ == '__main__':
    import sys
    sys.path.append(
        os.path.dirname( os.path.dirname( os.path.realpath(__file__) ) )
    )
    app = Flask(__name__)
    app.config.from_object('settings')
    name2label = no_prefix_wrapper(name2label, app.config.get('TABLE_PREFIX'))
    generate_models(app)


分享到:
评论

相关推荐

    sqlalchemy documentation sqlalchemy 文档

    - 如何在SQL表达式中使用JOIN。 #### 3.14 生成选择和转换入门 - 如何使用生成选择和转换。 #### 3.15 其他一切 - 包括绑定参数对象、函数、集合操作等。 #### 3.15.1 绑定参数对象 - 如何使用绑定参数。 #### ...

    SQLAlchemy技术文档(中文版)

    - **功能说明**:使用会话中的`query()`方法执行数据库查询。 - **示例代码**: ```python for instance in session.query(User).order_by(User.id): print(instance.name, instance.fullname) ``` - **关键点...

    sqlalchemy

    2. 安装方法:安装SQLAlchemy有多种方法,包括通过pip安装、使用setup.py安装、安装C扩展以及在Python 3环境下的安装。这使得SQLAlchemy可以在不同的平台和环境中工作。 3. 支持的平台:SQLAlchemy旨在提供跨平台的...

    SQLAlchemy

    SQLAlchemy是一种流行的ORM(对象关系映射)工具,它在Python社区中被广泛使用,特别是在数据库操作和管理方面。ORM工具的主要目的是将数据库中的数据表结构映射到编程语言中的对象模型,允许开发者使用更加面向对象...

    python flask在蓝图中集成sqlalchemy

    在这个主题中,“python flask在蓝图中集成sqlalchemy”,我们将探讨如何在Flask应用中使用蓝图(Blueprint)结构,并将SQLAlchemy整合其中,以便实现灵活且可扩展的后端开发。 首先,让我们了解Flask中的蓝图。...

    python sqlalchemy 使用手册 英文版

    本手册是一份关于Python编程语言中SQLAlchemy ORM库的使用指南。SQLAlchemy是一个用于Python编程语言的数据库工具包,它提供了一系列工具来处理关系数据库,并提供了一个完整的SQL表达语言。本手册涵盖了SQLAlchemy...

    sqlalchemy文档资料翻译

    - **Column and Data Types**:涵盖SQLAlchemy中的各种数据类型及其使用方法,以及如何自定义数据类型。 - **扩展与插件**:`sqlalchemy.ext` 包含了一系列的扩展和插件,这些插件可以增强SQLAlchemy的功能。 ###...

    FlaskSQLAlchemy下载及安装

    **Flask-SQLAlchemy**是Python微框架Flask的一个扩展,它使得在Flask应用中集成和使用SQLAlchemy变得更加方便。SQLAlchemy是Python的一个强大的ORM(对象关系映射)库,能够将数据库操作转化为面向对象的编程方式。...

    sqlalchemy的使用举例源码

    `SQLAlchemy`是Python中的一款强大的ORM(对象关系映射)库,它允许开发者使用Python类来定义数据库模式,并在Python代码中...通过学习这个例子,我们可以提升在Python项目中使用`SQLAlchemy`的能力,提高开发效率。

    Python使用SQLAlchemy.docx

    Python使用SQLAlchemy.docx

    SQLAlchemy v1.1官方文档

    - **使用EXISTS操作符**:在查询中使用SQL的EXISTS操作符,以提高某些查询的效率。 - **加载模式**:介绍SQLAlchemy的延迟加载(懒加载)和急切加载(即刻加载)模式,以及如何使用它们来优化应用程序的性能。 ### ...

    Python SQLAlchemy入门教程(基本用法)

    标题“Python SQLAlchemy入门教程(基本用法)”表明本文档是一份面向初学者的教程,旨在介绍如何使用Python中的SQLAlchemy库来进行数据库操作。SQLAlchemy是一个流行的对象关系映射(ORM)工具,它允许开发者通过...

    Python利用flask sqlalchemy实现分页效果

    在Flask应用中使用Flask-SQLAlchemy,可以帮助开发者更高效地进行数据库操作,例如查询、更新和删除等。 #### 创建Flask应用与数据库模型 首先,我们需要设置一个基本的Flask应用并连接到数据库。以下代码示例展示...

    SQLAlchemy详细教程

    * 提高代码可读性:使用 SQLAlchemy,可以将数据库操作封装在 Python 对象中,使代码更加简洁易读。 * 提高代码安全性:使用 SQLAlchemy,可以避免 SQL 注入攻击,提高代码安全性。 SQLAlchemy 的基本概念 --------...

    Python 进阶(三):Python使用ORM框架SQLAlchemy操作Oracle数据库

    本文主要介绍Python使用ORM框架SQLAlchemy操作Oracle数据库。 1. 安装Oracle Instant Client 2. 安装依赖库 使用以下命令来安装SQLAlchemy和cx_Oracle库: pip install SQLAlchemy pip install cx_Oracle 3.创建...

    sqlalchemy官方文档

    **SQLAlchemy ORM** 是 **SQLAlchemy** 中的核心模块之一,用于实现对象关系映射功能。这部分内容极为丰富,涵盖了以下主题: 1. **对象关系教程**: - **版本检查**:确保使用的 SQLAlchemy 版本与文档兼容。 - ...

Global site tag (gtag.js) - Google Analytics