faker制造大数据库脚本

faker用于生成数据，php和python都有此包

php版本

composer require fzaninotto/faker
composer require cakephp/orm

require './vendor/autoload.php';

use Cake\Datasource\ConnectionManager;
use Cake\ORM\TableRegistry;

set_time_limit(0);

function convert($size)
{
    //友好显示size
    $unit = array('b', 'kb', 'mb', 'gb', 'tb', 'pb');
    return @round($size / pow(1024, ($i = floor(log($size, 1024)))), 2) . ' ' . $unit[$i];
}

function microtime_float()
{
    //返回微妙
    list($usec, $sec) = explode(" ", microtime());
    return ((float) $usec + (float) $sec);
}

ConnectionManager::setConfig('default', [
    'className' => 'Cake\Database\Connection',
    'driver' => 'Cake\Database\Driver\Mysql',
    'persistent' => true,
    'host' => 'localhost',
    'username' => 'root',
    'password' => 'asdqwe123',
    'database' => 'test',
    'encoding' => 'utf8',
    'timezone' => 'UTC',
    'cacheMetadata' => false // If set to `true` you need to install the optional "cakephp/cache" package.
]);
$faker = Faker\Factory::create('zh_CN');

$Users = TableRegistry::get('User');
$m = 1000; //循环1000次
$start = microtime_float();
while ($m > 0) {
    $m--;
    $n = 1000;
    $data = [];
    while ($n > 0) {//每次循环批量插入1000条
        $n--;
        $data[] = [
            'username' => $faker->name,
            'age' => rand(10, 70),
            'sex' => rand(0, 1),
            'address' => $faker->address,
            'ctime' => $faker->dateTime
        ];
        echo 'speed memory:' . convert(memory_get_usage()) . "\r\n"; //打印内存占用情况
    }
    $entities = $Users->newEntities($data);
    $result = $Users->saveMany($entities);
}
$end = microtime_float();
$speed_time = $end - $start;
echo 'speed time:' . $speed_time . ' second' . "\r\n";

php create.php     #运行脚本

python版本

# coding=utf-8
from faker import Factory
import uniout
import logging
import threading
import random
import time
from datetime import datetime
from orator import DatabaseManager, Model
print "开始..."
fake = Factory.create('zh_CN')

# 创建 日志 对象
logger = logging.getLogger()
handler = logging.StreamHandler()
formatter = logging.Formatter(
    '%(asctime)s %(name)-12s %(levelname)-8s %(message)s')
handler.setFormatter(formatter)
logger.addHandler(handler)
logger.setLevel(logging.DEBUG)

# Connect to the database

config = {
    'mysql': {
        'driver': 'mysql',
        'host': 'localhost',
        'database': 'test',
        'user': 'root',
        'password': 'asdqwe123',
        'prefix': ''
    }
}

db = DatabaseManager(config)
Model.set_connection_resolver(db)


class User(Model):
    __table__ = 'user'
    pass


thread_nums = 10
ARR = (7, 9, 10, 5, 8, 4, 2, 1, 6, 3, 7, 9, 10, 5, 8, 4, 2)
LAST = ('1', '0', 'X', '9', '8', '7', '6', '5', '4', '3', '2')


def makeIdNo():
    u''' 随机生成新的18为身份证号码 '''
    t = time.localtime()[0]
    x = '%02d%02d%02d%04d%02d%02d%03d' % (random.randint(10, 99),
                                          random.randint(01, 99),
                                          random.randint(01, 99),
                                          random.randint(t - 80, t - 18),
                                          random.randint(1, 12),
                                          random.randint(1, 28),
                                          random.randint(1, 999))
    y = 0
    for i in range(17):
        y += int(x[i]) * ARR[i]

    return '%s%s' % (x, LAST[y % 11])


def insert_user(thread_name, nums):
    logger.info('开启线程%s' % thread_name)
    for _ in range(nums):
        email = fake.email()
        job = fake.job()
        address = fake.address().encode('utf-8')
        password = fake.md5()
        text = fake.text().encode('utf-8')
        username = str(fake.user_name())
        truename = fake.name().encode('utf-8')
        create_time = fake.date_time_between_dates(
            datetime_start=datetime(2017, 1, 1, 0, 0, 0),
            datetime_end=datetime(2017, 8, 12, 0, 0, 0, 0))
        create_time = str(create_time)
        company = fake.company().encode('utf-8')
        # Create a new record
        user = User()
        user.tel = fake.phone_number()
        user.username = username
        user.truename = truename
        user.job = job
        user.company = company
        user.password = password
        user.email = email
        user.id_no = makeIdNo()
        user.city_id = random.randint(1, 290)
        user.address = address
        user.summary = text
        user.gender = random.randint(1, 2)
        user.age = fake.random_int(10,50)
        user.site = fake.uri()
        user.uuid = fake.uuid4()
        user.save()


for i in range(thread_nums):
    t = threading.Thread(target=insert_user, args=(i, 400000))
    t.start()

CREATE TABLE `user` (
  `id` int(11) NOT NULL AUTO_INCREMENT,
  `tel` varchar(20) NOT NULL,
  `profile_id` int(11) NOT NULL DEFAULT '0',
  `username` varchar(20) NOT NULL,
  `truename` varchar(10) NOT NULL,
  `job` varchar(100) NOT NULL DEFAULT '',
  `company` varchar(100) NOT NULL DEFAULT '',
  `password` varchar(50) NOT NULL,
  `email` varchar(30) NOT NULL,
  `id_no` varchar(20) NOT NULL,
  `city_id` varchar(10) NOT NULL DEFAULT '',
  `address` varchar(50) NOT NULL,
  `summary` varchar(2000) NOT NULL,
  `gender` tinyint(4) NOT NULL,
  `age` tinyint(4) NOT NULL,
  `site` varchar(100) NOT NULL DEFAULT '',
  `uuid` varchar(50) NOT NULL,
  `created_at` datetime NOT NULL,
  `updated_at` datetime NOT NULL DEFAULT CURRENT_TIMESTAMP ON UPDATE CURRENT_TIMESTAMP,
  PRIMARY KEY (`id`)
) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8 COMMENT='测试用户表';