README

一种通用工具，用于从数据库中提取和清理选定的表作为测试数据。在json配置文件的控制下，将数据库的子集从一台数据库复制到另一台。然后可以将后者数据库导出到SQL文件作为测试数据文件。

为什么选择PHParrot？

小时候我们去探望祖母的时候，我和我哥哥会等到所有的成年人离开房间，然后不停地对我们祖母的宠物鹦鹉重复我们所知道的（虽然非常温和）脏话，希望它会模仿我们。

因为这个工具（鹦鹉）从一处复制（文本）到另一处（并且因为这似乎是给项目取一个包含“PHP”的项目名的绝佳机会），所以“PHParrot”似乎很合适。

顺便说一句：祖母的鹦鹉从未重复过我们试图教它的任何词，但它学会了打嗝（我叔叔的影响）。

用法

• 创建您希望填充的目标数据库。工具只会输出到现有数据库。目标数据库的内容将被删除。
• 使用数据库服务器配置创建 config/credentials.json
• 创建 config/*.db.json 文件以定义每个所需数据库的映射
• 运行 bin/dbsampler.php

配置格式

所有配置文件都位于 config 子目录下。文件 不能 包含注释，因为JSON格式不支持此功能，但作为惯例，如果可能，将忽略名为“comment”的字段。

credentials.json

"driver": "pdo_mysql" 目前假定，但这可能在将来改变。

MySQL

参见 config/credentials.dist.json

{
  "driver": "pdo_mysql",
  "dbUser": "root",
  "dbPassword": "SOMEPASSWORD",
  "dbHost": "127.0.0.1"
}

如果您需要不同的源和目标服务器，则变为：

{
  "source": {
    "driver": "pdo_mysql",
    "dbUser": "root",
    "dbPassword": "SOMEPASSWORD",
    "dbHost": "sourceDB.example.com"
  },

  "dest" : {
    "driver": "pdo_mysql",
    "dbUser": "root",
    "dbPassword": "SOMEPASSWORD",
    "dbHost": "127.0.0.1"
  }
}

如果您需要准备连接，请添加一个 initialSql 章节来准备连接

{
  "source": {
    "driver": "pdo_mysql",
    "dbUser": "root",
    "dbPassword": "SOMEPASSWORD",
    "dbHost": "sourceDB.example.com",
    "initialSql": [
      "SET NAMES UTF8"
    ]
  },
  "dest": {
    "driver": "pdo_mysql",
    "dbUser": "root",
    "dbPassword": "SOMEPASSWORD",
    "dbHost": "127.0.0.1",
    "initialSql": [
      "SET NAMES UTF8",
      "SET foreign_key_checks = 0"
    ]
  }
}        

Sqlite

参见 config/credentials.dist.json

{
    "driver": "pdo_sqlite",
    "directory": "..\/path\/to\/sqlite-dbs"
}

路径默认相对于配置文件，除非它们以“/”开头。要迁移的Sqlite数据库默认为此目录中的 *.sqlite 文件

dbname.db.json

{
  "name": "small-sqlite-test",          # Configuration name
  "sourceDb": "small-source",           # Name of the source DB
  "destDb": "small-dest",               # Name of the destination DB. This DB will get trashed
  "tables": {                           # A set of tables to be copied over. Each table is defined as "table": config
                                        # Every config stanza requires a sampler field. For now, look these up in 
                                        # \PHParrot\Parrot\MigrationConfigProcessor::$samplerMap
                                        # All other fields depend on the specific sampler being used; these should 
                                        # all be documented in their own class files in src/Sample
    "fruits": {
      "sampler": "matched",
      "constraints": {
        "name": [
          "apple",
          "pear"
        ]
      },
      "remember": {
        "id": "fruit_ids"               # Cross-referencing is supported by "remember" stanzas
                                        # These take the field name of which the values are to be remembered
                                        # matched to a variable name in which the values will be stored
                                        # Note: Variable declarations do not include a '$' symbol 
                                        # References MUST be 'remember'ed before being used, there is no
      }                                 # dependency resolution here, so order your config appropriately
    },
    "vegetables": {
      "sampler": "NewestById",
      "idField": "id",
      "quantity": 2
    },
    "fruit_x_basket": {
      "sampler": "matched",
      "constraints": {
        "fruit_id": "$fruit_ids"        # Remembered variables, with $ sign, can be used as cross-references
                                        # This will expand to all ids of the fruits table matched above
      },
      "where" : [
        "basket_id > 1"                 # The matched sampler can also accept a list of arbitrary WHERE clauses
      ],
      "remember": {
        "basket_id": "basket_ids"
      }
    },
    "baskets": {
      "sampler": "matched",             # samplers support field cleaners that are defined in
                                        # \PHParrot\Parrot\FieldCleanerProvider::getCleanerByName
                                        # They modify or replace the content of the field that they are keyed to
      "constraints": {
        "id": "$basket_ids"
      },
      "cleanFields": {
        "name": "fakefullname"
      }
    }
  },
  "views": [                            # view support is experimental
    "some_view"                         # views are specified as name only but format may change
  ]                                     # The destination's CURRENT_USER() is used as the DEFINER for MySQL DBs
}

"Faker" 清理器

任何不需要参数的 'faker' (fzaninotto/faker) 生成器都可以通过使用 "name": "faker:GENERATOR" 在 cleanFields 章节中直接使用，例如：

 "cleanFields": {
   "ip": "faker:ipv4"
 },

扩展项目

该工具的设计主要是通过添加自定义采样器（必须实现 \PHParrot\Parrot\Sampler\Sampler）和清理器（必须实现 \PHParrot\Parrot\Cleaner\FieldCleaner 并使用 \PHParrot\Parrot\Cleaner\RowCleaner::registerCleaner 进行注册）来扩展。

目前，只支持mysql和sqlite数据库，但这也可能被扩展。

此工具不转换数据库驱动程序。这意味着如果您的源数据库是MySQL，则目标数据库也必须是MySQL。