迁移文件中所有额外的垃圾是什么？答案

【问题标题】：What is with all the extra junk in the migration file?迁移文件中所有额外的垃圾是什么？
【发布时间】：2011-09-01 02:17:06
【问题描述】：

我今天开始测试 grails 数据库迁移插件。我与域进行了初始同步，然后基于运行 diff 命令创建了一个新的更改日志。这应该像重命名列和（重新）将索引分配给该列一样简单。所以我希望看到以下...

changeSet(author: "gdboling (generated)", id: "1306242441630-1") {
    addColumn(tableName: "manuscript_review_reviewer") {
        column(name: "reviewer_id", type: "bigint") {
            constraints(nullable: "false")
        }
    }
}
changeSet(author: "gdboling (generated)", id: "1306242441630-2") {
    addPrimaryKey(columnNames: "manuscript_review_id, reviewer_id", constraintName: "manuscript_rePK", tableName: "manuscript_review_reviewer")
}
changeSet(author: "gdboling (generated)", id: "1306242441630-3") {
    dropPrimaryKey(tableName: "manuscript_review_reviewer")
}
changeSet(author: "gdboling (generated)", id: "1306242441630-4") {
    dropForeignKeyConstraint(baseTableName: "manuscript_review_reviewer", baseTableSchemaName: "manning_main_dev", constraintName: "FK9BEFBBB34863C659")
}
changeSet(author: "gdboling (generated)", id: "1306242441630-179") {
    addForeignKeyConstraint(baseColumnNames: "reviewer_id", baseTableName: "manuscript_review_reviewer", constraintName: "FK9BEFBBB3A28B41DF", deferrable: "false", initiallyDeferred: "false", referencedColumnNames: "id", referencedTableName: "user", referencesUniqueColumn: "false")
}
changeSet(author: "gdboling (generated)", id: "1306242441630-212") {
    dropColumn(columnName: "user_id", tableName: "manuscript_review_reviewer")
}

一切都很好。但是，我还看到很多与我的更改无关的 dropIndex 和 createIndex 块。有人知道为什么会这样吗？

【问题讨论】：

标签： grails database-migration liquibase

【解决方案1】：

在http://jira.grails.org/browse/GPDATABASEMIGRATION 上写一个问题，提供要重现的信息（数据库供应商，最好是一些代码和您的数据库架构），我会看看可以做什么。最坏的情况是您需要将多个元素合并为一个 - 这是自动化的棘手工作，通常需要一些人工干预。

【讨论】：