无需编程，基于甲骨文oracle数据库零代码生成CRUD增删改查RESTful API接口

回顾

通过之前一篇文章 无需编程，基于PostgreSQL零代码生成CRUD增删改查RESTful API接口 的介绍，采用抽象工厂设计模式，已经支持了大象数据库PostgreSQL。之前通过字符串拼接生成DDL SQL语句，比较繁琐。本文开始，引入了FreeMarker模版引擎，通过配置模版实现创建和修改物理表结构SQL语句，简化了大量代码，提高了效率，并且通过配置oracle数据库SQL模版，基于oracle数据库，零代码实现crud增删改查。

FreeMarker简介

FreeMarker是一款模板引擎： 即一种基于模板和要改变的数据，并用来生成输出文本(HTML网页，电子邮件，配置文件，源代码等)的通用工具。 它不是面向最终用户的，而是一个Java类库，是一款程序员可以嵌入他们所开发产品的组件。模板编写为FreeMarker Template Language (FTL)。它是简单的，专用的语言， 不是像PHP那样成熟的编程语言。 那就意味着要准备数据在真实编程语言中来显示，比如数据库查询和业务运算，之后模板显示已经准备好的数据。在模板中，你可以专注于如何展现数据，而在模板之外可以专注于要展示什么数据。

UI界面

通过产品对象为例，无需编程，基于Oracle数据库，通过配置零代码实现CRUD增删改查RESTful API接口和管理UI。

创建产品

编辑产品数据

产品数据列表

通过Oracle SQL Developer查询Oracle数据

定义元数据对象模型

元数据表ca_meta_table

元数据表ca_meta_table，用于记录表的基本信息。

TableEntity对象

TableEntity为“元数据表”对象，和ca_meta_table字段对应

public class TableEntity {
    private Long id;

    private String name;

    private String caption;

    private String description;

    private Timestamp createdDate;

    private Timestamp lastModifiedDate;

    private String pluralName;

    private String tableName;

    private EngineEnum engine;

    private Boolean createPhysicalTable;

    private Boolean reverse;

    private Boolean systemable;

    private Boolean readOnly;

    private List<ColumnEntity> columnEntityList;

    private List<IndexEntity> indexEntityList;
}

元数据列ca_meta_column

元数据列ca_meta_column，用于记录表字段信息，比如类型，长度，默认值等。

ColumnEntity对象

ColumnEntity为“元数据列”对象，和ca_meta_column字段对应

public class ColumnEntity {
  private Long id;

  private String name;

  private String caption;

  private String description;

  private Timestamp createdDate;

  private Timestamp lastModifiedDate;

  private Integer displayOrder;

  private DataTypeEnum dataType;

  private IndexTypeEnum indexType;

  private IndexStorageEnum indexStorage;

  private String indexName;

  private Integer length;

  private Integer precision;

  private Integer scale;

  private String defaultValue;

  private Long seqId;

  private Boolean unsigned;

  private Boolean autoIncrement;

  private Boolean nullable;

  private Boolean insertable;

  private Boolean updatable;

  private Boolean queryable;

  private Boolean displayable;

  private Boolean systemable;

  private Long tableId;
}

元数据索引ca_meta_index

元数据索引ca_meta_index，用于记录表联合索引信息，比如索引类型，名称等。

IndexEntity对象

IndexEntity为“元数据索引”对象，和ca_meta_index字段对应

public class IndexEntity {
  private Long id;

  private String name;

  private String caption;

  private String description;

  private Timestamp createdDate;

  private Timestamp lastModifiedDate;

  private IndexTypeEnum indexType;

  private IndexStorageEnum indexStorage;

  private Long tableId;

  private List<IndexLineEntity> indexLineEntityList;
}

元数据索引行ca_meta_index_line

元数据索引行ca_meta_index_line，用于记录表联合索引行信息，一个联合索引可以对应多个联合索引行，表示由多个字段组成。

IndexLineEntity对象

IndexLineEntity“元数据索行”对象，和ca_meta_index_line字段对应

public class IndexLineEntity {
  private Long id;

  private Long columnId;

  private ColumnEntity columnEntity;

  private Long indexId;
}

定义FreeMarker模版

创建表create-table.sql.ftl

CREATE TABLE "${tableName}" (
<#list columnEntityList as columnEntity>
  <#if columnEntity.dataType == "BOOL">
    "${columnEntity.name}" NUMBER(1)<#if columnEntity.defaultValue??> DEFAULT <#if columnEntity.defaultValue == "true">1<#else>0</#if></#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "INT">
    "${columnEntity.name}" INT<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity.indexType?? && columnEntity.indexType == "PRIMARY"> PRIMARY KEY</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "BIGINT">
    "${columnEntity.name}" INT<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity.indexType?? && columnEntity.indexType == "PRIMARY"> PRIMARY KEY</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "FLOAT">
    "${columnEntity.name}" FLOAT<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "DOUBLE">
    "${columnEntity.name}" REAL<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "DECIMAL">
    "${columnEntity.name}" DECIMAL<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "DATE">
    "${columnEntity.name}" DATE<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "TIME">
    "${columnEntity.name}" CHAR(8)<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "DATETIME">
    "${columnEntity.name}" DATE<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "TIMESTAMP">
    "${columnEntity.name}" TIMESTAMP<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "CHAR">
    "${columnEntity.name}" CHAR(${columnEntity.length})<#if columnEntity.defaultValue??> DEFAULT '${columnEntity.defaultValue}'</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity.indexType?? && columnEntity.indexType == "PRIMARY"> PRIMARY KEY</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "VARCHAR">
    "${columnEntity.name}" VARCHAR(${columnEntity.length})<#if columnEntity.defaultValue??> DEFAULT '${columnEntity.defaultValue}'</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity.indexType?? && columnEntity.indexType == "PRIMARY"> PRIMARY KEY</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "PASSWORD">
    "${columnEntity.name}" VARCHAR(200)<#if columnEntity.defaultValue??> DEFAULT '${columnEntity.defaultValue}'</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "ATTACHMENT">
    "${columnEntity.name}" VARCHAR(4000)<#if columnEntity.defaultValue??> DEFAULT '${columnEntity.defaultValue}'</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "TEXT">
    "${columnEntity.name}" VARCHAR(4000)<#if columnEntity.defaultValue??> DEFAULT '${columnEntity.defaultValue}'</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "LONGTEXT">
    "${columnEntity.name}" LONG<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "BLOB">
    "${columnEntity.name}" BLOB<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#elseif columnEntity.dataType == "LONGBLOB">
    "${columnEntity.name}" BLOB<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity_has_next>,</#if>
  <#else>
    "${columnEntity.name}" VARCHAR(200)<#if columnEntity.defaultValue??> DEFAULT ${columnEntity.defaultValue}</#if><#if columnEntity.nullable != true> NOT NULL</#if><#if columnEntity.indexType?? && columnEntity.indexType == "PRIMARY"> PRIMARY KEY</#if><#if columnEntity_has_next>,</#if>
  </#if>
</#list>
);

<#list columnEntityList as columnEntity>
  <#if columnEntity.indexType?? && columnEntity.indexType == "UNIQUE">
    ALTER TABLE "${tableName}" ADD CONSTRAINT "${columnEntity.indexName}" UNIQUE("${columnEntity.name}");
  </#if>

  <#if columnEntity.indexType?? && (columnEntity.indexType == "INDEX" || columnEntity.indexType == "FULLTEXT")>
    CREATE INDEX "${columnEntity.indexName}" ON "${tableName}" ("${columnEntity.name}");
  </#if>
</#list>

<#if indexEntityList??>
  <#list indexEntityList as indexEntity>
    <#if indexEntity.indexType?? && indexEntity.indexType == "UNIQUE">
      ALTER TABLE "${tableName}" ADD CONSTRAINT "${indexEntity.name}" UNIQUE(<#list indexEntity.indexLineEntityList as indexLineEntity>"${indexLineEntity.columnEntity.name}"<#if indexLineEntity_has_next>,</#if></#list>);
    </#if>

    <#if indexEntity.indexType?? && (indexEntity.indexType == "INDEX" || indexEntity.indexType == "FULLTEXT")>
      CREATE INDEX "${indexEntity.name}" ON "${tableName}" (<#list indexEntity.indexLineEntityList as indexLineEntity>"${indexLineEntity.columnEntity.name}"<#if indexLineEntity_has_next>,</#if></#list>);
    </#if>
  </#list>
</#if>

COMMENT ON TABLE "${tableName}" IS '${caption}';

<#list columnEntityList as columnEntity>
  COMMENT ON COLUMN "${tableName}"."${columnEntity.name}" IS '${columnEntity.caption}';
</#list>

模版解析SQL

首先保存元数据信息，下一步传递模版名称和元数据model，动态解析成创建表SQL语句，然后创建物理表，这样元数据和物理表就关联上了。运行时通过解析元数据动态生成insert，select，update，delete等SQL语句，零代码实现业务数据crud功能。

public String processTemplateToString(String database, String templateName, Object dataModel) {
    String str = null;
    StringWriter stringWriter = new StringWriter();
    try {
        Configuration config = new Configuration(Configuration.VERSION_2_3_31);
        config.setNumberFormat("#");
        String templateValue = getTemplate(database, templateName);
        if (templateValue == null) {
          return str;
        }

        Template template = new Template(templateName, templateValue, config);
        template.process(dataModel, stringWriter);

        str = stringWriter.getBuffer().toString().trim();
        log.info(str);
    } catch (Exception e) {
        e.printStackTrace();
        throw new BusinessException(ApiErrorCode.DEFAULT_ERROR, e.getMessage());
    }

    return str;
}

public List<String> toCreateTableSql(TableEntity tableEntity) {
  String createTableSql = processTemplateToString("create-table.sql.ftl", tableEntity);

  if (createTableSql == null) {
    throw new BusinessException(ApiErrorCode.DEFAULT_ERROR, "create-table.sql is empty!");
  }

  List<String> sqls = new ArrayList<String>();
  String[] subSqls = createTableSql.split(";");
  for (String t : subSqls) {
    String subSql = t.trim();
    if (!subSql.isEmpty()) {
      sqls.add(t);
    }
  }

  return sqls;
}

public Long create(TableDTO tableDTO) {
  TableEntity tableEntity = tableMapper.toEntity(tableDTO);
  //TODO
  Long tableId = crudService.create(TABLE_TABLE_NAME, tableEntity);
  List<String> sqlList = crudService.toCreateTableSql(tableEntity);
  for (String sql: sqlList) {
    execute(sql);
  }
  //TODO
  return tableId;
}

修改表

包括表结构和索引的修改，删除等，和创建表原理类似。

application.properties

需要根据需要配置数据库连接驱动，无需重新发布，就可以切换不同的数据库。

#oracle
spring.datasource.url=jdbc:oracle:thin:@//localhost:1521/XEPDB1
spring.datasource.driverClassName=oracle.jdbc.OracleDriver
spring.datasource.username=crudapi
spring.datasource.password=crudapi
spring.datasource.initialization-mode=always
spring.datasource.schema=classpath:schema.sql

小结

本文主要介绍了crudapi支持oracle数据库实现原理，并且以产品对象为例，零代码实现了CRUD增删改查RESTful API，后续介绍更多的数据库，比如MSSQL Server，Mongodb等。

实现方式	代码量	时间	稳定性
传统开发	1000行左右	2天/人	5个bug左右
crudapi系统	0行	1分钟	基本为0

综上所述，利用crudapi系统可以极大地提高工作效率和节约成本，让数据处理变得更简单！