如何使用 FlywayDB 处理 Oracle 的 UTF-8 字符集

Question

这里是flyway maven插件配置：

<plugins>
    <plugin>
        <groupId>org.flywaydb</groupId>
        <artifactId>flyway-maven-plugin</artifactId>
        <version>3.2.1</version>
        <configuration>
            <sqlMigrationSuffix>.oracle.sql</sqlMigrationSuffix>
            <placeholderReplacement>true</placeholderReplacement>
            <placeholderPrefix>#[</placeholderPrefix>
            <placeholderSuffix>]</placeholderSuffix>
            <encoding>UTF-8</encoding>
            <table>T00M001</table>
            <locations>
                <location>classpath:/META-INF/flyway/oracle</location>
                <location>classpath:/com/chorke/dbms/jdbcmi/oracle</location>
            </locations>
            <resolvers>
                <resolver>com.chorke.dbms.flyway.resolvers.FlywayResolverImpl</resolver>
            </resolvers>
            <callbacks>
                <callback>com.chorke.dbms.flyway.callbacks.AfterEachMigrate</callback>
                <callback>com.chorke.dbms.flyway.callbacks.BeforeEachMigrate</callback>
            </callbacks>
             <serverId>chorke.flyway.oracle</serverId>
             <url>jdbc:oracle:thin:@127.0.0.1:1521:xe</url>
        </configuration>
        <dependencies>
            <dependency>
                <groupId>com.chorke.dbms</groupId>
                <artifactId>chorke-dbms-flyway</artifactId>
                <version>1.0.00-SNAPSHOT</version>
            </dependency>
            <dependency>
                <groupId>com.chorke.dbms</groupId>
                <artifactId>chorke-dbms-oracle</artifactId>
                <version>1.0.00-SNAPSHOT</version>
            </dependency>
            <dependency>
                <groupId>com.chorke.dbms</groupId>
                <artifactId>chorke-dbms-jdbcmi</artifactId>
                <version>1.0.00-SNAPSHOT</version>
            </dependency>
            <dependency>
                <groupId>com.oracle</groupId>
                <artifactId>ojdbc6</artifactId>
                <version>11.2.0.3</version>
            </dependency>
        </dependencies>
    </plugin>
</plugins>

这里是flyway.properties

flyway.encoding: UTF-8
flyway.placeholders.key: value
flyway.placeholders.name: nome
flyway.placeholders.prop: value

这是迁移脚本的片段：

-- DDL OF T01I001
-- -----------------------------------------------------------------------------

CREATE TABLE t01i001
(
  t_msg_code VARCHAR2(5 BYTE) NOT NULL 
, t_lang1_msg VARCHAR2(80 BYTE) 
, t_lang2_msg VARCHAR2(80 BYTE) 
, t_entry_user VARCHAR2(4 BYTE) 
, t_entry_date DATE 
, t_upd_user VARCHAR2(4 BYTE) 
, t_upd_date DATE 
);

--  DML OF T01I001
-- -----------------------------------------------------------------------------

INSERT INTO t01i001 (t_msg_code,t_lang1_msg,t_lang2_msg,t_entry_user
,t_entry_date,t_upd_user,t_upd_date) VALUES ('00','يوجد سجل أخر لليوم والوقت المحدد'
,'Record already exists for the specified Date and Time.','2'
,to_date('13-10-2005','DD-MM-YYYY'),'C',to_date('27-11-2005','DD-MM-YYYY'));

INSERT INTO t01i001 (t_msg_code,t_lang1_msg,t_lang2_msg,t_entry_user
,t_entry_date,t_upd_user,t_upd_date) VALUES ('01'
,'البيانات قد أدخلت من قبل  لرقم الزيارة هذا'
,'Data has already been entered for this Visit Number.','2'
,to_date('13-10-2005','DD-MM-YYYY'),'C',to_date('27-11-2005','DD-MM-YYYY'));

这里是 oracle 11g XE R2 字符集的注册表：

NLS_LANG =AMERICAN_AMERICA.UTF8

这是 Maven 目标：

mvn clean install flyway:migrate flyway:info

出现了两类问题：

t_lang1_msg尺寸超过
导入的字符已更改

虽然我们增加了 t_lang1_msg 大小，但迁移工作正常。但它会自动转换 يوجد سول أخر لليوم والوقت المحدد 到 ÙŠÙцØ¬Ø¯ Ø³Ø¬Ù„ Ø£Ø®Ø± Ù„Ù„ÙŠÙцÙ... §Ù^ „Ù^Ù，Øª Ø§Ù„Ù...ØØ¯Ø¯。哪一个对我们来说是不可读的。

关于这个问题的任何解决方案？感谢您对解决此问题的回答。

Answer 1

其实对flyway-maven-plugin、flyway.properties以及sql 脚本及其编码。但是 ant 脚本目标内部存在问题。在每个 sprint 的基础上，有一项任务是 concat sql 文件的特定顺序。如下：

<concat destfile="${file.flyway.sql.concat}" outputencoding="UTF-8"append="true">
    <sort xmlns:rcmp="antlib:org.apache.tools.ant.types.resources.comparators">
        <fileset refid="dir.set.flyway.db.version.build"/>
        <rcmp:name />
    </sort>
</concat>

将 encoding="UTF-8" 属性添加到 concat 任务后，它工作正常。

<concat destfile="${file.flyway.sql.concat}" 
    outputencoding="UTF-8" encoding="UTF-8" append="true">
    <sort xmlns:rcmp="antlib:org.apache.tools.ant.types.resources.comparators">
        <fileset refid="dir.set.flyway.db.version.build"/>
        <rcmp:name />
    </sort>
</concat>

这就解决了这个问题。

如何使用 FlywayDB 处理 Oracle 的 UTF-8 字符集

How to handle UTF-8 charset with Oracle using FlywayDB

java

oracle

plsql

maven

flyway