打算抄袭一下MyBatis的Java注解方式配置,添加到在正在开发的持久层工具里,只需要定义一个接口方法,就可以用一个动态代理来使用它,这种方式的优点是可以利用Java的导舤功能快速定位SQL,比文本方式保存的模板定位方便,而且方法名和参数都是Java强类型,支持重构:
@Select("select * from users where id = #{id}") User getUserById(Integer id);
但还是老问题,Java对多行文本的支持太差,MyBatis的解决方法个人不是太喜欢,破坏了SQL的可读性:
private String selectPersonSql() {
return new SQL() {{
SELECT("P.ID, P.USERNAME, P.PASSWORD, P.FULL_NAME");
SELECT("P.LAST_NAME, P.CREATED_ON, P.UPDATED_ON");
FROM("PERSON P");
FROM("ACCOUNT A");
INNER_JOIN("DEPARTMENT D on D.ID = P.DEPARTMENT_ID");
INNER_JOIN("COMPANY C on D.COMPANY_ID = C.ID");
WHERE("P.ID = A.ID");
WHERE("P.FIRST_NAME like ?");
OR();
WHERE("P.LAST_NAME like ?");
GROUP_BY("P.ID");
HAVING("P.LAST_NAME like ?");
OR();
HAVING("P.FIRST_NAME like ?");
ORDER_BY("P.ID");
ORDER_BY("P.FULL_NAME");
}}.toString();
}
本人以前发过一篇文章,试图用Java注释的方式来支持多行文本,以方便SQL的定位(见https://my.oschina.net/drinkjava2/blog/892309),但是那篇文章中的方法有三个问题,一是Java文件在布署时,要手工拷贝一份到resource目录中去,不方便;二是不小心注释会被编辑器给格式化了;三是需要手工在程序中给出Java源文件的绝对路径。最近绞尽脑汁,又找到一个好一点的方法,依然是利用Java注释,但是没有以上三个缺点了:
做法:
1. 在pom.xml中添加一个build-helper-maven-plugin插件,将resource目录加入编译范围,插件会将所有在resources下的所有Java文件当作资源文件打包,这样就不用再手工拷贝一份到resouces目录下去了。而且经实测,这个插件在运行 mvn eclipse:eclipse 命令时能正确添加resources目录到eclipse中的编译目录:
<plugin>
<groupId>org.codehaus.mojo</groupId>
<artifactId>build-helper-maven-plugin</artifactId>
<version>3.0.0</version>
<executions>
<execution>
<id>add-source</id>
<phase>generate-sources</phase>
<goals>
<goal>add-source</goal>
</goals>
<configuration>
<sources>
<source>src/main/resources</source>
</sources>
</configuration>
</execution>
<execution>
<id>add-test-source</id>
<phase>generate-sources</phase>
<goals>
<goal>add-test-source</goal>
</goals>
<configuration>
<sources>
<source>src/test/resources</source>
</sources>
</configuration>
</execution>
</executions>
</plugin>
2. 写一个工具类,用来读取Java源码, 在Eclipse环境下时,从project/src/main或project/src/text下读,在布署环境下时,从类根目录下读取(因为resources下的Java文件被当作资源文件打包,路径变成了从根目录算起):
public abstract class TextUtils {// NOSONAR
private static final Map<Class<?>, String> javaFileCache = new ConcurrentHashMap<Class<?>, String>();
@SuppressWarnings("all")
public static String getJavaSourceCodeUTF8(Class<?> clazz) {
return getJavaSourceCode(clazz, "UTF-8");
}
@SuppressWarnings("all")
public static String getJavaSourceCode(Class<?> clazz, String encoding) {
if (javaFileCache.containsKey(clazz))
return javaFileCache.get(clazz);
String classPathName = StrUtils.substringBefore(clazz.getName(), "$");// aa.bb.Cc
classPathName = "/" + StrUtils.replace(classPathName, ".", "/");// /aa/bb/Cc
String fileName = classPathName + ".java";// /aa/bb/Cc.java
InputStream inputStream = null;
try {
inputStream = TextUtils.class.getResourceAsStream(fileName);
if (inputStream == null) {// Not found, it means in eclipse
File file = new File(clazz.getResource(classPathName + ".class").getFile());
String absPath = file.getAbsolutePath();
absPath = StrUtils.replace(absPath, "\\", "/");
String projectFolder = StrUtils.substringBefore(absPath, "/target/");
String realFile = projectFolder + "/src/main/resources" + classPathName + ".java";
file = new File(realFile);
if (!file.exists()) {
realFile = projectFolder + "/src/test/resources" + classPathName + ".java";
file = new File(realFile);
}
if (!file.exists())
throw new IOException("Can not find '" + realFile + "' in resources folder");
inputStream = new FileInputStream(file);
if (inputStream == null)
throw new IOException("Error happen when open file '" + realFile + "'");
}
ByteArrayOutputStream result = new ByteArrayOutputStream();
byte[] buffer = new byte[1024];
int length;
while ((length = inputStream.read(buffer)) != -1)
result.write(buffer, 0, length);
String javaSrc = result.toString(encoding);
javaFileCache.put(clazz, javaSrc);
return javaSrc;
} catch (IOException e) {
throw new RuntimeException(e);
} finally {
if (inputStream != null)
try {
inputStream.close();
} catch (IOException e) {
throw new RuntimeException(e);
}
}
}
}
3. 在src/main/resources或src/test/resources目录下添加需要访问源码的Java文件如:
public interface TextSample1 {
@SQL("select *from User")
public <T> T retrieveAllUsers();
@Handler(MapListHandler.class) //DbUtils自带的
public <T> T retrieveUserById(String firstName, String lastName);
/*-
SELECT P.ID, P.USERNAME, P.PASSWORD, P.FULL_NAME
P.LAST_NAME,P.CREATED_ON, P.UPDATED_ON
FROM PERSON P, ACCOUNT A
INNER JOIN DEPARTMENT D on D.ID = P.DEPARTMENT_ID
INNER JOIN COMPANY C on D.COMPANY_ID = C.ID
WHERE (P.ID = A.ID AND P.FIRST_NAME like #{firstName})
OR (P.LAST_NAME like #{lastName})
GROUP BY P.ID
HAVING (P.LAST_NAME like #{lastName})
OR (P.FIRST_NAME like #{firstName})
ORDER BY P.ID, P.FULL_NAME
*/
}
4.在程序的任何地方,就可以用以下代码获取java的源码,以便做进一步处理了(抽取出对应方法的多行文本,这一步我还没有做,但1到4都测试过了,所以想当然是没问题的)。
TextUtils.getJavaSourceCodeUTF8(TextSample1.class)
对了,差点忘了说,所有注释必须用/*-开头,这样就不会被Eclipse给误格式化了。
以上方法只在Eclipse+Maven 开发方式下实测过,其它编辑器不在考虑范围内。