Maven 配置文件过滤与列表

Question

I'm using Maven Filtering to create a property file that would have specialized properties for a given stack.

I have a application.properties file with a property like

elasticsearch.members=${elasticsearch.members}

I have two config files

config.ppd.properties which has

elasticsearch.members=ppd-es-01:9700;ppd-es-02:9700;ppd-es-03:9700

config.prod.properties which has

elasticsearch.members=prod-es-01:9700;prod-es-02:9700;prod-es-03:9700

I have 2 sets of servers with the aliases ppd-es-nn and prod-es-nn

This all works great. When I generate the artifacts using the ppd profile, I get the correct property in the application.properties file.

But what I'd really like to do is have this concatenation somehow done in the application.properties file. So, I could specify the number of elasticsearch servers and a pattern in the application.properties file.

elasticsearch.members=${environment}-es-[index]:9700;...

Is there some way I can achieve that? Perhaps subclassing the class that does the filtering? This example is slightly contrived since I wanted to keep it simple. But one of the things I want to do is append the deployment mode (blue or green) to each server alias. And while I can do that in the filter file, it would be better to do it in the application.properties file so that no typo in the filter class could mess up the alias.

Answer 1

I like the groovy-maven-plugin for tasks like this. The below is untested, but should give you the idea.

<properties>
  <!-- these may be overridden on the cmdline -->
  <target.env>ppd</target.env>
  <es.server.count>2</es.server.count>
<properties>

<plugin>
  <groupId>org.codehaus.gmaven</groupId>
  <artifactId>groovy-maven-plugin</artifactId>
  <version><latestPluginVersionHere></version>
  <executions>
    <execution>
      <id>build-es-host-list</id>
      <!-- phase must be earlier than process-resources -->
      <phase>validate</phase>
      <goals>
        <goal>execute</goal>
      </goals>
      <configuration>
        <source>
           <![CDATA[
            def sb = new StringBuilder()
            def targetEnv = project.properties.getProperty('targetEnv')
            def numServers = project.properties.getProperty('es.server.count')
            (1..numServers).eachWithIndex { it, idx -> 
              sb.append("${targetEnv}-es-${idx}:9700")
              if (idx < numServers) {
                sb.append(';')
              }
            }
            project.properties.setProperty('elasticsearch.members', sb.toString())
          ]]>
        </source>
      </configuration>
    </execution>
  </executions>
  <dependencies>
    <dependency>
      <groupId>org.codehaus.groovy</groupId>
      <artifactId>groovy-all</artifactId>
      <version>2.4.12</version>
    </dependency>
  </dependencies>
</plugin>

Then, the application.properties file line needs to use the property:

elasticsearch.members=${elasticsearch.members}

The POM needs to enable resource filtering for the calculated value to be used in the file.