概述###
做报表时，有时需要根据不同的业务生成不同的报表。这样，需要能够动态地配置列字段，并根据列字段来输出对应的报表。使用函数接口结合枚举可以比较优雅地实现配置式编程。
问题描述如下：
假设有对象 Student, Teacher ，它们均有属性 Id, Name, Able 。 要求：(1) 打印这些 Student, Teacher 的字段 (Id, Name) 的内容 ; (2) 打印这些 Student, Teacher 的字段 (Name, Able) 的内容。
Java代码示例###
直接上代码。应该能看懂。 需要 Java1.8 才能正常运行。
接口定义####
package zzz.study.function;
/**
* Created by shuqin on 17/3/30.
*/
public interface Person {
String getId();
String getName();
String able();
}
对象定义####
类 Student ：
package zzz.study.function;
/**
* Created by shuqin on 17/3/30.
*/
public class Student implements Person {
private String studentId;
private String name;
private String able;
public Student(String studentId, String name, String able) {
this.studentId = studentId;
this.name = name;
this.able = able;
}
@Override
public String getId() {
return studentId;
}
@Override
public String getName() {
return name;
}
@Override
public String able() {
return able;
}
}
类 Teacher :
package zzz.study.function;
/**
* Created by shuqin on 17/3/30.
*/
public class Teacher implements Person {
private String teacherId;
private String name;
private String able;
public Teacher(String teacherId, String name, String able) {
this.teacherId = teacherId;
this.name = name;
this.able = able;
}
@Override
public String getId() {
return teacherId;
}
@Override
public String getName() {
return name;
}
@Override
public String able() {
return able;
}
}
字段定义配置####
字段定义配置是核心。 这里结合了枚举和函数式接口。这里之所以写成 FieldEnum(fieldName, fieldTitle, fieldValueGetMethod) 的定义方式，是为了便于管理。同样可以采用两个 Map 来实现：Map<fieldName, fieldTitle>, Map<fieldName, fieldValueGetMethod>，这样更适用于 Java1.6 , 不过要把两个 Map 拼起来才是完整的字段定义视图。 Person::getName 是方法引用，(Person p) -> p.getName() 的简写形式。
package zzz.study.function;
import java.util.function.Function;
/**
* Created by shuqin on 17/3/30.
*/
public enum FieldConf {
Id("Id", "编号", Person::getId),
Name("Name", "姓名", Person::getName),
Able("Able", "能力", Person::able);
private String name;
private String title;
private Function<Person, String> method;
FieldConf(String name, String title, Function<Person,String> method) {
this.name = name;
this.title = title;
this.method = method;
}
public String getName() {
return name;
}
public String getTitle() {
return title;
}
public Function<Person, String> getMethod() {
return method;
}
}
字段定义伴生类####
FieldConfAccompany 是 FieldConf 的伴生类， 从 Scala 的伴生对象借鉴而来，体现了 类变量、方法 与 实例变量、方法 分离的设计思想，使得两者各司其责， 都比较简洁。
package zzz.study.function;
import java.util.*;
/**
* Created by shuqin on 17/3/30.
* FieldConf 的伴生对象, 从Scala借鉴而来
*/
public class FieldConfAccompany {
private static Map<String, FieldConf> fieldConfMap = new HashMap<String, FieldConf>();
private static List<String> allFields = new ArrayList<>();
static {
for (FieldConf fc: FieldConf.values()) {
fieldConfMap.put(fc.name(), fc);
allFields.add(fc.getName());
}
}
public static FieldConf getInstance(String name) {
return fieldConfMap.get(name);
}
public static List<String> getAllFields() {
return Collections.unmodifiableList(allFields);
}
}
客户端使用####
这里使用了 java8 Stream api 。 并没有什么特别的，只是针对列表的批量流式处理，具备延迟计算特性。
package zzz.study.function;
import java.util.Arrays;
import java.util.List;
import java.util.stream.Collectors;
/**
* Created by shuqin on 17/3/30.
*/
public class Report {
public static void main(String[] args) {
report(Arrays.asList(new String[] {"Id", "Name"}), getPersons());
report(Arrays.asList(new String[] {"Name", "Able"}), getPersons());
}
public static void report(List<String> fields, List<Person> persons) {
String reportTitles = fields.stream().map(
field -> FieldConfAccompany.getInstance(field).getTitle()
).collect(Collectors.joining(","));
List<String> rows = persons.stream().map(
p -> fields.stream().map(
field -> FieldConfAccompany.getInstance(field).getMethod().apply(p)
).collect(Collectors.joining(","))
).collect(Collectors.toList());
System.out.println(reportTitles);
System.out.println(String.join("\n",rows));
}
private static List<Person> getPersons() {
Person s1 = new Student("s1", "liming", "Study");
Person s2 = new Student("s2", "xueying", "Piano");
Person t1 = new Teacher("t1", "Mr.Q", "Swim");
Person t2 = new Teacher("t2", "Mrs.L", "Dance");
return Arrays.asList(new Person[] {s1, s2, t1, t2});
}
}
输出：
编号,姓名
s1,liming
s2,xueying
t1,Mr.Q
t2,Mrs.L
姓名,能力
liming,Study
xueying,Piano
Mr.Q,Swim
Mrs.L,Dance
Scala代码示例###
业务类####
同样先定义业务类 Person, Student, Teacher 。 可以看到 Scala 的类定义比 Java 类定义的语法形式简洁不少，相当于语言层面实现了 lombok 的功能。
package scalastudy.extend
/**
* Created by shuqin on 17/4/10.
*/
trait Person {
def getId: String
def getName: String
def getAble: String
}
class Student(studentId: String, name:String, able:String) extends Person {
override def getId: String = studentId
override def getName: String = name
override def getAble: String = able
}
class Teacher(teacherId: String, name:String, able:String) extends Person {
override def getId: String = teacherId
override def getName: String = name
override def getAble: String = able
}
字段配置####
Scala 没有直接支持枚举类型，而是提供了 Enumeration 助手类。 可是这个类也不支持定义方法字段，因此，采用样例对象来模拟枚举功能。这里 apply 实现了静态构造器的功能，通过指定名称获取对应的样例对象。Scala 的 Case 功能非常强大，可以匹配常量、变量、容器结构及元素、类对象、正则表达式等各种对象，并赋值给相应的变量。
package scalastudy.extend
/**
* Created by shuqin on 17/4/10.
* Scala 实现枚举; 由于 Enumeration 不支持枚举含有方法字段，因此采用样例对象模拟实现。
*/
sealed class FieldConf(name:String, title:String, able: (Person)=>String) {
def getTitle = title
def getAble = able
}
object FieldConf {
def apply(name: String): FieldConf = {
name match {
case "Id" => Id
case "Name" => Name
case "Able" => Able
case _ => Unknown
}
}
}
case object Id extends FieldConf("Id", "编号", p => p.getId)
case object Name extends FieldConf("Name", "姓名", p => p.getName)
case object Able extends FieldConf("Able", "能力", p => p.getAble)
case object Unknown extends FieldConf("Unknown", "未知", p => "")
报表输出####
Scala 提供了相当多的助手方法，可以方便地实现常用功能，比如对列表拼接字符串。 在 Java 中就要难受地一次次编写无聊的 new StringBuilder , append, return sb.toString 这种套话， 而在 Scala 只要使用 mkString 即可，类似于 Python 的 join 方法。 Scala 的 lambda 表达式也很简洁，如果只有单变量的话，不必显式写出 p => doFor(p) 的形式， 而是直接可写成 doFor(_) 。
package scalastudy.extend
/**
* Created by shuqin on 17/4/10.
*/
object ExtendedReport extends App {
launch()
def launch(): Unit = {
report(List("Id", "Name"), getPersons())
report(List("Name", "Able"), getPersons())
}
def report(fields:List[String], persons:List[Person]):Unit = {
val titles = fields.map(FieldConf(_).getTitle).mkString(",")
println(titles)
val rows = persons.map(
p => fields.map(FieldConf(_).getAble.apply(p)).mkString(",")
).mkString("\n")
println(rows)
}
def getPersons():List[Person] = {
List(new Student("s1", "liming", "Study"), new Student("s2", "xueying", "Piano"),
new Teacher("t1", "Mr.Q", "Swim"), new Teacher("t2", "Mrs.L", "Dance"))
}
}
