需求
指标基本格式:
clm.{type}.{hostId}.$metricItem
示例1:
// 待匹配表达式:<hostId: 为36位的UUID>
summarize(clm.pm.xxx.yyy.uuu.zzz.0193f1b3-7bcc-4374-8546-8e87b7276003.agg.cpu.percent-avg.system,'10min','avg',true)
希望匹配出:
- type: pm.xxx.yyy.uuu.zzz
- hostId: 0193f1b3-7bcc-4374-8546-8e87b7276003
返回的Map为:
{hostId=0193f1b3-7bcc-4374-8546-8e87b7276003, type=pm.xxx.yyy.uuu.zzz}
示例2:
// 待匹配表达式:<hostId: 为36位的UUID>
summarize(clm.pm.0193f1b3-7bcc-4374-8546-8e87b7276003.agg.cpu.percent-avg.system,'10min','avg',true)
希望匹配出:
- type: pm
- hostId: 0193f1b3-7bcc-4374-8546-8e87b7276003
返回的Map为:
{hostId=0193f1b3-7bcc-4374-8546-8e87b7276003, type=pm}
示例3:
// 待匹配表达式:<hostId: 使用 "_"分割的IP地址>
summarize(clm.pm.xxx.yyy.uuu.zzz.10_144_202_141.agg.cpu.percent-avg.system,'10min','avg',true)
希望匹配出:
- type: pm.xxx.yyy.uuu.zzz
- hostId: 0193f1b3-7bcc-4374-8546-8e87b7276003
返回的Map为:
{hostId=10_144_202_141, type=pm.xxx.yyy.uuu.zzz}
基本思路
使用正则表达式的group()去获取结果;
- group()正则表示中,使用小括号
()
表示不同的group,有多少个开括号(
,就会有多少个组,表示为:group(0)、group(1)、group(N); - group()正则表示中,可以为不同的group添加名称,如
<type>
或<hostId>
,方便匹配;
对于示例1
和示例2
,可使用下面的group()表达式去匹配:
// hostId: 为36位的UUID
"^.*clm\.(?<type>\w+\.*.*)\.(?<hostId>\w{8}(?:-\w{4}){3}-\w{12}?)"
解析:
上面包括多个组,下面仅介绍感兴趣的组,比如:
1. (?<type>\w+\.*.*) : 别名为 type,匹配此()内的值;
2. (?<hostId>\w{8}(?:-\w{4}){3}-\w{12}?): 别名为 hostId,匹配此()内的值(注:为最外层的小括号)
对于示例3
,可使用下面的group()表达式去匹配:
// hostId: 使用 "_"分割的IP地址
"^.*clm\.(?<type>\w+\.*.*)\.(?<hostId>\d{0,3}_\d{0,3}_\d{0,3}_\d{0,3})"
代码实现
RegexGroupUtils.java
import com.google.common.base.Strings;
import com.google.common.collect.Maps;
import org.springframework.util.CollectionUtils;
import org.springframework.util.ObjectUtils;
import java.util.Map;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
public class RegexGroupUtils {
private static final Pattern groupAlias = Pattern.compile("\<(.*?)\>");
/**
* 使用正则表达式的group,查找出String中的参数值;<br/>
* 示例如下:
* <pre>{@code
* // 格式:
* clm.{type}.{hostId}.$metricItem
*
* // 待匹配表达式:<hostId: 为36位的UUID>
* summarize(clm.pm.xxx.yyy.uuu.zzz.0193f1b3-7bcc-4374-8546-8e87b7276003.agg.cpu.percent-avg.system,'10min','avg',true)
*
* 希望匹配出:
* - type: pm.xxx.yyy.uuu.zzz
* - hostId: 0193f1b3-7bcc-4374-8546-8e87b7276003
*
* 返回的Map为:
* {hostId=0193f1b3-7bcc-4374-8546-8e87b7276003, type=pm.xxx.yyy.uuu.zzz}
* }</pre>
*
* @param sourceString 待匹配的字符串
* @param pattern 正则表达式
* @return
*/
public static Map<String, String> findParams(String pattern, String sourceString) {
Map<String, String> mapResult = Maps.newConcurrentMap();
if (Strings.isNullOrEmpty(sourceString) || ObjectUtils.isEmpty(pattern)) {
return mapResult;
}
// 找出pattern中以"<xxxx>"表示的组别名
Set<String> groupNames = PlaceHolderReplaceUtils.findPlaceHolderKeys(pattern, groupAlias);
if (CollectionUtils.isEmpty(groupNames)) {
return mapResult;
}
// 创建 Pattern 对象
Pattern r = Pattern.compile(pattern);
// 现在创建 matcher 对象
Matcher m = r.matcher(sourceString);
if (m.find()) {
for (String param : groupNames) {
String value = m.group(param);
if (!Strings.isNullOrEmpty(value)) {
mapResult.put(param, value);
}
}
}
return mapResult;
}
}
PlaceHolderReplaceUtils.java
import com.google.common.base.Strings;
import com.google.common.collect.Sets;
import org.springframework.util.CollectionUtils;
import org.springframework.util.ObjectUtils;
import java.beans.PropertyDescriptor;
import java.lang.reflect.Method;
import java.util.Map;
import java.util.Set;
import java.util.regex.Matcher;
import java.util.regex.Pattern;
/**
* Desc: 占位符替换, 占位符表示为:{@code {placeholder}};
* <p>
* 示例:替换如下{xxx}占位符中的内容
* <pre>"名字:{name},年龄:{age},学校:{school}"</pre>
*/
public class PlaceHolderReplaceUtils {
private static final Pattern pattern = Pattern.compile("\{(.*?)\}");
private static Matcher matcher;
/**
* 查找String中的占位符keys;<br/>
* 示例: "名字:{name},年龄:{age},学校:{school}", 则返回:Set[name,age,school]
* <p>
* pattern示例:
* <pre> {@code
* // 尖括号:<placeHolder> 表示为占位符
* Pattern pattern = Pattern.compile("\<(.*?)\>");
*
* // 大括号:{placeHolder} 表示为占位符, 上面的示例中就使用{}作为占位符
* Pattern pattern = Pattern.compile("\{(.*?)\}");
* }
* </pre>
*
* @param sourceString
* @param pattern
* @return
*/
public static Set<String> findPlaceHolderKeys(String sourceString, Pattern pattern) {
Set<String> placeHolderSet = Sets.newConcurrentHashSet();
if (Strings.isNullOrEmpty(sourceString) || ObjectUtils.isEmpty(pattern)) {
return placeHolderSet;
}
String targetString = sourceString;
matcher = pattern.matcher(sourceString);
while (matcher.find()) {
String key = matcher.group(); //示例: {name}
String placeHolder = key.substring(1, key.length() - 1).trim(); //示例: name
placeHolderSet.add(placeHolder);
}
return placeHolderSet;
}
}
测试代码
import org.junit.Test;
public class RegexGroupUtilsTest {
@Test
public void findParams() {
//-----------start 测试匹配 uuid(32-36位)
String line1 = "summarize(clm.pm.xxx.yyy.uuu.zzz.0193f1b3-7bcc-4374-8546-8e87b7276003.agg.cpu.percent-avg.system,'10min','avg'," +
"true)";
String line2 = "summarize(clm.pm.0193f1b3-7bcc-4374-8546-8e87b7276003.agg.cpu.percent-avg.system,'10min','avg',true)";
String pattern = "^.*clm\.(?<type>\w+\.*.*)\.(?<hostId>\w{8}(?:-\w{4}){3}-\w{12}?)"; //OK
System.out.println("-----line1--------");
System.out.println(RegexGroupUtils.findParams(pattern, line1).toString());
System.out.println("-----line2--------");
System.out.println(RegexGroupUtils.findParams(pattern, line2).toString());
//-----------end 测试匹配 uuid(32-36位)
//-----------start 测试匹配 ip(下划线分割)
System.out.println("----------------IP------------");
String lineIp = "summarize(clm.pm.xxx.yyy.uuu.zzz.10_144_202_141.agg.cpu.percent-avg.system,'10min','avg'," +
"true)";
String patternIp = "^.*clm\.(?<type>\w+\.*.*)\.(?<hostId>\d{0,3}_\d{0,3}_\d{0,3}_\d{0,3})";
System.out.println(RegexGroupUtils.findParams(patternIp, lineIp).toString());
//-----------start 测试匹配 ip(下划线分割)
}
/**
* 输出:
*
* -----line1--------
* {hostId=0193f1b3-7bcc-4374-8546-8e87b7276003, type=pm.xxx.yyy.uuu.zzz}
* -----line2--------
* {hostId=0193f1b3-7bcc-4374-8546-8e87b7276003, type=pm}
* ----------------IP------------
* {hostId=10_144_202_141, type=pm.xxx.yyy.uuu.zzz}
*/
}