Java 正则表达式 es查询参数range匹配和替换
目录
在es查询的时候,处理了es参数查询中常见的内容。如果是有range的情况,要怎么处理呢? Range是带三层的花括号的内容:
{
"range": {
"age": {
"gte": 25,
"lt": 65
}
}
}
匹配单个range
要求:
如何把
{filter": [{"range":{"log_date":{"gte":"$start","lte":"$param.end","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}}]}
参数:
Map<String,Object> single = new HashMap<>();
single.put("paramName","me");
single.put("start","2021-03-08");
single.put("ends","2021-03-10");
转换为:
{filter": []}
处理思路:
1,先匹配有值的情况
2,匹配完后还有带参数的情况,整个置空
代码:
public static void main(String[] args) {
Map<String,Object> single = new HashMap<>();
single.put("paramName","me");
single.put("start","2021-03-08");
single.put("end","2021-03-10");
List<String> one = Lists.newArrayList("paramName","start");
String content = "{filter\": [{\"range\":{\"log_date\":{\"gte\":\"$start\",\"lte\":\"$param.end\",\"time_zone\":\"+08:00\",\"format\":\"yyyy-MM-dd HH:mm:ss\"}}}]}";
System.out.println("ori: "+content);
for (String e : ListUtils.emptyIfNull(one)) {
String value = MapUtils.getString(single, e);
// 先匹配现有的内容
content = matchReplaceWithCondition(content, e, value);
}
getReplaceRange(content);
}
public static String matchReplaceWithCondition( String content,String condition,String value)
{
String pattern = "\\$([a-zA-Z0-9_.]*)" + condition;
Pattern p = Pattern.compile(pattern);
Matcher m = p.matcher(content);
StringBuffer sb = new StringBuffer();
while (m.find()) {
String group = m.group();
m.appendReplacement(sb, group == null ? "" : ("\"").concat(value).concat("\""));
}
m.appendTail(sb);
return sb.toString();
}
public static String getReplaceRange(String content) {
String pattern = "\\{\"range\":\\{\"([a-zA-Z_]*)\":\\{\"([a-zA-Z_]*)\":(.*)}}}";
Pattern p = Pattern.compile(pattern);
Matcher m = p.matcher(content);
StringBuffer sb = new StringBuffer();
while (m.find()) {
String value = m.group();
System.out.println("match: "+value);
m.appendReplacement(sb, value.contains("$")? "" : value);
}
m.appendTail(sb);
System.out.println("getReplaceRange: "+sb.toString());
return sb.toString();
}
pattern = "\\{\"range\":\\{\"([a-zA-Z_]*)\":\\{\"([a-zA-Z_]*)\":(.*)}}}"; range里面后面有多个字段,用 (.*) 全匹配
结果:
ori: {filter": [{"range":{"log_date":{"gte":"$start","lte":"$param.end","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}}]}
match: {"range":{"log_date":{"gte":""2021-03-08"","lte":"$param.end","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}}
getReplaceRange: {filter": []}
多个的情况:
如果是多个的话,会是什么情况呢?
String content = "{\"filter\": [{\"range\":{\"log_date\":{\"gte\":\"$start\",\"lte\":\"$param.end\",\"time_zone\":\"+08:00\",\"format\":\"yyyy-MM-dd HH:mm:ss\"}}}" +
",{\"range\":{\"age\":{\"gte\":\"11\",\"lte\":\"25\"}}}]}";
结果:
ori: {"filter": [{"range":{"log_date":{"gte":"$ff","lte":"$param.time","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}},{"range":{"age":{"gte":"11","lte":"25"}}}]}
match: {"range":{"log_date":{"gte":"$ff","lte":"$param.time","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}},{"range":{"age":{"gte":"11","lte":"25"}}}
getReplaceRange: {"filter": []}
思考:
匹配的时候,都匹配到了,导致整个都被替换了。那该怎么处理呢? 考虑限制匹配次数吗?
再去分析下多个的情况: 多个的话,之间是逗号隔开的,A,B,C;情况基本是两种,单个的,只有本身,多个的都是多个:单个带逗号 和最后的单个结合。所以可以用或“|”的关系去写;
匹配多个range
代码:
public static String getReplaceRange(String content) {
String patternReg = "\\{\"range\":\\{\"([a-zA-Z_]*)\":\\{\"([a-zA-Z_]*)\":(.*)}}}";
String pattern = patternReg+",|"+patternReg;
Pattern p = Pattern.compile(pattern);
Matcher m = p.matcher(content);
StringBuffer sb = new StringBuffer();
while (m.find()) {
String value = m.group();
System.out.println("match: "+value);
m.appendReplacement(sb, value.contains("$")? "" : value);
}
m.appendTail(sb);
System.out.println("getReplaceRange: "+sb.toString());
return sb.toString();
}
结果:
public static String getReplaceRange(String content) {
String patternReg = "\\{\"range\":\\{\"([a-zA-Z_]*)\":\\{\"([a-zA-Z_]*)\":(.*)}}}";
String pattern = patternReg+",|"+patternReg;
Pattern p = Pattern.compile(pattern);
Matcher m = p.matcher(content);
StringBuffer sb = new StringBuffer();
while (m.find()) {
String value = m.group();
System.out.println("match: "+value);
m.appendReplacement(sb, value.contains("$")? "" : value);
}
m.appendTail(sb);
System.out.println("getReplaceRange: "+sb.toString());
return sb.toString();
}
这样就匹配到了
整合:
整合es查询参数匹配和替换的内容,进行测试
public static void main(String[] args) {
Map<String,Object> single = new HashMap<>();
single.put("paramName","me");
single.put("mean","wonders");
single.put("detail","yan");
single.put("start","2020-03-05");
single.put("end","2020-03-05");
List<String> one = Lists.newArrayList("paramName","detail","start");
String content = "{\"bool\":{\"must\":[{\"match\":{\"flowname\":$param.paramName}},{\"match\":{\"flowId\":$param.paramId}}]," +
"\"must_not\":[{\"term\":{\"reasonCode\":$mean}},{\"term\":{\"reasons\":$param.detail}}]" +
",\"filter\": [{\"range\":{\"log_date\":{\"gte\":\"$start\",\"lte\":\"$param.end\",\"time_zone\":\"+08:00\",\"format\":\"yyyy-MM-dd HH:mm:ss\"}}}," +
"{\"range\":{\"age\":{\"gte\":\"11\",\"lte\":\"25\"}}}]}}";
System.out.println("ori: "+content);
for (String e : ListUtils.emptyIfNull(one)) {
String value = MapUtils.getString(single, e);
// 先匹配现有的内容
content = matchReplaceWithCondition(content, e, value);
}
// 匹配参数内容为空的情况
content = getReplaceJson(content);
// 处理range的情况
content = getReplaceRange(content);
}
结果:
ori: {"bool":{"must":[{"match":{"flowname":$param.paramName}},{"match":{"flowId":$param.paramId}}],"must_not":[{"term":{"reasonCode":$mean}},{"term":{"reasons":$param.detail}}],"filter": [{"range":{"log_date":{"gte":"$start","lte":"$param.end","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}},{"range":{"age":{"gte":"11","lte":"25"}}}]}}
getReplaceJson: {"bool":{"must":[{"match":{"flowname":"me"}}],"must_not":[{"term":{"reasons":"yan"}}],"filter": [{"range":{"log_date":{"gte":""2020-03-05"","lte":"$param.end","time_zone":"+08:00","format":"yyyy-MM-dd HH:mm:ss"}}},{"range":{"age":{"gte":"11","lte":"25"}}}]}}
getReplaceRange: {"bool":{"must":[{"match":{"flowname":"me"}}],"must_not":[{"term":{"reasons":"yan"}}],"filter": [{"range":{"age":{"gte":"11","lte":"25"}}}]}}
总结:
匹配多个的时候,可以先考虑单个的匹配,然后把多个情况进行分解,找出可能的情形,这样写pattern就容易一些。逐个分解,逐个击破。
除了aggs统计内容,加了range的处理,查询条件基本上是OK了。后面有碰到其它的情况,再进行处理。