org.keycloak.common.util.UriUtils

我不得不在Keycloak扩展中解析uri和查询参数，并发现这个实用工具类非常有用:

org.keycloak.common.util.UriUtils:
static MultivaluedHashMap<String,String> decodeQueryString(String queryString) 

还有一个有用的方法可以删除一个查询参数:

static String   stripQueryParam(String url, String name)

解析URL有 org.keycloak.common.util.KeycloakUriBuilder:

KeycloakUriBuilder  uri(String uriTemplate)
String  getQuery()

还有很多其他的好东西。

一种现成的URI查询部分解码解决方案(包括解码和多参数值)

评论

我对https://stackoverflow.com/a/13592567/1211082中@Pr0gr4mm3r提供的代码不满意。基于流的解决方案不做URLDecoding，可变版本的笨拙。

因此，我阐述了一个解决方案

Can decompose a URI query part into a Map<String, List<Optional<String>>> Can handle multiple values for the same parameter name Can represent parameters without a value properly (Optional.empty() instead of null) Decodes parameter names and values correctly via URLdecode Is based on Java 8 Streams Is directly usable (see code including imports below) Allows for proper error handling (here via turning a checked exception UnsupportedEncodingExceptioninto a runtime exception RuntimeUnsupportedEncodingException that allows interplay with stream. (Wrapping regular function into functions throwing checked exceptions is a pain. And Scala Try is not available in the Java language default.)

Java代码

import java.io.UnsupportedEncodingException;
import java.net.URLDecoder;
import java.util.*;
import static java.util.stream.Collectors.*;

public class URIParameterDecode {
    /**
     * Decode parameters in query part of a URI into a map from parameter name to its parameter values.
     * For parameters that occur multiple times each value is collected.
     * Proper decoding of the parameters is performed.
     * 
     * Example
     *   <pre>a=1&b=2&c=&a=4</pre>
     * is converted into
     *   <pre>{a=[Optional[1], Optional[4]], b=[Optional[2]], c=[Optional.empty]}</pre>
     * @param query the query part of an URI 
     * @return map of parameters names into a list of their values.
     *         
     */
    public static Map<String, List<Optional<String>>> splitQuery(String query) {
        if (query == null || query.isEmpty()) {
            return Collections.emptyMap();
        }

        return Arrays.stream(query.split("&"))
                    .map(p -> splitQueryParameter(p))
                    .collect(groupingBy(e -> e.get0(), // group by parameter name
                            mapping(e -> e.get1(), toList())));// keep parameter values and assemble into list
    }

    public static Pair<String, Optional<String>> splitQueryParameter(String parameter) {
        final String enc = "UTF-8";
        List<String> keyValue = Arrays.stream(parameter.split("="))
                .map(e -> {
                    try {
                        return URLDecoder.decode(e, enc);
                    } catch (UnsupportedEncodingException ex) {
                        throw new RuntimeUnsupportedEncodingException(ex);
                    }
                }).collect(toList());

        if (keyValue.size() == 2) {
            return new Pair(keyValue.get(0), Optional.of(keyValue.get(1)));
        } else {
            return new Pair(keyValue.get(0), Optional.empty());
        }
    }

    /** Runtime exception (instead of checked exception) to denote unsupported enconding */
    public static class RuntimeUnsupportedEncodingException extends RuntimeException {
        public RuntimeUnsupportedEncodingException(Throwable cause) {
            super(cause);
        }
    }

    /**
     * A simple pair of two elements
     * @param <U> first element
     * @param <V> second element
     */
    public static class Pair<U, V> {
        U a;
        V b;

        public Pair(U u, V v) {
            this.a = u;
            this.b = v;
        }

        public U get0() {
            return a;
        }

        public V get1() {
            return b;
        }
    }
}

Scala代码

．.． 为了完整起见，我忍不住要用Scala提供简洁美观的解决方案

import java.net.URLDecoder

object Decode {
  def main(args: Array[String]): Unit = {
    val input = "a=1&b=2&c=&a=4";
    println(separate(input))
  }

  def separate(input: String) : Map[String, List[Option[String]]] = {
    case class Parameter(key: String, value: Option[String])

    def separateParameter(parameter: String) : Parameter =
      parameter.split("=")
               .map(e => URLDecoder.decode(e, "UTF-8")) match {
      case Array(key, value) =>  Parameter(key, Some(value))
      case Array(key) => Parameter(key, None)
    }

    input.split("&").toList
      .map(p => separateParameter(p))
      .groupBy(p => p.key)
      .mapValues(vs => vs.map(p => p.value))
  }
}

以下是我的解决方案与减少和可选:

private Optional<SimpleImmutableEntry<String, String>> splitKeyValue(String text) {
    String[] v = text.split("=");
    if (v.length == 1 || v.length == 2) {
        String key = URLDecoder.decode(v[0], StandardCharsets.UTF_8);
        String value = v.length == 2 ? URLDecoder.decode(v[1], StandardCharsets.UTF_8) : null;
        return Optional.of(new SimpleImmutableEntry<String, String>(key, value));
    } else
        return Optional.empty();
}

private HashMap<String, String> parseQuery(URI uri) {
    HashMap<String, String> params = Arrays.stream(uri.getQuery()
            .split("&"))
            .map(this::splitKeyValue)
            .filter(Optional::isPresent)
            .map(Optional::get)
            .reduce(
                // initial value
                new HashMap<String, String>(), 
                // accumulator
                (map, kv) -> {
                     map.put(kv.getKey(), kv.getValue()); 
                     return map;
                }, 
                // combiner
                (a, b) -> {
                     a.putAll(b); 
                     return a;
                });
    return params;
}

我忽略重复的参数(我取最后一个)。 我使用Optional<SimpleImmutableEntry<String, String>>稍后忽略垃圾 还原从一个空映射开始，然后在每个SimpleImmutableEntry上填充它

如果你问，reduce在最后一个参数中需要这个奇怪的组合器，它只在并行流中使用。它的目标是合并两个中间结果(这里是HashMap)。

我有一个Kotlin版本，看看这是如何在谷歌的顶部结果。

@Throws(UnsupportedEncodingException::class)
fun splitQuery(url: URL): Map<String, List<String>> {

    val queryPairs = LinkedHashMap<String, ArrayList<String>>()

    url.query.split("&".toRegex())
            .dropLastWhile { it.isEmpty() }
            .map { it.split('=') }
            .map { it.getOrEmpty(0).decodeToUTF8() to it.getOrEmpty(1).decodeToUTF8() }
            .forEach { (key, value) ->

                if (!queryPairs.containsKey(key)) {
                    queryPairs[key] = arrayListOf(value)
                } else {

                    if(!queryPairs[key]!!.contains(value)) {
                        queryPairs[key]!!.add(value)
                    }
                }
            }

    return queryPairs
}

还有扩展方法

fun List<String>.getOrEmpty(index: Int) : String {
    return getOrElse(index) {""}
}

fun String.decodeToUTF8(): String { 
    URLDecoder.decode(this, "UTF-8")
}

org.apache.http.client.utils.URLEncodedUtils

是否有一个知名的库可以帮你做到这一点

import org.apache.hc.client5.http.utils.URLEncodedUtils

String url = "http://www.example.com/something.html?one=1&two=2&three=3&three=3a";

List<NameValuePair> params = URLEncodedUtils.parse(new URI(url), Charset.forName("UTF-8"));

for (NameValuePair param : params) {
  System.out.println(param.getName() + " : " + param.getValue());
}

输出

one : 1
two : 2
three : 3
three : 3a

只是Java 8版本的更新

public Map<String, List<String>> splitQuery(URL url) {
    if (Strings.isNullOrEmpty(url.getQuery())) {
        return Collections.emptyMap();
    }
    return Arrays.stream(url.getQuery().split("&"))
            .map(this::splitQueryParameter)
            .collect(Collectors.groupingBy(SimpleImmutableEntry::getKey, LinkedHashMap::new, **Collectors**.mapping(Map.Entry::getValue, **Collectors**.toList())));
}

mapping和toList()方法必须用于顶部答案中没有提到的collector。否则它会在IDE中抛出编译错误