试试Freemarker，模板库。

public static String format(String format, Map<String, Object> values) {
    StringBuilder formatter = new StringBuilder(format);
    List<Object> valueList = new ArrayList<Object>();

    Matcher matcher = Pattern.compile("\\$\\{(\\w+)}").matcher(format);

    while (matcher.find()) {
        String key = matcher.group(1);

        String formatKey = String.format("${%s}", key);
        int index = formatter.indexOf(formatKey);

        if (index != -1) {
            formatter.replace(index, index + formatKey.length(), "%s");
            valueList.add(values.get(key));
        }
    }

    return String.format(formatter.toString(), valueList.toArray());
}

例子:

String format = "My name is ${1}. ${0} ${1}.";

Map<String, Object> values = new HashMap<String, Object>();
values.put("0", "James");
values.put("1", "Bond");

System.out.println(format(format, values)); // My name is Bond. James Bond.

我的答案是:

a)尽可能使用StringBuilder

b)保持“占位符”的位置(以任何形式:整数是最好的，特殊字符如dollar宏等)，然后使用StringBuilder.insert()(参数的几个版本)。

当StringBuilder内部转换为String时，使用外部库似乎有些过度，而且我认为会显著降低性能。

这是一个旧的线程，但只是为了记录，你也可以使用Java 8风格，像这样:

public static String replaceParams(Map<String, String> hashMap, String template) {
    return hashMap.entrySet().stream().reduce(template, (s, e) -> s.replace("%(" + e.getKey() + ")", e.getValue()),
            (s, s2) -> s);
}

用法:

public static void main(String[] args) {
    final HashMap<String, String> hashMap = new HashMap<String, String>() {
        {
            put("foo", "foo1");
            put("bar", "bar1");
            put("car", "BMW");
            put("truck", "MAN");
        }
    };
    String res = replaceParams(hashMap, "This is '%(foo)' and '%(foo)', but also '%(bar)' '%(bar)' indeed.");
    System.out.println(res);
    System.out.println(replaceParams(hashMap, "This is '%(car)' and '%(foo)', but also '%(bar)' '%(bar)' indeed."));
    System.out.println(replaceParams(hashMap, "This is '%(car)' and '%(truck)', but also '%(foo)' '%(bar)' + '%(truck)' indeed."));
}

输出将是:

This is 'foo1' and 'foo1', but also 'bar1' 'bar1' indeed.
This is 'BMW' and 'foo1', but also 'bar1' 'bar1' indeed.
This is 'BMW' and 'MAN', but also 'foo1' 'bar1' + 'MAN' indeed.

你可以使用StringTemplate库，它提供了你想要的和更多。

import org.antlr.stringtemplate.*;

final StringTemplate hello = new StringTemplate("Hello, $name$");
hello.setAttribute("name", "World");
System.out.println(hello.toString());

可以使用Apache Commons StringSubstitutor。注意，StrSubstitutor已弃用。

import org.apache.commons.text.StringSubstitutor;
// ...
Map<String, String> values = new HashMap<>();
values.put("animal", "quick brown fox");
values.put("target", "lazy dog");
StringSubstitutor sub = new StringSubstitutor(values);
String result = sub.replace("The ${animal} jumped over the ${target}.");
// "The quick brown fox jumped over the lazy dog."

这个类支持为变量提供默认值。

String result = sub.replace("The number is ${undefined.property:-42}.");
// "The number is 42."

要使用递归变量替换，调用setEnableSubstitutionInVariables(true);。

Map<String, String> values = new HashMap<>();
values.put("b", "c");
values.put("ac", "Test");
StringSubstitutor sub = new StringSubstitutor(values);
sub.setEnableSubstitutionInVariables(true);
String result = sub.replace("${a${b}}");
// "Test"