前提
前面写过一篇关于Environment属性加载的源码分析和扩展，里面提到属性的占位符解析和类型转换是相对复杂的，这篇文章就是要分析和解读这两个复杂的问题。关于这两个问题，选用一个比较复杂的参数处理方法PropertySourcesPropertyResolver#getProperty，解析占位符的时候依赖到PropertySourcesPropertyResolver#getPropertyAsRawString：
protected String getPropertyAsRawString(String key) {
return getProperty(key, String.class, false);
}
protected <T> T getProperty(String key, Class<T> targetValueType, boolean resolveNestedPlaceholders) {
if (this.propertySources != null) {
for (PropertySource<?> propertySource : this.propertySources) {
if (logger.isTraceEnabled()) {
logger.trace("Searching for key '" + key + "' in PropertySource '" +
propertySource.getName() + "'");
}
Object value = propertySource.getProperty(key);
if (value != null) {
if (resolveNestedPlaceholders && value instanceof String) {
//解析带有占位符的属性
value = resolveNestedPlaceholders((String) value);
}
logKeyFound(key, propertySource, value);
//需要时转换属性的类型
return convertValueIfNecessary(value, targetValueType);
}
}
}
if (logger.isDebugEnabled()) {
logger.debug("Could not find key '" + key + "' in any property source");
}
return null;
}
属性占位符解析
属性占位符的解析方法是PropertySourcesPropertyResolver的父类AbstractPropertyResolver#resolveNestedPlaceholders：
protected String resolveNestedPlaceholders(String value) {
return (this.ignoreUnresolvableNestedPlaceholders ?
resolvePlaceholders(value) : resolveRequiredPlaceholders(value));
}
ignoreUnresolvableNestedPlaceholders属性默认为false，可以通过AbstractEnvironment#setIgnoreUnresolvableNestedPlaceholders(boolean ignoreUnresolvableNestedPlaceholders)设置，当此属性被设置为true，解析属性占位符失败的时候(并且没有为占位符配置默认值)不会抛出异常，返回属性原样字符串，否则会抛出IllegalArgumentException。我们这里只需要分析AbstractPropertyResolver#resolveRequiredPlaceholders：
//AbstractPropertyResolver中的属性：
//ignoreUnresolvableNestedPlaceholders=true情况下创建的PropertyPlaceholderHelper实例
@Nullable
private PropertyPlaceholderHelper nonStrictHelper;
//ignoreUnresolvableNestedPlaceholders=false情况下创建的PropertyPlaceholderHelper实例
@Nullable
private PropertyPlaceholderHelper strictHelper;
//是否忽略无法处理的属性占位符，这里是false，也就是遇到无法处理的属性占位符且没有默认值则抛出异常
private boolean ignoreUnresolvableNestedPlaceholders = false;
//属性占位符前缀，这里是"${"
private String placeholderPrefix = SystemPropertyUtils.PLACEHOLDER_PREFIX;
//属性占位符后缀，这里是"}"
private String placeholderSuffix = SystemPropertyUtils.PLACEHOLDER_SUFFIX;
//属性占位符解析失败的时候配置默认值的分隔符，这里是":"
@Nullable
private String valueSeparator = SystemPropertyUtils.VALUE_SEPARATOR;
public String resolveRequiredPlaceholders(String text) throws IllegalArgumentException {
if (this.strictHelper == null) {
this.strictHelper = createPlaceholderHelper(false);
}
return doResolvePlaceholders(text, this.strictHelper);
}
//创建一个新的PropertyPlaceholderHelper实例，这里ignoreUnresolvablePlaceholders为false
private PropertyPlaceholderHelper createPlaceholderHelper(boolean ignoreUnresolvablePlaceholders) {
return new PropertyPlaceholderHelper(this.placeholderPrefix, this.placeholderSuffix, this.valueSeparator, ignoreUnresolvablePlaceholders);
}
//这里最终的解析工作委托到PropertyPlaceholderHelper#replacePlaceholders完成
private String doResolvePlaceholders(String text, PropertyPlaceholderHelper helper) {
return helper.replacePlaceholders(text, this::getPropertyAsRawString);
}
最终只需要分析PropertyPlaceholderHelper#replacePlaceholders，这里需要重点注意：
- 注意到这里的第一个参数text就是属性值的源字符串，例如我们需要处理的属性为myProperties: ${server.port}-${spring.application.name}，这里的text就是${server.port}-${spring.application.name}。
- replacePlaceholders方法的第二个参数placeholderResolver，这里比较巧妙，这里的方法引用this::getPropertyAsRawString相当于下面的代码：
//PlaceholderResolver是一个函数式接口
@FunctionalInterface
public interface PlaceholderResolver {
@Nullable
String resolvePlaceholder(String placeholderName);
}
//this::getPropertyAsRawString相当于下面的代码
return new PlaceholderResolver(){
@Override
String resolvePlaceholder(String placeholderName){
//这里调用到的是PropertySourcesPropertyResolver#getPropertyAsRawString，有点绕
return getPropertyAsRawString(placeholderName);
}
}
接着看PropertyPlaceholderHelper#replacePlaceholders的源码：
//基础属性
//占位符前缀，默认是"${"
private final String placeholderPrefix;
//占位符后缀，默认是"}"
private final String placeholderSuffix;
//简单的占位符前缀，默认是"{"，主要用于处理嵌套的占位符如${xxxxx.{yyyyy}}
private final String simplePrefix;
//默认值分隔符号，默认是":"
@Nullable
private final String valueSeparator;
//替换属性占位符
public String replacePlaceholders(String value, PlaceholderResolver placeholderResolver) {
Assert.notNull(value, "'value' must not be null");
return parseStringValue(value, placeholderResolver, new HashSet<>());
}
//递归解析带占位符的属性为字符串
protected String parseStringValue(
String value, PlaceholderResolver placeholderResolver, Set<String> visitedPlaceholders) {
StringBuilder result = new StringBuilder(value);
int startIndex = value.indexOf(this.placeholderPrefix);
while (startIndex != -1) {
//搜索第一个占位符后缀的索引
int endIndex = findPlaceholderEndIndex(result, startIndex);
if (endIndex != -1) {
//提取第一个占位符中的原始字符串，如${server.port}->server.port
String placeholder = result.substring(startIndex + this.placeholderPrefix.length(), endIndex);
String originalPlaceholder = placeholder;
//判重
if (!visitedPlaceholders.add(originalPlaceholder)) {
throw new IllegalArgumentException(
"Circular placeholder reference '" + originalPlaceholder + "' in property definitions");
}
// Recursive invocation, parsing placeholders contained in the placeholder key.
// 递归调用，实际上就是解析嵌套的占位符，因为提取的原始字符串有可能还有一层或者多层占位符
placeholder = parseStringValue(placeholder, placeholderResolver, visitedPlaceholders);
// Now obtain the value for the fully resolved key...
// 递归调用完毕后，可以确定得到的字符串一定是不带占位符，这个时候调用getPropertyAsRawString获取key对应的字符串值
String propVal = placeholderResolver.resolvePlaceholder(placeholder);
// 如果字符串值为null，则进行默认值的解析，因为默认值有可能也使用了占位符，如${server.port:${server.port-2:8080}}
if (propVal == null && this.valueSeparator != null) {
int separatorIndex = placeholder.indexOf(this.valueSeparator);
if (separatorIndex != -1) {
String actualPlaceholder = placeholder.substring(0, separatorIndex);
// 提取默认值的字符串
String defaultValue = placeholder.substring(separatorIndex + this.valueSeparator.length());
// 这里是把默认值的表达式做一次解析，解析到null，则直接赋值为defaultValue
propVal = placeholderResolver.resolvePlaceholder(actualPlaceholder);
if (propVal == null) {
propVal = defaultValue;
}
}
}
// 上一步解析出来的值不为null，但是它有可能是一个带占位符的值，所以后面对值进行递归解析
if (propVal != null) {
// Recursive invocation, parsing placeholders contained in the
// previously resolved placeholder value.
propVal = parseStringValue(propVal, placeholderResolver, visitedPlaceholders);
// 这一步很重要，替换掉第一个被解析完毕的占位符属性，例如${server.port}-${spring.application.name} -> 9090--${spring.application.name}
result.replace(startIndex, endIndex + this.placeholderSuffix.length(), propVal);
if (logger.isTraceEnabled()) {
logger.trace("Resolved placeholder '" + placeholder + "'");
}
// 重置startIndex为下一个需要解析的占位符前缀的索引，可能为-1，说明解析结束
startIndex = result.indexOf(this.placeholderPrefix, startIndex + propVal.length());
}
else if (this.ignoreUnresolvablePlaceholders) {
// 如果propVal为null并且ignoreUnresolvablePlaceholders设置为true，直接返回当前的占位符之间的原始字符串尾的索引，也就是跳过解析
// Proceed with unprocessed value.
startIndex = result.indexOf(this.placeholderPrefix, endIndex + this.placeholderSuffix.length());
}
else {
// 如果propVal为null并且ignoreUnresolvablePlaceholders设置为false，抛出异常
throw new IllegalArgumentException("Could not resolve placeholder '" +
placeholder + "'" + " in value \"" + value + "\"");
}
// 递归结束移除判重集合中的元素
visitedPlaceholders.remove(originalPlaceholder);
}
else {
// endIndex = -1说明解析结束
startIndex = -1;
}
}
return result.toString();
}
//基于传入的起始索引，搜索第一个占位符后缀的索引，兼容嵌套的占位符
private int findPlaceholderEndIndex(CharSequence buf, int startIndex) {
//这里index实际上就是实际需要解析的属性的第一个字符，如${server.port}，这里index指向s
int index = startIndex + this.placeholderPrefix.length();
int withinNestedPlaceholder = 0;
while (index < buf.length()) {
//index指向"}"，说明有可能到达占位符尾部或者嵌套占位符尾部
if (StringUtils.substringMatch(buf, index, this.placeholderSuffix)) {
//存在嵌套占位符，则返回字符串中占位符后缀的索引值
if (withinNestedPlaceholder > 0) {
withinNestedPlaceholder--;
index = index + this.placeholderSuffix.length();
}
else {
//不存在嵌套占位符，直接返回占位符尾部索引
return index;
}
}
//index指向"{"，记录嵌套占位符个数withinNestedPlaceholder加1，index更新为嵌套属性的第一个字符的索引
else if (StringUtils.substringMatch(buf, index, this.simplePrefix)) {
withinNestedPlaceholder++;
index = index + this.simplePrefix.length();
}
else {
//index不是"{"或者"}"，则进行自增
index++;
}
}
//这里说明解析索引已经超出了原字符串
return -1;
}
//StringUtils#substringMatch，此方法会检查原始字符串str的index位置开始是否和子字符串substring完全匹配
public static boolean substringMatch(CharSequence str, int index, CharSequence substring) {
if (index + substring.length() > str.length()) {
return false;
}
for (int i = 0; i < substring.length(); i++) {
if (str.charAt(index + i) != substring.charAt(i)) {
return false;
}
}
return true;
}
上面的过程相对比较复杂，因为用到了递归，我们举个实际的例子说明一下整个解析过程，例如我们使用了四个属性项，我们的目标是获取server.desc的值：
application.name=spring
server.port=9090
spring.application.name=${application.name}
server.desc=${server.port-${spring.application.name}}:${description:"hello"}
属性类型转换
在上一步解析属性占位符完毕之后，得到的是属性字符串值，可以把字符串转换为指定的类型，此功能由AbstractPropertyResolver#convertValueIfNecessary完成：
protected <T> T convertValueIfNecessary(Object value, @Nullable Class<T> targetType) {
if (targetType == null) {
return (T) value;
}
ConversionService conversionServiceToUse = this.conversionService;
if (conversionServiceToUse == null) {
// Avoid initialization of shared DefaultConversionService if
// no standard type conversion is needed in the first place...
// 这里一般只有字符串类型才会命中
if (ClassUtils.isAssignableValue(targetType, value)) {
return (T) value;
}
conversionServiceToUse = DefaultConversionService.getSharedInstance();
}
return conversionServiceToUse.convert(value, targetType);
}
实际上转换的逻辑是委托到DefaultConversionService的父类方法GenericConversionService#convert：
public <T> T convert(@Nullable Object source, Class<T> targetType) {
Assert.notNull(targetType, "Target type to convert to cannot be null");
return (T) convert(source, TypeDescriptor.forObject(source), TypeDescriptor.valueOf(targetType));
}
public Object convert(@Nullable Object source, @Nullable TypeDescriptor sourceType, TypeDescriptor targetType) {
Assert.notNull(targetType, "Target type to convert to cannot be null");
if (sourceType == null) {
Assert.isTrue(source == null, "Source must be [null] if source type == [null]");
return handleResult(null, targetType, convertNullSource(null, targetType));
}
if (source != null && !sourceType.getObjectType().isInstance(source)) {
throw new IllegalArgumentException("Source to convert from must be an instance of [" +
sourceType + "]; instead it was a [" + source.getClass().getName() + "]");
}
// 从缓存中获取GenericConverter实例，其实这一步相对复杂，匹配两个类型的时候，会解析整个类的层次进行对比
GenericConverter converter = getConverter(sourceType, targetType);
if (converter != null) {
// 实际上就是调用转换方法
Object result = ConversionUtils.invokeConverter(converter, source, sourceType, targetType);
// 断言最终结果和指定类型是否匹配并且返回
return handleResult(sourceType, targetType, result);
}
return handleConverterNotFound(source, sourceType, targetType);
}
上面所有的可用的GenericConverter的实例可以在DefaultConversionService的addDefaultConverters中看到，默认添加的转换器实例已经超过20个，有些情况下如果无法满足需求可以添加自定义的转换器，实现GenericConverter接口添加进去即可。
小结
SpringBoot在抽象整个类型转换器方面做的比较好，在SpringMVC应用中，采用的是org.springframework.boot.autoconfigure.web.format.WebConversionService，兼容了Converter、Formatter、ConversionService等转换器类型并且对外提供一套统一的转换方法。
（本文完）