SpringBoot之OriginTrackedPropertiesLoader类源码学习

源码解析

/*** 作用是从给定的资源（如文件或输入流）中加载 .properties 文件，* 并将属性键值对转换为带有来源信息（origin）的 OriginTrackedValue 对象。*/
public class OriginTrackedPropertiesLoader {private final Resource resource;/*** Create a new {@link OriginTrackedPropertiesLoader} instance.* @param resource the resource of the {@code .properties} data*/public OriginTrackedPropertiesLoader(Resource resource) {Assert.notNull(resource, "Resource must not be null");this.resource = resource;}/*** Load {@code .properties} data and return a map of {@code String} ->* {@link OriginTrackedValue}.* @return the loaded properties* @throws IOException on read error*/public Map<String, OriginTrackedValue> load() throws IOException {return load(true);}/*** 加载配置数据到一个Map中，根据指定的资源文件。* 如果expandLists为true，则将列表类型的配置项展开。** @param expandLists 是否展开列表类型的配置项* @return 包含配置数据的Map，键为配置项名称，值为配置项的值和来源信息* @throws IOException 如果读取资源文件时发生错误*/Map<String, OriginTrackedValue> load(boolean expandLists) throws IOException {//创建CharacterReader对象，用于逐行读取资源文件内容。try (CharacterReader reader = new CharacterReader(this.resource)) {//初始化结果MapMap<String, OriginTrackedValue> result = new LinkedHashMap<>();//初始化字符串缓冲区StringBuilder buffer = new StringBuilder();//循环读取逐行资源文件内容，直到文件末尾while (reader.read()) {//读取配置项的键名，并去除前后空格String key = loadKey(buffer, reader).trim();//如果配置项是列表类型且需要展开列表，则处理列表配置项if (expandLists && key.endsWith("[]")) {//去除列表配置项的"[]"后缀key = key.substring(0, key.length() - 2);//初始化列表索引int index = 0;//循环读取列表中的每个配置项do {//读取配置项的值OriginTrackedValue value = loadValue(buffer, reader, true);//将配置项添加到结果Map中，键名格式为key[index]put(result, key + "[" + (index++) + "]", value);//如果当前行不是行尾，则继续读取if (!reader.isEndOfLine()) {reader.read();}} while (!reader.isEndOfLine());}else {//读取非列表配置项的值OriginTrackedValue value = loadValue(buffer, reader, false);//将配置项添加到结果Map中put(result, key, value);}}//返回包含所有配置项的Mapreturn result;}}/*** 从输入中加载属性键* 该方法负责读取输入直到遇到属性分隔符或行尾，同时忽略前导和尾随的空白字符** @param buffer 用于存储读取的键的字符串构建器* @param reader 提供输入字符的字符阅读器* @return 返回读取的属性键的字符串表示* @throws IOException 如果读取过程中发生I/O错误*/private String loadKey(StringBuilder buffer, CharacterReader reader) throws IOException {// 清空缓冲区以准备读取新的键buffer.setLength(0);// 初始化前一个字符是否为空白字符的标志boolean previousWhitespace = false;// 循环读取直到行尾while (!reader.isEndOfLine()) {// 如果遇到属性分隔符，读取下一个字符并返回当前缓冲区的内容if (reader.isPropertyDelimiter()) {reader.read();return buffer.toString();}// 如果当前字符不是空白字符，但前一个字符是空白字符，则返回当前缓冲区的内容if (!reader.isWhiteSpace() && previousWhitespace) {return buffer.toString();}// 更新前一个字符是否为空白字符的标志previousWhitespace = reader.isWhiteSpace();// 将当前字符追加到缓冲区中buffer.append(reader.getCharacter());// 读取下一个字符reader.read();}// 如果到达行尾，返回当前缓冲区的内容return buffer.toString();}/*** 从输入中加载一行值，可以选择性地在列表分隔符处分割** @param buffer 缓存区，用于存储从输入中读取的值* @param reader 字符读取器，用于从输入源读取字符* @param splitLists 指示是否应在遇到列表分隔符时分割的布尔值* @return 返回一个包含值及其来源信息的OriginTrackedValue对象* @throws IOException 如果在读取过程中发生I/O错误*/private OriginTrackedValue loadValue(StringBuilder buffer, CharacterReader reader, boolean splitLists)throws IOException {// 清空缓冲区以准备读取新的值buffer.setLength(0);// 跳过行首的空白字符，直到遇到非空白字符或行尾while (reader.isWhiteSpace() && !reader.isEndOfLine()) {reader.read();}// 记录当前读取位置，用于后续创建Origin对象Location location = reader.getLocation();// 读取字符直到行尾或（如果splitLists为真）遇到列表分隔符while (!reader.isEndOfLine() && !(splitLists && reader.isListDelimiter())) {buffer.append(reader.getCharacter());reader.read();}// 创建一个表示值来源的Origin对象Origin origin = new TextResourceOrigin(this.resource, location);// 使用缓冲区中的字符串和其来源信息创建并返回一个OriginTrackedValue对象return OriginTrackedValue.of(buffer.toString(), origin);}/*** Reads characters from the source resource, taking care of skipping comments,* handling multi-line values and tracking {@code '\'} escapes.* 用于逐字符读取文件内容，处理注释、转义字符、多行值等特殊情况。* 提供了多种辅助方法来跳过空白字符、处理注释、读取转义字符等。*/private static class CharacterReader implements Closeable{//省略......}}

案例

test-properties.properties配置文件

   # foo
blah   =   hello world
bar   foo=baz
hello   world
proper\\ty=test
foo
bat = a\\
bling = a=b#commented-property=test
test=properties
test-unicode=properties\u0026test# comment ending \
test\=property=helloworld
test-colon-separator: my-property
test-tab-property=foo\tbar
test-return-property=foo\rbar
test-newline-property=foo\nbar
test-form-feed-property=foo\fbar
test-whitespace-property   =   foo   bar
test-multiline= a\b\\\c
foods[]=Apple,\
Orange,\
Strawberry,\
Mango
languages[perl]=Elite
languages[python]=Elite
language[pascal]=Lame
test-multiline-immediate=\
foo
!commented-two=bang\
test-bang-property=foo!
another=bar
test-property-value-comment=foo \
!bar #foo
test-multiline-immediate-bang=\
!foo#test ISO 8859-1
test-iso8859-1-chars=����������test-with-trailing-space= trailing 

	private ClassPathResource resource;private Map<String, OriginTrackedValue> properties;@Testvoid compareToJavaProperties() throws Exception {String path = "test-properties.properties";this.resource = new ClassPathResource(path, getClass());this.properties = new OriginTrackedPropertiesLoader(this.resource).load();Properties java = PropertiesLoaderUtils.loadProperties(this.resource);Properties ours = new Properties();new OriginTrackedPropertiesLoader(this.resource).load(false).forEach((k, v) -> System.out.println(k+":"+v.getValue()));}

运行结果

blah:hello world
bar:foo=baz
hello:world
proper\ty:test
foo:
bat:a\
bling:a=b
test:properties
test-unicode:properties&test
test=property:helloworld
test-colon-separator:my-property
test-tab-property:foo	bar
bar
test-newline-property:foo
bar
test-form-feed-property:foobar
test-whitespace-property:foo   bar
test-multiline:ab\c
foods[]:Apple,Orange,Strawberry,Mango
languages[perl]:Elite
languages[python]:Elite
language[pascal]:Lame
test-multiline-immediate:foo
test-bang-property:foo!
another:bar
test-property-value-comment:foo !bar #foo
test-multiline-immediate-bang:!foo
test-iso8859-1-chars:����������
test-with-trailing-space:trailing