springboot启动流程(十)springboot自动配置机制

所有文章

https://www.cnblogs.com/lay2017/p/11478237.html

 

正文

第七篇文章中我们了解到,refresh过程将会调用ConfigurationClassPostProcessor这个后置处理器,而这个后置处理器将会去调用ConfigurationClassParser这个配置类的解析器,而第一个被解析的配置类就是我们main方法所在的主类(主类是在refresh之前的,prepareRefresh方法加载成为BeanDefinition到BeanFactory中的)。

而后,在第八篇文章中我们主要看了看ConfigurationClassParser是怎么解析配置类的@ComponentScan这个注解的。

那么本文将继续从ConfigurationClassParser这个过程开始,看看parse的处理过程关于自动配置的内容。

 

自动配置入口

首先,我们回到ConfigurationClassParser的parse方法。

public void parse(Set<BeanDefinitionHolder> configCandidates) {
    for (BeanDefinitionHolder holder : configCandidates) {
        BeanDefinition bd = holder.getBeanDefinition();
        try {
            if (bd instanceof AnnotatedBeanDefinition) {
                // 解析主类的入口
                parse(((AnnotatedBeanDefinition) bd).getMetadata(), holder.getBeanName());
            } else if (bd instanceof AbstractBeanDefinition && ((AbstractBeanDefinition) bd).hasBeanClass()) {
                parse(((AbstractBeanDefinition) bd).getBeanClass(), holder.getBeanName());
            } else {
                parse(bd.getBeanClassName(), holder.getBeanName());
            }
        } catch (BeanDefinitionStoreException ex) {
            throw ex;
        } catch (Throwable ex) {
            throw new BeanDefinitionStoreException(
                    "Failed to parse configuration class [" + bd.getBeanClassName() + "]", ex);
        }
    }
    // 处理自动配置的入口
    this.deferredImportSelectorHandler.process();
}

springboot解析过程将从解析main方法所在的主类开始,所以我们先跟进parse方法

protected final void parse(AnnotationMetadata metadata, String beanName) throws IOException {
    processConfigurationClass(new ConfigurationClass(metadata, beanName));
}

再跟进processConfigurationClass方法

protected void processConfigurationClass(ConfigurationClass configClass) throws IOException {
    // 判断当前配置类是否应该跳过解析
    if (this.conditionEvaluator.shouldSkip(configClass.getMetadata(), ConfigurationPhase.PARSE_CONFIGURATION)) {
        return;
    }

    // 省略

    // 向父类递归解析
    SourceClass sourceClass = asSourceClass(configClass);
    do {
        // 解析当前配置类的核心逻辑
        sourceClass = doProcessConfigurationClass(configClass, sourceClass);
    }
    while (sourceClass != null);

    //
}

这里先做了一个判断,是否跳过当前配置类的解析(后面会提及)。而后就是对配置类的递归解析,如果有父类将会递归解析。

 

跟进doProcessConfigurationClass方法,我们省略其它内容

protected final SourceClass doProcessConfigurationClass(ConfigurationClass configClass, SourceClass sourceClass)
        throws IOException {

    // 

    // 处理@Import注解
    processImports(configClass, sourceClass, getImports(sourceClass), true);

    //
    return null;
}

我们看到,解析逻辑中包含着一个processImports方法,用于处理@Import注解。我们知道,每一个springboot程序将会注解一个@SpringBootApplication注解,这个注解是一个组合注解,我们看看该注解。

@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Inherited
@SpringBootConfiguration
@EnableAutoConfiguration
@ComponentScan(excludeFilters = { @Filter(type = FilterType.CUSTOM, classes = TypeExcludeFilter.class),
        @Filter(type = FilterType.CUSTOM, classes = AutoConfigurationExcludeFilter.class) })
public @interface SpringBootApplication {
    // 省略
}

@SpringBootApplication注解组合了@EnableAutoConfiguration注解,我们再看看@EnableAutoConfiguration注解

@Target(ElementType.TYPE)
@Retention(RetentionPolicy.RUNTIME)
@Documented
@Inherited
@AutoConfigurationPackage
@Import(AutoConfigurationImportSelector.class)
public @interface EnableAutoConfiguration {
    // 省略
}

可以看到,@Import注解导入了一个AutoConfigurationImportSelector类。

 

我们再回到doProcessConfigurationClass方法

protected final SourceClass doProcessConfigurationClass(ConfigurationClass configClass, SourceClass sourceClass)
        throws IOException {

    // 

    // 处理@Import注解
    processImports(configClass, sourceClass, getImports(sourceClass), true);

    //
    return null;
}

processImports之前,会调用getImports方法获取当前配置类的@Import,也包含组合的注解。所以@SpringBootApplication组合的@Import注解导入的配置类AutoConfigurationImportSelector将在这里被获取。

 

我们跟进getImport方法看看

private Set<SourceClass> getImports(SourceClass sourceClass) throws IOException {
    Set<SourceClass> imports = new LinkedHashSet<>();
    Set<SourceClass> visited = new LinkedHashSet<>();
    collectImports(sourceClass, imports, visited);
    return imports;
}

再跟进collectImports看看是怎么搜集类的

private void collectImports(SourceClass sourceClass, Set<SourceClass> imports, Set<SourceClass> visited)
        throws IOException {

    if (visited.add(sourceClass)) {
        // 获取所有注解
        for (SourceClass annotation : sourceClass.getAnnotations()) {
            String annName = annotation.getMetadata().getClassName();
            // 非@Import注解的,递归看看有没有组合@Import注解
            if (!annName.equals(Import.class.getName())) {
                collectImports(annotation, imports, visited);
            }
        }
        imports.addAll(sourceClass.getAnnotationAttributes(Import.class.getName(), "value"));
    }
}

很显然,搜集过程将对所有注解递归处理,这样我们就获得了main方法所在主类的所有@Import导入的类。

 

再回到doProcessConfigurationClass方法

protected final SourceClass doProcessConfigurationClass(ConfigurationClass configClass, SourceClass sourceClass)
        throws IOException {

    // 

    // 处理@Import注解
    processImports(configClass, sourceClass, getImports(sourceClass), true);

    //
    return null;
}

跟进processImports方法

private void processImports(ConfigurationClass configClass, SourceClass currentSourceClass,
        Collection<SourceClass> importCandidates, boolean checkForCircularImports) {
    //

    if (checkForCircularImports && isChainedImportOnStack(configClass)) {
        //
    }
    else {
        //
        try {
            for (SourceClass candidate : importCandidates) {
                if (candidate.isAssignable(ImportSelector.class)) {
                    Class<?> candidateClass = candidate.loadClass();
                    ImportSelector selector = BeanUtils.instantiateClass(candidateClass, ImportSelector.class);
                    //
                    if (selector instanceof DeferredImportSelector) {
                        this.deferredImportSelectorHandler.handle(configClass, (DeferredImportSelector) selector);
                    }
                    else {
                        //
                    }
                }
                else if (candidate.isAssignable(ImportBeanDefinitionRegistrar.class)) {
                    //
                }
                else {
                    // 
                    processConfigurationClass(candidate.asConfigClass(configClass));
                }
            }
        }
        // 
    }
}

这里遍历了我们getImports方法获取到的类,但是目前我们还只有AutoConfiguratonImportSelector这个类。这个类实现了DeferredImportSelector接口,所以我们继续跟进handle方法

public void handle(ConfigurationClass configClass, DeferredImportSelector importSelector) {
    DeferredImportSelectorHolder holder = new DeferredImportSelectorHolder(configClass, importSelector);
    if (this.deferredImportSelectors == null) {
        //
    } else {
        // 添加到集合
        this.deferredImportSelectors.add(holder);
    }
}

AutoConfigurationImportSelector被包装已经添加到了集合中。那么这个被添加到集合中的类是什么时候被处理的呢?

我们回到本文最开始的代码片段,parse方法

public void parse(Set<BeanDefinitionHolder> configCandidates) {
    for (BeanDefinitionHolder holder : configCandidates) {
        BeanDefinition bd = holder.getBeanDefinition();
        try {
            if (bd instanceof AnnotatedBeanDefinition) {
                // 解析主类的入口
                parse(((AnnotatedBeanDefinition) bd).getMetadata(), holder.getBeanName());
            } else if (bd instanceof AbstractBeanDefinition && ((AbstractBeanDefinition) bd).hasBeanClass()) {
                parse(((AbstractBeanDefinition) bd).getBeanClass(), holder.getBeanName());
            } else {
                parse(bd.getBeanClassName(), holder.getBeanName());
            }
        } catch (BeanDefinitionStoreException ex) {
            throw ex;
        } catch (Throwable ex) {
            throw new BeanDefinitionStoreException(
                    "Failed to parse configuration class [" + bd.getBeanClassName() + "]", ex);
        }
    }
    // 处理自动配置的入口
    this.deferredImportSelectorHandler.process();
}

可以看到,parse方法的最后一行,调用了process方法,将会对这些类进行处理,这也就是自动配置的入口方法了。

 

处理AutoConfigurationImportSelector

我们跟进process方法,看看处理过程

public void process() {
    List<DeferredImportSelectorHolder> deferredImports = this.deferredImportSelectors;
    this.deferredImportSelectors = null;
    try {
        if (deferredImports != null) {
            DeferredImportSelectorGroupingHandler handler = new DeferredImportSelectorGroupingHandler();
            deferredImports.sort(DEFERRED_IMPORT_COMPARATOR);
            // 遍历导入的类,将这些类注册到handler中
            deferredImports.forEach(handler::register);
            // 处理handler中的导入类
            handler.processGroupImports();
        }
    } finally {
        //
    }
}

这里先将导入类调用handler的register方法进行注册,然后集中处理。我们稍微瞄一眼register方法

public void register(DeferredImportSelectorHolder deferredImport) {
    Class<? extends Group> group = deferredImport.getImportSelector().getImportGroup();
    DeferredImportSelectorGrouping grouping = this.groupings.computeIfAbsent((group != null ? group : deferredImport), key -> new DeferredImportSelectorGrouping(createGroup(group)));
    // 添加到group中
    grouping.add(deferredImport);
    this.configurationClasses.put(deferredImport.getConfigurationClass().getMetadata(), deferredImport.getConfigurationClass());
}

跟进add方法

private final List<DeferredImportSelectorHolder> deferredImports = new ArrayList<>();

public void add(DeferredImportSelectorHolder deferredImport) {
    this.deferredImports.add(deferredImport);
}

添加到一个集合中存放起来

 

我们回到process方法

public void process() {
    List<DeferredImportSelectorHolder> deferredImports = this.deferredImportSelectors;
    this.deferredImportSelectors = null;
    try {
        if (deferredImports != null) {
            DeferredImportSelectorGroupingHandler handler = new DeferredImportSelectorGroupingHandler();
            deferredImports.sort(DEFERRED_IMPORT_COMPARATOR);
            // 遍历导入的类,将这些类注册到handler中
            deferredImports.forEach(handler::register);
            // 处理handler中的导入类
            handler.processGroupImports();
        }
    } finally {
        //
    }
}

register完毕以后,将会processGroupImports,我们跟进processGroupImports方法

public void processGroupImports() {
    for (DeferredImportSelectorGrouping grouping : this.groupings.values()) {
        // 获取所有AutoConfigurationImportSelector返回的待处理的配置类,并遍历
        grouping.getImports().forEach(entry -> {
            ConfigurationClass configurationClass = this.configurationClasses.get(
                    entry.getMetadata());
            try {
                // 处理所有配置类
                processImports(configurationClass, asSourceClass(configurationClass),
                        asSourceClasses(entry.getImportClassName()), false);
            }
            catch (BeanDefinitionStoreException ex) {
                throw ex;
            }
            catch (Throwable ex) {
                throw new BeanDefinitionStoreException(
                        "Failed to process import candidates for configuration class [" +
                                configurationClass.getMetadata().getClassName() + "]", ex);
            }
        });
    }
}

AutoConfigurationImportSelector需要进行自动配置的类将会在这里的getImports方法中返回,而后processImports方法将会处理所有这些需要自动配置的类

 

我们先跟进getImports方法,看看是怎么获取所有待处理的配置类的

public Iterable<Group.Entry> getImports() {
    for (DeferredImportSelectorHolder deferredImport : this.deferredImports) {
        // 处理生成结果
        this.group.process(deferredImport.getConfigurationClass().getMetadata(), deferredImport.getImportSelector());
    }
    // 返回结果
    return this.group.selectImports();
}

我们主要看process方法,进入到AutoConfigurationGroup的process方法

public void process(AnnotationMetadata annotationMetadata, DeferredImportSelector deferredImportSelector) {
    AutoConfigurationEntry autoConfigurationEntry = ((AutoConfigurationImportSelector) deferredImportSelector).getAutoConfigurationEntry(getAutoConfigurationMetadata(), annotationMetadata);
    this.autoConfigurationEntries.add(autoConfigurationEntry);
    for (String importClassName : autoConfigurationEntry.getConfigurations()) {
        this.entries.putIfAbsent(importClassName, annotationMetadata);
    }
}

再跟进getAutoConfigurationEntry

protected AutoConfigurationEntry getAutoConfigurationEntry(AutoConfigurationMetadata autoConfigurationMetadata,
        AnnotationMetadata annotationMetadata) {
    //
    AnnotationAttributes attributes = getAttributes(annotationMetadata);
    List<String> configurations = getCandidateConfigurations(annotationMetadata, attributes);
    //

    // 过滤
    configurations = filter(configurations, autoConfigurationMetadata);
    //
    return new AutoConfigurationEntry(configurations, exclusions);
}

继续跟进getCandidateConfigurations

protected List<String> getCandidateConfigurations(AnnotationMetadata metadata, AnnotationAttributes attributes) {
    List<String> configurations = SpringFactoriesLoader.loadFactoryNames(getSpringFactoriesLoaderFactoryClass(),
            getBeanClassLoader());
    //
    return configurations;
}

我们看到一个熟悉的方法loadFactoryNames(不熟悉的话,请阅读辅助内容),看看getSpringFactoriesLoaderFactoryClass返回什么

protected Class<?> getSpringFactoriesLoaderFactoryClass() {
    return EnableAutoConfiguration.class;
}

我们可以随机打开一个spring.factories看看EnableAutoConfiguration作为key的配置

可以看到spring.factories中将需要进行自动配置的类作为value配置在这里,所以getCandidateConfigurations方法将会把这些配置类返回。我们再回到getAutoConfigurationEntry方法

protected AutoConfigurationEntry getAutoConfigurationEntry(AutoConfigurationMetadata autoConfigurationMetadata,
        AnnotationMetadata annotationMetadata) {
    //
    AnnotationAttributes attributes = getAttributes(annotationMetadata);
    List<String> configurations = getCandidateConfigurations(annotationMetadata, attributes);
    //

    // 过滤
    configurations = filter(configurations, autoConfigurationMetadata);
    //
    return new AutoConfigurationEntry(configurations, exclusions);
}

获取完configurations后,将会进行一次过滤操作,这样可以避免大量的不需要配置的类被加载。

再回到AutoConfiguration的process方法

private final Map<String, AnnotationMetadata> entries = new LinkedHashMap<>();

public void process(AnnotationMetadata annotationMetadata, DeferredImportSelector deferredImportSelector) {
    AutoConfigurationEntry autoConfigurationEntry = ((AutoConfigurationImportSelector) deferredImportSelector).getAutoConfigurationEntry(getAutoConfigurationMetadata(), annotationMetadata);
    this.autoConfigurationEntries.add(autoConfigurationEntry);
    for (String importClassName : autoConfigurationEntry.getConfigurations()) {
        this.entries.putIfAbsent(importClassName, annotationMetadata);
    }
}

我们调用getAutoConfigurationEntry获得需要自动配置的类,然后再这里会被添加到一个Map集合中存放起来。

到这里,AutoConfigurationImportSelector的getImports方法的process过程就结束了。我们回到getImports方法

public Iterable<Group.Entry> getImports() {
    for (DeferredImportSelectorHolder deferredImport : this.deferredImports) {
        // 处理生成结果
        this.group.process(deferredImport.getConfigurationClass().getMetadata(), deferredImport.getImportSelector());
    }
    // 返回结果
    return this.group.selectImports();
}

process方法获取了Imports,而selectImports将返回结果。

 

到这里,我们的getImports方法就获取了可能需要进行自动配置的类,回到DeferredImportSelectorGroupingHandler类的processGroupImports方法

public void processGroupImports() {
    for (DeferredImportSelectorGrouping grouping : this.groupings.values()) {
        // 获取所有AutoConfigurationImportSelector返回的待处理的配置类,并遍历
        grouping.getImports().forEach(entry -> {
            ConfigurationClass configurationClass = this.configurationClasses.get(
                    entry.getMetadata());
            try {
                // 处理所有配置类
                processImports(configurationClass, asSourceClass(configurationClass),
                        asSourceClasses(entry.getImportClassName()), false);
            }
            catch (BeanDefinitionStoreException ex) {
                throw ex;
            }
            catch (Throwable ex) {
                throw new BeanDefinitionStoreException(
                        "Failed to process import candidates for configuration class [" +
                                configurationClass.getMetadata().getClassName() + "]", ex);
            }
        });
    }
}

跟进processImports

private void processImports(ConfigurationClass configClass, SourceClass currentSourceClass,
        Collection<SourceClass> importCandidates, boolean checkForCircularImports) {
    //
    if (checkForCircularImports && isChainedImportOnStack(configClass)) {
        //
    }
    else {
        //
        try {
            for (SourceClass candidate : importCandidates) {
                if (candidate.isAssignable(ImportSelector.class)) {
                    //
                }
                else if (candidate.isAssignable(ImportBeanDefinitionRegistrar.class)) {
                    //
                }
                else {
                    // 处理配置类
                    processConfigurationClass(candidate.asConfigClass(configClass));
                }
            }
        }
        //
    }
}

导入的类作为配置类来处理,跟进processConfigurationClass方法

protected void processConfigurationClass(ConfigurationClass configClass) throws IOException {
    // 是否要进行配置解析
    if (this.conditionEvaluator.shouldSkip(configClass.getMetadata(), ConfigurationPhase.PARSE_CONFIGURATION)) {
        return;
    }

    //

    SourceClass sourceClass = asSourceClass(configClass);
    do {
        // 解析的逻辑
        sourceClass = doProcessConfigurationClass(configClass, sourceClass);
    }
    while (sourceClass != null);
    // 
}

我们久违的processConfigurationClass方法,一开始我们关注的是doProcessConfigurationClass看它解析过程的。现在我们来看看shouldSkip方法,看看是怎么判断当前配置类是否要进行解析的。

 

跟进shouldSkip方法

public boolean shouldSkip(@Nullable AnnotatedTypeMetadata metadata, @Nullable ConfigurationPhase phase) {
    //

    List<Condition> conditions = new ArrayList<>();
    // 获取配置类中所有@Conditional以及组合@Conditional的条件
    for (String[] conditionClasses : getConditionClasses(metadata)) {
        for (String conditionClass : conditionClasses) {
            Condition condition = getCondition(conditionClass, this.context.getClassLoader());
            conditions.add(condition);
        }
    }

    //

    // 遍历这些条件
    for (Condition condition : conditions) {
        ConfigurationPhase requiredPhase = null;
        if (condition instanceof ConfigurationCondition) {
            requiredPhase = ((ConfigurationCondition) condition).getConfigurationPhase();
        }
        // 判断条件是否不匹配
        if ((requiredPhase == null || requiredPhase == phase) && !condition.matches(this.context, metadata)) {
            return true;
        }
    }

    return false;
}

shouldSkip方法,将会拿到当前配置类的所有@Conditional或者组合了@Conditional的注解,并将注解生成Condition条件,再遍历这些条件看是否有不满足条件的将返回true。

 

总结

springboot的自动配置将从解析main方法所在的主类开始,ConfigurationClassParser在解析@Import的时候会获取到AutoConfigurationImportSelector类。AutoConfigurationImportSelector将会获取到所有可能需要进行自动配置的类,而后每个配置类将被和main方法所在的主类一样准备解析,在解析之前会根据像@Conditional或者组合@Conditional的注解来生成判断条件Condition,根据是否满足Condition来决定是否要进行自动配置。

 

posted @ 2019-09-14 00:56  __lay  阅读(2941)  评论(0编辑  收藏  举报