Logback 架构与输出流程

Logback 主要由三个 jar 一起组成

• slf4j-api
• logback-core
• logback-classic

<dependencies>
    <dependency>
        <groupId>ch.qos.logback</groupId>
        <artifactId>logback-classic</artifactId>
        <version>1.2.3</version>
    </dependency>
    <dependency>
        <groupId>ch.qos.logback</groupId>
        <artifactId>logback-core</artifactId>
        <version>1.2.3</version>
    </dependency>
    <dependency>
        <groupId>org.slf4j</groupId>
        <artifactId>slf4j-api</artifactId>
        <version>1.7.25</version>
    </dependency>
</dependencies>
复制代码

import org.slf4j.Logger;
import org.slf4j.LoggerFactory;

public class Main {
    public static void main(String[] args) {
        Logger logger = LoggerFactory.getLogger(Main.class);
        logger.debug("hello world");
    }
}
复制代码

当不存在默认配置时、Logback 会在 root logger 中新增一个 ConsoleAppender

Logback 可以打印其内部的状态信息、通过 StatusPrinter 和 StatusManager

private static void test2(){
    LoggerContext loggerContext  = (LoggerContext) LoggerFactory.getILoggerFactory();
    StatusManager statusManager = loggerContext.getStatusManager();
    StatusPrinter.print(statusManager);
}
复制代码

打印如下

21:20:30,270 |-INFO in ch.qos.logback.classic.LoggerContext[default] - Could NOT find resource [logback-test.xml]
21:20:30,270 |-INFO in ch.qos.logback.classic.LoggerContext[default] - Could NOT find resource [logback.groovy]
21:20:30,271 |-INFO in ch.qos.logback.classic.LoggerContext[default] - Could NOT find resource [logback.xml]
21:20:30,274 |-INFO in ch.qos.logback.classic.BasicConfigurator@6d8a00e3 - Setting up default configuration.
复制代码

没有找到三个默认的配置文件

• logback-test.xml
• logback.groovy
• logback.xml

只能使用默认的配置、将日志输出到控制台。

Appender 作为输出的目的地

• console
• file
• db
• MQ

如果发生了错误、Logback 会在控制台打印自己内部的状态

三大组件

• Logger
• Appender
• Layous

Logger 作为 logback-classic 模块的一部分、Appender 与 Layouts 接口作为 logback-core 的一部分、作为一个通用的模块、logback-core 没有 Logger 的概念

LoggerContext

实现了 slf4j 的 ILoggerFactory 接口、用于产生 Logger

LoggerContext 主要的成员变量

扫描二维码关注公众号，回复： 13739103 查看本文章

• 第一个是 root logger
• 第二个就是一个缓存、所以对于 Logger 的名字来说、它大小写敏感
• TurboFilter 作为一个全局的过滤器

层级结构

Logback 的 Logger 以 name 的 . 作为层级结构的划分

名为 com.demo.Main 的 logger 它的父 logger 是 com.demo logger、再往上就是 com 在网上就是 root

public final class Logger implements org.slf4j.Logger, LocationAwareLogger, AppenderAttachable<ILoggingEvent>, Serializable {
    private static final long serialVersionUID = 5454405123156820674L; // 8745934908040027998L;
    public static final String FQCN = ch.qos.logback.classic.Logger.class.getName();
    private String name;
    transient private Level level;
    transient private int effectiveLevelInt;
    transient private Logger parent;
    transient private List<Logger> childrenList;
复制代码

字段 parent 就是用来存放父级 logger 的、而 childrenList 则是存放子 logger 的、当父 logger 的等级发生改变的时候就是依靠该字段去通知子 logger 。

有效等级

我们知道如果我们 logger 设置的等级大于调用的方法、那么该日志不会被输出、显然这种等值的比较、可以依靠数值来实现。而这个值在 Logger 类中就是 effectiveLevelInt 这个字段

public static final int OFF_INT = Integer.MAX_VALUE;
public static final int ERROR_INT = 40000;
public static final int WARN_INT = 30000;
public static final int INFO_INT = 20000;
public static final int DEBUG_INT = 10000;
public static final int TRACE_INT = 5000;
public static final int ALL_INT = Integer.MIN_VALUE;
    public static final Level OFF = new Level(OFF_INT, "OFF");
    public static final Level ERROR = new Level(ERROR_INT, "ERROR");
    public static final Level WARN = new Level(WARN_INT, "WARN");
    public static final Level INFO = new Level(INFO_INT, "INFO");
    public static final Level DEBUG = new Level(DEBUG_INT, "DEBUG");
    public static final Level TRACE = new Level(TRACE_INT, "TRACE");
    public static final Level ALL = new Level(ALL_INT, "ALL");
复制代码

我们可以看到

public LoggerContext() {
    super();
    this.loggerCache = new ConcurrentHashMap<String, Logger>();

    this.loggerContextRemoteView = new LoggerContextVO(this);
    this.root = new Logger(Logger.ROOT_LOGGER_NAME, null, this);
    this.root.setLevel(Level.DEBUG);
    loggerCache.put(Logger.ROOT_LOGGER_NAME, root);
    initEvaluatorMap();
    size = 1;
    this.frameworkPackages = new ArrayList<String>();
}
复制代码

默认情况下 root logger 的 level 为 debug、effectiveLevelInt 的值为 10000

默认情况下、如果我们创建使用的 logger 没有指定 level 的话、它会继承它的父类的 effectiveLevelInt

Logger createChildByName(final String childName) {
		........
    Logger childLogger;
    childLogger = new Logger(childName, this, this.loggerContext);
    childrenList.add(childLogger);
    childLogger.effectiveLevelInt = this.effectiveLevelInt;
    return childLogger;
} 
复制代码

this 就是父 logger

当 父 logger level 发生变化时、会调用方法通知子 logger

Logger createChildByName(final String childName) {
    int i_index = LoggerNameUtil.getSeparatorIndexOf(childName, this.name.length() + 1);
    if (i_index != -1) {
        throw new IllegalArgumentException("For logger [" + this.name + "] child name [" + childName
                        + " passed as parameter, may not include '.' after index" + (this.name.length() + 1));
    }

    if (childrenList == null) {
        childrenList = new CopyOnWriteArrayList<Logger>();
    }
    Logger childLogger;
    childLogger = new Logger(childName, this, this.loggerContext);
    childrenList.add(childLogger);
    childLogger.effectiveLevelInt = this.effectiveLevelInt;
    return childLogger;
}
复制代码

demo 如下

private static void test3(){
    ch.qos.logback.classic.Logger logger = (ch.qos.logback.classic.Logger) LoggerFactory.getLogger("x.y");
    logger.setLevel(Level.INFO);
    logger.info("info");
    logger.debug("debug");
    ch.qos.logback.classic.Logger childLogger = (ch.qos.logback.classic.Logger) LoggerFactory.getLogger("x.y.z");
    childLogger.info("info");
    childLogger.debug("debug");
    logger.setLevel(Level.TRACE);
    logger.trace("trace");
    logger.debug("debug");
    childLogger.trace("trace");
    childLogger.debug("debug");
}
复制代码

输出的话、你自己想想

Appender

一个 Logger 可以有多个 Appender、比如说日志既可以输出到控制台也可以到日志文件

// Logger 类中使用 AppenderAttachableImpl<ILoggingEvent> aai 保存、内部使用 COWList 保存
public synchronized void addAppender(Appender<ILoggingEvent> newAppender) {
    if (aai == null) {
        aai = new AppenderAttachableImpl<ILoggingEvent>();
    }
    aai.addAppender(newAppender);
}
复制代码

Appender 默认也是具有继承的特性、比如说 root 的 Appender 是 Console、而子 logger 的 Appender 是 file、那么使用子 logger 打印日志、日志即会输出到日志文件也会输出到 console 中

// Logger
public void callAppenders(ILoggingEvent event) {
    int writes = 0;
    for (Logger l = this; l != null; l = l.parent) {
        writes += l.appendLoopOnAppenders(event);
        if (!l.additive) {
            break;
        }
    }
    // No appenders in hierarchy
    if (writes == 0) {
        loggerContext.noAppenderDefinedWarning(this);
    }
}
复制代码

// AppenderAttachableImpl
public int appendLoopOnAppenders(E e) {
    int size = 0;
    final Appender<E>[] appenderArray = appenderList.asTypedArray();
    final int len = appenderArray.length;
    for (int i = 0; i < len; i++) {
        appenderArray[i].doAppend(e);
        size++;
    }
    return size;
}
复制代码

先调用自己的 Appender 然后调用父类的。

可以将 additive 设置为 false、那么子 logger 不再对父 logger 有继承 Appender

流程

过滤链

public void info(String msg) {
    filterAndLog_0_Or3Plus(FQCN, null, Level.INFO, msg, null, null);
}

private void filterAndLog_0_Or3Plus(final String localFQCN, final Marker marker, final Level level, final String msg, final Object[] params,
                                    final Throwable t) {

  final FilterReply decision = loggerContext.getTurboFilterChainDecision_0_3OrMore(marker, this, level, msg, params, t);

  if (decision == FilterReply.NEUTRAL) {
    if (effectiveLevelInt > level.levelInt) {
      return;
    }
  } else if (decision == FilterReply.DENY) {
    return;
  }

  buildLoggingEventAndAppend(localFQCN, marker, level, msg, params, t);
}
复制代码

从 LoggerContext 中获取 TurboFilterList

final FilterReply getTurboFilterChainDecision_0_3OrMore(final Marker marker, final Logger logger, final Level level, final String format,
                final Object[] params, final Throwable t) {
    if (turboFilterList.size() == 0) {
        return FilterReply.NEUTRAL;
    }
    return turboFilterList.getTurboFilterChainDecision(marker, logger, level, format, params, t);
}
复制代码

默认情况是没有 turboFilter 、所以会返回 NEUTRAL

final FilterReply getTurboFilterChainDecision_0_3OrMore(final Marker marker, final Logger logger, final Level level, final String format,
                final Object[] params, final Throwable t) {
    if (turboFilterList.size() == 0) {
        return FilterReply.NEUTRAL;
    }
    return turboFilterList.getTurboFilterChainDecision(marker, logger, level, format, params, t);
}
复制代码

public FilterReply getTurboFilterChainDecision(final Marker marker, final Logger logger, final Level level, final String format, final Object[] params,
                final Throwable t) {

    final int size = size();
    // if (size == 0) {
    // return FilterReply.NEUTRAL;
    // }
    if (size == 1) {
        try {
            TurboFilter tf = get(0);
            return tf.decide(marker, logger, level, format, params, t);
        } catch (IndexOutOfBoundsException iobe) {
            return FilterReply.NEUTRAL;
        }
    }

    Object[] tfa = toArray();
    final int len = tfa.length;
    for (int i = 0; i < len; i++) {
        // for (TurboFilter tf : this) {
        final TurboFilter tf = (TurboFilter) tfa[i];
        final FilterReply r = tf.decide(marker, logger, level, format, params, t);
        if (r == FilterReply.DENY || r == FilterReply.ACCEPT) {
            return r;
        }
    }
    return FilterReply.NEUTRAL;
}
复制代码

如果存在多个、如果明确是 DENY 或者 ACCPET 的话、直接返回、如果不确定、那么遍历到最后返回 NEUTRAL

如果返回到是 NEUTRAL 则判断调用的方法的级别是否大于 logger 设置的 level 如果小于则不打印直接返回。

同理、返回 DENY 也是直接返回、只有 ACCEPT 才能继续往下走

LoggingEvent 对象

private void buildLoggingEventAndAppend(final String localFQCN, final Marker marker, final Level level, final String msg, final Object[] params,
                final Throwable t) {
    LoggingEvent le = new LoggingEvent(localFQCN, this, level, msg, t, params);
    le.setMarker(marker);
    callAppenders(le);
}
复制代码

LoggingEvent 对象，这个对象包含了日志请求所有相关的参数，请求的 logger，日志请求的级别，日志信息，与日志一同传递的异常信息，当前时间，当前线程，以及当前类的各种信息和 MDC。

调用 Appender

public void callAppenders(ILoggingEvent event) {
    int writes = 0;
    for (Logger l = this; l != null; l = l.parent) {
        writes += l.appendLoopOnAppenders(event);
        if (!l.additive) {
            break;
        }
    }
    // No appenders in hierarchy
    if (writes == 0) {
        loggerContext.noAppenderDefinedWarning(this);
    }
}
复制代码

这个在上面谈及过

public int appendLoopOnAppenders(E e) {
    int size = 0;
    final Appender<E>[] appenderArray = appenderList.asTypedArray();
    final int len = appenderArray.length;
    for (int i = 0; i < len; i++) {
        appenderArray[i].doAppend(e);
        size++;
    }
    return size;
}
复制代码

public synchronized void doAppend(E eventObject) {
    // WARNING: The guard check MUST be the first statement in the
    // doAppend() method.

    // prevent re-entry.
    if (guard) {
        return;
    }

    try {
        guard = true;

        if (!this.started) {
            if (statusRepeatCount++ < ALLOWED_REPEATS) {
                addStatus(new WarnStatus("Attempted to append to non started appender [" + name + "].", this));
            }
            return;
        }

        if (getFilterChainDecision(eventObject) == FilterReply.DENY) {
            return;
        }

        // ok, we now invoke derived class' implementation of append
        this.append(eventObject);

    } catch (Exception e) {
        if (exceptionCount++ < ALLOWED_REPEATS) {
            addError("Appender [" + name + "] failed to append.", e);
        }
    } finally {
        guard = false;
    }
}
复制代码

这里会涉及到自定义的过滤器 getFilterChainDecision

格式化输出

// SyslogAppenderBase
@Override
protected void append(E eventObject) {
    if (!isStarted()) {
        return;
    }

    try {
        String msg = layout.doLayout(eventObject);
        if (msg == null) {
            return;
        }
        if (msg.length() > maxMessageSize) {
            msg = msg.substring(0, maxMessageSize);
        }
        sos.write(msg.getBytes(charset));
        sos.flush();
        postProcess(eventObject, sos);
    } catch (IOException ioe) {
        addError("Failed to send diagram to " + syslogHost, ioe);
    }
}
复制代码

这个会借助 Layout 对象、将 EventObject 转换成一个字符串。

但是也存在一些 Appender 不需要对其进行格式化、比如说输出到MQ、Socket 之类的

private void dispatchEvents(ObjectWriter objectWriter) throws InterruptedException, IOException {
    while (true) {
        E event = deque.takeFirst();
        postProcessEvent(event);
        Serializable serializableEvent = getPST().transform(event);
        try {
            objectWriter.write(serializableEvent);
        } catch (IOException e) {
            tryReAddingEventToFrontOfQueue(event);
            throw e;
        }
    }
}
复制代码

它们将其序列化如何写入输出流中

    public Serializable transform(ILoggingEvent event) {
        if (event == null) {
            return null;
        }
        if (event instanceof LoggingEvent) {
            return LoggingEventVO.build(event);
        } else if (event instanceof LoggingEventVO) {
            return (LoggingEventVO) event;
        } else {
            throw new IllegalArgumentException("Unsupported type " + event.getClass().getName());
        }
    }

}
复制代码

发送 LoggingEvent

当日志事件被完全格式化之后将会通过每个 appender 发送到具体的目的地。