RocketMQ源码 Broker-ConsumerFilterManager 消费者数据过滤管理组件源码分析

news2024/9/29 11:30:55

前言

ConsumerFilterManager 继承了ConfigManager配置管理组件,拥有将内存数据持久化到磁盘文件consumerFilter.json的能力。它主要负责,对在消费者拉取消息时,进行消息数据过滤,且只针对使用表达式过滤的消费者有效。


源码版本:4.9.3

源码架构图

核心数据结构

可以看到内存中维护了 topic -> consumer group -> ConsumerFilterData 映射关系的数据结构。

/**
 * Consumer filter data manager.Just manage the consumers use expression filter.
 * 消费者过滤数据管理组件。只管理使用表达式过滤的消费者。
 */
public class ConsumerFilterManager extends ConfigManager {

    // 核心数据结构:topic -> consumer group -> ConsumerFilterData
    private ConcurrentMap<String/*Topic*/, FilterDataMapByTopic>
        filterDataByTopic = new ConcurrentHashMap<String/*Topic*/, FilterDataMapByTopic>(256);

    private transient BrokerController brokerController;

    // 布隆过滤器
    private transient BloomFilter bloomFilter;
}

深入看下 FilterDataMapByTopic 类,是上面数据结构的一个子集,维护了 消费组 -> 消费组过滤数据映射关系。

    public static class FilterDataMapByTopic {

        // 核心数据结构:consumer group -> ConsumerFilterData
        private ConcurrentMap<String/*consumer group*/, ConsumerFilterData>
            groupFilterData = new ConcurrentHashMap<String, ConsumerFilterData>();

        private String topic;
    }

在深入一步,看下 ConsumerFilterData 的数据结构,包含了全部与消费者过滤有关的关键信息。

/**
 * Filter data of consumer.
 */
public class ConsumerFilterData {
    // 消费组
    private String consumerGroup;
    // 主题
    private String topic;
    // 过滤器表达式
    private String expression;
    // 过滤器类型
    private String expressionType;
    // 过滤器编译后的表达式
    private transient Expression compiledExpression;
    // 过滤器创建时间
    private long bornTime;
    // 过滤器过期时间
    private long deadTime = 0;
    // 过滤器版本
    private long version;
    // 布隆过滤器数据
    private BloomFilterData bloomFilterData;
    // 客户端版本
    private long clientVersion;
}

核心数据行为

从下面代码可以看到,ConsumerFilterManager的行为主要是注册订阅、取消订阅、清理过期订阅、序列化、反序列化等维护内存元数据的行为。过滤行为不在这个组件里体现,在其他调用方法中会有具体使用方式。

/**
 * Consumer filter data manager.Just manage the consumers use expression filter.
 * 消费者过滤数据管理组件。只管理使用表达式过滤的消费者。
 */
public class ConsumerFilterManager extends ConfigManager {

    private static final InternalLogger log = InternalLoggerFactory.getLogger(LoggerName.FILTER_LOGGER_NAME);

    private static final long MS_24_HOUR = 24 * 3600 * 1000;

    // 核心数据结构:topic -> consumer group -> ConsumerFilterData
    private ConcurrentMap<String/*Topic*/, FilterDataMapByTopic>
        filterDataByTopic = new ConcurrentHashMap<String/*Topic*/, FilterDataMapByTopic>(256);

    private transient BrokerController brokerController;
    // 布隆过滤器
    private transient BloomFilter bloomFilter;

    public ConsumerFilterManager() {
        // just for test
        this.bloomFilter = BloomFilter.createByFn(20, 64);
    }

    public ConsumerFilterManager(BrokerController brokerController) {
        this.brokerController = brokerController;
        this.bloomFilter = BloomFilter.createByFn(
            brokerController.getBrokerConfig().getMaxErrorRateOfBloomFilter(),
            brokerController.getBrokerConfig().getExpectConsumerNumUseFilter()
        );
        // then set bit map length of store config.
        brokerController.getMessageStoreConfig().setBitMapLengthConsumeQueueExt(
            this.bloomFilter.getM()
        );
    }

    /**
     * Build consumer filter data.Be care, bloom filter data is not included.
     *
     * @return maybe null
     */
    public static ConsumerFilterData build(final String topic, final String consumerGroup,
        final String expression, final String type,
        final long clientVersion) {
        if (ExpressionType.isTagType(type)) {
            return null;
        }

        ConsumerFilterData consumerFilterData = new ConsumerFilterData();
        consumerFilterData.setTopic(topic);
        consumerFilterData.setConsumerGroup(consumerGroup);
        consumerFilterData.setBornTime(System.currentTimeMillis());
        consumerFilterData.setDeadTime(0);
        consumerFilterData.setExpression(expression);
        consumerFilterData.setExpressionType(type);
        consumerFilterData.setClientVersion(clientVersion);
        try {
            consumerFilterData.setCompiledExpression(
                FilterFactory.INSTANCE.get(type).compile(expression)
            );
        } catch (Throwable e) {
            log.error("parse error: expr={}, topic={}, group={}, error={}", expression, topic, consumerGroup, e.getMessage());
            return null;
        }

        return consumerFilterData;
    }

    /**
     * 在指定消费组注册消费者过滤数据
     * @param consumerGroup
     * @param subList
     */
    public void register(final String consumerGroup, final Collection<SubscriptionData> subList) {
        for (SubscriptionData subscriptionData : subList) {
            register(
                subscriptionData.getTopic(),
                consumerGroup,
                subscriptionData.getSubString(),
                subscriptionData.getExpressionType(),
                subscriptionData.getSubVersion()
            );
        }

        // make illegal topic dead.
        Collection<ConsumerFilterData> groupFilterData = getByGroup(consumerGroup);

        Iterator<ConsumerFilterData> iterator = groupFilterData.iterator();
        while (iterator.hasNext()) {
            ConsumerFilterData filterData = iterator.next();

            boolean exist = false;
            for (SubscriptionData subscriptionData : subList) {
                if (subscriptionData.getTopic().equals(filterData.getTopic())) {
                    exist = true;
                    break;
                }
            }

            if (!exist && !filterData.isDead()) {
                filterData.setDeadTime(System.currentTimeMillis());
                log.info("Consumer filter changed: {}, make illegal topic dead:{}", consumerGroup, filterData);
            }
        }
    }

    public boolean register(final String topic, final String consumerGroup, final String expression,
        final String type, final long clientVersion) {
        // 不支持tag类型
        if (ExpressionType.isTagType(type)) {
            return false;
        }

        if (expression == null || expression.length() == 0) {
            return false;
        }

        // 获取topic对应的消费者过滤数据
        FilterDataMapByTopic filterDataMapByTopic = this.filterDataByTopic.get(topic);

        if (filterDataMapByTopic == null) {
            FilterDataMapByTopic temp = new FilterDataMapByTopic(topic);
            FilterDataMapByTopic prev = this.filterDataByTopic.putIfAbsent(topic, temp);
            filterDataMapByTopic = prev != null ? prev : temp;
        }
        // 创建布隆过滤器数据
        BloomFilterData bloomFilterData = bloomFilter.generate(consumerGroup + "#" + topic);
        // 注册过滤数据到topic
        return filterDataMapByTopic.register(consumerGroup, expression, type, bloomFilterData, clientVersion);
    }

    // 取消注册消费者过滤数据
    public void unRegister(final String consumerGroup) {
        for (Entry<String, FilterDataMapByTopic> entry : filterDataByTopic.entrySet()) {
            entry.getValue().unRegister(consumerGroup);
        }
    }

    public ConsumerFilterData get(final String topic, final String consumerGroup) {
        if (!this.filterDataByTopic.containsKey(topic)) {
            return null;
        }
        if (this.filterDataByTopic.get(topic).getGroupFilterData().isEmpty()) {
            return null;
        }

        return this.filterDataByTopic.get(topic).getGroupFilterData().get(consumerGroup);
    }

    // 获取消费组下所有过滤数据
    public Collection<ConsumerFilterData> getByGroup(final String consumerGroup) {
        Collection<ConsumerFilterData> ret = new HashSet<ConsumerFilterData>();

        Iterator<FilterDataMapByTopic> topicIterator = this.filterDataByTopic.values().iterator();
        while (topicIterator.hasNext()) {
            FilterDataMapByTopic filterDataMapByTopic = topicIterator.next();

            Iterator<ConsumerFilterData> filterDataIterator = filterDataMapByTopic.getGroupFilterData().values().iterator();

            while (filterDataIterator.hasNext()) {
                ConsumerFilterData filterData = filterDataIterator.next();

                if (filterData.getConsumerGroup().equals(consumerGroup)) {
                    ret.add(filterData);
                }
            }
        }

        return ret;
    }

    // 获取topic下所有过滤数据
    public final Collection<ConsumerFilterData> get(final String topic) {
        if (!this.filterDataByTopic.containsKey(topic)) {
            return null;
        }
        if (this.filterDataByTopic.get(topic).getGroupFilterData().isEmpty()) {
            return null;
        }

        return this.filterDataByTopic.get(topic).getGroupFilterData().values();
    }

    public BloomFilter getBloomFilter() {
        return bloomFilter;
    }

    @Override
    public String encode() {
        return encode(false);
    }

    @Override
    public String configFilePath() {
        if (this.brokerController != null) {
            // 配置存储路径 config/consumerFilter.json
            return BrokerPathConfigHelper.getConsumerFilterPath(
                this.brokerController.getMessageStoreConfig().getStorePathRootDir()
            );
        }
        return BrokerPathConfigHelper.getConsumerFilterPath("./unit_test");
    }

    // 将json字符串反序列化为ConsumerFilterManager对象
    @Override
    public void decode(final String jsonString) {
        ConsumerFilterManager load = RemotingSerializable.fromJson(jsonString, ConsumerFilterManager.class);
        if (load != null && load.filterDataByTopic != null) {
            boolean bloomChanged = false;
            for (Entry<String, FilterDataMapByTopic> entry : load.filterDataByTopic.entrySet()) {
                FilterDataMapByTopic dataMapByTopic = entry.getValue();
                if (dataMapByTopic == null) {
                    continue;
                }

                for (Entry<String, ConsumerFilterData> groupEntry : dataMapByTopic.getGroupFilterData().entrySet()) {

                    ConsumerFilterData filterData = groupEntry.getValue();

                    if (filterData == null) {
                        continue;
                    }

                    try {
                        filterData.setCompiledExpression(
                                FilterFactory.INSTANCE.get(filterData.getExpressionType()).compile(filterData.getExpression())
                        );
                    } catch (Exception e) {
                        log.error("load filter data error, " + filterData, e);
                    }

                    // check whether bloom filter is changed
                    // if changed, ignore the bit map calculated before.
                    if (!this.bloomFilter.isValid(filterData.getBloomFilterData())) {
                        bloomChanged = true;
                        log.info("Bloom filter is changed!So ignore all filter data persisted! {}, {}", this.bloomFilter, filterData.getBloomFilterData());
                        break;
                    }

                    log.info("load exist consumer filter data: {}", filterData);

                    if (filterData.getDeadTime() == 0) {
                        // we think all consumers are dead when load
                        long deadTime = System.currentTimeMillis() - 30 * 1000;
                        filterData.setDeadTime(
                                deadTime <= filterData.getBornTime() ? filterData.getBornTime() : deadTime
                        );
                    }
                }
            }

            if (!bloomChanged) {
                this.filterDataByTopic = load.filterDataByTopic;
            }
        }
    }

    // 将ConsumerFilterManager对象序列化为json字符串
    @Override
    public String encode(final boolean prettyFormat) {
        // clean
        {
            clean();
        }
        return RemotingSerializable.toJson(this, prettyFormat);
    }

    // 清理过期的过滤数据
    public void clean() {
        Iterator<Map.Entry<String, FilterDataMapByTopic>> topicIterator = this.filterDataByTopic.entrySet().iterator();
        while (topicIterator.hasNext()) {
            Map.Entry<String, FilterDataMapByTopic> filterDataMapByTopic = topicIterator.next();

            Iterator<Map.Entry<String, ConsumerFilterData>> filterDataIterator
                = filterDataMapByTopic.getValue().getGroupFilterData().entrySet().iterator();

            while (filterDataIterator.hasNext()) {
                Map.Entry<String, ConsumerFilterData> filterDataByGroup = filterDataIterator.next();

                ConsumerFilterData filterData = filterDataByGroup.getValue();
                if (filterData.howLongAfterDeath() >= (this.brokerController == null ? MS_24_HOUR : this.brokerController.getBrokerConfig().getFilterDataCleanTimeSpan())) {
                    log.info("Remove filter consumer {}, died too long!", filterDataByGroup.getValue());
                    filterDataIterator.remove();
                }
            }

            if (filterDataMapByTopic.getValue().getGroupFilterData().isEmpty()) {
                log.info("Topic has no consumer, remove it! {}", filterDataMapByTopic.getKey());
                topicIterator.remove();
            }
        }
    }

    public ConcurrentMap<String, FilterDataMapByTopic> getFilterDataByTopic() {
        return filterDataByTopic;
    }

    public void setFilterDataByTopic(final ConcurrentHashMap<String, FilterDataMapByTopic> filterDataByTopic) {
        this.filterDataByTopic = filterDataByTopic;
    }

    public static class FilterDataMapByTopic {

        // 核心数据结构:consumer group -> ConsumerFilterData
        private ConcurrentMap<String/*consumer group*/, ConsumerFilterData>
            groupFilterData = new ConcurrentHashMap<String, ConsumerFilterData>();

        private String topic;

        public FilterDataMapByTopic() {
        }

        public FilterDataMapByTopic(String topic) {
            this.topic = topic;
        }

        // 取消注册某个消费组的过滤器
        public void unRegister(String consumerGroup) {
            if (!this.groupFilterData.containsKey(consumerGroup)) {
                return;
            }

            ConsumerFilterData data = this.groupFilterData.get(consumerGroup);

            if (data == null || data.isDead()) {
                return;
            }

            long now = System.currentTimeMillis();

            log.info("Unregister consumer filter: {}, deadTime: {}", data, now);

            data.setDeadTime(now);
        }

        public boolean register(String consumerGroup, String expression, String type, BloomFilterData bloomFilterData,
            long clientVersion) {
            ConsumerFilterData old = this.groupFilterData.get(consumerGroup);

            if (old == null) {
                // 构建过滤器数据
                ConsumerFilterData consumerFilterData = build(topic, consumerGroup, expression, type, clientVersion);
                if (consumerFilterData == null) {
                    return false;
                }
                // 设置布隆过滤器
                consumerFilterData.setBloomFilterData(bloomFilterData);

                // 放入内存数据结构
                old = this.groupFilterData.putIfAbsent(consumerGroup, consumerFilterData);
                if (old == null) {
                    log.info("New consumer filter registered: {}", consumerFilterData);
                    return true;
                } else {
                    if (clientVersion <= old.getClientVersion()) {
                        if (!type.equals(old.getExpressionType()) || !expression.equals(old.getExpression())) {
                            log.warn("Ignore consumer({} : {}) filter(concurrent), because of version {} <= {}, but maybe info changed!old={}:{}, ignored={}:{}",
                                consumerGroup, topic,
                                clientVersion, old.getClientVersion(),
                                old.getExpressionType(), old.getExpression(),
                                type, expression);
                        }
                        if (clientVersion == old.getClientVersion() && old.isDead()) {
                            reAlive(old);
                            return true;
                        }

                        return false;
                    } else {
                        this.groupFilterData.put(consumerGroup, consumerFilterData);
                        log.info("New consumer filter registered(concurrent): {}, old: {}", consumerFilterData, old);
                        return true;
                    }
                }
            } else {
                // 当前版本号小于旧的版本号
                if (clientVersion <= old.getClientVersion()) {
                    if (!type.equals(old.getExpressionType()) || !expression.equals(old.getExpression())) {
                        log.info("Ignore consumer({}:{}) filter, because of version {} <= {}, but maybe info changed!old={}:{}, ignored={}:{}",
                            consumerGroup, topic,
                            clientVersion, old.getClientVersion(),
                            old.getExpressionType(), old.getExpression(),
                            type, expression);
                    }
                    if (clientVersion == old.getClientVersion() && old.isDead()) {
                        reAlive(old);
                        return true;
                    }

                    return false;
                }
                // 新版本号大于旧的版本号
                boolean change = !old.getExpression().equals(expression) || !old.getExpressionType().equals(type);
                if (old.getBloomFilterData() == null && bloomFilterData != null) {
                    change = true;
                }
                if (old.getBloomFilterData() != null && !old.getBloomFilterData().equals(bloomFilterData)) {
                    change = true;
                }

                // if subscribe data is changed, or consumer is died too long.
                if (change) {
                    // 构建过滤器数据
                    ConsumerFilterData consumerFilterData = build(topic, consumerGroup, expression, type, clientVersion);
                    if (consumerFilterData == null) {
                        // new expression compile error, remove old, let client report error.
                        this.groupFilterData.remove(consumerGroup);
                        return false;
                    }
                    consumerFilterData.setBloomFilterData(bloomFilterData);
                    // 设置过滤器数据
                    this.groupFilterData.put(consumerGroup, consumerFilterData);

                    log.info("Consumer filter info change, old: {}, new: {}, change: {}",
                        old, consumerFilterData, change);

                    return true;
                } else {
                    // 版本号一致,更新过滤器数据
                    old.setClientVersion(clientVersion);
                    if (old.isDead()) {
                        reAlive(old);
                    }
                    return true;
                }
            }
        }

        protected void reAlive(ConsumerFilterData filterData) {
            long oldDeadTime = filterData.getDeadTime();
            filterData.setDeadTime(0);
            log.info("Re alive consumer filter: {}, oldDeadTime: {}", filterData, oldDeadTime);
        }

        public final ConsumerFilterData get(String consumerGroup) {
            return this.groupFilterData.get(consumerGroup);
        }

        public final ConcurrentMap<String, ConsumerFilterData> getGroupFilterData() {
            return this.groupFilterData;
        }

        public void setGroupFilterData(final ConcurrentHashMap<String, ConsumerFilterData> groupFilterData) {
            this.groupFilterData = groupFilterData;
        }

        public String getTopic() {
            return topic;
        }

        public void setTopic(final String topic) {
            this.topic = topic;
        }
    }
}

本文来自互联网用户投稿,该文观点仅代表作者本人,不代表本站立场。本站仅提供信息存储空间服务,不拥有所有权,不承担相关法律责任。如若转载,请注明出处:http://www.coloradmin.cn/o/1309481.html

如若内容造成侵权/违法违规/事实不符,请联系多彩编程网进行投诉反馈,一经查实,立即删除!

相关文章

SFP3006-ASEMI大电流快恢复二极管SFP3006

编辑&#xff1a;ll SFP3006-ASEMI大电流快恢复二极管SFP3006 型号&#xff1a;SFP3006 品牌&#xff1a;ASEMI 封装&#xff1a;TO-247 最大平均正向电流&#xff1a;30A 最大重复峰值反向电压&#xff1a;600V 产品引线数量&#xff1a;3 产品内部芯片个数&#xff1…

多项式回归

多项式回归 多项式回归 是非线性回归的一种&#xff0c;之前讨论的线性回归都是直线&#xff0c;而多项式回归则是曲线的回归&#xff0c;它通过引入原始预测变量的高阶项&#xff08;如平方、立方等&#xff09;来拟合数据的非线性模式。在多项式回归中&#xff0c;虽然模型对…

06 python 文件基础操作

6.1 .1文件读取操作 演示对文件的读取 # 打开文件 import timef open(02_word.txt, r, encoding"UTF-8") print(type(f))# #读取文件 - read() # print(f读取10个字节的结果{f.read(10)}) # print(f读取全部字节的结果{f.read()})# #读取文件 - readLines() # lines…

头歌-Python 基础

第1关&#xff1a;建模与仿真 1、 建模过程&#xff0c;通常也称为数学优化建模(Mathematical Optimization Modeling)&#xff0c;不同之处在于它可以确定特定场景的特定的、最优化或最佳的结果。这被称为诊断一个结果&#xff0c;因此命名为▁▁▁。 填空1答案&#xff1a;决…

Talk | UCSB博士生王丹青: 大语言模型的协作学习以及个性化生成评估

本期为TechBeat人工智能社区第555期线上Talk。 北京时间12月13日(周三)20:00&#xff0c;加州大学圣塔芭芭拉分校博士生—王丹青的Talk已准时在TechBeat人工智能社区开播&#xff01; 她与大家分享的主题是: “大语言模型的协作学习以及个性化生成评估”&#xff0c;介绍了她的…

re:Invent 产品体验分享:Amazon ElastiCache Serverless 缓存即时扩展功能与感受

授权说明&#xff1a;本篇文章授权活动官方亚马逊云科技文章转发、改写权&#xff0c;包括不限于在亚马逊云科技开发者社区、 知乎、自媒体平台、第三方开发者媒体等亚马逊云科技官方渠道&#xff09;。 文章目录 前言产品介绍产品使用步骤1.创建缓存服务2.安全组开放访问权限…

高效营销系统集成:百度营销的API无代码解决方案,提升电商与广告效率

百度营销API连接&#xff1a;构建无代码开发的高效集成体系 在数字营销的高速发展时代&#xff0c;企业追求的是快速响应市场的能力以及提高用户运营的效率。百度营销API连接正是为此而生&#xff0c;它通过无代码开发的方式&#xff0c;实现了电商平台、营销系统和CRM的一站式…

uniapp用picker实现自定义三级联动(children)

数据大概得结构&#xff1a; html: <view class"invite"><u-cell title"点击选则" isLink :value"value" click"showstrue"></u-cell><u-picker :show"shows" ref"uPicker" :columns"…

微服务最佳实践:构建可扩展且高效的系统

微服务架构彻底改变了现代软件开发&#xff0c;提供了无与伦比的敏捷性、可扩展性和可维护性。然而&#xff0c;有效实施微服务需要深入了解最佳实践&#xff0c;以充分发挥微服务的潜力&#xff0c;同时避免常见的陷阱。在这份综合指南中&#xff0c;我们将深入研究微服务的关…

【Vue】vue增加导航标签

系列文章 【Vue】vue增加导航标签 本文链接&#xff1a;https://blog.csdn.net/youcheng_ge/article/details/134965353 【Vue】Element开发笔记 本文链接&#xff1a;https://blog.csdn.net/youcheng_ge/article/details/133947977 【Vue】vue&#xff0c;在Windows IIS平台…

关于pytorch中的dim的理解

今天碰到一个代码看起来很简单&#xff0c;但是细究原理又感觉好像不太通不太对劲&#xff0c;就是多维tensor数据的操作&#xff0c;比如&#xff1a;y.sum(dim2)&#xff0c;乍一看很简单数据相加操作&#xff0c;但是仔细一想&#xff0c;这里在第3维度的数据到底是横向相加…

Python中的继承:概念、用法与示例

目录 一、引言 二、继承的概念 三、继承的用法 1、继承父类的属性和方法 2、添加新的属性和方法 3、覆盖父类的方法 四、示例代码展示 五、继承中的多态性 六、继承中的封装和抽象 七、继承中的多重继承 总结 一、引言 面向对象编程&#xff08;OOP&#xff09;是一…

循环神经网络(1)循环神经网络的记忆能力实验

循环神经网络&#xff08;Recurrent Neural Network&#xff0c;RNN&#xff09;是一类具有短期记忆能力的神经网络&#xff0e;在循环神经网络中&#xff0c;神经元不但可以接受其他神经元的信息&#xff0c;也可以接受自身的信息&#xff0c;形成具有环路的网络结构&#xff…

node加密集合(前端加密、后台解密)

文章目录 一、crypto 加解密生成私密钥公钥加密&#xff08;也可私钥加密&#xff09;私钥解密&#xff08;也可公钥解密&#xff09; 二、node-rsa加解密生成公私秘钥使用公钥加密&#xff08;也可私钥加密&#xff09;使用私钥解密&#xff08;也可公钥解密&#xff09; 三、…

Shell三剑客:正则表达式(元字符)

一、定义&#xff1a;元字符字符是这样一类字符&#xff0c;它们表达的是不同字面本身的含义 二、分类&#xff1a; 1、基本正则表达式元字符 # ^ 行首定位 [rootlocalhost ~]# grep root /etc/passwd root:x:0:0:root:/root:/bin/bash operator:x:11:0:operator:/root:/…

洁丽雅×实在RPA丨用取数宝获取全域经营数据,成本降低80%以上

12月6日&#xff0c;实在智能“实力派数智先锋访谈节”邀请到洁丽雅数据部负责人韩宇轩&#xff0c;围绕“「取数宝」为企业带来效率提升和成本降低”的话题&#xff0c;为线上观众分享了洁丽雅在数字化转型过程中的实践和经验。 洁丽雅组建于2003年&#xff0c;前身是1986年成…

智能客服的应用——政务领域

#本文来源清华大学数据治理研究中心政务热线数智化发展报告 &#xff0c;如有侵权&#xff0c;请联系删除。 面对地方政务热线发展所面临的挑战&#xff0c;数智化转型已经成为了热线系统突破当前发展瓶颈、实现整体提质增效的关键手段。《意见》中也明确指出&#xff0c;政务…

用友 U8总账凭证打印设置

总账--凭证打印——设置 是设置凭证打印显示的格子框&#xff0c;勾上就有框&#xff0c;去掉就没有框。

打印机怎么扫描文件到电脑?6个步骤!轻松完成!

“在工作时我经常需要用到打印机&#xff0c;有时候需要将部分文件扫描到电脑。但是我不是很清楚应该如何操作&#xff0c;有什么方法可以让打印机快速传输文件到电脑的方法吗&#xff1f;” 在人们的工作和学习中&#xff0c;打印机成了很多用户的必备工具。人们可以用它来打印…

swing快速入门(八)

注释很详细&#xff0c;直接上代码 上一篇 新增内容 cardLayout布局管理器 事件监听器的创建与绑定 多种布局与容器的结合使用 import java.awt.*; import java.awt.event.ActionEvent; import java.awt.event.ActionListener;public class swing_test_6 {public static v…