新版实现AGG BY TO_MAP(cls,score) as (data)返回的数据类型是内建MAP类型，因此可以用键值对的方式直接访问，简单高效。

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

* [Feature-946][alarm] Alarm after task monitoring retry

* Format code
Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

1.引入泛型，以自适应多种数据类型；
2.直接返回Map<K,V>,避免数据转换

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

* add a table, name: dlink_upload_file_record

* File upload implement

* Modify function name and javadoc

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

* [Bug] fix the banner of dlink admin app

* remove banners

* [Fix][dlink-gateway]修复k8s Application、Yarn Application、Yarn PerJob提交任务时未指定集群配置MasterMemoryMB、TaskManagerMemoryMB、SlotsPerTaskManager

* [Fix][dlink-gateway]修复k8s Application、Yarn Application、Yarn PerJob提交任务时未指定集群配置MasterMemoryMB、TaskManagerMemoryMB、SlotsPerTaskManager

* [Fix][dlink-gateway]修复k8s application模式提交失败，优化增加获取JobId等待时间，

* [Fix][dlink-common]扩展系统配置参数添加jobid的最大等待时间

优化PG库"schema_name"查询

解决KafkaSinkBuilder没有序列化会报错的问题。

单目标使用 'sink.*'
多目标使用 'sink[N].*', N为0开始的index索引.
Co-authored-by: 金鑫 <jinyanhui@huansi.net>

[Fix][dlink-gateway]修复k8s Application、Yarn Application、Yarn PerJob提交任务时未指定集群配置MasterMemoryMB、TaskManagerMemoryMB、SlotsPerTaskManager (#970)

* [Fix][dlink-gateway]修复k8s Application、Yarn Application、Yarn PerJob提交任务时未指定集群配置MasterMemoryMB、TaskManagerMemoryMB、SlotsPerTaskManager

* sqlserver增加date类型数据转换规则.

* sqlserver增加date类型数据转换规则.
Co-authored-by: 金鑫 <jinyanhui@huansi.net>

* Add swagger api doc

* update datasource config info in the application file

* Configure global checkstyle validation.

* Configure global checkstyle validation.

* The bug that the cluster information cannot be obtained in the savepoint save list

* Add taskId in case dlink maintains sql

* format code

* enhance manage logic

* 1. fix taskId is 0
2. reformat code

* reformat code

* refactor imports

* 增加sqlserver-cdc支持,可以通过手写finksql完成sqlservercdc同步,但是不支持cdcsource模式.

* 修改报名\类名为sqlserver

* 增加sqlserver-cdc支持,可以通过手写finksql完成sqlservercdc同步,但是不支持cdcsource模式.

* 优化trycatch异常日志打印代码

* 支持sqlservercdc功能
Co-authored-by: 金鑫 <jinyanhui@huansi.net>

* 修改点：
1、dlink-flink1.14增加postgres cdc connector，支持flinksql开发。
2、整库同步增加支持postgres数据库。

* 修改点：
1、METADATA_TYPE与dlink-meta中保持一致，避免Driver找不到lib的情况；
2、url中/后改为databasename，原来代码中填schema名是不正确的。

* postgresCDCBuilder改为首字母大写。

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

* [Feature-923][*] Build column lineage base on flink logical plan

* [Feature-923][*] Dynamic add getColumnOrigins method to class RelMdColumnOrigins by javassist
Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

* Added dlink-core module code style.

* fix AbstractTrans error

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>

Co-authored-by: wenmo <32723967+wenmo@users.noreply.github.com>