Merge branch 'main' of https://github.com/ModelEngine-Group/DataMate

Author: o0Shark0o

frontend/src/pages/DataCollection/Create/CreateTask.tsx

@@ -81,19 +81,7 @@ export default function CollectionTaskCreate() {
   const handleSubmit = async () => {
     try {
       await form.validateFields();
-
-      const values = form.getFieldsValue(true);
-      const payload = {
-        name: values.name,
-        description: values.description,
-        syncMode: values.syncMode,
-        scheduleExpression: values.scheduleExpression,
-        timeoutSeconds: values.timeoutSeconds,
-        templateId: values.templateId,
-        config: values.config,
-      };
-
-      await createTaskUsingPost(payload);
+      await createTaskUsingPost(newTask);
       message.success("任务创建成功");
       navigate("/data/collection");
     } catch (error) {

runtime/datax/nfsreader/src/main/java/com/modelengine/edatamate/plugin/reader/nfsreader/NfsReader.java

@@ -69,12 +69,14 @@ public static class Task extends Reader.Task {
         private Configuration jobConfig;
         private String mountPoint;
         private Set<String> fileType;
+        private List<String> files;
 
         @Override
         public void init() {
             this.jobConfig = super.getPluginJobConf();
             this.mountPoint = this.jobConfig.getString("mountPoint");
             this.fileType = new HashSet<>(this.jobConfig.getList("fileType", Collections.emptyList(), String.class));
+            this.files = this.jobConfig.getList("files", Collections.emptyList(), String.class);
         }
 
         @Override
@@ -83,6 +85,7 @@ public void startRead(RecordSender recordSender) {
                 List<String> files = stream.filter(Files::isRegularFile)
                         .filter(file -> fileType.isEmpty() || fileType.contains(getFileSuffix(file)))
                         .map(path -> path.getFileName().toString())
+                        .filter(fileName -> this.files.isEmpty() || this.files.contains(fileName))
                         .collect(Collectors.toList());
                 files.forEach(filePath -> {
                     Record record = recordSender.createRecord();

scripts/db/data-collection-init.sql

@@ -73,5 +73,5 @@ CREATE TABLE t_dc_collection_templates (
 ) COMMENT='数据归集模板配置表';
 
 INSERT IGNORE INTO t_dc_collection_templates(id, name, description, source_type, source_name, target_type, target_name, template_content, built_in, created_by, updated_by)
-VALUES ('1', 'NAS归集模板', '将NAS存储上的文件归集到DataMate平台上。', 'nfsreader', 'nfsreader', 'nfswriter', 'nfswriter', '{"parameter": {}, "reader": {}, "writer": {}}', True, 'system', 'system'),
-       ('2', 'OBS归集模板', '将OBS存储上的文件归集到DataMate平台上。', 'obsreader', 'obsreader', 'obswriter', 'obswriter', '{"parameter": {"endpoint": {"name": "服务地址","description": "OBS的服务地址。","type": "input"},"bucket": {"name": "存储桶名称","description": "OBS存储桶名称。","type": "input"},"accessKey": {"name": "访问密钥","description": "OBS访问密钥。","type": "input"},"secretKey": {"name": "密钥","description": "OBS密钥。","type": "input"},"prefix": {"name": "匹配前缀","description": "按照匹配前缀去选中OBS中的文件进行归集。","type": "input"}}, "reader": {}, "writer": {}}', True, 'system', 'system');
+VALUES ('1', 'NAS归集模板', '将NAS存储上的文件归集到DataMate平台上。', 'nfsreader', 'nfsreader', 'nfswriter', 'nfswriter', '{"parameter": {"ip": {"name": "NAS地址","description": "NAS服务的地址，可以为IP或者域名。","type": "input", "required": true}, "path": {"name": "共享路径","description": "NAS服务的共享路径。","type": "input", "required": true}, "files": {"name": "文件列表","description": "指定文件列表进行归集。","type": "select", "required": false}}, "reader": {}, "writer": {}}', True, 'system', 'system'),
+       ('2', 'OBS归集模板', '将OBS存储上的文件归集到DataMate平台上。', 'obsreader', 'obsreader', 'obswriter', 'obswriter', '{"parameter": {"endpoint": {"name": "服务地址","description": "OBS的服务地址。","type": "input", "required": true},"bucket": {"name": "存储桶名称","description": "OBS存储桶名称。","type": "input", "required": true},"accessKey": {"name": "AK","description": "OBS访问密钥。","type": "input", "required": true},"secretKey": {"name": "SK","description": "OBS密钥。","type": "password", "required": true},"prefix": {"name": "匹配前缀","description": "按照匹配前缀去选中OBS中的文件进行归集。","type": "input", "required": true}}, "reader": {}, "writer": {}}', True, 'system', 'system');

scripts/images/backend-python/Dockerfile

@@ -17,19 +17,18 @@ FROM python:3.12-slim
 # Note: to use the cache mount syntax you must build with BuildKit enabled:
 # DOCKER_BUILDKIT=1 docker build . -f scripts/images/datamate-python/Dockerfile -t datamate-backend-python
 
-RUN apt-get update \
-    && apt-get install -y --no-install-recommends openjdk-21-jre-headless \
-    && rm -rf /var/lib/apt/lists/*
+RUN apt-get update && \
+    apt-get install -y --no-install-recommends vim openjdk-21-jre nfs-common rsync && \
+    rm -rf /var/lib/apt/lists/*
 
 ENV PYTHONDONTWRITEBYTECODE=1 \
     PYTHONUNBUFFERED=1 \
-    # Poetry configuration
     POETRY_VERSION=2.2.1 \
     POETRY_NO_INTERACTION=1 \
     POETRY_VIRTUALENVS_CREATE=false \
     POETRY_CACHE_DIR=/tmp/poetry_cache
 
-ENV JAVA_HOME=/usr/lib/jvm/java-21-openjdk-amd64
+ENV JAVA_HOME=/usr/lib/jvm/java-21-openjdk
 
 ENV PATH="/root/.local/bin:$JAVA_HOME/bin:$PATH"
 

scripts/images/backend/Dockerfile

@@ -1,16 +1,3 @@
-FROM maven:3-eclipse-temurin-8 AS datax-builder
-
-RUN apt-get update && \
-    apt-get install -y git && \
-    git clone https://github.com/alibaba/DataX.git
-
-COPY runtime/datax/ DataX/
-
-RUN cd DataX && \
-    sed -i "s/com.mysql.jdbc.Driver/com.mysql.cj.jdbc.Driver/g" \
-    plugin-rdbms-util/src/main/java/com/alibaba/datax/plugin/rdbms/util/DataBaseType.java && \
-    mvn -U clean package assembly:assembly -Dmaven.test.skip=true
-
 FROM maven:3-eclipse-temurin-21 AS builder
 
 COPY backend/ /opt/backend
@@ -22,12 +9,11 @@ RUN cd /opt/backend/services && \
 FROM eclipse-temurin:21-jdk
 
 RUN apt-get update && \
-    apt-get install -y vim wget curl nfs-common rsync python3 python3-pip python-is-python3 dos2unix && \
+    apt-get install -y vim wget curl rsync python3 python3-pip python-is-python3 dos2unix && \
     apt-get clean && \
     rm -rf /var/lib/apt/lists/*
 
 COPY --from=builder /opt/backend/services/main-application/target/datamate.jar /opt/backend/datamate.jar
-COPY --from=datax-builder /DataX/target/datax/datax /opt/datax
 
 COPY scripts/images/backend/start.sh /opt/backend/start.sh