feature: multiple ratio configurations can be set for the data set. (#103)

feature: multiple ratio configurations can be set for the data set.

backend/services/data-management-service/src/main/java/com/datamate/datamanagement/domain/model/dataset/FileTag.java

@@ -20,7 +20,7 @@ import java.util.Map;
 @JsonIgnoreProperties(ignoreUnknown = true)
 @JsonNaming(PropertyNamingStrategies.SnakeCaseStrategy.class)
 public class FileTag {
-    private Map<String, Object> value;
+    private Map<String, Object> values;
 
     private String type;
 
@@ -30,7 +30,7 @@ public class FileTag {
 
     public List<String> getTags() {
         List<String> tags = new ArrayList<>();
-        Object tagValues = value.get(type);
+        Object tagValues = values.get(type);
         if (tagValues instanceof List) {
             for (Object tag : (List<?>) tagValues) {
                 if (tag instanceof String) {

frontend/src/pages/RatioTask/Create/CreateRatioTask.tsx

@@ -38,7 +38,7 @@ export default function CreateRatioTask() {
       const totals = String(values.totalTargetCount);
       const config = ratioTaskForm.ratioConfigs.map((c) => {
         return {
-          datasetId: c.id,
+          datasetId: c.source,
           counts: String(c.quantity ?? 0),
           filterConditions: { label: c.labelFilter, dateRange: String(c.dateRange ?? 0)},
         };

frontend/src/pages/RatioTask/Create/components/RatioConfig.tsx

@@ -1,11 +1,16 @@
-import React, { useMemo, useState } from "react";
+import React, { useMemo, useState, useEffect, FC } from "react";
-import { Badge, Card, Input, Progress, Button, DatePicker, Select } from "antd";
+import {
-import { BarChart3, Filter, Clock } from "lucide-react";
+  Badge,
+  Card,
+  Progress,
+  Button,
+  Select,
+  Table,
+  InputNumber,
+  Space,
+} from "antd";
+import { BarChart3, Filter } from "lucide-react";
 import type { Dataset } from "@/pages/DataManagement/dataset.model.ts";
-import dayjs from 'dayjs';
-
-const { RangePicker } = DatePicker;
-const { Option } = Select;
 
 const TIME_RANGE_OPTIONS = [
   { label: '最近1天', value: 1 },
@@ -21,9 +26,9 @@ interface RatioConfigItem {
   type: "dataset" | "label";
   quantity: number;
   percentage: number;
-  source: string;
+  source: string; // dataset id
   labelFilter?: string;
-  dateRange?: string;
+  dateRange?: number;
 }
 
 interface RatioConfigProps {
@@ -35,169 +40,113 @@ interface RatioConfigProps {
   onChange?: (configs: RatioConfigItem[]) => void;
 }
 
-const RatioConfig: React.FC<RatioConfigProps> = ({
+const genId = (datasetId: string) =>
-                                                   ratioType,
+  `${datasetId}-${Math.random().toString(36).slice(2, 9)}`;
-                                                   selectedDatasets,
+
-                                                   datasets,
+const RatioConfig: FC<RatioConfigProps> = ({
-                                                   totalTargetCount,
+                                             ratioType,
-                                                   distributions,
+                                             selectedDatasets,
-                                                   onChange,
+                                             datasets,
-                                                 }) => {
+                                             totalTargetCount,
-  const [ratioConfigs, setRatioConfigs] = useState<RatioConfigItem[]>([]);
+                                             distributions,
-  const [datasetFilters, setDatasetFilters] = useState<Record<string, {
+                                             onChange,
-    labelFilter?: string;
+                                           }) => {
-    dateRange?: string;
+  const [ratioConfigs, setRatioConfigs] = useState<RatioConfigItem[]>([]);
-  }>>({});
 
-  // 配比项总数
   const totalConfigured = useMemo(
     () => ratioConfigs.reduce((sum, c) => sum + (c.quantity || 0), 0),
     [ratioConfigs]
   );
 
-  // 获取数据集的标签列表
   const getDatasetLabels = (datasetId: string): string[] => {
     const dist = distributions[String(datasetId)] || {};
     return Object.keys(dist);
   };
 
-  // 自动平均分配
+  const addConfig = (datasetId: string) => {
+    const dataset = datasets.find((d) => String(d.id) === datasetId);
+    const newConfig: RatioConfigItem = {
+      id: genId(datasetId),
+      name: dataset?.name || datasetId,
+      type: ratioType,
+      quantity: 0,
+      percentage: 0,
+      source: datasetId,
+    };
+    const newConfigs = [...ratioConfigs, newConfig];
+    setRatioConfigs(newConfigs);
+    onChange?.(newConfigs);
+  };
+
+  const removeConfig = (configId: string) => {
+    const newConfigs = ratioConfigs.filter((c) => c.id !== configId);
+    const adjusted = recomputePercentages(newConfigs);
+    setRatioConfigs(adjusted);
+    onChange?.(adjusted);
+  };
+
+  const updateConfig = (
+    configId: string,
+    updates: Partial<
+      Pick<RatioConfigItem, "quantity" | "labelFilter" | "dateRange">
+    >
+  ) => {
+    const newConfigs = ratioConfigs.map((c) =>
+      c.id === configId ? { ...c, ...updates } : c
+    );
+    const adjusted = recomputePercentages(newConfigs);
+    setRatioConfigs(adjusted);
+    onChange?.(adjusted);
+  };
+
+  const recomputePercentages = (configs: RatioConfigItem[]) => {
+    return configs.map((c) => ({
+      ...c,
+      percentage:
+        totalTargetCount > 0
+          ? Math.round((c.quantity / totalTargetCount) * 100)
+          : 0,
+    }));
+  };
+
   const generateAutoRatio = () => {
     const selectedCount = selectedDatasets.length;
     if (selectedCount === 0) return;
     const baseQuantity = Math.floor(totalTargetCount / selectedCount);
     const remainder = totalTargetCount % selectedCount;
-    const newConfigs = selectedDatasets.map((datasetId, index) => {
+
+    let newConfigs: RatioConfigItem[] = ratioConfigs.filter(
+      (c) => !selectedDatasets.includes(c.source)
+    );
+
+    selectedDatasets.forEach((datasetId, index) => {
       const dataset = datasets.find((d) => String(d.id) === datasetId);
       const quantity = baseQuantity + (index < remainder ? 1 : 0);
-      return {
+      const config: RatioConfigItem = {
-        id: datasetId,
+        id: genId(datasetId),
         name: dataset?.name || datasetId,
         type: ratioType,
         quantity,
         percentage: Math.round((quantity / totalTargetCount) * 100),
         source: datasetId,
-        labelFilter: datasetFilters[datasetId]?.labelFilter,
-        dateRange: datasetFilters[datasetId]?.dateRange,
       };
+      newConfigs.push(config);
     });
+
     setRatioConfigs(newConfigs);
     onChange?.(newConfigs);
   };
 
-  // 更新数据集配比项
+  useEffect(() => {
-  const updateDatasetQuantity = (datasetId: string, quantity: number) => {
+    const keep = ratioConfigs.filter((c) =>
-    setRatioConfigs((prev) => {
+      selectedDatasets.includes(c.source)
-      const existingIndex = prev.findIndex(
-        (config) => config.source === datasetId
-      );
-      const totalOtherQuantity = prev
-        .filter((config) => config.source !== datasetId)
-        .reduce((sum, config) => sum + config.quantity, 0);
-
-      const dataset = datasets.find((d) => String(d.id) === datasetId);
-      const newConfig: RatioConfigItem = {
-        id: datasetId,
-        name: dataset?.name || datasetId,
-        type: ratioType,
-        quantity: Math.min(quantity, totalTargetCount - totalOtherQuantity),
-        percentage: Math.round((quantity / totalTargetCount) * 100),
-        source: datasetId,
-        labelFilter: datasetFilters[datasetId]?.labelFilter,
-        dateRange: datasetFilters[datasetId]?.dateRange,
-      };
-
-      let newConfigs;
-      if (existingIndex >= 0) {
-        newConfigs = [...prev];
-        newConfigs[existingIndex] = newConfig;
-      } else {
-        newConfigs = [...prev, newConfig];
-      }
-      onChange?.(newConfigs);
-      return newConfigs;
-    });
-  };
-
-  // 更新筛选条件
-  const updateFilters = (datasetId: string, updates: {
-    labelFilter?: string;
-    dateRange?: [string, string];
-  }) => {
-    setDatasetFilters(prev => ({
-      ...prev,
-      [datasetId]: {
-        ...prev[datasetId],
-        ...updates,
-      }
-    }));
-  };
-
-  // 渲染筛选器
-  const renderFilters = (datasetId: string) => {
-    const labels = getDatasetLabels(datasetId);
-    const config = ratioConfigs.find(c => c.source === datasetId);
-    const filters = datasetFilters[datasetId] || {};
-
-    return (
-      <div className="mb-3 p-2 bg-gray-50 rounded">
-        <div className="flex items-center gap-2 mb-2">
-          <Filter size={14} className="text-gray-400" />
-          <span className="text-xs font-medium">筛选条件</span>
-        </div>
-
-        <div className="grid grid-cols-1 md:grid-cols-2 gap-3">
-          <div>
-            <div className="text-xs text-gray-500 mb-1">标签筛选</div>
-            <Select
-              style={{ width: '100%' }}
-              placeholder="选择标签"
-              value={filters.labelFilter}
-              onChange={(value) => updateFilters(datasetId, { labelFilter: value })}
-              allowClear
-              onClear={() => updateFilters(datasetId, { labelFilter: undefined })}
-            >
-              {labels.map(label => (
-                <Option key={label} value={label}>{label}</Option>
-              ))}
-            </Select>
-          </div>
-
-          <div>
-            <div className="text-xs text-gray-500 mb-1">标签更新时间</div>
-            <Select
-              style={{ width: '100%' }}
-              placeholder="选择标签更新时间"
-              value={filters.dateRange}
-              onChange={(dates) => updateFilters(datasetId, { dateRange: dates })}
-              allowClear
-              onClear={() => updateFilters(datasetId, { dateRange: undefined })}
-            >
-              {TIME_RANGE_OPTIONS.map(option => (
-                <Option key={option.value} value={option.value}>
-                  {option.label}
-                </Option>
-              ))}
-            </Select>
-          </div>
-        </div>
-      </div>
     );
-  };
+    if (keep.length !== ratioConfigs.length) {
-
+      const adjusted = recomputePercentages(keep);
-  // 选中数据集变化时，初始化筛选条件
+      setRatioConfigs(adjusted);
-  React.useEffect(() => {
+      onChange?.(adjusted);
-    const initialFilters: Record<string, any> = {};
+    }
-    selectedDatasets.forEach(datasetId => {
+    // eslint-disable-next-line react-hooks/exhaustive-deps
-      const config = ratioConfigs.find(c => c.source === datasetId);
-      if (config) {
-        initialFilters[datasetId] = {
-          labelFilter: config.labelFilter,
-          dateRange: config.dateRange,
-        };
-      }
-    });
-    setDatasetFilters(prev => ({ ...prev, ...initialFilters }));
   }, [selectedDatasets]);
 
   return (
@@ -209,15 +158,18 @@ const RatioConfig: React.FC<RatioConfigProps> = ({
             (已配置:{totalConfigured}/{totalTargetCount}条)
           </span>
         </span>
-        <Button
+        <div className="flex items-center gap-2">
-          type="link"
+          <Button
-          size="small"
+            type="link"
-          onClick={generateAutoRatio}
+            size="small"
-          disabled={selectedDatasets.length === 0}
+            onClick={generateAutoRatio}
-        >
+            disabled={selectedDatasets.length === 0}
-          平均分配
+          >
-        </Button>
+            平均分配
+          </Button>
+        </div>
       </div>
+
       {selectedDatasets.length === 0 ? (
         <div className="text-center py-8 text-gray-500">
           <BarChart3 className="w-12 h-12 mx-auto mb-2 text-gray-300" />
@@ -225,7 +177,6 @@ const RatioConfig: React.FC<RatioConfigProps> = ({
         </div>
       ) : (
         <div className="flex-overflow-auto gap-4 p-4">
-          {/* 配比预览 */}
           {ratioConfigs.length > 0 && (
             <div className="p-3 bg-gray-50 rounded-lg mb-4">
               <div className="grid grid-cols-2 gap-4 text-sm">
@@ -250,54 +201,146 @@ const RatioConfig: React.FC<RatioConfigProps> = ({
           <div className="flex-1 overflow-auto space-y-4">
             {selectedDatasets.map((datasetId) => {
               const dataset = datasets.find((d) => String(d.id) === datasetId);
-              const config = ratioConfigs.find((c) => c.source === datasetId);
-              const currentQuantity = config?.quantity || 0;
-
               if (!dataset) return null;
 
+              const datasetConfigs = ratioConfigs.filter(
+                (c) => c.source === datasetId
+              );
+
+              const labels = getDatasetLabels(datasetId);
+              const usedLabels = datasetConfigs
+                .map((c) => c.labelFilter)
+                .filter(Boolean) as string[];
+
+              const columns = [
+                {
+                  title: "配比项",
+                  dataIndex: "id",
+                  key: "id",
+                  render: (_: any, record: RatioConfigItem) => (
+                    <Space>
+                      <Filter size={14} className="text-gray-400" />
+                      <span className="text-sm">{record.name}</span>
+                    </Space>
+                  ),
+                },
+                {
+                  title: "标签筛选",
+                  dataIndex: "labelFilter",
+                  key: "labelFilter",
+                  render: (_: any, record: RatioConfigItem) => {
+                    const availableLabels = labels
+                      .map((l) => ({ label: l, value: l }))
+                      .filter(
+                        (opt) =>
+                          opt.value === record.labelFilter ||
+                          !usedLabels.includes(opt.value)
+                      );
+                    return (
+                      <Select
+                        style={{ width: "160px" }}
+                        placeholder="选择标签"
+                        value={record.labelFilter}
+                        options={availableLabels}
+                        allowClear
+                        onChange={(value) =>
+                          updateConfig(record.id, {
+                            labelFilter: value || undefined,
+                          })
+                        }
+                      />
+                    );
+                  },
+                },
+                {
+                  title: "标签更新时间",
+                  dataIndex: "dateRange",
+                  key: "dateRange",
+                  render: (_: any, record: RatioConfigItem) => (
+                    <Select
+                      style={{ width: "140px" }}
+                      placeholder="选择标签更新时间"
+                      value={record.dateRange}
+                      options={TIME_RANGE_OPTIONS}
+                      allowClear
+                      onChange={(value) =>
+                        updateConfig(record.id, {
+                          dateRange: value || undefined,
+                        })
+                      }
+                    />
+                  ),
+                },
+                {
+                  title: "数量",
+                  dataIndex: "quantity",
+                  key: "quantity",
+                  render: (_: any, record: RatioConfigItem) => (
+                    <InputNumber
+                      min={0}
+                      max={Math.min(dataset.fileCount || 0, totalTargetCount)}
+                      value={record.quantity}
+                      onChange={(v) =>
+                        updateConfig(record.id, { quantity: Number(v || 0) })
+                      }
+                    />
+                  ),
+                },
+                {
+                  title: "占比",
+                  dataIndex: "percentage",
+                  key: "percentage",
+                  render: (_: any, record: RatioConfigItem) => (
+                    <div style={{ minWidth: 140 }}>
+                      <div className="text-xs mb-1">
+                        {record.percentage ?? 0}%
+                      </div>
+                      <Progress
+                        percent={record.percentage ?? 0}
+                        size="small"
+                        showInfo={false}
+                      />
+                    </div>
+                  ),
+                },
+                {
+                  title: "操作",
+                  dataIndex: "actions",
+                  key: "actions",
+                  render: (_: any, record: RatioConfigItem) => (
+                    <Button danger size="small" onClick={() => removeConfig(record.id)}>
+                      删除
+                    </Button>
+                  ),
+                },
+              ];
+
               return (
                 <Card key={datasetId} size="small" className="mb-4">
                   <div className="flex items-center justify-between mb-3">
                     <div className="flex items-center gap-2">
-                      <span className="font-medium text-sm">
+                      <span className="font-medium text-sm">{dataset.name}</span>
-                        {dataset.name}
-                      </span>
                       <Badge color="gray">{dataset.fileCount}条</Badge>
                     </div>
                     <div className="text-xs text-gray-500">
-                      {config?.percentage || 0}%
+                      {datasetConfigs.reduce((s, c) => s + (c.percentage || 0), 0)}%
                     </div>
                   </div>
 
-                  {/* 筛选条件 */}
+                  <Table
-                  {renderFilters(datasetId)}
+                    dataSource={datasetConfigs}
-
+                    columns={columns}
-                  <div className="flex items-center gap-2 mb-2">
+                    pagination={false}
-                    <span className="text-xs">数量:</span>
+                    rowKey="id"
-                    <Input
-                      type="number"
-                      value={currentQuantity}
-                      onChange={(e) =>
-                        updateDatasetQuantity(
-                          datasetId,
-                          Number(e.target.value)
-                        )
-                      }
-                      style={{ width: 100 }}
-                      min={0}
-                      max={Math.min(
-                        dataset.fileCount || 0,
-                        totalTargetCount
-                      )}
-                    />
-                    <span className="text-xs text-gray-500">条</span>
-                  </div>
-                  <Progress
-                    percent={Math.round(
-                      (currentQuantity / totalTargetCount) * 100
-                    )}
                     size="small"
+                    locale={{ emptyText: "暂无配比项，请添加" }}
                   />
+
+                  <div className="flex justify-end mt-3">
+                    <Button size="small" onClick={() => addConfig(datasetId)}>
+                      添加配比项
+                    </Button>
+                  </div>
                 </Card>
               );
             })}

frontend/src/pages/RatioTask/Detail/RatioTaskDetail.tsx

@@ -177,11 +177,6 @@ export default function RatioTaskDetail() {
         <Badge color={ratioTask.status?.color} text={ratioTask.status?.label} />
       ),
     },
-    {
-      key: "type",
-      label: "配比方式",
-      children: ratioTask.type || "未知",
-    },
     {
       key: "createdBy",
       label: "创建者",

frontend/src/pages/RatioTask/Home/RatioTask.tsx

@@ -95,12 +95,6 @@ export default function RatioTasksPage() {
         );
       },
     },
-    {
-      title: "配比方式",
-      dataIndex: "ratio_method",
-      key: "ratio_method",
-      width: 120,
-    },
     {
       title: "目标数量",
       dataIndex: "totals",

runtime/datamate-python/app/db/models/ratio_task.py

@@ -28,7 +28,6 @@ class RatioInstance(Base):
     name = Column(String(64), nullable=True, comment="名称")
     description = Column(Text, nullable=True, comment="描述")
     target_dataset_id = Column(String(64), nullable=True, comment="模板数据集ID")
-    ratio_method = Column(String(50), nullable=True, comment="配比方式，按标签（TAG）,按数据集（DATASET）")
     ratio_parameters = Column(JSON, nullable=True, comment="配比参数")
     merge_method = Column(String(50), nullable=True, comment="合并方式")
     status = Column(String(20), nullable=True, comment="状态")
@@ -39,7 +38,7 @@ class RatioInstance(Base):
     updated_by = Column(String(255), nullable=True, comment="更新者")
 
     def __repr__(self) -> str:
-        return f"<RatioInstance(id={self.id}, name={self.name}, method={self.ratio_method}, status={self.status})>"
+        return f"<RatioInstance(id={self.id}, name={self.name}, status={self.status})>"
 
 
 class RatioRelation(Base):

runtime/datamate-python/app/module/dataset/schema/dataset_file.py

@@ -27,15 +27,15 @@ class PagedDatasetFileResponse(BaseModel):
     size: int = Field(..., description="每页大小")
 
 class DatasetFileTag(BaseModel):
-    id: str = Field(..., description="标签ID")
+    id: str = Field(None, description="标签ID")
-    type: str = Field(..., description="类型")
+    type: str = Field(None, description="类型")
-    from_name: str = Field(..., description="标签名称")
+    from_name: str = Field(None, description="标签名称")
-    value: dict = Field(..., description="标签值")
+    values: dict = Field(None, description="标签值")
 
     def get_tags(self) -> List[str]:
         tags = []
-        # 如果 value 是字典类型，根据 type 获取对应的值
+        # 如果 values 是字典类型，根据 type 获取对应的值
-        tag_values = self.value.get(self.type, [])
+        tag_values = self.values.get(self.type, [])
 
         # 处理标签值
         if isinstance(tag_values, list):

runtime/datamate-python/app/module/synthesis/interface/ratio_task.py

@@ -170,7 +170,6 @@ async def list_ratio_tasks(
                     description=i.description,
                     status=i.status,
                     totals=i.totals,
-                    ratio_method=i.ratio_method,
                     target_dataset_id=i.target_dataset_id,
                     target_dataset_name=(ds.name if ds else None),
                     created_at=str(i.created_at) if getattr(i, "created_at", None) else None,
@@ -330,7 +329,6 @@ async def get_ratio_task(
                 description=instance.description,
                 status=instance.status or "UNKNOWN",
                 totals=instance.totals or 0,
-                ratio_method=instance.ratio_method or "",
                 config=config,
                 target_dataset=target_dataset_info,
                 created_at=instance.created_at,

runtime/datamate-python/app/module/synthesis/schema/ratio_task.py

@@ -88,7 +88,6 @@ class RatioTaskItem(BaseModel):
     description: Optional[str] = None
     status: Optional[str] = None
     totals: Optional[int] = None
-    ratio_method: Optional[str] = None
     target_dataset_id: Optional[str] = None
     target_dataset_name: Optional[str] = None
     created_at: Optional[str] = None
@@ -110,7 +109,6 @@ class RatioTaskDetailResponse(BaseModel):
     description: Optional[str] = Field(None, description="任务描述")
     status: str = Field(..., description="任务状态")
     totals: int = Field(..., description="目标总数")
-    ratio_method: str = Field(..., description="配比方式")
     config: List[Dict[str, Any]] = Field(..., description="配比配置")
     target_dataset: Dict[str, Any] = Field(..., description="目标数据集信息")
     created_at: Optional[datetime] = Field(None, description="创建时间")

runtime/datamate-python/app/module/synthesis/service/ratio_task.py

@@ -1,3 +1,4 @@
+from datetime import datetime
 from typing import List, Optional, Dict, Any
 import random
 import json
@@ -173,7 +174,7 @@ class RatioTaskService:
     @staticmethod
     async def handle_selected_file(existing_paths: set[Any], f, session, target_ds: Dataset):
         src_path = f.file_path
-        dst_prefix = f"/dataset/{target_ds.id}"
+        dst_prefix = f"/dataset/{target_ds.id}/"
         file_name = RatioTaskService.get_new_file_name(dst_prefix, existing_paths, f)
 
         new_path = dst_prefix + file_name
@@ -181,18 +182,20 @@ class RatioTaskService:
         await asyncio.to_thread(os.makedirs, dst_dir, exist_ok=True)
         await asyncio.to_thread(shutil.copy2, src_path, new_path)
 
-        new_file = DatasetFiles(
+        file_data = {
-            dataset_id=target_ds.id,  # type: ignore
+            "dataset_id": target_ds.id,  # type: ignore
-            file_name=file_name,
+            "file_name": file_name,
-            file_path=new_path,
+            "file_path": new_path,
-            file_type=f.file_type,
+            "file_type": f.file_type,
-            file_size=f.file_size,
+            "file_size": f.file_size,
-            check_sum=f.check_sum,
+            "check_sum": f.check_sum,
-            tags=f.tags,
+            "tags": f.tags,
-            dataset_filemetadata=f.dataset_filemetadata,
+            "tags_updated_at": datetime.now(),
-            status="ACTIVE",
+            "dataset_filemetadata": f.dataset_filemetadata,
-        )
+            "status": "ACTIVE",
-        session.add(new_file)
+        }
+        file_record = {k: v for k, v in file_data.items() if v is not None}
+        session.add(DatasetFiles(**file_record))
         existing_paths.add(new_path)
 
     @staticmethod

scripts/db/data-ratio-init.sql

@@ -6,7 +6,6 @@ CREATE TABLE IF NOT EXISTS t_st_ratio_instances
     name        varchar(64) COMMENT '名称',
     description TEXT COMMENT '描述',
     target_dataset_id     varchar(64) COMMENT '模板数据集ID',
-    ratio_method      varchar(50) COMMENT '配比方式，按标签（TAG）,按数据集（DATASET）',
     ratio_parameters     JSON COMMENT '配比参数',
     merge_method     varchar(50) COMMENT '合并方式',
     status    varchar(20) COMMENT '状态',