fix the ratio task config (#224)

* fix: fix the dataset card icon

* fix: fix the dataset file tag distribution and ratio task

* refactor: change dateRange config from latest to start-end

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

@@ -12,20 +12,25 @@ class LabelFilter(BaseModel):
     value: Optional[str] = Field(None, description="标签值")
 
 class FilterCondition(BaseModel):
-    date_range: Optional[str] = Field(None, description="数据范围", alias="dateRange")
+    date_range: Optional[List[str]] = Field(None, description="数据范围", alias="dateRange")
     label: Optional[LabelFilter] = Field(None, description="标签")
 
     @field_validator("date_range")
     @classmethod
-    def validate_date_range(cls, v: Optional[str]) -> Optional[str]:
-        # ensure it's a numeric string if provided
-        if not v:
-            return v
+    def validate_date_range(cls, date_range: Optional[List[str]]) -> Optional[List[str]]:
+        # ensure it's a date range if provided
+        if not date_range or len(date_range) == 0:
+            return date_range
+        if len(date_range) != 2:
+            raise ValueError("date_range must be a list of two date strings: [start, end]")
         try:
-            int(v)
-            return v
+            start = datetime.fromisoformat(date_range[0])
+            end = datetime.fromisoformat(date_range[1])
+            if start > end:
+                raise ValueError("date_range start must be earlier than or equal to end")
+            return date_range
         except (ValueError, TypeError) as e:
-            raise ValueError("date_range must be a numeric string")
+            raise ValueError("date_range items must be ISO date strings (e.g. YYYY-MM-DD)")
 
     class Config:
         # allow population by field name when constructing model programmatically

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

@@ -271,14 +271,13 @@ class RatioTaskService:
         logger.info(f"start filter file: {file}, conditions: {conditions}")
 
         # Check data range condition if provided
-        if conditions.date_range:
+        if conditions.date_range and len(conditions.date_range) == 2:
             try:
                 from datetime import datetime, timedelta
-                data_range_days = int(conditions.date_range)
-                if data_range_days > 0:
-                    cutoff_date = datetime.now() - timedelta(days=data_range_days)
-                    if file.tags_updated_at and file.tags_updated_at < cutoff_date:
-                        return False
+                start_at = datetime.fromisoformat(conditions.date_range[0])
+                end_at = datetime.fromisoformat(conditions.date_range[1])
+                if file.tags_updated_at and (file.tags_updated_at < start_at or file.tags_updated_at > end_at):
+                    return False
             except (ValueError, TypeError) as e:
                 logger.warning(f"Invalid data_range value: {conditions.date_range}", e)
                 return False
@@ -294,7 +293,7 @@ class RatioTaskService:
                 for tag in all_tags:
                     if conditions.label.label and tag.get("label") != conditions.label.label:
                         continue
-                    if conditions.label.value is None:
+                    if conditions.label.value is None or len(conditions.label.value) == 0:
                         return True
                     if tag.get("value") == conditions.label.value:
                         return True