MongoTemplate返回未格式化对象问题及聚合查询代码求助

Hey there! Let's work through this MongoTemplate aggregation issue you're hitting where you're getting unformatted objects back, and fix up your method to return the List<String> of relative IDs you need.

First, Let's Diagnose the Core Issue #

Your current code cuts off at mongoTemplate.aggregate(aggregation, Master.class.getSimpleName(), ...) — the missing third parameter is the key problem here. Without specifying the correct output type, MongoTemplate can't properly map the aggregation results to a structured format, leading to those unformatted objects.

Solution 1: Map Results to Document & Extract IDs in Application Layer #

This approach uses MongoDB's Document class to directly receive the aggregation output, then extracts the merged IDs from the result documents.

import org.springframework.data.mongodb.core.aggregation.Aggregation;
import org.springframework.data.mongodb.core.aggregation.AggregationResults;
import org.springframework.data.mongodb.core.aggregation.MatchOperation;
import org.springframework.data.mongodb.core.aggregation.ProjectionOperation;
import org.bson.Document;
import java.util.ArrayList;
import java.util.List;

public List<String> fetchRelativeIds(String externalId) {
    // Projection: Merge ancestors._id and descendants._id into a single array, keep the document's _id
    ProjectionOperation projectionOperation = Aggregation.project()
            .and("ancestors._id").concatArrays("descendants._id").as("weNeedRelations")
            .andInclude("_id");

    // Match documents by the external ID
    MatchOperation filterByIds = Aggregation.match(new Criteria("externalIds.fenergo").is(externalId));

    // Build the aggregation pipeline
    Aggregation aggregation = Aggregation.newAggregation(filterByIds, projectionOperation);

    // Execute aggregation and map results to Document
    AggregationResults<Document> results = mongoTemplate.aggregate(
            aggregation, 
            Master.class.getSimpleName(), 
            Document.class
    );

    // Extract all relative IDs from the results
    List<String> relativeIds = new ArrayList<>();
    for (Document doc : results.getMappedResults()) {
        // Add merged IDs from weNeedRelations array
        List<String> mergedIds = (List<String>) doc.get("weNeedRelations");
        if (mergedIds != null) {
            relativeIds.addAll(mergedIds);
        }
        // Add the current document's own _id if needed
        String currentDocId = doc.getString("_id");
        if (currentDocId != null) {
            relativeIds.add(currentDocId);
        }
    }

    return relativeIds;
}

Solution 2: Optimize Aggregation to Return IDs Directly (Database-Side Processing) #

For better performance, we can extend the aggregation pipeline to handle merging, unwinding, and deduplicating IDs directly in MongoDB, so we get a clean list of unique IDs without extra application-layer work:

import org.springframework.data.mongodb.core.aggregation.*;
import org.bson.Document;
import java.util.Arrays;
import java.util.Collections;
import java.util.List;

public List<String> fetchRelativeIds(String externalId) {
    MatchOperation filterByIds = Aggregation.match(new Criteria("externalIds.fenergo").is(externalId));

    // Merge ancestors and descendants IDs into one array
    ProjectionOperation mergeIds = Aggregation.project()
            .and("ancestors._id").concatArrays("descendants._id").as("allIds")
            .andInclude("_id");

    // Add the current document's _id to the array
    ProjectionOperation includeCurrentId = Aggregation.project()
            .and("allIds").concatArrays(Arrays.asList("$id")).as("allIds");

    // Unwind the array to process each ID individually
    UnwindOperation unwindIds = Aggregation.unwind("allIds");

    // Group all IDs and deduplicate them
    GroupOperation groupIds = Aggregation.group().addToSet("allIds").as("uniqueRelativeIds");

    // Build and execute the full pipeline
    Aggregation aggregation = Aggregation.newAggregation(
            filterByIds, 
            mergeIds, 
            includeCurrentId, 
            unwindIds, 
            groupIds
    );

    AggregationResults<Document> results = mongoTemplate.aggregate(
            aggregation, 
            Master.class.getSimpleName(), 
            Document.class
    );

    // Extract the final list of unique IDs
    Document result = results.getUniqueMappedResult();
    return result != null ? (List<String>) result.get("uniqueRelativeIds") : Collections.emptyList();
}

Important Notes to Avoid Unformatted/Incorrect Results #

• Handle ObjectId to String Conversion: If _id fields are stored as MongoDB ObjectId instead of strings, use ConvertOperators.ToString to convert them in the projection:

  ProjectionOperation projectionOperation = Aggregation.project()
          .and(ConvertOperators.ToString.toString("$ancestors._id"))
          .concatArrays(ConvertOperators.ToString.toString("$descendants._id"))
          .as("weNeedRelations")
          .andInclude(ConvertOperators.ToString.toString("$id"));
  

• Verify Field Paths: Double-check that externalIds.fenergo, ancestors._id, and descendants._id match the actual structure of your Master collection documents — mismatched paths will return empty results.

内容的提问来源于stack exchange，提问作者Vinay Prajapati