MongoDB聚合查询：如何调整‘今日已评测游戏’结果格式？

Absolutely, you can handle this transformation directly within MongoDB's aggregation pipeline—no need to post-process with JavaScript! Let's fix that array-wrapping issue with a couple of straightforward steps.

The core problem here is that the $lookup stage always returns an array for the game field, even when there’s only one matching document in the Game collection. That’s why you’re seeing the nested [{}] structure in your results. Here are two ways to fix this right in your pipeline:

Option 1: Unwrap the array with $unwind #

If you just want to convert the game array into a direct object (keeping the nested game key), add a $unwind stage immediately after $lookup:

db.Review.aggregate([
 { $match: { publishedDate: { $gte: cutoff } } },
 { $sortByCount: '$game.id' },
 { $limit: 10 },
 { $lookup: { from: 'Game', localField: '_id', foreignField: 'id', as: 'game' } },
 { $unwind: '$game' }, // Converts the single-element array to a plain object
 { $project: { 
    'game.id': 1, 
    'game.name': 1, 
    'game.topCriticScore': 1, 
    'game.firstReleaseDate': 1, 
    'game.tier': 1, 
    '_id': 0 
  } }
]);

This will change your output to:

[ { "game": { /* game document */ } }, { "game": { /* game document */ } }, ... ]

Option 2: Flatten the structure entirely #

If you’d prefer to have the game’s fields as top-level properties (and keep the review count from $sortByCount), use $arrayElemAt with $replaceRoot to merge the game data and review count into a single flat document:

db.Review.aggregate([
 { $match: { publishedDate: { $gte: cutoff } } },
 { $sortByCount: '$game.id' },
 { $limit: 10 },
 { $lookup: { from: 'Game', localField: '_id', foreignField: 'id', as: 'game' } },
 { $addFields: {
    game: { $arrayElemAt: ['$game', 0] }, // Extract the single game from the array
    reviewCount: '$count' // Rename the count to a more descriptive field
  } },
 { $replaceRoot: {
    newRoot: { $mergeObjects: ['$game', { reviewCount: '$reviewCount' }] } // Merge game data and count
  } },
 { $project: {
    id: 1,
    name: 1,
    topCriticScore: 1,
    firstReleaseDate: 1,
    tier: 1,
    reviewCount: 1
  } }
]);

This will give you clean, flat results like:

[
  { 
    "id": "game_123", 
    "name": "Epic Adventure Game", 
    "topCriticScore": 89, 
    "firstReleaseDate": ISODate("2024-01-15T00:00:00Z"), 
    "tier": "A", 
    "reviewCount": 7 
  },
  // More results...
]

Pick the approach that best fits your desired output structure—both work entirely within MongoDB’s aggregation framework, so you can skip the client-side mapping step.

内容的提问来源于stack exchange，提问作者MattEnth