Interroger des données dans Azure Cosmos DB for MongoDB à l’aide de JavaScript

Article
08/15/2024

S’APPLIQUE À : MongoDB

Utilisez des requêtes et des pipelines d’agrégation pour rechercher et manipuler des documents dans une collection.

Notes

Les exemples d’extraits de code sont disponibles sur GitHub sous la forme d’un projet JavaScript.

Documentation de référence de l’API pour MongoDB | Package MongoDB (npm)

Rechercher des documents

Pour rechercher des documents, utilisez une requête pour définir la façon dont les documents sont trouvés.

// assume doc exists

const product = {
    _id: ObjectId("62b1f43a9446918500c875c5"),
    category: "gear-surf-surfboards",
    name: "Yamba Surfboard 7",
    quantity: 12,
    sale: false
};


// For unsharded database: use id
const query1 = { _id: ObjectId(product._id) };
const foundById = await client.db("adventureworks").collection('products').findOne(query1);
console.log(`Read doc:\t\n${Object.keys(foundById).map(key => `\t${key}: ${foundById[key]}\n`)}`);

// For sharded database: point read doc from collection using the id and partitionKey
const query2 = { _id: ObjectId(product._id), category: product.category };
const foundByIdAndPartitionKey = await client.db("adventureworks").collection('products').findOne(query2);
console.log(`Read doc 2:\t\n${Object.keys(foundByIdAndPartitionKey).map(key => `\t${key}: ${foundByIdAndPartitionKey[key]}\n`)}`);


// Find one by unique doc property value
const query3 = { name: product.name};
const foundByUniqueValue = await client.db("adventureworks").collection('products').findOne(query3);
console.log(`Read doc 3:\t\n${Object.keys(foundByUniqueValue).map(key => `\t${key}: ${foundByUniqueValue[key]}\n`)}`);

// Find one (with many that match query) still returns one doc
const query4 = { category: product.category };
const foundByNonUniqueValue = await client.db("adventureworks").collection('products').findOne(query4);
console.log(`Read doc 4:\t\n${Object.keys(foundByNonUniqueValue).map(key => `\t${key}: ${foundByNonUniqueValue[key]}\n`)}`);

// Find all that match query
const query5 = { category: product.category };
const foundAll = await client.db("adventureworks").collection('products').find(query5).sort({_id: 1}).toArray();
console.log(`Matching all in product category:\n${foundAll.map(doc => `\t${doc._id}: ${doc.name}\n`)}`);

// Find all in collection with empty query {}
const foundAll2 = await client.db("adventureworks").collection('products').find({}).toArray();
console.log(`All docs:\n${foundAll2.map(doc => `\t${doc._id}: ${doc.name}\n`)}`);

// Pagination - next 5 docs
// sort by name require an index on name
const nextFiveDocs = await client.db("adventureworks").collection('products').find({}).sort({name: 1}).skip(5).limit(5).toArray();
console.log(`All docs:\n${foundAll2.map(doc => `\t${doc._id}: ${doc.name}\n`)}`);

L’extrait de code précédent affiche l’exemple de sortie de console suivant :

Read doc:
        _id: 62b1f43a9446918500c875c5
,       name: Yamba Surfboard-13
,       category: gear-surf-surfboards
,       quantity: 20
,       sale: false
,       discontinued: true

Read doc 2:
        _id: 62b1f43a9446918500c875c5
,       name: Yamba Surfboard-13
,       category: gear-surf-surfboards
,       quantity: 20
,       sale: false
,       discontinued: true

Read doc 3:
        _id: 62b23a371a09ed6441e5ee31
,       category: gear-surf-surfboards
,       name: Yamba Surfboard 7
,       quantity: 5
,       sale: true
,       discontinued: true

Read doc 4:
        _id: 62b1f43a9446918500c875c5
,       name: Yamba Surfboard-13
,       category: gear-surf-surfboards
,       quantity: 20
,       sale: false
,       discontinued: true

Matching all in product category:
        62b1f43a9446918500c875c5: Yamba Surfboard-13
,       62b1f4670c7af8c2942b7c10: Yamba Surfboard-3
,       62b1f46fa6546d2afb5715ac: Yamba Surfboard-90
,       62b1f474e4b43498c05d295b: Yamba Surfboard-9

All docs:
        62b1f43a9446918500c875c5: Yamba Surfboard-13
,       62b1f4670c7af8c2942b7c10: Yamba Surfboard-3
,       62b1f46fa6546d2afb5715ac: Yamba Surfboard-90
,       62b1f474e4b43498c05d295b: Yamba Surfboard-9
,       62b1f47896aa8cfa280edf2d: Yamba Surfboard-55
,       62b1f47dacbf04e86c8abf25: Yamba Surfboard-11
,       62b1f4804ee53f4c5c44778c: Yamba Surfboard-97
,       62b1f492ff69395b30a03169: Yamba Surfboard-93
,       62b23a371a09ed6441e5ee30: Yamba Surfboard 3
,       62b23a371a09ed6441e5ee31: Yamba Surfboard 7

All docs:
        62b1f43a9446918500c875c5: Yamba Surfboard-13
,       62b1f4670c7af8c2942b7c10: Yamba Surfboard-3
,       62b1f46fa6546d2afb5715ac: Yamba Surfboard-90
,       62b1f474e4b43498c05d295b: Yamba Surfboard-9
,       62b1f47896aa8cfa280edf2d: Yamba Surfboard-55
,       62b1f47dacbf04e86c8abf25: Yamba Surfboard-11
,       62b1f4804ee53f4c5c44778c: Yamba Surfboard-97
,       62b1f492ff69395b30a03169: Yamba Surfboard-93
,       62b23a371a09ed6441e5ee30: Yamba Surfboard 3
,       62b23a371a09ed6441e5ee31: Yamba Surfboard 7

done

Pipelines d’agrégation

Les pipelines d’agrégation sont utiles pour isoler sur votre serveur Azure Cosmos DB les calculs de requête, les transformations et d’autres traitements coûteux, au lieu d’effectuer ces opérations sur le client.

Pour obtenir une prise en charge spécifique du pipeline d’agrégation, reportez-vous à ce qui suit :

Syntaxe des pipelines d’agrégation

Un pipeline est un tableau comportant une série de phases représentées par des objets JSON.

const pipeline = [
    stage1,
    stage2
]

Syntaxe des phases de pipeline

Une phase définit l’opération ainsi que les données auxquelles elle s’applique, par exemple :

$match : rechercher des documents
$addFields : ajouter un champ au curseur, provenant généralement de l’étape précédente
$limit : limiter le nombre de résultats retournés dans le curseur
$project : passer des champs nouveaux ou existants (peuvent être des champs calculés)
$group : regrouper les résultats dans un ou plusieurs champs du pipeline
$sort : trier les résultats

// reduce collection to relative documents
const matchStage = {
    '$match': {
        'categoryName': { $regex: 'Bikes' },
    }
}

// sort documents on field `name`
const sortStage = { 
    '$sort': { 
        "name": 1 
    } 
},

Agréger le pipeline pour obtenir un curseur itérable

Le pipeline est agrégé pour produire un curseur itérable.

const db = 'adventureworks';
const collection = 'products';

const aggCursor = client.db(databaseName).collection(collectionName).aggregate(pipeline);

await aggCursor.forEach(product => {
    console.log(JSON.stringify(product));
});

Utiliser un pipeline d’agrégation en JavaScript

Utilisez un pipeline pour maintenir le traitement des données sur le serveur avant de revenir au client.

Exemple de données de produit

Les agrégations ci-dessous utilisent l’exemple de regroupement de produits avec des données ayant la forme suivante :

[
    {
        "_id": "08225A9E-F2B3-4FA3-AB08-8C70ADD6C3C2",
        "categoryId": "75BF1ACB-168D-469C-9AA3-1FD26BB4EA4C",
        "categoryName": "Bikes, Touring Bikes",
        "sku": "BK-T79U-50",
        "name": "Touring-1000 Blue, 50",
        "description": "The product called \"Touring-1000 Blue, 50\"",
        "price": 2384.0700000000002,
        "tags": [
        ]
    },
    {
        "_id": "0F124781-C991-48A9-ACF2-249771D44029",
        "categoryId": "56400CF3-446D-4C3F-B9B2-68286DA3BB99",
        "categoryName": "Bikes, Mountain Bikes",
        "sku": "BK-M68B-42",
        "name": "Mountain-200 Black, 42",
        "description": "The product called \"Mountain-200 Black, 42\"",
        "price": 2294.9899999999998,
        "tags": [
        ]
    },
    {
        "_id": "3FE1A99E-DE14-4D11-B635-F5D39258A0B9",
        "categoryId": "26C74104-40BC-4541-8EF5-9892F7F03D72",
        "categoryName": "Components, Saddles",
        "sku": "SE-T924",
        "name": "HL Touring Seat/Saddle",
        "description": "The product called \"HL Touring Seat/Saddle\"",
        "price": 52.640000000000001,
        "tags": [
        ]
    },
]

Exemple 1 : Sous-catégories de produits, nombre de produits et prix moyen

Utilisez l’exemple de code suivant pour indiquer le prix moyen dans chaque sous-catégorie de produit.

// Goal: Find the average price of each product subcategory with 
// the number of products in that subcategory.
// Sort by average price descending.

// Read .env file and set environment variables
require('dotenv').config();

// Use official mongodb driver to connect to the server
const { MongoClient } = require('mongodb');

// New instance of MongoClient with connection string
// for Cosmos DB
const url = process.env.COSMOS_CONNECTION_STRING;
const client = new MongoClient(url);

async function main() {

  try {

    // Use connect method to connect to the server
    await client.connect();

    // Group all products by category
    // Find average price of each category
    // Count # of products in each category
    const groupByCategory = {
      '$group': {
        '_id': '$categoryName',
        'averagePrice': {
          '$avg': '$price'
        },
        'countOfProducts': {
          '$sum': 1
        }
      },
    };

    // Round price to 2 decimal places
    // Don't return _id
    // Rename category name help in `_id` to `categoryName`
    // Round prices to 2 decimal places
    // Rename property for countOfProducts to nProducts
    const additionalTransformations = {
      '$project': {
        '_id': 0,
        'category': '$_id',
        'nProducts':'$countOfProducts',
        'averagePrice': { '$round': ['$averagePrice', 2] }
      }
    };

    // Sort by average price descending
    const sort = { '$sort': { '$averagePrice': -1 } };

    // stages execute in order from top to bottom
    const pipeline = [
      groupByCategory,
      additionalTransformations,
      sort
    ];

    const db = 'adventureworks';
    const collection = 'products';

    // Get iterable cursor
    const aggCursor = client.db(db).collection(collection).aggregate(pipeline);

    // Display each item in cursor
    await aggCursor.forEach(product => {
      console.log(JSON.stringify(product));
    });

    return 'done';
  } catch (err) {
    console.log(JSON.stringify(err));
  }
}

main()
  .then(console.log)
  .catch(console.error)
  .finally(() => {
    // Close the db and its underlying connections
    client.close()
  });

// Results:
// {"averagePrice":51.99,"category":"Clothing, Jerseys","nProducts":8}
// {"averagePrice":1683.36,"category":"Bikes, Mountain Bikes","nProducts":32}
// {"averagePrice":1597.45,"category":"Bikes, Road Bikes","nProducts":43}
// {"averagePrice":20.24,"category":"Components, Chains","nProducts":1}
// {"averagePrice":25,"category":"Accessories, Locks","nProducts":1}
// {"averagePrice":631.42,"category":"Components, Touring Frames","nProducts":18}
// {"averagePrice":9.25,"category":"Clothing, Socks","nProducts":4}
// {"averagePrice":125,"category":"Accessories, Panniers","nProducts":1}
// ... remaining fields ...

Exemple 2 : Types de vélos avec gamme de prix

Utilisez l’exemple de code suivant pour créer des rapports à partir de la sous-catégorie Bikes.

// Goal: Find the price range for the different bike subcategories. 

// Read .env file and set environment variables
require('dotenv').config();

// Use official mongodb driver to connect to the server
const { MongoClient } = require('mongodb');

// New instance of MongoClient with connection string
// for Cosmos DB
const url = process.env.COSMOS_CONNECTION_STRING;
const client = new MongoClient(url);

async function main() {

  try {

    // Use connect method to connect to the server
    await client.connect();

    const categoryName = 'Bikes';

    const findAllBikes = {
      '$match': {
        'categoryName': { $regex:  categoryName},
      }
    };

    // Convert 'Bikes, Touring Bikes' to ['Bikes', 'Touring Bikes']
    const splitStringIntoCsvArray = {
      $addFields: {
        'categories': { '$split': ['$categoryName', ', '] }
      }
    };

    // Remove first element from array
    // Converts ['Bikes', 'Touring Bikes'] to ['Touring Bikes']
    const removeFirstElement = {
      $addFields: {
        'subcategory': { '$slice': ['$categories', 1, { $subtract: [{ $size: '$categories' }, 1] }] }
      }
    }

    // Group items by book subcategory, and find min, avg, and max price
    const groupBySubcategory = {
      '$group': {
        '_id': '$subcategory',
        'maxPrice': {
          '$max': '$price'
        },
        'averagePrice': {
          '$avg': '$price'
        },
        'minPrice': {
          '$min': '$price'
        },
        'countOfProducts': {
          '$sum': 1
        }
      },
    };

    // Miscellaneous transformations
    // Don't return _id
    // Convert subcategory from array of 1 item to string in `name`
    // Round prices to 2 decimal places
    // Rename property for countOfProducts to nProducts
    const additionalTransformations = {
      '$project': {
        '_id': 0,
        'name': { '$arrayElemAt': ['$_id', 0]},
        'nProducts': '$countOfProducts',
        'min': { '$round': ['$minPrice', 2] },
        'avg': { '$round': ['$averagePrice', 2] },
        'max': { '$round': ['$maxPrice', 2] }
      }
    };

    // Sort by subcategory
    const sortBySubcategory = { '$sort': 
        { 'name': 1 } 
    };

    // stages execute in order from top to bottom
    const pipeline = [
      findAllBikes,
      splitStringIntoCsvArray,
      removeFirstElement,
      groupBySubcategory,
      additionalTransformations,
      sortBySubcategory
    ];

    const db = 'adventureworks';
    const collection = 'products';

    // Get iterable cursor
    const aggCursor = client.db(db).collection(collection).aggregate(pipeline);

    // Display each item in cursor
    await aggCursor.forEach(product => {
      console.log(JSON.stringify(product));
    });

    return 'done';
  } catch (err) {
    console.log(JSON.stringify(err));
  }
}

main()
  .then(console.log)
  .catch(console.error)
  .finally(() => {
    // Close the db and its underlying connections
    client.close();
  });

// Results: 
// {'name':'Mountain Bikes','nProducts':32,'min':539.99,'avg':1683.37,'max':3399.99}
// {'name':'Road Bikes','nProducts':43,'min':539.99,'avg':1597.45,'max':3578.27}
// {'name':'Touring Bikes','nProducts':22,'min':742.35,'avg':1425.25,'max':2384.07}

Partager via