Programmatic resolvers
While Configuration-based (declarative) resolvers explains how additionalResolvers can shape and augment the unified schema with configuration changes, programmatic resolvers shape the schema programmatically using JavaScript.
You can also use custom resolvers to batch repeated queries and get better response times. For more information, see Batching.
The additionalResolvers config allows you to upload a custom resolver as a JavaScript file to the Mesh.
Programmatic additionalResolvers
In this example, we will use additionalResolvers to apply a set of discounts to products in Adobe Commerce. The following example uses two handlers:
-
The
Veniahandler is a publicly available Adobe Commerce GraphQL endpoint for the Venia sample storefront. -
The
DiscountsAPIhandler points to a.jsonfile that contains key-value pairs of sample item names and their corresponding discount percentages.
{
"meshConfig": {
"sources": [
{
"name": "Venia",
"handler": {
"graphql": {
"endpoint": "https://venia.magento.com/graphql",
"useGETForQueries": true
}
}
},
{
"name": "DiscountsAPI",
"handler": {
"JsonSchema": {
"baseUrl": "https://raw.githubusercontent.com/AdobeDocs/graphql-mesh-gateway/refs/heads/main/static/examples",
"operations": [
{
"type": "Query",
"field": "discounts",
"path": "/discounts-api.json",
"method": "GET",
"responseSample": "https://raw.githubusercontent.com/AdobeDocs/graphql-mesh-gateway/main/static/examples/discounts-api.json"
}
]
}
}
}
],
"additionalResolvers": [
"./additional-resolvers.js"
]
}
}
Create a JavaScript file named additional-resolvers.js in the same directory as your mesh. Add the following contents to the file:
data-slots=heading, code
data-repeat=1
data-languages=js
additional-resolvers.js
module.exports = {
resolvers: {
ConfigurableProduct: {
special_price: {
selectionSet: '{ name price_range { maximum_price { final_price { value } } } }',
resolve: (root, args, context, info) => {
let max = 0;
try {
max = root.price_range.maximum_price.final_price.value;
} catch (e) {
// set a default value
max = 0;
}
return context.DiscountsAPI.Query.discounts({
root,
args,
context,
info,
selectionSet: '{ name discount }',
})
.then(response => {
let discount = 0;
const discountConfig = response.find(discount => discount.name === root.name);
if (discountConfig) {
discount = discountConfig.discount;
}
return max * ((100 - discount) / 100);
})
.catch(() => {
return null;
});
},
},
},
},
};
This javascript file targets the special_price field on ConfigurableProduct that uses the maximum_price for a product and then applies the discount listed for that product in the discountsapi.json file. The following arguments define how the custom resolver interacts with your mesh:
-
special_price- the name of the field we are adding a custom resolver to -
selectionSet- the information from the parent field,ConfigurableProduct, that thespecial_pricefield needs to resolve -
root,context,args, andinfo- are the parameters provided to the new resolver function
Running the following query results in a response that lists the original maximum_price value and the special_price that was calculated using the DiscountsAPI file. In this example, we are searching for "sweater", but you could modify it to search for any products.
In the following response, you can see that the "Roxana Cropped Sweater" and the "Hanna Sweater" we specified in our discounts-api.json file have a special_price that is 10% less than their value.
data-variant=info
data-slots=text
data-slots=heading, code
data-repeat=2
data-languages=graphql, json
GraphQL Query
{
products(search: "sweater") {
items {
name
special_price
price_range {
maximum_price {
final_price {
value
}
}
}
}
}
}
Response
{
"data": {
"products": {
"items": [
{
"name": "Juno Sweater",
"special_price": 54.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 54.4
}
}
}
},
{
"name": "Hanna Sweater",
"special_price": 70.56,
"price_range": {
"maximum_price": {
"final_price": {
"value": 78.4
}
}
}
},
{
"name": "Echo Sweater",
"special_price": 62.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 62.4
}
}
}
},
{
"name": "Corina Lace-Back Sweater",
"special_price": 86.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 86.4
}
}
}
},
{
"name": "Roxana Cropped Sweater",
"special_price": 56.16,
"price_range": {
"maximum_price": {
"final_price": {
"value": 62.4
}
}
}
},
{
"name": "Helena Cardigan",
"special_price": 78.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 78.4
}
}
}
},
{
"name": "Rosalina Cardigan",
"special_price": 78.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 78.4
}
}
}
},
{
"name": "Brigid Boucle Cardigan",
"special_price": 94.4,
"price_range": {
"maximum_price": {
"final_price": {
"value": 94.4
}
}
}
}
]
}
},
"extensions": {}
}
fetch with programmatic resolvers
Instead of adding another source handler, edge meshes can interact with third-party services using the fetch method.
The script below uses a programmatic resolver to fetch discounts from a remote server. In this case, we are using the Discounts API discussed in Programmatic additionalResolvers.
module.exports = {
resolvers: {
ConfigurableProduct: {
special_price: {
selectionSet:
"{ name price_range { maximum_price { final_price { value } } } }",
resolve: (root, args, context) => {
let max = 0;
try {
max = root.price_range.maximum_price.final_price.value;
} catch (e) {
// ignore
}
context.logger.log("Fetching discounts from remote server");
return globalThis
.fetch(
"https://raw.githubusercontent.com/AdobeDocs/graphql-mesh-gateway/main/static/examples/discounts-api.json"
)
.then((response) => {
if (response) {
return response.json();
}
return [];
})
.then((discounts) => {
const discountConfig = discounts.find(
(discount) => discount.name === root.name
);
if (discountConfig) {
return max * ((100 - discountConfig.discount) / 100);
} else {
return max;
}
})
.catch((err) => {
context.logger.log(err.message);
return err.message;
});
},
},
},
},
};
context.logger
data-variant=info
data-slots=text
context.logger is available in both declarative and programmatic resolvers.data-src=../../../../context-logger.md
Example
The following example resolver calculates a product's final price by adding a 10% markup to the base price.
module.exports = {
resolvers: {
Product: {
price: {
resolve: (root, args, context) => {
context.logger.log("Calculating product price");
try {
const price = root.basePrice * 1.1; // Add 10% markup
context.logger.log("Price calculated successfully");
return price;
} catch (error) {
context.logger.error("Price calculation failed");
return root.basePrice;
}
},
},
},
},
};