Amazon Selling Partner API client for Node version 8x
npm install amazon-sp-api-node8Just don't ask why!
The client handles calls to the Amazon Selling Partner API. It wraps up all the necessary stuff such as requesting access tokens and providing (a simplified!) way of calling the API, but also provides some convenience, i.e. a wrapper for requesting and downloading reports and an internal handling of rate limits when calls are throttled.
``diff`
- Please note: There are a few breaking changes if you are
- upgrading from version 0.x.x to 1.x.x. Please see the link below.
List of breaking changes when upgrading to version 1.x.x
- Prerequisites
- Installation
- Getting Started
- Setting credentials from environment variables
- Setting credentials from file
- Setting credentials from constructor config object
- Usage
- Config params
- Use a proxy agent
- Exchange an authorization code for a refresh token
- Request access token
- Call the API
- Examples
- Endpoints
- Versions
- Version specific operation implementations
- Defining endpoints versions on class level
- Fallback
- Unsupported endpoints/versions/operations
- Grantless operations
- Restore rates
- Timeouts
- Download reports
- Upload feeds
- TypeScript Support
- Sandbox mode
- Known Issues
- Seller Support
- Breaking Changes
Make sure that you followed the Selling Partner API Developer Guide and have successfully completed the steps Registering as a developer, Registering your application and have a valid refresh token (if you use the client only for your own seller account the easiest way is using the self authorization as described in the developer guide).
`bash`
npm install amazon-sp-api
Before you can use the client you need to add your app client and secret.
- SELLING_PARTNER_APP_CLIENT_ID=SELLING_PARTNER_APP_CLIENT_SECRET
- =
Instead of setting the credentials via environment variables you may load them from a credentials file. The default path to the file is ~/.amzspapi/credentials (path can be changed when creating a client) and you add the credentials one per line:
`bash`
SELLING_PARTNER_APP_CLIENT_ID=
SELLING_PARTNER_APP_CLIENT_SECRET=
Although the most convenient and recommended way of setting the credentials is via environment variables or config file it is also possible to pass the credentials inside the config object when creating an instance of the client (i.e. if you have no means of using env vars or a config file). The structure of the constructor config object will be explained below.
Require library:
`javascript
// commonjs
const SellingPartner = require("amazon-sp-api");
// esm
import { SellingPartner } from "amazon-sp-api";
`
Create client and call API:
`javascript`
(async () => {
try {
const spClient = new SellingPartner({
region: "eu", // The region to use for the SP-API endpoints ("eu", "na" or "fe")
refresh_token: "
});
let res = await spClient.callAPI({
operation: "getMarketplaceParticipations",
endpoint: "sellers"
});
console.log(res);
} catch (e) {
console.log(e);
}
})();
The class constructor takes a config object with the following structure as input:
`javascript`
{
region:'
refresh_token:'
access_token:'
endpoints_versions:{
...
},
credentials:{
SELLING_PARTNER_APP_CLIENT_ID:'
SELLING_PARTNER_APP_CLIENT_SECRET:'
},
options:{
credentials_path:'~/.amzspapi/credentials',
auto_request_tokens:true,
auto_request_throttled:true,
version_fallback:true,
use_sandbox:false,
only_grantless_operations:false,
user_agent:'amazon-sp-api/
debug_log:false,
timeouts:{
...
},
retry_remote_timeout:true,
https_proxy_agent:
}
}
Valid properties of the config object:
| Name | Type | Default | Description |
| :----------------------------------- | :----: | :-----: | :---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| region
_required_ | string | - | The region to use for the SP-API endpoints.
Must be one of: eu, na or fe |only_grantless_operations
| refresh_token
_optional_ | string | - | The refresh token of your app user.
Required if option is set to false. |"reports":"2021-06-30"
| access_token
_optional_ | string | - | The temporary access token requested with the refresh token of the app user. |
| endpoints_versions
_optional_ | object | - | Defines the version to use for an endpoint as key/value pairs, i.e. . If none given the client is using the first (meaning the oldest) version for an endpoint..endpoints
Call on class instance to retrieve a complete list of all endpoints, versions and operations supported by the client. |SELLING_PARTNER_APP_CLIENT_ID
| credentials
_optional_ | object | - | The app client credentials. Must include the two credentials properties and SELLING_PARTNER_APP_CLIENT_SECRET
NOTE: Should only be used if you have no means of using environment vars or credentials file! |
| options
_optional_ | object | - | Additional options, see table below for all possible options properties. |
Valid properties of the config options:
| Name | Type | Default | Description |
| :------------------------------------------ | :-----: | :---------------------------------------------------------------------------------------------------: | -------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| credentials_path
_optional_ | string | ~/.amzspapi/credentials | A custom absolute path to your credentials file location. |
| auto_request_tokens
_optional_ | boolean | true | Whether or not the client should retrieve new access token if non given or expired. |
| auto_request_throttled
_optional_ | boolean | true | Whether or not the client should automatically retry a request when throttled. |
| version_fallback
_optional_ | boolean | true | Whether or not the client should try to use an older version of an endpoint if the operation is not defined for the desired version. |
| use_sandbox
_optional_ | boolean | false | Whether or not to use the sandbox endpoint. |
| only_grantless_operations
_optional_ | boolean | false | Whether or not to only use grantless operations. |
| user_agent
_optional_ | string | amazon-sp-api/
| debug_log
_optional_ | boolean | false | Whether or not the client should print console logs for debugging purposes. |
| timeouts
_optional_ | object | - | Allows to set timeouts for requests. Valid keys are response, idle and deadline. Please see detailed information in the Timeouts section. |
| retry_remote_timeout
_optional_ | boolean | true | Whether or not the client should retry a request to the remote server that failed with an ETIMEDOUT error |
| https_proxy_agent
_optional_ | object | - | Possibility to add your own HTTPS Proxy Agent. Please see detailed information in the Using Proxy Agent section. |
If you are behind a firewall and would like to use a proxy server then you can pass a custom proxy agent to the options object. See the following example:
`javascript`
const { HttpsProxyAgent } = require("hpagent");
const agent = new HttpsProxyAgent({ proxy: "http://x.x.x.x:zzzz" });
const spClient = new SellingPartner({
region: "eu",
refresh_token: "
options: {
https_proxy_agent: agent
}
});
If you already have a refresh token you can skip this step. If you only want to use the API for your own seller account you can just use the self authorization to obtain a valid refresh token.
If you want to exchange an authorization code of a seller you can use the .exchange() function of the client. The neccessary authorization code is returned to your callback URI as spapi_oauth_code when a seller authorizes your application (see authorization workflow in docs).
Once you have obtained the authorization_code you can exchange it for a refresh token:
`javascript`
const spClient = new SellingPartner({
region: "eu",
options: {
only_grantless_operations: true
}
});
let res = await spClient.exchange("
console.log(res.refresh_token);
NOTE: You will have to create a new class instance once you have obtained the refresh_token and pass it inside the constructor in order to make calls to the API.
If you only provide the region and refresh_token parameters the client will automatically request an access_token for you (with a TTL of 1 hour) and reuse it for future api calls for the class instance.
Instead of having the client handle the access_token requests automatically, you may also refresh them manually:
`javascript`
const spClient = new SellingPartner({
region: "eu",
refresh_token: "
options: {
auto_request_tokens: false
}
});
await spClient.refreshAccessToken();
If you want to use the same credentials for multiple instances you can retrieve them via getters and use them as input for a new instance:
`javascript
let access_token = spClient.access_token;
const spClient = new SellingPartner({
region: "eu",
refresh_token: "
access_token: access_token
});
`
All calls to the SP-API will be triggered by using the .callAPI() function, which takes an object with the following structure as input:
`javascript`
{
operation:'
endpoint:'
path:{
...
},
query:{
...
},
body:{
...
},
api_path:'
method:'GET',
headers:{
...
},
restricted_data_token:'
options:{
version:'
restore_rate:'
raw_result:false,
timeouts:{
...
}
}
}
Valid properties of the object:
| Name | Type | Default | Description |
| :-------------------------------------- | :----: | :-----: | :---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| operation
_optional_ | string | - | The operation you want to request, see SP API Developer Guide.
May also include endpoint as shorthand dot notation.
Call .endpoints on class instance to retrieve a complete list of all endpoints, versions and operations supported by the client.api_path
Required if is not defined. |.endpoints
| endpoint
_optional_ | string | - | The endpoint of the operation, (see Endpoints).
Call on class instance to retrieve a complete list of all endpoints, versions and operations supported by the client.operation
Required if endpoint is not part of as shorthand dot notation and api_path is not defined. |operation
| path
_optional_ | object | - | The input paramaters added to the path of the operation. |
| query
_optional_ | object | - | The input paramaters added to the query string of the operation. |
| body
_optional_ | object | - | The input paramaters added to the body of the operation. |
| api_path
_optional_ | string | - | The full api path of an operation. Can be used to call operations that are not yet supported or have a new version that is not yet supported by the client.
Required if is not defined. |api_path
| method
_optional_ | string | - | The HTTP method to use.
Required only if is defined.GET
Must be one of: , POST, PUT,DELETE or PATCH. |createRestrictedDataToken
| headers
_optional_ | object | - | Additional headers that will be added to the call. |
| restricted_data_token
_optional_ | string | - | A token received from a operation. Neccessary to include PII (Personally Identifiable Informaton) for some restricted operations, see Tokens API use case guide for a list of restricted operations.
NOTE: Your developer account must be approved for PII by Amazon in order to be able to receive PII, otherwise the token will have no effect, meaning the result of restricted operations will not include PII. |
| options
_optional_ | object | - | Additional options, see table below for all possible options properties. |
Valid properties of the config options:
| Name | Type | Default | Description |
| :----------------------------- | :-----: | :-----: | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| version
_optional_ | string | - | The endpoint's version that should be used when calling the operation. Will be preferred over an endpoints_versions setting.version_fallback
NOTE: The call might still use an older version of the endpoint if the operation is not available for the specified version and is set to true. |raw_result
| restore_rate
_optional_ | number | - | The restore rate (in seconds) that should be used when calling the operation. Will be preferred over the default restore rate of the operation. |
| raw_result
_optional_ | boolean | false | Whether or not the client should return the "raw" result, which will include the raw body, buffer chunks, statuscode and headers of the result and also the request object with method, hostname, path, headers and body of the request.
NOTE: Setting to true will skip the internal formatting or error checking, but might be helpful when you need additional information besides the payload or when the client encounters JSON.parse errors. |response
| timeouts
_optional_ | object | - | Allows to set timeouts for requests. Valid keys are , idle and deadline. Please see detailed information in the Timeouts section. |
To call an operation of an API endpoint you pass in the operation and the endpoint it belongs to. See the following example:
`javascript`
let res = await spClient.callAPI({
operation: "getMarketplaceParticipations",
endpoint: "sellers"
});
Instead of using the endpoint property you may also prepend the endpoint to the operation as shorthand dot notation:
`javascript`
let res = await spClient.callAPI({
operation: "sellers.getMarketplaceParticipations"
});
Here are a few examples that use some more properties:
`javascript`
let res = await spClient.callAPI({
operation: "getOrderMetrics",
endpoint: "sales",
query: {
marketplaceIds: ["A1PA6795UKMFR9"],
interval: "2020-10-01T00:00:00-07:00--2020-10-01T20:00:00-07:00",
granularity: "Hour"
}
});
`javascript`
let res = await spClient.callAPI({
operation: "catalogItems.getCatalogItem",
path: {
asin: "B084J4QQFT"
},
query: {
marketplaceIds: ["A1PA6795UKMFR9"]
},
options: {
version: "2022-04-01"
}
});
`javascript`
let res = await spClient.callAPI({
operation: "createReport",
endpoint: "reports",
body: {
reportType: "GET_FLAT_FILE_OPEN_LISTINGS_DATA",
marketplaceIds: ["A1PA6795UKMFR9"]
}
});
`javascript`
let res = await spClient.callAPI({
operation: "finances.listFinancialEvents",
query: {
PostedAfter: "2020-03-01T00:00:00-07:00",
PostedBefore: "2020-03-02T00:00:00-07:00"
},
options: {
raw_result: true
}
});
`javascript`
try {
let res = await spClient.callAPI({
operation: "getCompetitivePricing",
endpoint: "productPricing",
query: {
Asins: ["B00Z7T970I", "B01BHHE9VK"],
ItemType: "Asin",
MarketplaceId: "A1PA6795UKMFR9"
},
options: {
version: "v0",
raw_result: true,
timeouts: {
response: 5000,
idle: 10000,
deadline: 30000
}
}
});
} catch (err) {
if (err.code) {
if (err.code === "API_RESPONSE_TIMEOUT")
console.log(
"SP-API ERROR: response timeout: " + err.timeout + "ms exceeded.",
err.message
);
if (err.code === "API_IDLE_TIMEOUT")
console.log(
"SP-API ERROR: idle timeout: " + err.timeout + "ms exceeded.",
err.message
);
if (err.code === "API_DEADLINE_TIMEOUT")
console.log(
"SP-API ERROR: deadline timeout: " + err.timeout + "ms exceeded.",
err.message
);
}
}
The exact endpoint's name of an operation will be the references name (see SP API Developer Guide) without API and all spaces removed and continued with a capital letter. So the Catalog Items API endpoint's name will be catalogItems, Fulfillment Inbound API will be fulfillmentInbound, Sellers API will be sellers and so on. You can also retrieve the endpoint names and their operations and versions by calling spClient.endpoints.
Every operation belongs to an endpoint that consists of one or more versions and each version consists of one or more operations. You will find a complete list of the endpoints with all versions and operations in the SP API Developer Guide. For a complete list of all currently by the client supported endpoints with versions and operations you can just call spClient.endpoints.
#### Version specific operation implementations
The client uses the first (in fact the oldest) endpoint version if no version is provided since new versions of some operations are not backward compatible. So in order to prevent breaking changes we can't enable latest endpoint versions by default. I.e. the two different implementations of the getCatalogItem operation (see catalogItemsV0 vs. catalogItems_2020-12-01) expect different input parameters and return different results.
The implementation of the getCatalogItem operation in the v0 version expects an asin and a MarketplaceId as input:
`javascript`
let res = await spClient.callAPI({
operation: "getCatalogItem",
endpoint: "catalogItems",
query: {
MarketplaceId: "A1PA6795UKMFR9"
},
path: {
asin: "B084DWG2VQ"
},
options: {
version: "v0"
}
});
In contrast, the implementation of the getCatalogItem operation in the 2020-12-01 version expects an asin, a marketplaceIds array and an includedData array as input:
`javascript`
let res = await spClient.callAPI({
operation: "getCatalogItem",
endpoint: "catalogItems",
query: {
marketplaceIds: ["A1PA6795UKMFR9"],
includedData: [
"identifiers",
"images",
"productTypes",
"salesRanks",
"summaries",
"variations"
]
},
path: {
asin: "B084DWG2VQ"
},
options: {
version: "2020-12-01"
}
});
Trying to call the new 2020-12-01 version without explicitly setting it would result in an InvalidInput error as the required MarketplaceId parameter is missing.
#### Defining endpoints versions on class level
There are different ways of specifying the version to use for endpoints and their corresponding operations. You can specify the version directly inside the options object of the .callAPI() function as seen in the examples above. But you can also enable a newer version for all operations of an endpoint by using the endpoints_versions setting in the constructor config object.2020-12-01
I.e. you can tell the class instance to use the new version for the catalogItems endpoint and thus enabling it for all operations of the endpoint throughout the class instance like this:
`javascript`
const spClient = new SellingPartner({
region: "eu",
refresh_token: "
endpoints_versions: {
catalogItems: "2020-12-01"
}
});
By doing so you can skip setting the version inside the options object each time when you are using .callAPI() with the new version of the getCatalogItem operation.
#### Fallback
If trying to call an operation that is not part of the endpoint's version you specified, the client will automatically try to find the operation in an earlier endpoint's version and use that implementation if version_fallback is set to true (which is the default).listCatalogCategories
I.e. the operation is not part of the new catalogItems endpoint version. So if the new version was set as in the example code above, the following call would still work, because it will automatically fallback to the operation's implementation in version v0:
`javascript`
let res = await spClient.callAPI({
operation: "listCatalogCategories",
endpoint: "catalogItems",
query: {
MarketplaceId: "A1PA6795UKMFR9",
ASIN: "B084DWG2VQ"
}
});
The newest client version should always have full support for all endpoints, versions and operations on release, however it might lack support for very recently added new endpoints, versions or operations. If you need an endpoint/version/operation that is not yet supported you can still call it by using the api_path parameter. I.e. if the new catalogItems version 2020-12-01 would not be supported yet we could still use the new implementation of the getCatalogItem operation by using the api_path and method properties:
`javascript`
let res = await spClient.callAPI({
api_path: "/catalog/2020-12-01/items/B084DWG2VQ",
method: "GET",
query: {
marketplaceIds: ["A1PA6795UKMFR9"],
includedData: [
"identifiers",
"images",
"productTypes",
"salesRanks",
"summaries",
"variations"
]
}
});
NOTE: If your api_path includes special characters that require encoding (i.e. an SKU that contains UTF-8 characters) you will have to encode these characters manually before passing your api_path to .callAPI().
Some operations don't require an explicit authorization by a seller, see list of grantless operations. A grantless operation needs another access token than other operations and as such a grantless token is NOT the access_token you can provide in the constructor config object. However if the auto_request_tokens option is set to true the client should handle everything for you.
If you do the token request manually you need to create a grantless token by calling refreshAccessToken with the scope of the corresponding endpoint. Currently there are only two different scopes: sellingpartnerapi::notifications for notifications endpoint and sellingpartnerapi::client_credential:rotation for application management endpoint.
If you don't need or have a refresh token you may use the client with the only_grantless_operations option set to true which allows you to create an instance of the client without a refresh_token.
To sum up, please see the following example that will return the destinations for your notifications.
First create a class instance that only allows to call grantless operations (no refresh_token included):
`javascript`
const spClient = new SellingPartner({
region: "eu",
options: {
auto_request_tokens: false,
only_grantless_operations: true
}
});
Then request a grantless token with the scope needed for the operation you want to call:
`javascript`
await spClient.refreshAccessToken("sellingpartnerapi::notifications");
Finally call the grantless operation:
`javascript`
let res = await spClient.callAPI({
operation: "getDestinations",
endpoint: "notifications"
});
If you set the auto_request_throttled option in the class constructor config object to true (which is the default), the client will automatically retry the call if its throttled. It will either use the restore rate from the result header field x-amzn-ratelimit-limit if given (see Usage Plans and Rate Limits), or the value of restore_rate option in .callAPI() function if given, or otherwise use the default restore rate of the operation. For testing purposes you can also set debug_log to true, which will trigger a console log every time the client retries a call. If you set auto_request_throttled to false the client will throw a QuotaExceeded error when a request is throttled.
NOTE: If you are using the same operation with the same seller account across multiple class instances the restore rate logic might NOT work correct or, even worse, result in an infinite quota exceeded loop. So if you're planning to do that you should probably set auto_request_throttled to false, catch the QuotaExceeded errors and handle the restore rate logic on your own.
You may set timeouts to stop requests, i.e. to prevent scripts from "hanging" forever because a request is not finishing. The three different timeout types are response, idle and deadline. You may set these inside the class constructor config options to be used for all requests started via .callAPI() or via the config options of the .callAPI() method for that specific call only. The latter will override the timeouts set via class config options.
NOTE:
The .download() method will NOT use the timeouts defined in class constructor config options. You have to provide the timeouts to each .download() call inside its options object.
See the table below for valid properties of the timeouts object:
| Name | Type | Default | Description |
| :------------------------- | :----: | :-----: | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| response
_optional_ | number | - | Timeout (in milliseconds) until a response timeout is fired. If exceeded the request will abort with an API_RESPONSE_TIMEOUT error. Response timeout is the time between sending the request and receiving the first byte of the response. Includes DNS and connection time. |API_IDLE_TIMEOUT
| idle
_optional_ | number | - | Timeout (in milliseconds) until an idle timeout is fired. if exceeded the request will abort with an error. Idle is the time between receiving the last chunk of the reponse and waiting for the next chunk to be received. Might be fired if a request is stalled before finished (i.e. when internet connection is lost). |API_DEADLINE_TIMEOUT
| deadline
_optional_ | number | - | Timeout (in milliseconds) until a deadline timeout is fired. If exceeded the request will abort with an error. Deadline is the time from the start of the request to receiving the response body in full. If the deadline is too short large responses may not load at all on slow connections. |
The easiest way of downloading a report is to use the .downloadReport() function that will wrap up all operations needed to request and retrieve a report in a single call. The function internally calls the operations createReport, getReport, getReportDocument and the .download() function that will download the final report document in sequence.
The function takes a config object with the following parameters as input:
| Name | Type | Default | Description |
| :----------------------------- | :----: | :--------: | ---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| body
_required_ | object | - | Includes the parameters necessary to request the report. These are the parameters usually passed in to the createReport operation (see createReport 2021-06-30). The possible values will be described below. |getReport
| version
_optional_ | string | 2021-06-30 | The report endpoint’s version that should be used when retrieving the report. |
| interval
_optional_ | number | 10000 | The request interval (in milliseconds) that should be used for re-requesting the operation when the report is still queued or in progress. |interval
| cancel_after
_optional_ | number | - | Cancels a report request after the specified number of retries. Each re-request defined by the value counts as one retry. |
| download
_optional_ | object | - | Includes optional parameters for the download of the report, i.e. to enable a json result or to additionally save the report to a file. The possible values will be described below. |
The body object may include the following properties:
| Name | Type | Default | Description |
| :------------------------------- | :--------------: | :-----: | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| reportType
_required_ | string | - | The report type. |
| marketplaceIds
_required_ | < string > array | - | A list of marketplace identifiers. The report document's contents will contain data for all of the specified marketplaces, unless the report type indicates otherwise. |
| dataStartTime
_optional_ | string | - | The start of a date and time range, in ISO 8601 date time format, used for selecting the data to report. The default is now. The value must be prior to or equal to the current date and time. Not all report types make use of this. |
| dataEndTime
_optional_ | string | - | The end of a date and time range, in ISO 8601 date time format, used for selecting the data to report. The default is now. The value must be prior to or equal to the current date and time. Not all report types make use of this. |
| reportOptions
_optional_ | object | - | Additional information passed to reports. This varies by report type. |
The download object may include the following properties:
| Name | Type | Default | Description |
| :------------------------- | :-----: | :-----: | -------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| json
_optional_ | boolean | false | Whether or not the content should be transformed to json before returning it (from tab delimited flat-file or XML). |
| unzip
_optional_ | boolean | true | Whether or not the content should be unzipped before returning it. |
| file
_optional_ | string | - | The absolute file path to save the report to.
NOTE: Even when saved to disk the report content is still returned. |
| charset
_optional_ | string | utf8 | The charset to use for decoding the content. If not defined, it uses per default the charset returned in content-type header or utf8 if no charset found in content-type header.unzip
NOTE: Is ignored when content is compressed and is set to false. |response
| timeouts
_optional_ | object | - | Allows to set timeouts for download requests. Valid keys are , idle and deadline. Please see detailed information in the Timeouts section. |
Please see the following example that will request a GET_FLAT_FILE_OPEN_LISTINGS_DATA report for the current report endpoint version 2021-06-30, re-request it every 8 seconds and, once its finished, will download the report, transform it to json and save it to disk:
`javascript`
let res = await sellingPartner.downloadReport({
body: {
reportType: "GET_FLAT_FILE_OPEN_LISTINGS_DATA",
marketplaceIds: ["A1PA6795UKMFR9"]
},
version: "2021-06-30",
interval: 8000,
download: {
json: true,
file: "
}
});
Instead of using the .downloadReport() function you may as well call the necessary operations on your own and use the .download() function to retrieve the final report data. Please see the following information below:
The .download() function takes the download details received from a getReportDocument operation as input, downloads the content, unzips it (if result is compressed), decrypts it and returns it.
Retrieve the download details from a getReportDocument operation:
`javascript`
let report_document = await spClient.callAPI({
operation: "getReportDocument",
endpoint: "reports",
path: {
reportDocumentId: "
}
});
The structure of the returned report_document should look like this:
`javascript`
{
reportDocumentId:'
compressionAlgorithm:'GZIP', // Only included if report is compressed
url: '
}
Call the .download() function to receive the content of the report. The default without any config options will download, decrypt and unzip the content and return it without reformatting or saving it to the disk:
`javascript`
let report = await spClient.download(report_document);
You may also include an options object as a 2nd parameter to the .download() function, i.e. to enable a json result or to additionally save the report to a file. The possible parameters are the same as for the download object for the .downloadReport() function already documented above.
The following call will download the report, transform it to json and save it to disk:
`javascript`
let report = await spClient.download(report_document, {
json: true,
file: "
});
Some reports may have an encoding other than UTF-8 and require special decoding with a different charset, i.e. the GET_MERCHANT_LISTINGS_ALL_DATA report is encoded as cp1252 for eu region marketplaces. The right charset to use for decoding is taken from the return header content-type, but you may force the use of a specific charset for decoding by passing in the optional charset property:
`javascript`
let report = await spClient.download(report_document, {
charset: "cp1252"
});
The .upload() function takes the feed upload details received from a createFeedDocument operation, the feed content and its content type to upload as input and uploads it.
Start by creating a feed object with a contentType and the content either as a string or a file path to a document:
| Name | Type | Default | Description |
| :---------------------------- | :----: | :-----: | ----------------------------------------------------------------------------------------------------------------------------------------------------------------------- |
| content
_optional_ | string | - | The content to upload as a string.
Required if file is not provided. |content
| file
_optional_ | string | - | The absolute file path to the feed content document to upload.
Required if is not provided. |text/xml
| contentType
_required_ | string | - | The contentType of the content to upload.
Should be one of or text/tab-separated-values and the charset of the content, i.e. text/xml; charset=utf-8. |
This will create an inventory feed (POST_INVENTORY_AVAILABILITY_DATA) that will update the quantity of a given SKU to 10:
`javascript
let feed = {
content:
,`
contentType: "text/xml; charset=utf-8"
};
Before you can upload the feed you need to retrieve the feed upload details from a createFeedDocument operation:
`javascript`
let feed_upload_details = await spClient.callAPI({
operation: "createFeedDocument",
endpoint: "feeds",
body: {
contentType: feed.contentType
}
});
Call the .upload() function to upload the content of the feed:
`javascript`
let res = await spClient.upload(feed_upload_details, feed);
After uploading the feed you have to trigger the processing of the feed by calling the createFeed operation with the necessary params (marketplaceIds, feedType and inputFeedDocumentId):
`javascript`
let feed_creation_infos = await spClient.callAPI({
operation: "createFeed",
endpoint: "feeds",
body: {
marketplaceIds: ["A1PA6795UKMFR9"],
feedType: "POST_INVENTORY_AVAILABILITY_DATA",
inputFeedDocumentId: feed_upload_details.feedDocumentId // retrieve the feedDocumentId from the "createFeedDocument" operation
}
});
NOTE: Although uploading and creating the feed was successful it doesn't mean that the processing of the feed itself was also successful. You can check the result of the feed once it has been processed by downloading the processing result with the .download() function quite similar as how to download reports. Use the feedId returned by the createFeed operation and call the getFeed operation, which will include a resultFeedDocumentId if feed processing is already done. The resultFeedDocumentId can be used with a getFeedDocument operation that will return the feed download details needed for the feed result download.
All TypeScript related information can be found in lib/typings. Currently types are not yet defined for all operations and/or params, so feel free to add new types following the readme. You are also welcome to create a pull request.
You can easily enable sandbox mode by setting use_sandbox in the constructor config options to true. General information on sandbox setup and behaviour can be found in the corresponding section in the Selling Partner API Developer Guide.
When using the sandbox you have to make sure to use the correct request parameters for the operation you want to test. You can find these inside the api models definitions in the docs by searching the corresponding json file for x-amzn-api-sandbox.
For example, this will test the getPricing operation in sandbox mode:
`javascript`
let res = await spClient.callAPI({
operation: "getPricing",
endpoint: "productPricing",
query: {
MarketplaceId: "TEST_CASE_400"
}
});
There is an issue with values of arrays as part of the query, when a value contains a ,. Due to Amazon expecting array values in query separated by , it will wrongfully split up values containing a , into two separate values. This is already a known issue communicated to Amazon.
We might be able to support you with everything else that can't be done with the API, i.e. a detailed sales dashboard, review management, automated reimbursements and more. Feel free to visit us at https://getarthy.com.
- Removed refreshRoleCredentials function and the getter for role_credentials. As Amazon has removed the neccessity for signing requests to the SP API, role credentials are not needed anymore.
- Trying to call an operation without specifying an endpoint now results in a NO_ENDPOINT_GIVEN error. Although deprecated since version 0.4.0 it was still possible to call an operation without an endpoint. This possibility has now been removed. However, it is still perfectly fine to omit the endpoint parameter and add it directly to the operation parameter via shorthand dot notation (i.e. operation: "sellers.getMarketplaceParticipations")
- The reports and feeds endpoint's version 2020-09-04 is deprecated since 27th June 2023 and has been removed. As a result, encryptionDetails` for downloading of reports and uploading of feeds is not returned anymore and all en-/decryption logic for reports and feeds has been removed as well.