evrythng-cli-plugin-csv-loader

EVRYTHNG CLI plugin to upsert resources from a CSV file. Each use of the load command performs the following:

• Read the input data CSV file.
• Validate columns match the input JSON Schema.
• Map each CSV record to an EVRYTHNG resource using the mapping file.
• Validate output resources with JSON Schema before they are created.
• Create resources, or update by a nominated key if they already exist.

Installation

Install alongside evrythng-cli, usually globally:

$ npm i -g evrythng-cli-plugin-csv-loader

Usage

Initialise a set of example files:

$ evrythng csv-loader init

Load data from some CSV file using a set of config files:

$ evrythng csv-loader load $configPath $csvPath

Quick Start

To start quickly, use the init command to generate a set of example files to modify for your particular data set:

$ evrythng csv-loader init

This will create a directory csv-loader-config containing:

• config.json - Example config file (see below for format)
• input.schema.json - Example JSON Schema file describing the expected format of the input data CSV file.
• output.schema.json - Example JSON Schema file describing the expected format of the output EVRYTHNG resources.
• mapping.json - Example JSON file describing how to map each CSV field to a field on the EVRYTHNG resource (see below for format)

Configuration Format

A configuration file is used to specify all other relevant files. Strict schemas and exhaustive mapping can help ensure data integrity.

• input
  • schema - JSONSchema for CSV records after they are read.
• output
  • schema - JSONSchema for each processed EVRYTHNG resource before it is created.
  • mapping - JSON file that maps CSV column headers to EVRYTHNG resource field names. Sub-objects and arrays are also supported with dot and square bracket notation respectively.
  • type - Type of EVRYTHNG resource to create.
  • updateKey - Either name or some identifiers key to be used for updates.
  • projectName - Name of the EVRYTHNG project to use.
  • defaultRedirectUrl - (optional) If set, used as the product/Thng redirection URL. Must contain {shortId} or {thngId}/{productId} somewhere depending on the resource type.
  • defaultShortDomain - (optional) If set, used as the short domain for the creation of the redirection URLs. Needs to be set when the domain is different than the default one (tn.gg).

An example configuration is shown below:

{
  "input": {
    "schema": "./example-config/input.schema.json"
  },
  "output": {
    "schema": "./example-config/output.schema.json",
    "mapping": "./example-config/mapping.json",
    "type": "product",
    "updateKey": "gs1:01",
    "projectName": "Example PoC",
    "defaultRedirectUrl": "https://google.com?id={shortId}",
    "defaultShortDomain": "tn.gg"
  }
}

Mapping Format

The mapping.json file describes how to map each field in a CSV record from its header name to the EVRYTHNG resource field name.

Array fields (i.e: tags) are supported with square bracket notation (e.g: tags[0]).

Sub-objects (i.e: identifiers or customFields) are suported with dot notation (e.g: identifier.gs1:01).

A redirection can be specified using the special redirection mapping value. This option overrides defaultRedirectUrl if it is set in the config file.

An example mapping file is shown below:

{
  "Brand": "brand",
  "SkuName": "name",
  "ProductCode": "identifiers.gs1:01",
  "ManufactureDate": "customFields.ManufactureDate",
  "BatchId": "tags[0]",
  "PhotoUrl": "photos[0]",
  "SecretCode": "",
  "ProductPage": "redirection"
}

All CSV column headers must be accounted for, but they can be skipped by leaving the mapped value as an empty string (like SecretCode shown above).

Tests

Run unit tests using mocha:

npm test

Run the example config, schemas, and data in example:

$ evrythng csv-loader load ./example/config.json ./example/data.csv