Wrapper library for EMSI skills API: https://api.lightcast.io/apis/skills#overview
This EMSI API wrapper exposes the following functions to interact with EMSI api:
- getApiToken() - Generate a jwt token to use with subsequent EMSI API calls
- getSkillById(skillId, token) - Get the EMSI skill details by emsi skill id
- extractSkillsFromText(text, token) - Extract the EMSI skills from the given text
- getRelatedSkills(skillIds, token) - Get the skills related to given skills ids
- searchSkills(criteria, token) - Get all EMSI skills matching the given criteria, the search criteria is an object which has the same fields as the url parameters of the GET /skills endpoint see query parameters names at https://api.lightcast.io/apis/skills#get-list-all-skills
Sample criteria object for searching for skills is:
const criteria = {
q: 'Java',
typeIds: 'ST1,ST2',
fields: 'id,name',
limit:5
}The configuration of the application using this library should have the following structure:
EMSI_SKILLS_API: {
AUTH_URL: process.env.EMSI_AUTH_URL || 'https://auth.emsicloud.com/connect/token',
BASE_URL: process.env.EMSI_SKILLS_BASE_URL || 'https://emsiservices.com/skills',
CLIENT_ID: process.env.EMSI_SKILLS_API_CLIENT_ID,
CLIENT_SECRET: process.env.EMSI_SKILLS_API_CLIENT_SECRET,
GRANT_TYPE: process.env.EMSI_AUTH_GRANT_TYPE || 'client_credentials',
SCOPE: process.env.EMSI_AUTH_SCOPE || 'emsi_open',
VERSION: process.env.EMSI_SKILLS_API_VERSION || 'latest',
SKILLS_EXTRACTION_CONFIDENCE_THRESHOLD: process.env.EMSI_SKILLS_EXTRACTION_CONFIDENCE_THRESHOLD || '0.6',
MAXIMUM_RELATED_SKILLS_TO_RETRIEVE_PER_REQUEST: process.env.MAXIMUM_RELATED_SKILLS_TO_RETRIEVE_PER_REQUEST || 5
}The above configuration parameters can be used for accessing the EMSI API, only need to additionally set the following parameters in the application environment
export EMSI_SKILLS_API_CLIENT_ID=<client_id>
export EMSI_SKILLS_API_CLIENT_SECRET=<client_secret>- In order to use this wrapper, it needs to be included in the application
package.json
"emsi-api-wrapper": "topcoder-platform/emsi-api-wrapper.git#develop"
- Create an instance of the wrapper:
const emsiApiWrapper = require('emsi-api-wrapper')
const config = require('config')
const client = emsiApiWrapper(config);- Call the wrapper functions:
const token = await client.getApiToken()
const text = 'Java Backend API'
const emsiSkills = await client.extractSkillsFromText(text, token)
for(const emsiSkill of emsiSkills) {
console.log(`Get the skill details, skillId = ${emsiSkill.skillId} `)
const details = await client.getSkillById(emsiSkill.skillId, token)
console.log(`Skill details = ${JSON.stringify(details, null, 2)}`)
}
console.log('Search all skills with Javascript name, limited to 5 skills')
const searchCriteria = {
q: 'Javascript',
limit: 5
}
const searchSkillsResult = await client.searchSkills(searchCriteria, token)
console.log(`The searched skills are = ${JSON.stringify(searchSkillsResult, null, 2)}`)
/**
* This extraction from file API only supoports utf-8 encoded PDF and docx file
*/
console.log('Extract skills from file')
const file = fs.readFileSync('./resume.docx')
const skillsFromFile = await client.extractSkillsFromFile(Buffer.from(file, 'utf-8'), 'docx', token)
console.log(`The extracted skill are: ${JSON.stringify(skillsFromFile)}`)A sample application which shows how to use the wrapper is provided at sample-app