Skip to main content
The DescriptionFormat enum specifies the desired output format for job descriptions returned by JobSpy. Different formats are useful for different use cases, from direct rendering to text analysis.

​
Enum Values

​
MARKDOWN

Value: "markdown" Returns job descriptions formatted in Markdown.
  • Clean, readable format with preserved structure
  • Easy to render in documentation and web applications
  • Preserves headings, lists, and basic formatting
  • Ideal for storing in content management systems
  • Can be easily converted to other formats
Best for: Documentation sites, CMSs, formatted displays, GitHub READMEs

​
HTML

Value: "html" Returns job descriptions as HTML markup.
  • Rich formatting with full styling preserved
  • Includes tags like <p>, <ul>, <strong>, <a>, etc.
  • Ready for direct rendering in web browsers
  • May include inline styles and classes
  • Most detailed formatting option
Best for: Web applications, email templates, rich text displays

​
PLAIN

Value: "plain" Returns job descriptions as plain text with no formatting.
  • All HTML tags and markdown syntax removed
  • Simple, clean text content
  • Whitespace normalized
  • Ideal for text analysis and search indexing
  • Smallest data size
Best for: Text analysis, search engines, AI processing, SMS/notifications

​
Usage Examples

​
Setting Format in scrapeJobs()

import { scrapeJobs, Site, DescriptionFormat } from 'jobspy-js';

// Get descriptions in Markdown format
const markdownResults = await scrapeJobs({
  site_name: Site.LINKEDIN,
  search_term: 'product designer',
  location: 'San Francisco, CA',
  description_format: DescriptionFormat.MARKDOWN,
  results_wanted: 20
});

for (const job of markdownResults.jobs) {
  console.log(job.title);
  console.log(job.description); // Markdown-formatted text
}

​
String Alternative

// You can also pass format as a string
const results = await scrapeJobs({
  site_name: Site.INDEED,
  search_term: 'data scientist',
  description_format: 'html', // equivalent to DescriptionFormat.HTML
  results_wanted: 30
});

​
Format Comparison

import { scrapeJobs, Site, DescriptionFormat } from 'jobspy-js';

// Fetch the same jobs in different formats
const searchParams = {
  site_name: Site.GLASSDOOR,
  search_term: 'software engineer',
  location: 'Seattle, WA',
  results_wanted: 10
};

const htmlResults = await scrapeJobs({
  ...searchParams,
  description_format: DescriptionFormat.HTML
});

const markdownResults = await scrapeJobs({
  ...searchParams,
  description_format: DescriptionFormat.MARKDOWN
});

const plainResults = await scrapeJobs({
  ...searchParams,
  description_format: DescriptionFormat.PLAIN
});

console.log('HTML:', htmlResults.jobs[0].description);
console.log('Markdown:', markdownResults.jobs[0].description);
console.log('Plain:', plainResults.jobs[0].description);

​
Format-Specific Processing

​
HTML Format - Direct Rendering

import { scrapeJobs, DescriptionFormat } from 'jobspy-js';

const results = await scrapeJobs({
  site_name: Site.LINKEDIN,
  search_term: 'frontend developer',
  description_format: DescriptionFormat.HTML,
  results_wanted: 50
});

// Render directly in a web page
for (const job of results.jobs) {
  document.getElementById('job-list').innerHTML += `
    <div class="job-card">
      <h2>${job.title}</h2>
      <h3>${job.company_name}</h3>
      <div class="job-description">
        ${job.description} <!-- HTML content rendered directly -->
      </div>
    </div>
  `;
}

​
Markdown Format - Convert to HTML

import { scrapeJobs, DescriptionFormat } from 'jobspy-js';
import { marked } from 'marked'; // npm install marked

const results = await scrapeJobs({
  site_name: Site.INDEED,
  search_term: 'devops engineer',
  description_format: DescriptionFormat.MARKDOWN,
  results_wanted: 25
});

// Convert markdown to HTML for rendering
for (const job of results.jobs) {
  const htmlDescription = marked(job.description || '');
  console.log(htmlDescription);
}

​
Plain Format - Text Analysis

import { scrapeJobs, DescriptionFormat } from 'jobspy-js';

const results = await scrapeJobs({
  site_name: Site.LINKEDIN,
  search_term: 'machine learning',
  description_format: DescriptionFormat.PLAIN,
  results_wanted: 100
});

// Analyze plain text for keywords
function analyzeSkills(description: string): string[] {
  const skills = ['Python', 'TensorFlow', 'PyTorch', 'scikit-learn', 'AWS', 'Docker'];
  return skills.filter(skill => 
    description.toLowerCase().includes(skill.toLowerCase())
  );
}

for (const job of results.jobs) {
  const requiredSkills = analyzeSkills(job.description || '');
  console.log(`${job.title}: ${requiredSkills.join(', ')}`);
}

​
Format for Different Use Cases

import { scrapeJobs, DescriptionFormat } from 'jobspy-js';

// For a job board website (HTML)
const webJobs = await scrapeJobs({
  site_name: Site.INDEED,
  search_term: 'nurse',
  description_format: DescriptionFormat.HTML,
  results_wanted: 50
});

// For a documentation site (Markdown)
const docJobs = await scrapeJobs({
  site_name: Site.GLASSDOOR,
  search_term: 'technical writer',
  description_format: DescriptionFormat.MARKDOWN,
  results_wanted: 30
});

// For AI/ML processing (Plain)
const aiJobs = await scrapeJobs({
  site_name: Site.LINKEDIN,
  search_term: 'analyst',
  description_format: DescriptionFormat.PLAIN,
  results_wanted: 200
});

// Feed plain text to AI model
const embeddings = await generateEmbeddings(aiJobs.jobs.map(j => j.description));

​
Fetching Individual Jobs

import { fetchLinkedInJob, DescriptionFormat } from 'jobspy-js';

// Fetch a single LinkedIn job in different formats
const jobId = '3823456789';

const htmlJob = await fetchLinkedInJob(jobId, DescriptionFormat.HTML);
const markdownJob = await fetchLinkedInJob(jobId, DescriptionFormat.MARKDOWN);
const plainJob = await fetchLinkedInJob(jobId, DescriptionFormat.PLAIN);

console.log('Same job, different formats:');
console.log('HTML length:', htmlJob.description?.length);
console.log('Markdown length:', markdownJob.description?.length);
console.log('Plain length:', plainJob.description?.length);
// Plain will typically be shortest, HTML longest

​
Format Characteristics

​
Size Comparison

Typical relative sizes for the same description: 
HTML:     ~4,500 bytes  (largest - includes tags, attributes, styling)
Markdown: ~3,200 bytes  (medium - includes markdown syntax)
Plain:    ~2,800 bytes  (smallest - text only)

​
Formatting Preservation

FeatureHTMLMarkdownPlain
Headings✅ <h1>, <h2>✅ #, ##❌ Lost
Bold/Italic✅ <strong>, <em>✅ **bold**, *italic*❌ Lost
Links✅ <a href>✅ [text](url)⚠️ URL text only
Lists✅ <ul>, <ol>✅ -, 1.⚠️ Basic structure
Line breaks✅ <br>, <p>✅ Double newline✅ Preserved
Tables✅ <table>✅ Pipe syntax❌ Lost
Styling✅ Colors, fonts❌ Not supported❌ Not supported

​
Processing Speed

Relative processing times: 
Plain:    Fastest  (minimal processing)
Markdown: Medium   (HTML to Markdown conversion)
HTML:     Slowest  (full HTML parsing, may include cleanup)

​
Default Behavior

If description_format is not specified: 
const results = await scrapeJobs({
  site_name: Site.LINKEDIN,
  search_term: 'engineer'
  // description_format not specified
});

// Default format depends on the scraper implementation
// Check the returned description to determine format

​
Type Definition

enum DescriptionFormat {
  MARKDOWN = "markdown",
  HTML = "html",
  PLAIN = "plain",
}

Build docs developers (and LLMs) love

Get started for freeTalk to us