LinkedIn scraper API

# 1. Create a run. Keep this key and body unchanged for retries.
IDEMPOTENCY_KEY=$(uuidgen)
curl --fail-with-body $PINEAPI_BASE_URL/v1/profile-searches \
  -H "Authorization: Bearer $PINEAPI_KEY" \
  -H "Idempotency-Key: $IDEMPOTENCY_KEY" \
  -H "Content-Type: application/json" \
  -d '{"current_job_titles":["Head of Talent"],"locations":["Spain"],"max_items":100}'
# HTTP 202: { "run_id": "<uuid>", "status": "queued", ... }

# 2. Copy run_id from that response. Poll until succeeded or partial.
RUN_ID='<uuid>'
curl --fail-with-body "$PINEAPI_BASE_URL/v1/runs/$RUN_ID" \
  -H "Authorization: Bearer $PINEAPI_KEY"
# queued / starting / running / processing are in progress.
# failed / cancelled are terminal: inspect failure before retrying.

# 3. Download profiles after success. Follow next_cursor while has_more is true.
curl --fail-with-body "$PINEAPI_BASE_URL/v1/runs/$RUN_ID/profiles?limit=200" \
  -H "Authorization: Bearer $PINEAPI_KEY"
# { "data": [...], "has_more": false, "next_cursor": null }

import { randomUUID } from 'node:crypto';

const base = process.env.PINEAPI_BASE_URL + '/v1';
const auth = { Authorization: `Bearer ${process.env.PINEAPI_KEY}` };
async function json(path, options = {}) {
  const response = await fetch(base + path, { ...options,
    headers: { ...auth, ...options.headers }, signal: AbortSignal.timeout(30_000) });
  if (!response.ok) throw new Error(`HTTP ${response.status}: ${await response.text()}`);
  return response.json();
}
// Persist this key and body if your application retries the creation request.
const key = randomUUID();
const { run_id } = await json('/profile-searches', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json', 'Idempotency-Key': key },
  body: JSON.stringify({ current_job_titles: ['Head of Talent'], locations: ['Spain'], max_items: 100 }),
});
const pending = new Set(['queued', 'starting', 'running', 'processing']);
const started = Date.now();
const deadline = Date.now() + 30 * 60_000;
let run;
do {
  if (Date.now() > deadline) throw new Error(`Polling timeout; resume with run_id ${run_id}`);
  await new Promise(resolve => setTimeout(resolve, Date.now() - started < 60_000 ? 5_000 : 20_000));
  run = await json(`/runs/${run_id}`);
} while (pending.has(run.status));
if (!['succeeded', 'partial'].includes(run.status)) throw new Error(JSON.stringify(run.failure ?? run.status));

let cursor;
do {
  const query = new URLSearchParams({ limit: '200' });
  if (cursor) query.set('cursor', cursor);
  const page = await json(`/runs/${run_id}/profiles?${query}`);
  for (const profile of page.data) console.log(profile.full_name);
  cursor = page.has_more ? page.next_cursor : null;
} while (cursor);

import os, time, uuid, requests

base = os.environ["PINEAPI_BASE_URL"].rstrip("/") + "/v1"
session = requests.Session()
session.headers["Authorization"] = f"Bearer {os.environ['PINEAPI_KEY']}"

def get(path, **kwargs):
    response = session.get(base + path, timeout=30, **kwargs)
    response.raise_for_status()
    return response.json()

# Keep this key and the same body if retrying the creation request.
key = str(uuid.uuid4())
response = session.post(base + "/profile-searches", timeout=30,
    headers={"Idempotency-Key": key},
    json={"current_job_titles": ["Head of Talent"], "locations": ["Spain"], "max_items": 100})
response.raise_for_status()
run_id = response.json()["run_id"]
started = time.monotonic()
deadline = started + 1800
while True:
    if time.monotonic() > deadline:
        raise TimeoutError(f"Resume polling run_id {run_id}")
    run = get(f"/runs/{run_id}")
    if run["status"] not in {"queued", "starting", "running", "processing"}:
        break
    time.sleep(5 if time.monotonic() - started < 60 else 20)
if run["status"] not in {"succeeded", "partial"}:
    raise RuntimeError(run.get("failure") or run["status"])

params = {"limit": 200}
while True:
    page = get(f"/runs/{run_id}/profiles", params=params)
    for profile in page["data"]:
        print(profile["full_name"])
    if not page["has_more"]:
        break
    params["cursor"] = page["next_cursor"]