Refactor apply handlers: generic with extensions + domain auto-routing
Generic handler now accepts options (transformUrl, formDetector, submitSelector, resumeSelector, beforeSubmit, verifySelector, etc.). Each ATS handler passes its overrides instead of reimplementing. Registry resolves handlers by: apply_type -> URL domain -> generic fallback. New ATS handlers only need to export SUPPORTED_TYPES and an apply() that calls genericApply with platform-specific options. Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
This commit is contained in:
@@ -1,8 +1,19 @@
|
||||
/**
|
||||
* generic.mjs — Generic external ATS handler
|
||||
* Best-effort form filler for any career page with a standard HTML form.
|
||||
* Handles single-page and multi-step flows (up to 5 steps).
|
||||
* Skips pages that require account creation or have CAPTCHAs.
|
||||
* generic.mjs — Configurable external ATS handler
|
||||
*
|
||||
* Base apply logic for any career page. ATS-specific handlers pass
|
||||
* an options object to customize URL transformation, selectors,
|
||||
* resume targeting, and submission verification.
|
||||
*
|
||||
* Options:
|
||||
* transformUrl(url) — modify the apply URL before navigation
|
||||
* formDetector — CSS selector to detect if form is already loaded
|
||||
* applyButtonSelector — selector for the "Apply" button on listing pages
|
||||
* resumeSelector — CSS selector for the resume file input
|
||||
* submitSelector — selector for the submit button (use locator syntax)
|
||||
* verifySelector — CSS selector to check if form is still present after submit
|
||||
* beforeSubmit(page) — async hook to run before clicking submit (e.g. upload resume)
|
||||
* closedTexts — extra strings to detect closed listings
|
||||
*/
|
||||
import {
|
||||
NAVIGATION_TIMEOUT, PAGE_LOAD_WAIT, FORM_FILL_WAIT, SUBMIT_WAIT
|
||||
@@ -12,58 +23,85 @@ export const SUPPORTED_TYPES = ['unknown_external'];
|
||||
|
||||
const MAX_STEPS = 5;
|
||||
|
||||
export async function apply(page, job, formFiller) {
|
||||
const url = job.apply_url;
|
||||
const DEFAULT_APPLY_BUTTONS = [
|
||||
'a:has-text("Apply Now")',
|
||||
'button:has-text("Apply Now")',
|
||||
'a:has-text("Apply for this job")',
|
||||
'button:has-text("Apply for this job")',
|
||||
'a:has-text("Apply")',
|
||||
'button:has-text("Apply")',
|
||||
].join(', ');
|
||||
|
||||
const DEFAULT_SUBMIT_BUTTONS = [
|
||||
'button:has-text("Submit Application")',
|
||||
'button:has-text("Submit your application")',
|
||||
'button:has-text("Apply Now")',
|
||||
'button:has-text("Apply for this job")',
|
||||
'input[type="submit"]:not([disabled])',
|
||||
'button[type="submit"]:not([disabled])',
|
||||
].join(', ');
|
||||
|
||||
const DEFAULT_NEXT_BUTTONS = [
|
||||
'button:has-text("Next")',
|
||||
'button:has-text("Continue")',
|
||||
'button:has-text("Save and Continue")',
|
||||
'a:has-text("Next")',
|
||||
].join(', ');
|
||||
|
||||
const CLOSED_TEXTS = [
|
||||
'no longer accepting', 'position has been filled',
|
||||
'this job is no longer', 'job not found',
|
||||
'this position is closed', 'listing has expired',
|
||||
'no longer available', 'page you are looking for',
|
||||
'job may be no longer', 'does not exist',
|
||||
'this role has been filled', 'posting has closed',
|
||||
];
|
||||
|
||||
export async function apply(page, job, formFiller, opts = {}) {
|
||||
const url = opts.transformUrl ? opts.transformUrl(job.apply_url) : job.apply_url;
|
||||
if (!url) return { status: 'no_button', meta: { title: job.title, company: job.company } };
|
||||
|
||||
const meta = { title: job.title, company: job.company };
|
||||
|
||||
await page.goto(url, { waitUntil: 'domcontentloaded', timeout: NAVIGATION_TIMEOUT });
|
||||
await page.waitForTimeout(PAGE_LOAD_WAIT);
|
||||
|
||||
const meta = {
|
||||
title: job.title,
|
||||
company: job.company,
|
||||
};
|
||||
|
||||
// Detect blockers: login walls, CAPTCHAs, closed listings
|
||||
const pageCheck = await page.evaluate(() => {
|
||||
// Detect blockers
|
||||
const extraClosed = opts.closedTexts || [];
|
||||
const pageCheck = await page.evaluate((extraClosed) => {
|
||||
const text = (document.body.innerText || '').toLowerCase();
|
||||
const hasLogin = !!(document.querySelector('input[type="password"]') ||
|
||||
(text.includes('sign in') && text.includes('create account')) ||
|
||||
(text.includes('log in') && text.includes('register')));
|
||||
// Only block on visible CAPTCHAs — invisible reCAPTCHA (size=invisible) fires on submit and usually passes
|
||||
const captchaFrames = Array.from(document.querySelectorAll('iframe[src*="recaptcha"], iframe[src*="captcha"]'));
|
||||
const hasVisibleCaptcha = captchaFrames.some(f => {
|
||||
const hasCaptcha = captchaFrames.some(f => {
|
||||
if (f.src.includes('size=invisible')) return false;
|
||||
const rect = f.getBoundingClientRect();
|
||||
return rect.width > 50 && rect.height > 50;
|
||||
});
|
||||
const hasCaptcha = hasVisibleCaptcha;
|
||||
const isClosed = text.includes('no longer accepting') || text.includes('position has been filled') ||
|
||||
text.includes('this job is no longer') || text.includes('job not found') ||
|
||||
text.includes('this position is closed') || text.includes('listing has expired') ||
|
||||
text.includes('no longer available') || text.includes('page you are looking for') ||
|
||||
text.includes('job may be no longer') || text.includes('does not exist') ||
|
||||
text.includes('this role has been filled') || text.includes('posting has closed') ||
|
||||
document.title.toLowerCase().includes('404');
|
||||
const closedTexts = [
|
||||
'no longer accepting', 'position has been filled',
|
||||
'this job is no longer', 'job not found',
|
||||
'this position is closed', 'listing has expired',
|
||||
'no longer available', 'page you are looking for',
|
||||
'job may be no longer', 'does not exist',
|
||||
'this role has been filled', 'posting has closed',
|
||||
...extraClosed,
|
||||
];
|
||||
const isClosed = closedTexts.some(t => text.includes(t)) || document.title.toLowerCase().includes('404');
|
||||
return { hasLogin, hasCaptcha, isClosed };
|
||||
}).catch(() => ({}));
|
||||
}, extraClosed).catch(() => ({}));
|
||||
|
||||
if (pageCheck.isClosed) return { status: 'closed', meta };
|
||||
if (pageCheck.hasLogin) return { status: 'skipped_login_required', meta };
|
||||
if (pageCheck.hasCaptcha) return { status: 'skipped_captcha', meta };
|
||||
|
||||
// Some pages land directly on the form; others need an Apply button click
|
||||
// Check if we landed directly on a form (with or without <form> wrapper)
|
||||
const hasFormAlready = await page.$('input[type="text"], input[type="email"], textarea');
|
||||
// Check if form is already loaded
|
||||
const formSelector = opts.formDetector || 'input[type="text"], input[type="email"], textarea';
|
||||
const hasFormAlready = await page.$(formSelector);
|
||||
if (!hasFormAlready) {
|
||||
const applyBtn = page.locator([
|
||||
'a:has-text("Apply Now")',
|
||||
'button:has-text("Apply Now")',
|
||||
'a:has-text("Apply for this job")',
|
||||
'button:has-text("Apply for this job")',
|
||||
'a:has-text("Apply")',
|
||||
'button:has-text("Apply")',
|
||||
].join(', ')).first();
|
||||
const applySelector = opts.applyButtonSelector || DEFAULT_APPLY_BUTTONS;
|
||||
const applyBtn = page.locator(applySelector).first();
|
||||
|
||||
if (await applyBtn.count() === 0) return { status: 'no_button', meta };
|
||||
|
||||
@@ -74,11 +112,9 @@ export async function apply(page, job, formFiller) {
|
||||
]);
|
||||
|
||||
if (newPage) {
|
||||
// Apply opened a new tab — switch to it
|
||||
await newPage.waitForLoadState('domcontentloaded').catch(() => {});
|
||||
await newPage.waitForTimeout(PAGE_LOAD_WAIT);
|
||||
// Recursively handle the new page (but return result to caller)
|
||||
return applyOnPage(newPage, job, formFiller, meta);
|
||||
return fillAndSubmit(newPage, job, formFiller, meta, opts);
|
||||
}
|
||||
|
||||
await page.waitForTimeout(FORM_FILL_WAIT);
|
||||
@@ -93,67 +129,56 @@ export async function apply(page, job, formFiller) {
|
||||
if (postClick.hasCaptcha) return { status: 'skipped_captcha', meta };
|
||||
}
|
||||
|
||||
return applyOnPage(page, job, formFiller, meta);
|
||||
return fillAndSubmit(page, job, formFiller, meta, opts);
|
||||
}
|
||||
|
||||
async function applyOnPage(page, job, formFiller, meta) {
|
||||
async function fillAndSubmit(page, job, formFiller, meta, opts) {
|
||||
for (let step = 0; step < MAX_STEPS; step++) {
|
||||
// Fill the current page/step
|
||||
const unknowns = await formFiller.fill(page, formFiller.profile.resume_path);
|
||||
|
||||
if (unknowns[0]?.honeypot) return { status: 'skipped_honeypot', meta };
|
||||
if (unknowns.length > 0) return { status: 'needs_answer', pending_question: unknowns[0], meta };
|
||||
|
||||
// Look for submit button — try specific text first, then generic type="submit"
|
||||
const submitBtn = await page.$([
|
||||
'button:has-text("Submit Application")',
|
||||
'button:has-text("Submit your application")',
|
||||
'button:has-text("Apply Now")',
|
||||
'button:has-text("Apply for this job")',
|
||||
'input[type="submit"]:not([disabled])',
|
||||
'button[type="submit"]:not([disabled])',
|
||||
].join(', '));
|
||||
// Hook: before submit (e.g. targeted resume upload)
|
||||
if (opts.beforeSubmit) await opts.beforeSubmit(page, formFiller);
|
||||
|
||||
// Look for Next/Continue button (multi-step forms)
|
||||
const nextBtn = !submitBtn ? await page.$([
|
||||
'button:has-text("Next")',
|
||||
'button:has-text("Continue")',
|
||||
'button:has-text("Save and Continue")',
|
||||
'a:has-text("Next")',
|
||||
].join(', ')) : null;
|
||||
// Find submit button
|
||||
const submitSelector = opts.submitSelector || DEFAULT_SUBMIT_BUTTONS;
|
||||
const submitBtn = page.locator(submitSelector).first();
|
||||
const hasSubmit = await submitBtn.count() > 0;
|
||||
|
||||
if (submitBtn) {
|
||||
if (hasSubmit) {
|
||||
await submitBtn.click();
|
||||
await page.waitForTimeout(SUBMIT_WAIT);
|
||||
|
||||
const postSubmit = await page.evaluate(() => {
|
||||
const verifySelector = opts.verifySelector || 'form button[type="submit"]:not([disabled])';
|
||||
const postSubmit = await page.evaluate((vs) => {
|
||||
const text = (document.body.innerText || '').toLowerCase();
|
||||
return {
|
||||
hasSuccess: text.includes('application submitted') || text.includes('successfully applied') ||
|
||||
text.includes('thank you') || text.includes('application received') ||
|
||||
text.includes('application has been') || text.includes('we received your'),
|
||||
hasForm: !!document.querySelector('form button[type="submit"]:not([disabled])'),
|
||||
hasForm: !!document.querySelector(vs),
|
||||
};
|
||||
}).catch(() => ({ hasSuccess: false, hasForm: false }));
|
||||
}, verifySelector).catch(() => ({ hasSuccess: false, hasForm: false }));
|
||||
|
||||
if (postSubmit.hasSuccess || !postSubmit.hasForm) {
|
||||
return { status: 'submitted', meta };
|
||||
}
|
||||
|
||||
console.log(` [generic] Submit clicked but form still present — may not have submitted`);
|
||||
return { status: 'incomplete', meta };
|
||||
}
|
||||
|
||||
// Multi-step: Next/Continue
|
||||
const nextBtn = await page.$(DEFAULT_NEXT_BUTTONS);
|
||||
if (nextBtn) {
|
||||
await nextBtn.click();
|
||||
await page.waitForTimeout(FORM_FILL_WAIT);
|
||||
continue; // Fill next step
|
||||
continue;
|
||||
}
|
||||
|
||||
// No submit or next button found
|
||||
return { status: 'no_submit', meta };
|
||||
}
|
||||
|
||||
console.log(` [generic] Exceeded ${MAX_STEPS} form steps`);
|
||||
return { status: 'incomplete', meta };
|
||||
}
|
||||
|
||||
Reference in New Issue
Block a user