Harbour ATS Jobs API.

import requests

# Provider-owned tenant: https://{tenant}.amal.harbourats.com/vacancies/  (also .atom / .srs)
# Verified custom host:  https://apply.tlt.com/vacancies/
base_url = "https://apply.tlt.com/vacancies/"

resp = requests.get(base_url, timeout=15, headers={"User-Agent": "jobo-bot/1.0"})
resp.raise_for_status()
html = resp.text

import re
from urllib.parse import urljoin, urlparse
from bs4 import BeautifulSoup

JOB_PATH = re.compile(r"^/vacancies/(\d+)(?:/[^/?#]+)?/?$")

def text(node):
    return re.sub(r"\s+", " ", node.get_text()).strip() if node else None

def parse_cards(html, page_url):
    soup = BeautifulSoup(html, "html.parser")
    host = urlparse(page_url).netloc
    cards = {}
    for a in soup.select("a[href*='/vacancies/']"):
        detail = urljoin(page_url, a.get("href", "")).split("?")[0].split("#")[0]
        u = urlparse(detail)
        m = JOB_PATH.match(u.path)
        if not m or u.netloc != host:
            continue
        job_id = m.group(1)
        card = a.find_parent(class_="bio") or a.find_parent(class_="vacancy-details") or a
        cards.setdefault(job_id, {
            "id": job_id,
            "url": detail.rstrip("/"),
            "title": text(card.select_one(".vacancy_title, .vacancy-title, strong")) or text(a),
            "location": text(card.select_one(".value_location")),
            "category": text(card.select_one(".value_job_category")),
            "contract_type": text(card.select_one(".value_contract_type")),
            "salary_text": text(card.select_one(".value_salary")),
            "closing_date": text(card.select_one(".value_closing_date")),
        })
    return cards

NEXT_PAGE = re.compile(r"^/vacancies/page/\d+/?$", re.IGNORECASE)

def next_page_url(html, page_url):
    soup = BeautifulSoup(html, "html.parser")
    host = urlparse(page_url).netloc
    for a in soup.select("a[href]"):
        label = (a.get_text() or "").strip()
        classes = " ".join(a.get("class") or []).lower()
        is_next = "next" in (a.get("rel") or []) or "next" in classes or label.lower().startswith("next")
        href = a.get("href", "")
        if is_next and re.search(r"/vacancies/page/\d+", href, re.IGNORECASE):
            nxt = urljoin(page_url, href).split("#")[0]
            u = urlparse(nxt)
            if u.netloc == host and NEXT_PAGE.match(u.path):
                return nxt
    return None

def crawl_board(base_url):
    all_cards, url = {}, base_url
    while url:
        html = requests.get(url, timeout=15).text
        page = parse_cards(html, url)
        if not set(page) - set(all_cards):  # no new IDs -> fleet exhausted
            break
        all_cards.update(page)
        url = next_page_url(html, url)
    return all_cards

import json

def job_posting(soup):
    for tag in soup.find_all("script", type="application/ld+json"):
        try:
            data = json.loads(tag.string or "")
        except (json.JSONDecodeError, TypeError):
            continue
        for node in (data if isinstance(data, list) else [data]):
            if isinstance(node, dict) and node.get("@type") == "JobPosting":
                return node
    return {}

def fetch_detail(card):
    html = requests.get(card["url"], timeout=15).text
    soup = BeautifulSoup(html, "html.parser")
    jp = job_posting(soup)
    salary = (jp.get("baseSalary") or {}).get("value") or {}
    return {
        "id": card["id"],
        "url": card["url"],
        "title": jp.get("title") or card["title"],
        "company": (jp.get("hiringOrganization") or {}).get("name"),
        "description": jp.get("description") or text(soup.select_one(
            ".value_job_advert_description, .vacancy_description, [itemprop='description']")),
        "posted_at": jp.get("datePosted"),
        "closes_at": card["closing_date"] or jp.get("validThrough"),
        "employment_type": jp.get("employmentType") or card["contract_type"],
        "salary_min": salary.get("minValue"),
        "salary_max": salary.get("maxValue"),
        "salary_currency": (jp.get("baseSalary") or {}).get("currency"),
    }

def apply_url(soup, card):
    u = urlparse(card["url"])
    same_host_default = f"{u.scheme}://{u.netloc}/vacancies/{card['id']}/apply/"
    for a in soup.select("a[href]"):
        label = (a.get_text() or "").strip().lower()
        if label not in ("apply", "apply now", "apply for this job"):
            continue
        href = urljoin(card["url"], a.get("href", ""))
        target = urlparse(href)
        # provider-hosted apply route on the same host
        if target.netloc == u.netloc and re.match(rf"^/vacancies/{card['id']}/apply/?$", target.path):
            return href, "provider_hosted"
        # explicit downstream apply on another HTTPS host
        if target.scheme == "https" and target.netloc != u.netloc:
            return href, "downstream"
    return same_host_default, "provider_hosted"