HrSmart Jobs API.

import requests
from bs4 import BeautifulSoup

tenant = "photobox"  # leading subdomain of {tenant}.mua.hrdepartment.com
base_url = f"https://{tenant}.mua.hrdepartment.com"

def listing_url(page: int, page_size: int = 100) -> str:
    return (
        f"{base_url}/hr/ats/JobSearch/viewAll"
        f"/jobSearchPaginationExternal_pageSize:{page_size}"
        f"/jobSearchPaginationExternal_page:{page}"
    )

import re

JOB_ID = re.compile(r"/hr/ats/Posting/view/(\d+)")

def parse_rows(html: str) -> list[dict]:
    soup = BeautifulSoup(html, "html.parser")
    jobs = []
    for row in soup.select("#jobSearchResultsGrid_table tbody tr"):
        anchor = row.select_one("a[href*='/hr/ats/Posting/view/']")
        if not anchor:
            continue
        m = JOB_ID.search(anchor.get("href", ""))
        if not m:
            continue
        job_id = m.group(1)
        cells = row.find_all("td")
        wide = len(cells) >= 5
        jobs.append({
            "external_id": job_id,
            "title": anchor.get_text(strip=True),
            # Canonicalize — HrSmart sometimes appends /0/13 for share links.
            "url": f"{base_url}/hr/ats/Posting/view/{job_id}",
            "reference_code": cells[0].get_text(strip=True) if wide else None,
            "employment_type": cells[2].get_text(strip=True) if wide else None,
            "location": cells[3].get_text(strip=True) if wide else None,
            "posted": cells[4].get_text(strip=True) if wide else None,
        })
    return jobs

COUNT_RE = re.compile(r"\bof\s+(\d+)\b", re.I)

def total_count(html: str) -> int | None:
    soup = BeautifulSoup(html, "html.parser")
    el = soup.select_one(".pagination_displaying")
    if not el:
        return None
    m = COUNT_RE.search(el.get_text(" ", strip=True))
    return int(m.group(1)) if m else None

def scrape_all(max_pages: int = 50) -> list[dict]:
    seen, out, total = set(), [], None
    for page in range(1, max_pages + 1):
        html = requests.get(listing_url(page), timeout=15).text
        if not html:
            break
        new = [j for j in parse_rows(html) if j["external_id"] not in seen]
        if page == 1:
            total = total_count(html)
        if not new:                       # no new records -> stop
            break
        for j in new:
            seen.add(j["external_id"])
            out.append(j)
        if total is not None and len(out) >= total:
            break
    return out

def field(soup, el_id):
    el = soup.find(id=el_id)
    return el.get_text(" ", strip=True) if el else None

def parse_detail(html: str) -> dict:
    soup = BeautifulSoup(html, "html.parser")
    desc = soup.find(id="job_details_ats_requisition_description")
    location = field(soup, "job_details_hua_location_id")
    sites = [s.split("(")[0].strip() for s in (location or "").split("\n") if s.strip()]
    return {
        "title": field(soup, "job_details_ats_requisition_title"),
        "description_html": desc.decode_contents() if desc else None,
        "locations": sites or ([location] if location else []),
        "employment_type": field(soup, "job_details_hua_job_type_id"),
        "department": field(soup, "job_details_hua_org_level_id"),
        "category": field(soup, "job_details_ats_requisition_category_id"),
        "education_level": field(soup, "job_details_ats_education_level_id"),
        "experience_level": field(soup, "job_details_ats_requisition_level_id"),
        "open_date": field(soup, "job_details_f_open_date_0"),
        "reference_code": field(soup, "job_details_ats_requisition_code"),
    }

import time

def fetch_detail(url: str) -> dict | None:
    resp = requests.get(url, timeout=15)
    if resp.status_code in (404, 410):
        return None                       # posting removed
    if resp.status_code in (403, 429):
        raise RuntimeError("Blocked / rate limited — back off and retry")
    resp.raise_for_status()
    time.sleep(0.5)                       # ~500ms between requests; <=3 concurrent
    return parse_detail(resp.text)