Oleeo / WCN Jobs API.

import re
from urllib.parse import urlsplit, unquote

def extract_site_id(vacancies_url: str) -> str | None:
    parts = urlsplit(vacancies_url)
    for chunk in (parts.query, parts.fragment):
        decoded = unquote(unquote(chunk))
        match = re.search(r"p_web_site_id=(\d+)", decoded)
        if match:
            return match.group(1)
    return None

# e.g. https://recruitment.victimsupport.org.uk/vacancies.html#filter=p_web_site_id%3D5695
site_id = extract_site_id(
    "https://recruitment.victimsupport.org.uk/vacancies.html#filter=p_web_site_id%3D5695"
)
print(site_id)  # 5695

import requests

def fetch_feed(origin: str, site_id: str) -> list[dict]:
    url = f"{origin}/utf8/ic_job_feeds.feed_engine"
    params = {
        "p_web_site_id": site_id,
        "p_published_to": "WWW",
        "p_language": "DEFAULT",
        "p_direct": "Y",
        "p_format": "MOBILE",
        "p_include_exclude_from_list": "N",
        "p_search": "",
        "p_summary": "Y",
    }
    resp = requests.get(url, params=params, headers={"Accept": "application/json"}, timeout=30)
    resp.raise_for_status()
    return resp.json().get("jobs", [])

origin = "https://recruitment.victimsupport.org.uk"
jobs = fetch_feed(origin, site_id)
print(f"{len(jobs)} vacancies (complete public set — no pagination)")

import re

VACANCY_RE = re.compile(r"/vacancy/(?:[^/?#]*-)?(?P<job_id>\d+)\.html(?:[?#]|$)", re.I)

def first_class(classifications: dict, name: str) -> str | None:
    for block in (classifications or {}).values():
        if block.get("name") == name:
            vals = [v.get("class_val", "").strip()
                    for v in block.get("values", []) if v.get("class_val")]
            if vals:
                return "; ".join(vals)
    return None

def map_listing(row: dict, site_id: str, origin: str) -> dict | None:
    job_id = str(row.get("id") or "")
    weblink = row.get("weblink") or ""
    if str(row.get("web_site_id")) != site_id or not weblink.startswith(origin):
        return None
    match = VACANCY_RE.search(weblink)
    if not match or match.group("job_id") != job_id:
        return None  # drop rows whose pretty URL ID doesn't match the feed ID

    cls = row.get("classifications", {})
    internet = (row.get("publication") or {}).get("internet", {})
    locations = row.get("locations") or [{}]
    return {
        "external_id": job_id,
        "title": (row.get("title") or "").strip(),
        "listing_url": weblink,
        "reference": row.get("refno"),
        "status": row.get("status"),
        "department": first_class(cls, "Area of Business"),
        "employment_type": first_class(cls, "Full time/Part time"),
        "contract_type": first_class(cls, "Contract type"),
        "salary": first_class(cls, "Basic Salary"),
        "location": first_class(cls, "Location or Function") or locations[0].get("city"),
        "published": internet.get("publish_date"),
        "closes": internet.get("closing_date"),
    }

listings = [m for row in jobs if (m := map_listing(row, site_id, origin))]

import requests
from bs4 import BeautifulSoup

def fetch_description(listing: dict) -> dict:
    resp = requests.get(listing["listing_url"], headers={"Accept": "text/html"}, timeout=30)
    resp.raise_for_status()
    soup = BeautifulSoup(resp.text, "html.parser")
    root = soup.select_one(".job_detail .job_description")
    if root is None:
        raise ValueError("vacancy HTML is missing the Oleeo job_description block")

    heading = root.select_one("h1")
    apply_link = root.select_one(".apply_now a[href]")
    for selector in ("h1", ".job_classifications", ".links"):
        for element in root.select(selector):
            element.decompose()

    listing["title"] = heading.get_text(strip=True) if heading else listing["title"]
    listing["apply_url"] = apply_link["href"] if apply_link else listing["listing_url"]
    listing["description_html"] = root.decode_contents().strip()
    return listing

# Cap detail concurrency and pace requests — the portal is a legacy PL/SQL runtime
for item in listings[:3]:
    fetch_description(item)