Oracle Cloud Jobs API.

import re
import requests
from urllib.parse import urlparse

SITE_NUMBER_PATTERNS = [
    re.compile(r"""siteNumber\s*[:=]\s*['"]([^'"]+)['"]""", re.IGNORECASE),
    re.compile(r'"siteNumber"\s*:\s*"([^"]+)"', re.IGNORECASE),
    re.compile(r"siteNumber=(CX_[^&\"'\s]+)", re.IGNORECASE),
]

def resolve_site_number(careers_url: str) -> str:
    """Resolve the Oracle Cloud siteNumber for a careers URL."""
    # A /sites/<segment> that starts with "CX" IS the siteNumber.
    path = urlparse(careers_url).path
    match = re.search(r"/sites/([^/?#]+)", path, re.IGNORECASE)
    if match and match.group(1).upper().startswith("CX"):
        return match.group(1)

    # Vanity slugs embed the real siteNumber in the careers-page HTML.
    html = requests.get(careers_url, timeout=30).text
    for pattern in SITE_NUMBER_PATTERNS:
        found = pattern.search(html)
        if found:
            return found.group(1)

    # Most production Oracle tenants default to CX_1.
    return "CX_1"

careers_url = "https://eeho.fa.us2.oraclecloud.com/hcmUI/CandidateExperience/en/sites/jobsearch/jobs"
site_number = resolve_site_number(careers_url)
print(f"Site number: {site_number}")

import uuid
import requests

DOMAIN = "eeho.fa.us2.oraclecloud.com"
SITE_NUMBER = "CX_1"
LISTINGS_PATH = "/hcmRestApi/resources/latest/recruitingCEJobRequisitions"
PAGE_SIZE = 200  # the CE API caps the finder-scoped list at 200 per request

HEADERS = {
    "ora-irc-cx-userid": str(uuid.uuid4()),  # any UUID works for anonymous access
    "ora-irc-language": "en",
    "content-type": "application/vnd.oracle.adf.resourceitem+json;charset=utf-8",
}

def build_listings_url(domain: str, site_number: str, offset: int, limit: int = PAGE_SIZE) -> str:
    # limit/offset MUST be inside the finder; the finder's ';' '=' ',' are literal.
    finder = f"findReqs;siteNumber={site_number},limit={limit},offset={offset}"
    expand = (
        "requisitionList.workLocation,requisitionList.otherWorkLocations,"
        "requisitionList.secondaryLocations,flexFieldsFacet.values,"
        "requisitionList.requisitionFlexFields"
    )
    query = f"onlyData=true&expand={expand}&finder={finder}"
    return f"https://{domain}{LISTINGS_PATH}?{query}"

url = build_listings_url(DOMAIN, SITE_NUMBER, offset=0)
data = requests.get(url, headers=HEADERS, timeout=30).json()

# The response holds a single "search" item; the jobs live in requisitionList.
search_item = data["items"][0]
jobs = search_item.get("requisitionList", [])
total = search_item.get("TotalJobsCount", 0)
print(f"Fetched {len(jobs)} of {total} jobs")

def build_job_url(domain: str, site_path: str, job_id: str) -> str:
    return f"https://{domain}/hcmUI/CandidateExperience/en/sites/{site_path}/job/{job_id}"

site_path = "jobsearch"  # the /sites/<segment> from the careers URL

for job in jobs[:5]:  # first 5 jobs
    print({
        "id": job.get("Id"),
        "title": job.get("Title"),
        "location": job.get("PrimaryLocation"),
        "secondary_locations": [loc.get("Name") for loc in job.get("secondaryLocations", [])],
        "posted_date": job.get("PostedDate"),
        "short_description": (job.get("ShortDescriptionStr") or "")[:100],
        "is_hot_job": job.get("HotJobFlag", False),
        "is_trending": job.get("TrendingFlag", False),
        "url": build_job_url(DOMAIN, site_path, job.get("Id")),
    })

import uuid
import requests

DETAILS_PATH = "/hcmRestApi/resources/latest/recruitingCEJobRequisitionDetails"

def get_job_details(domain: str, site_number: str, job_id: str) -> dict | None:
    """Fetch full job details from Oracle Cloud."""
    # Keep the finder's '"', ';', '=' and ',' literal — do not percent-encode them.
    finder = f'ById;Id="{job_id}",siteNumber={site_number}'
    query = f"expand=all&onlyData=true&finder={finder}"
    url = f"https://{domain}{DETAILS_PATH}?{query}"

    headers = {
        "ora-irc-cx-userid": str(uuid.uuid4()),
        "ora-irc-language": "en",
        "content-type": "application/vnd.oracle.adf.resourceitem+json;charset=utf-8",
    }

    data = requests.get(url, headers=headers, timeout=30).json()
    items = data.get("items") or []
    return items[0] if items else None

details = get_job_details(DOMAIN, SITE_NUMBER, "307750")
if details:
    print({
        "id": details.get("Id"),
        "title": details.get("Title"),
        "category": details.get("Category"),
        "department": details.get("Department"),
        "workplace_type": details.get("WorkplaceType"),
        "description": (details.get("ExternalDescriptionStr") or "")[:200],
        "qualifications": (details.get("ExternalQualificationsStr") or "")[:200],
        "responsibilities": (details.get("ExternalResponsibilitiesStr") or "")[:200],
        "skills": [s.get("Skill") for s in details.get("skills", [])],
        "flex_fields": {f.get("Prompt"): f.get("Value") for f in details.get("requisitionFlexFields", [])},
    })

import time
import uuid
import requests

def fetch_all_jobs(domain: str, site_number: str, page_size: int = PAGE_SIZE) -> list:
    """Fetch every job, paginating on TotalJobsCount."""
    all_jobs = []
    offset = 0
    total = None

    while True:
        url = build_listings_url(domain, site_number, offset, page_size)
        data = requests.get(url, headers=HEADERS, timeout=30).json()

        items = data.get("items") or []
        if not items:
            break

        search_item = items[0]
        jobs = search_item.get("requisitionList") or []
        if not jobs:
            break

        all_jobs.extend(jobs)
        if total is None:
            total = search_item.get("TotalJobsCount") or 0
        print(f"Fetched {len(all_jobs)} of {total} jobs...")

        # Page on TotalJobsCount; the top-level hasMore is unreliable.
        offset += len(jobs)
        if total and offset >= total:
            break
        if not total and len(jobs) < page_size:  # defensive stop when count is absent
            break

        time.sleep(0.3)  # matches the scraper's 300ms inter-request delay

    return all_jobs

all_jobs = fetch_all_jobs(DOMAIN, SITE_NUMBER)
print(f"Total jobs collected: {len(all_jobs)}")