# =============================================================================
# Job·Ka — Groupe KA
# Auteur  : Simon-Pierre Boucher
# Contact : contact@spboucher.ai
# Fichier : jobka/connectors/ultipro.py
# Rôle    : Classe de plateforme UKG Pro Recruiting (UltiPro) — API JSON
#           publique LoadSearchResults, un employeur = une sous-classe
# Créé    : 2026-08-18   Modifié : 2026-08-18
# =============================================================================
"""Plateforme UKG Pro Recruiting (recruiting.ultipro.com).

API JSON publique (celle du frontal) :
- liste  : POST https://recruiting.ultipro.com/<ORG>/JobBoard/<BOARD>/
                JobBoardView/LoadSearchResults
           body {opportunitySearch:{Top, Skip, OrderBy…}}
           -> {opportunities:[{Id, Title, RequisitionNumber, BriefDescription,
               FullTime, JobCategoryName, PostedDate, Locations:[{Address}]}],
               totalCount}
- détail : POST .../JobBoardView/LoadOpportunity {opportunityId} -> Description
           complète (HTML). Visité avec cache BD + budget.
"""
from __future__ import annotations

import os
import re

from ..schema import JobPosting, clean_html, is_quebec_location
from .base import BaseConnector

PAGE_SIZE = 50
MAX_DETAILS = int(os.environ.get("JOBKA_ULTIPRO_DETAIL_LIMIT", "80"))


class UltiProConnector(BaseConnector):
    """Base UltiPro — sous-classes : définir source_id, EMPLOYER, ORG, BOARD."""

    ats = "ultipro"
    request_delay = 0.8

    EMPLOYER = ""
    ORG = ""      # ex. MES1005MESR
    BOARD = ""    # GUID du JobBoard
    HOST = "recruiting.ultipro.com"   # tenants canadiens : recruiting.ultipro.ca
    quebec_only = True
    max_pages = 20

    @property
    def _board_url(self) -> str:
        return f"https://{self.HOST}/{self.ORG}/JobBoard/{self.BOARD}"

    @staticmethod
    def _loc_fields(loc: dict) -> tuple[str, str, str, str]:
        """(ville, province, code postal, libellé) d'un objet Location."""
        addr = loc.get("Address") or {}
        city = addr.get("City") or ""
        state = ((addr.get("State") or {}).get("Code")
                 if isinstance(addr.get("State"), dict)
                 else addr.get("State")) or ""
        postal = addr.get("PostalCode") or ""
        label = loc.get("LocalizedDescription") or city
        return city, str(state), postal, label

    def _keep(self, locations: list[dict]) -> bool:
        if not self.quebec_only:
            return True
        for loc in locations:
            city, state, postal, label = self._loc_fields(loc)
            if state.upper() == "QC" or re.match(r"^[GHJ]\d[A-Z]", postal.upper()):
                return True
            if is_quebec_location(f"{label} {city}"):
                return True
        return False

    _DESC_RE = re.compile(r'"Description"\s*:\s*"((?:[^"\\]|\\.)*)"')

    def _fetch_detail(self, opp_id: str) -> dict:
        """La page OpportunityDetail embarque l'objet JSON de l'offre —
        on en extrait la description complète (chaîne JSON échappée)."""
        html = self.get(f"{self._board_url}/OpportunityDetail",
                        params={"opportunityId": opp_id}).text
        best = ""
        for m in self._DESC_RE.finditer(html):
            try:
                import json as _json
                val = _json.loads(f'"{m.group(1)}"')
            except ValueError:
                continue
            if len(val) > len(best):
                best = val
        return {"description": clean_html(best)}

    def fetch(self) -> list[JobPosting]:
        url = f"{self._board_url}/JobBoardView/LoadSearchResults"
        out: list[JobPosting] = []
        details_used = 0
        skip = 0
        for _ in range(self.max_pages):
            body = {
                "opportunitySearch": {
                    "Top": PAGE_SIZE, "Skip": skip, "QueryString": "",
                    "OrderBy": [{"Value": "postedDateDesc",
                                 "PropertyName": "PostedDate",
                                 "Ascending": False}],
                    "Filters": [],
                },
                "matchCriteria": {"PreferredJobs": [], "Educations": [],
                                  "LicenseAndCertifications": [], "Skills": [],
                                  "hasNoLicenses": False, "SkippedSkills": []},
            }
            data = self.post(url, json=body,
                             headers={"Accept": "application/json"}).json()
            opps = data.get("opportunities") or []
            if not opps:
                break
            for o in opps:
                locations = o.get("Locations") or []
                if not self._keep(locations):
                    continue
                oid = str(o.get("Id") or "")
                if not oid:
                    continue
                city = label = postal = ""
                for loc in locations:
                    c, s, p, lbl = self._loc_fields(loc)
                    if s.upper() == "QC" or is_quebec_location(f"{lbl} {c}"):
                        city, postal, label = c or lbl, p, lbl
                        break
                job = JobPosting(
                    source=self.source_id, external_id=oid,
                    url=f"{self._board_url}/OpportunityDetail?opportunityId={oid}",
                    employer=self.EMPLOYER,
                    title=o.get("Title") or "",
                    city=city, postal_code=postal, location_label=label,
                    date_posted=o.get("PostedDate") or None,
                    ats=self.ats,
                )
                job.details["requisition"] = o.get("RequisitionNumber") or ""
                if o.get("JobCategoryName"):
                    job.details["team"] = o["JobCategoryName"]
                if o.get("FullTime") is True:
                    job.details["employment_label"] = "Temps plein"
                elif o.get("FullTime") is False:
                    job.details["employment_label"] = "Temps partiel"
                key = (o.get("PostedDate") or "") + (o.get("Title") or "")[:40]
                if details_used < MAX_DETAILS:
                    fresh = [False]

                    def _fn(i=oid, fresh=fresh):
                        fresh[0] = True
                        return self._fetch_detail(i)

                    d = self.detail(oid, key, _fn)
                    if fresh[0]:
                        details_used += 1
                    job.description = d.get("description", "")
                if not job.description:
                    job.description = clean_html(o.get("BriefDescription") or "")
                out.append(job)
            skip += PAGE_SIZE
            total = int(data.get("totalCount") or 0)
            if total and skip >= total:
                break
        return out