# =============================================================================
# Job·Ka — Groupe KA
# Auteur  : Simon-Pierre Boucher
# Contact : contact@spboucher.ai
# Fichier : jobka/connectors/wpjobmanager.py
# Rôle    : Classe de plateforme WordPress + WP Job Manager — flux job_feed
#           (contenu complet) + JSON-LD des pages poste (lieu, date limite).
#           Utilisée par la plateforme « carrieresante.gouv.qc.ca » du réseau
#           de la santé (un site WordPress par établissement Santé Québec).
# Créé    : 2026-08-19   Modifié : 2026-08-19
# =============================================================================
"""Plateforme WordPress / WP Job Manager.

- liste  : GET <BASE>/?feed=job_feed&posts_per_page=500
           -> RSS : <item> title / link (/poste/<slug>/) / pubDate /
              guid (?post_type=job_listing&p=<id>) / content:encoded (HTML
              complet de l'offre) — tout en une requête.
- détail : la page /poste/<slug>/ embarque un JSON-LD schema.org/JobPosting
           (validThrough, jobLocation, employmentType) — cache BD + budget,
           uniquement pour compléter lieu et date limite.
"""
from __future__ import annotations

import html as _html
import os
import re

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

MAX_DETAILS = int(os.environ.get("JOBKA_WPJM_DETAIL_LIMIT", "60"))

_ITEM_RE = re.compile(r"<item>(.*?)</item>", re.S | re.I)
_TAG_RE = {t: re.compile(rf"<{t}>(.*?)</{t}>", re.S | re.I)
           for t in ("title", "link", "pubDate", "guid")}
_CONTENT_RE = re.compile(
    r"<content:encoded><!\[CDATA\[(.*?)\]\]></content:encoded>", re.S | re.I)
_PID_RE = re.compile(r"[?&](?:amp;)?p=(\d+)")


class WPJobManagerConnector(BaseConnector):
    """Base WP Job Manager — sous-classes : définir source_id, EMPLOYER,
    BASE (racine du site WordPress) et DEFAULT_CITY au besoin."""

    ats = "wpjobmanager"
    request_delay = 1.0

    EMPLOYER = ""
    BASE = ""
    DEFAULT_CITY = ""
    PER_PAGE = 500

    def _cdata(self, s: str) -> str:
        s = (s or "").strip()
        if s.startswith("<![CDATA["):
            s = s[9:]
            if s.endswith("]]>"):
                s = s[:-3]
        return _html.unescape(s.strip())

    def _fetch_detail(self, url: str) -> dict:
        html = self.get(url).text
        node = _jsonld.extract_jobposting(html)
        return _jsonld.jobposting_fields(node) if node else {}

    def fetch(self) -> list[JobPosting]:
        xml = self.get(f"{self.BASE}/",
                       params={"feed": "job_feed",
                               "posts_per_page": str(self.PER_PAGE)}).text
        out: list[JobPosting] = []
        details_used = 0
        seen: set[str] = set()
        for raw in _ITEM_RE.findall(xml):
            def g(tag: str, raw=raw) -> str:
                m = _TAG_RE[tag].search(raw)
                return self._cdata(m.group(1)) if m else ""
            link, guid = g("link"), g("guid")
            m_id = _PID_RE.search(_html.unescape(guid))
            eid = m_id.group(1) if m_id else ""
            title = g("title")
            if not link or not title:
                continue
            if not eid:
                m_slug = re.search(r"/poste/([^/]+)/?$", link)
                eid = m_slug.group(1) if m_slug else ""
            if not eid or eid in seen:
                continue
            seen.add(eid)
            m_c = _CONTENT_RE.search(raw)
            job = JobPosting(
                source=self.source_id, external_id=eid, url=link,
                employer=self.EMPLOYER, title=title,
                description=clean_html(_html.unescape(m_c.group(1)))
                if m_c else "",
                city=self.DEFAULT_CITY,
                date_posted=g("pubDate") or None,
                ats=self.ats,
            )
            key = f"{title}|{job.date_posted or ''}"
            if details_used < MAX_DETAILS:
                fresh = [False]

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

                fields = self.detail(eid, key, _fn)
                if fresh[0]:
                    details_used += 1
            else:
                fields = self.stale_detail(eid)
            if fields:
                if fields.get("date_deadline"):
                    job.date_deadline = fields["date_deadline"]
                if fields.get("city"):
                    job.city = fields["city"]
                if fields.get("postal_code"):
                    job.postal_code = fields["postal_code"]
                if fields.get("employment_label"):
                    job.details["employment_label"] = fields["employment_label"]
                if fields.get("lat") is not None:
                    job.lat, job.lng = fields["lat"], fields["lng"]
            out.append(job)
        return out