from __future__ import annotations

from apps.sources.models import RawDocument

from .ats import (
    GreenhouseAdapter,
    LeverAdapter,
    RecruiteeAdapter,
    SmartRecruitersAdapter,
    WorkableAdapter,
)
from .base import ExtractionResult
from .generic_html import GenericHtmlAdapter
from .jsonld import JsonLdJobPostingAdapter
from .kempen import KempenEmployerAdapter
from .mol_region import MolRegionEmployerAdapter
from .regional import CordaCampusAdapter, LocalEmployerListingAdapter
from .rss import RssAdapter


class AdapterRegistry:
    def __init__(self) -> None:
        self.greenhouse = GreenhouseAdapter()
        self.lever = LeverAdapter()
        self.recruitee = RecruiteeAdapter()
        self.smartrecruiters = SmartRecruitersAdapter()
        self.workable = WorkableAdapter()
        self.corda_campus = CordaCampusAdapter()
        self.local_employers = LocalEmployerListingAdapter()
        self.mol_region_employers = MolRegionEmployerAdapter()
        self.kempen_employers = KempenEmployerAdapter()
        self.jsonld = JsonLdJobPostingAdapter()
        self.generic = GenericHtmlAdapter()
        self.rss = RssAdapter()
        self.providers = [
            self.corda_campus,
            self.local_employers,
            self.mol_region_employers,
            self.kempen_employers,
            self.greenhouse,
            self.lever,
            self.recruitee,
            self.smartrecruiters,
            self.workable,
        ]

    def extract(self, document: RawDocument) -> ExtractionResult:
        content = document.body_text
        url = document.final_url or document.url
        content_type = (document.content_type or "").lower()
        if document.kind == RawDocument.Kind.XML or "rss" in content_type or "atom" in content_type:
            return self.rss.extract(content, url=url)

        for provider in self.providers:
            if provider._supports_url(url):
                result = provider.extract(content, url=url)
                if any(
                    msg in result.warnings
                    for msg in (
                        "Geen parseerbare ATS-response",
                        "Geen herkenbare ATS-markup voor deze adapter",
                    )
                ):
                    continue
                return result

        jsonld_result = self.jsonld.extract(content, url=url)
        if jsonld_result.jobs:
            return jsonld_result
        generic_result = self.generic.extract(content, url=url)
        generic_result.warnings = jsonld_result.warnings + generic_result.warnings
        return generic_result


registry = AdapterRegistry()