from __future__ import annotations

from datetime import datetime
from typing import Literal

from pydantic import BaseModel, Field


# ── Request ───────────────────────────────────────────────────────────────

class AnalyzeRequest(BaseModel):
    resume_id: str
    job_description: str = Field(min_length=100, max_length=10_000)
    target_role: str = ""


# ── Response sub-types ────────────────────────────────────────────────────

class ScoreWeights(BaseModel):
    keyword_coverage: float = 0.30
    semantic_similarity: float = 0.25
    skills_overlap: float = 0.20
    experience_alignment: float = 0.15
    resume_quality: float = 0.10


class ComponentBreakdown(BaseModel):
    keyword_coverage: int
    semantic_similarity: int
    skills_overlap: int
    experience_alignment: int
    resume_quality: int


class SectionNote(BaseModel):
    section: str
    status: Literal["strong", "ok", "weak", "missing"]
    note: str
    score: int


class Recommendation(BaseModel):
    priority: Literal["high", "medium", "low"]
    category: str
    message: str


class BulletRewrite(BaseModel):
    original: str
    improved: str
    reason: str


# ── Full response ─────────────────────────────────────────────────────────

class AnalyzeResponse(BaseModel):
    analysis_id: str
    overall_score: int
    weights: ScoreWeights
    components: ComponentBreakdown
    matched_keywords: list[str]
    missing_keywords: list[str]
    weak_keywords: list[str]
    section_notes: list[SectionNote]
    recommendations: list[Recommendation]
    bullet_rewrites: list[BulletRewrite]   # populated in Phase 4 (LLM)
    role_fit_verdict: str
    role_fit_summary: str
    parsed_sections: dict                  # skills list, bullets, etc.
    disclaimer: str = (
        "ATS-style compatibility score (heuristic) — not an official ATS result."
    )
    latency_ms: int


class AnalysisDetail(BaseModel):
    analysis_id: str
    resume_id: str
    created_at: datetime
    result: AnalyzeResponse