admin / Strike

"""
StrikeXi Risk Assessment engine (V2).

Derives a qualitative risk assessment from the maturity scores produced by the
scoring engine. Lower maturity => higher residual cyber risk. This translates
the per-principle / per-objective maturity (0-100) into:

  * an overall risk rating (Critical / High / Medium / Low),
  * a per-objective risk breakdown,
  * a prioritised list of the highest-risk principles (key risk areas), and
  * a short narrative summary.

It reads only persisted score snapshots, so it is stable for completed
assessments and reflects exactly the responses the user gave.
"""
from sqlalchemy.orm import Session

from . import models

OBJ_TITLES = {
    "A": "Managing security risk",
    "B": "Protecting against cyber attack",
    "C": "Detecting cyber security events",
    "D": "Minimising the impact of incidents",
}


def _risk_band(score: float) -> dict:
    """Map a maturity score (0-100) to a residual-risk band."""
    if score < 40:
        return {"level": "Critical", "rank": 4, "colour": "#c0392b"}
    if score < 60:
        return {"level": "High", "rank": 3, "colour": "#e67e22"}
    if score < 80:
        return {"level": "Medium", "rank": 2, "colour": "#f1c40f"}
    return {"level": "Low", "rank": 1, "colour": "#27ae60"}


def _overall_band_from_rank(rank: float) -> dict:
    if rank >= 3.5:
        return {"level": "Critical", "colour": "#c0392b"}
    if rank >= 2.5:
        return {"level": "High", "colour": "#e67e22"}
    if rank >= 1.5:
        return {"level": "Medium", "colour": "#f1c40f"}
    return {"level": "Low", "colour": "#27ae60"}


def build_risk_summary(db: Session, assessment: models.Assessment) -> dict:
    """Return a structured risk assessment derived from persisted scores."""
    principles = {p.id: p for p in db.query(models.CafPrinciple).all()}

    pscores = (
        db.query(models.AssessmentPrincipleScore)
        .filter(models.AssessmentPrincipleScore.assessment_id == assessment.id).all()
    )
    oscores = {
        s.objective_id: float(s.score)
        for s in db.query(models.AssessmentObjectiveScore)
        .filter(models.AssessmentObjectiveScore.assessment_id == assessment.id).all()
    }

    # Count answered questions to express assessment coverage.
    answered = (
        db.query(models.AssessmentAnswer)
        .filter(models.AssessmentAnswer.assessment_id == assessment.id).count()
    )

    # Per-principle risk rows
    principle_risks = []
    for ps in pscores:
        pr = principles.get(ps.principle_id)
        score = float(ps.score)
        band = _risk_band(score)
        principle_risks.append({
            "principle_id": ps.principle_id,
            "principle_title": pr.title if pr else "",
            "objective_id": pr.objective_id if pr else "",
            "score": score,
            "risk_level": band["level"],
            "risk_rank": band["rank"],
            "colour": band["colour"],
        })

    # Per-objective risk
    objective_risks = []
    for oid in ["A", "B", "C", "D"]:
        if oid not in oscores:
            continue
        score = oscores[oid]
        band = _risk_band(score)
        objective_risks.append({
            "objective_id": oid,
            "title": OBJ_TITLES.get(oid, ""),
            "score": score,
            "risk_level": band["level"],
            "colour": band["colour"],
        })

    # Overall risk = mean of objective risk ranks (so it reflects the spread,
    # not just the average score, giving weak areas due weight).
    if objective_risks:
        mean_rank = sum(_risk_band(o["score"])["rank"] for o in objective_risks) / len(objective_risks)
    else:
        mean_rank = 1.0
    overall = _overall_band_from_rank(mean_rank)

    # Key risk areas = highest-risk principles first (lowest score), top 5.
    key_risks = sorted(principle_risks, key=lambda x: x["score"])[:5]

    counts = {"Critical": 0, "High": 0, "Medium": 0, "Low": 0}
    for r in principle_risks:
        counts[r["risk_level"]] += 1

    narrative = _narrative(overall["level"], counts, len(principle_risks))

    return {
        "overall_risk": overall["level"],
        "overall_colour": overall["colour"],
        "answered_questions": answered,
        "principles_assessed": len(principle_risks),
        "counts": counts,
        "objective_risks": objective_risks,
        "principle_risks": sorted(principle_risks,
                                  key=lambda x: (x["objective_id"], x["principle_id"])),
        "key_risks": key_risks,
        "narrative": narrative,
    }


def _narrative(overall_level: str, counts: dict, total: int) -> str:
    high = counts["Critical"] + counts["High"]
    if total == 0:
        return ("No principles have been scored yet, so a risk position cannot "
                "be determined. Complete the assessment to generate a risk profile.")
    base = (f"Based on the responses provided, the organisation's overall residual "
            f"cyber risk is assessed as {overall_level}. ")
    if counts["Critical"]:
        base += (f"{counts['Critical']} principle(s) fall into the Critical band and "
                 f"require urgent remediation. ")
    if high:
        base += (f"{high} of {total} assessed principles present a High or Critical "
                 f"residual risk and should be prioritised. ")
    else:
        base += ("No principles fall into the High or Critical risk bands. ")
    base += ("Addressing the key risk areas below, in priority order, will reduce "
             "residual risk and raise overall cyber maturity.")
    return base