"""Provenance of shared answer content, independent of Gemini corrections."""
from hashlib import sha256
from pathlib import Path

from .json_io import atomic_update_json, read_json


def _path(root):
    return Path(root) / ".copienator" / "answer_origins.json"


def origins_for(root, copy_id, label, pdf):
    entry = read_json(_path(root), default={}).get(str(copy_id), {}).get(label, {})
    return _valid_origins(entry, pdf)


def _valid_origins(entry, pdf):
    if not entry or not Path(pdf).is_file():
        return []
    if entry.get("sha256") != sha256(Path(pdf).read_bytes()).hexdigest():
        return []
    return entry.get("labels", [])


def populated_labels(root, copy_id, source_label):
    """Find current answer PDFs populated from this source's recorded content."""
    entries = read_json(_path(root), default={}).get(str(copy_id), {})
    copies = Path(root) / "Copies" / f"Copie{copy_id}"
    populated = []
    for label, entry in entries.items():
        if label == source_label or source_label not in entry.get("labels", []):
            continue
        pdf = copies / f"{label}.pdf"
        if not pdf.is_file():
            pdf = copies / f"{label}_new.pdf"
        if source_label in _valid_origins(entry, pdf):
            populated.append(label)
    return sorted(populated)


def save_origins(root, copy_id, label, pdf, origins):
    entry = {"labels": list(dict.fromkeys(origins)),
             "sha256": sha256(Path(pdf).read_bytes()).hexdigest()}

    def update(data):
        data.setdefault(str(copy_id), {})[label] = entry

    atomic_update_json(_path(root), update, default_factory=dict)