import json
from pathlib import Path


def convert_hq_entries(csv_path: Path, backup: bool = True) -> None:
    """Convert HQ1:,HQ2:,HQ3: columns in the desynthesis recipes CSV into a
    single JSON-like dictionary string stored in the HQ column.

    The CSV currently looks like::
        Item,Crystal,Ingredients,HQ,Cap
        Distilled Water x3,Lightning,Tahrongi Cactus,HQ1: Distilled Water x6,HQ2: Distilled Water x9,HQ3: Distilled Waterx12,2

    After conversion it will be::
        Distilled Water x3,Lightning,Tahrongi Cactus,{"HQ1":"Distilled Water x6","HQ2":"Distilled Water x9","HQ3":"Distilled Waterx12"},2
    """
    csv_path = Path(csv_path)
    if not csv_path.exists():
        raise FileNotFoundError(csv_path)

    text = csv_path.read_text(encoding="utf-8").splitlines()
    output_lines = []

    for line in text:
        # Keep skill category lines (e.g., "Alchemy") unchanged.
        if "," not in line:
            output_lines.append(line)
            continue

        parts = [p.strip() for p in line.split(",")]

        # The header line already has correct length (5).
        if parts[:5] == ["Item", "Crystal", "Ingredients", "HQ", "Cap"]:
            output_lines.append(",".join(parts))
            continue

        # If this row already has 5 columns, leave as-is.
        if len(parts) <= 5:
            output_lines.append(",".join(parts))
            continue

        # Otherwise consolidate HQ columns.
        item, crystal, ingredients = parts[:3]
        cap = parts[-1]
        hq_parts = parts[3:-1]

        hq_dict = {}
        unnamed_counter = 1
        for h in hq_parts:
            h = h.strip()
            if not h:
                continue
            if ":" in h:
                key, value = h.split(":", 1)
                hq_dict[key.strip()] = value.strip()
            else:
                # Handle unlabeled HQ values by assigning sequential keys.
                key = f"HQ{unnamed_counter}"
                unnamed_counter += 1
                hq_dict[key] = h

        # Build dictionary string with spaces after commas to match example formatting.
        hq_json_readable = "{" + ",".join([f'"{k}": "{v}"' if "\"" not in v else f'"{k}": {json.dumps(v)}' for k, v in hq_dict.items()]) + "}"

        new_line = ",".join([item, crystal, ingredients, hq_json_readable, cap])
        output_lines.append(new_line)

    # Backup original file.
    if backup:
        backup_path = csv_path.with_suffix(csv_path.suffix + ".bak")
        if not backup_path.exists():
            backup_path.write_text("\n".join(text), encoding="utf-8")

    csv_path.write_text("\n".join(output_lines) + "\n", encoding="utf-8")


if __name__ == "__main__":
    target = Path(__file__).resolve().parents[1] / "datasets" / "desythesis_recipes.csv"
    convert_hq_entries(target)