#!/usr/bin/env python3
"""Schema-aware JSON -> CSV exporter (Python 3 stdlib only).

Deterministic, dependency-free. Produces explicit, auditable CSV from JSON
with a stable column order, type-aware flattening, and header control.
"""
import csv, io, json, sys
from collections import OrderedDict

def _is_scalar(v):
    return v is None or isinstance(v, (str, int, float, bool))

def flatten(value, prefix=""):
    """Return (columns, row) mapping for a single object value, recursive."""
    cols = []
    row = []
    if isinstance(value, dict):
        for k, v in value.items():
            key = f"{prefix}.{k}" if prefix else k
            if isinstance(v, dict):
                c, r = flatten(v, key)
                cols.extend(c); row.extend(r)
            elif isinstance(v, list):
                cols.append(key); row.append(json.dumps(v, ensure_ascii=False, separators=(',', ':')))
            else:
                cols.append(key); row.append("" if v is None else v)
    elif isinstance(value, list):
        # array of scalars -> semicolon-joined; array of objects -> keep as one cell (JSON)
        cols.append(prefix or "value")
        row.append(json.dumps(value, ensure_ascii=False, separators=(',', ':')))
    else:
        cols.append(prefix or "value")
        row.append("" if value is None else value)
    return cols, row

def schema_aware_json_to_csv(input_path, output_path, header=True, encoding="utf-8"):
    with open(input_path, "r", encoding=encoding) as f:
        data = json.load(f)
    if isinstance(data, dict) and not _is_scalar(data) and all(isinstance(v, dict) for v in data.values()):
        records = list(data.values())
    elif isinstance(data, list):
        records = data
    else:
        records = [data]
    # Determine stable column order from first record + any extras
    columns = []
    for rec in records:
        c, _ = flatten(rec)
        for col in c:
            if col not in columns:
                columns.append(col)
    rows = []
    for rec in records:
        cols, row = flatten(rec)
        m = dict(zip(cols, row))
        rows.append([m.get(col, "") for col in columns])
    with open(output_path, "w", newline="", encoding=encoding) as f:
        w = csv.writer(f)
        if header:
            w.writerow(columns)
        w.writerows(rows)
    return columns, len(rows)

def json_to_csv_from_text(text, header=True):
    data = json.loads(text)
    out = io.StringIO()
    if isinstance(data, list):
        records = data
    elif isinstance(data, dict) and all(isinstance(v, dict) for v in data.values()):
        records = list(data.values())
    else:
        records = [data]
    columns = []
    for rec in records:
        c, _ = flatten(rec)
        for col in c:
            if col not in columns:
                columns.append(col)
    w = csv.writer(out)
    if header:
        w.writerow(columns)
    for rec in records:
        cols, row = flatten(rec)
        m = dict(zip(cols, row))
        w.writerow([m.get(col, "") for col in columns])
    return out.getvalue()

if __name__ == "__main__":
    if len(sys.argv) < 3:
        print("usage: jsontocsv.py <input.json> <output.csv>")
        sys.exit(2)
    cols, n = schema_aware_json_to_csv(sys.argv[1], sys.argv[2])
    print(f"columns={len(cols)} rows={n}")