"""Reference solution for Build-versus-buy analysis project: compare cash flows, timing and the outside option.
Original synthetic teaching data only. Python 3.10+, standard library.
Run python solve.py [path-to-project-inputs]. No network or credentials.
"""
from pathlib import Path
from datetime import datetime, date
import csv, json, math, statistics, sys

def read(root, name):
    with (Path(root)/name).open(newline='',encoding='utf-8') as f:
        reader=csv.DictReader(f)
        if not reader.fieldnames or len(reader.fieldnames)!=len(set(reader.fieldnames)):
            raise ValueError(f'{name}: missing or duplicate headers')
        rows=list(reader)
        if any(None in row or any(v is None for v in row.values()) for row in rows):
            raise ValueError(f'{name}: inconsistent row length')
        return rows

def rules(root): return json.loads((Path(root)/'rules.json').read_text())

def number(value):
    x=float(value)
    if not math.isfinite(x): raise ValueError('Non-finite input')
    return x

def nonnegative(value):
    x=number(value)
    if x<0:raise ValueError('Negative quantity or amount')
    return x

def flag(value):
    if value not in ('true','false'):raise ValueError('Boolean must be true or false')
    return value=='true'

def timestamp(value):
    d=datetime.fromisoformat(value.replace('Z','+00:00'))
    if d.tzinfo is None:raise ValueError('Timestamp requires an explicit time zone')
    return d

def unique(rows, keys, ignore=()):
    by={};duplicates=0
    for row in rows:
        key=tuple(row[k] for k in keys)
        if any(not k for k in key):raise ValueError('Missing record identity')
        if key in by:
            prior={k:v for k,v in by[key].items() if k not in ignore}
            new={k:v for k,v in row.items() if k not in ignore}
            if prior!=new:raise ValueError(f'Conflicting records at {key}')
            duplicates+=1
        else:by[key]=row
    return list(by.values()),duplicates

def rounded(value):
    if isinstance(value,float): return round(value,8)
    if isinstance(value,dict):return {k:rounded(v) for k,v in value.items()}
    if isinstance(value,list):return [rounded(v) for v in value]
    return value

def month_index(s):
    d=datetime.strptime(s,'%Y-%m')
    if d.strftime('%Y-%m')!=s:raise ValueError('Use YYYY-MM')
    return d.year*12+d.month

def build_buy(root):
    r=rules(root);rows,_=unique(read(root,'cash_flows.csv'),['option','year']);rate=number(r['discount_rate'])
    if rate<=-1:raise ValueError('Discount rate must be greater than -1')
    flows={}
    for x in rows:
        y=int(x['year'])
        if y<0 or str(y)!=x['year']:raise ValueError('Invalid year')
        flows.setdefault(x['option'],{})[y]=number(x['incremental_cash_flow'])
    if set(flows)!=set(['build','buy','partner']):raise ValueError('Option set changed')
    if any(set(v)!=set(range(5)) for v in flows.values()):raise ValueError('Missing year; zero must be explicit')
    npv={k:sum(v/(1+rate)**y for y,v in data.items()) for k,data in flows.items()}
    buy=flows['buy'];pvbenefits=sum(v/(1+rate)**y for y,v in buy.items() if y>0)
    if abs(buy[0]+r['buy_price']+r['buy_integration_now'])>1e-8:raise ValueError('Purchase and integration do not match year0')
    return {'npv':npv,'buy_operating_cash_flow_pv':pvbenefits,'buy_one_year_delay_npv':buy[0]+pvbenefits/(1+rate),'buy_cashflows_minus20pct_npv':buy[0]+.8*pvbenefits,'buy_cashflows_plus20pct_npv':buy[0]+1.2*pvbenefits,'maximum_buy_price_to_match_build':pvbenefits-r['buy_integration_now']-npv['build'],'maximum_buy_price_to_match_partner':pvbenefits-r['buy_integration_now']-npv['partner']}

def main():
    directory=Path(sys.argv[1]) if len(sys.argv)>1 else Path(__file__).resolve().parent
    try:
        result=build_buy(directory)
        print(json.dumps(rounded(result),indent=2,allow_nan=False))
    except (OSError,ValueError,KeyError,TypeError,ZeroDivisionError) as exc:
        print(f"Input could not be analyzed: {exc}",file=sys.stderr);return 2
    return 0
if __name__=='__main__':raise SystemExit(main())