import re, json, sys, io
from openpyxl import load_workbook
from unidecode import unidecode

sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8')

ps = open(r'D:\VNPT-M\Tool\audit_tool\NEW\Update\audit_lhdsin\audit_check_script\audit_cis_windows.ps1', 'r', encoding='utf-8').read()

# Strategy: find ALL lines with Write-CheckResult, regardless of conditional context
# Also handle the case where IsDomain branches produce different descriptions

all_checks = []

# Find all Write-CheckResult calls - including those inside conditionals
for m in re.finditer(r'Write-CheckResult\s+"([^"]+)"\s+"([^"]+)"', ps):
    check_id = m.group(1)
    desc = m.group(2)
    if not check_id.startswith('$'):
        all_checks.append((check_id, desc))

print(f'Direct calls (no variables): {len(all_checks)}')

# Now expand variable-based calls from Test-FirewallProfile and Test-EventLogPolicy
fw_calls = re.findall(r'Test-FirewallProfile\s+"(\w+)"\s+"([\d.]+)"\s+"(\w+)"', ps)
for profile, prefix, label in fw_calls:
    all_checks.append((f'{prefix}.1', f"Thiet lap trang thai 'Windows Firewall: {label} : Firewall state'"))
    all_checks.append((f'{prefix}.2', f"Thiet lap trang thai 'Windows Firewall: {label} : Inbound connections'"))
    all_checks.append((f'{prefix}.3', f"Thiet lap trang thai 'Windows Firewall: {label} : Outbound connections'"))

log_calls = re.findall(r'Test-EventLogPolicy\s+"(\w+)"\s+"([\d.]+)"\s+"(\w+)"', ps)
for logname, prefix, label in log_calls:
    all_checks.append((f'{prefix}.1', f"Thiet lap chinh sach '{label} : Control Event Log behavior when the log file reaches its maximum size'"))
    all_checks.append((f'{prefix}.2', f"Thiet lap chinh sach '{label} : Specify the maximum log file size (KB)'"))

# Build criteria list, removing duplicates (same CheckId + same Description)
seen = set()
criteria_unique = []
for cid, desc in all_checks:
    key = (cid, desc)
    if key not in seen:
        seen.add(key)
        criteria_unique.append((cid, desc))

criteria = criteria_unique
print(f'Total unique criteria: {len(criteria)}')

# Build ps_criteria lookup
ps_criteria = {}
for cid, desc in criteria:
    full = f"{cid}. {desc}"
    key = unidecode(full)
    ps_criteria[key] = full

# Load Excel
wb = load_workbook(r'D:\VNPT-M\Tool\audit_tool\NEW\Update\audit_lhdsin\config\Windows_Checklist.xlsx')
ws = wb.worksheets[0]

excel_rows = {}

# Pass 1: match by full unidecoded text
for row in range(1, ws.max_row + 1):
    val_a = str(ws.cell(row=row, column=1).value or "").strip()
    val_b = str(ws.cell(row=row, column=2).value or "").strip()
    combined = f"{val_a} {val_b}".strip()
    norm = unidecode(combined)
    if norm in ps_criteria:
        excel_rows[row] = ps_criteria[norm]
        del ps_criteria[norm]

# Pass 2: match by CheckId prefix
still = dict(ps_criteria)
for cid_key, full_text in still.items():
    cid = full_text.split('. ')[0]
    for row in range(1, ws.max_row + 1):
        if row in excel_rows:
            continue
        val_a = str(ws.cell(row=row, column=1).value or "").strip()
        if val_a and cid in val_a:
            excel_rows[row] = full_text
            del ps_criteria[cid_key]
            break

# Pass 3: match column B alone
still2 = dict(ps_criteria)
for cid_key, full_text in still2.items():
    cid = full_text.split('. ')[0]
    for row in range(1, ws.max_row + 1):
        if row in excel_rows:
            continue
        val_b = str(ws.cell(row=row, column=2).value or "").strip()
        if val_b:
            norm_b = unidecode(val_b)
            parts = norm_b.split()
            if parts and cid in parts[0]:
                excel_rows[row] = full_text
                del ps_criteria[cid_key]
                break

print(f'\nMatched rows: {len(excel_rows)}')
print(f'Unmatched: {len(ps_criteria)}')

if ps_criteria:
    print('\nUnmatched check IDs:')
    for k in sorted(ps_criteria.keys()):
        cid = ps_criteria[k].split('. ')[0]
        print(f'  {cid}: {ps_criteria[k][:80]}')

# Build new config
new_data = []
for row in sorted(excel_rows.keys()):
    new_data.append({str(row): excel_rows[row]})

new_config = {"data": new_data}

output_path = r'D:\VNPT-M\Tool\audit_tool\NEW\Update\audit_lhdsin\config\windows_config.json'
with open(output_path, 'w', encoding='utf-8') as f:
    json.dump(new_config, f, ensure_ascii=False, indent=4)

print(f'\nNew config: {output_path}')
print(f'Total entries: {len(new_data)}')

# Verify against actual output
out_file = r'D:\VNPT-M\Tool\audit_tool\NEW\Update\audit_lhdsin\output\MEVAS-NTL-DATAN_2026_07_29-05_18_36.txt'
with open(out_file, 'r', encoding='utf-8', errors='replace') as f:
    lines = f.read().splitlines()

audit_results = {}
for line in lines:
    line = line.strip()
    if line.startswith('{') and line.endswith('}'):
        try:
            parsed = json.loads(line)
            key = unidecode(str(list(parsed.keys())[0]))
            audit_results[key] = (str(list(parsed.values())[0]), line)
        except:
            pass

cfg_names = {unidecode(str(list(item.values())[0])) for item in new_config['data']}
output_names = set(audit_results.keys())

matched = cfg_names & output_names
output_only = output_names - cfg_names

print(f'Output lines: {len(output_names)}')
print(f'Config matches output: {len(matched)}/{len(output_names)}')
print(f'Output not in config: {len(output_only)}')
if output_only:
    print('Missing from config:')
    for n in sorted(list(output_only))[:5]:
        print(f'  {n[:80]}')