# -*- coding: utf-8 -*-
import sys, io, re
sys.stdout = io.TextIOWrapper(sys.stdout.buffer, encoding='utf-8', errors='replace')

TRACE_FILE = r'c:\Program Files\Firebird\Firebird_2_5_CGM\default_trace.log'
TIME_FROM = '2026-03-29T13:20:00'
TIME_TO   = '2026-03-29T13:20:15'

# Hledame dotazy ktere pravdepodobne patri k eDavky oknu:
# - zminka o tabulkach s davkami/podanim na pojistovnu
# - sloupce jako ZP, PODACI, ODESLANO, STAV v kontextu davek
KEYWORDS = [
    'EDAVKY', 'EDAVKA', 'DAVKY_POJ', 'REGISTR_POJ',
    'PODACI_CISLO', 'PODACI',
    'FAKDAV', 'BB_DAVKA', 'BB_FAKTURA',
    'STAV_PODANI', 'ID_PODANI',
    'DAVKA_ROK', 'DAVKA_DISK', 'DAVKA_CASTKA',
]

# Tabulky ktere NECHCEME (HPN = neschopenky, RECEPT = recepty, atd.)
EXCLUDE_TABLES = ['FROM HPN', 'FROM NES', 'FROM RECEPT', 'FROM CLICKDOC',
                  'FROM UZIVATEL', 'FROM ZARIZENI', 'FROM ODDEL', 'FROM PRACOVISTE']

print(f"Čtu trace: {TRACE_FILE}")
print(f"Časové okno: {TIME_FROM} .. {TIME_TO}")
print(f"Hledám klíčová slova: {KEYWORDS}")
print("=" * 70)

ts_re = re.compile(r'^\d{4}-\d{2}-\d{2}T\d{2}:\d{2}:\d{2}')

in_window = False
current_ts = ''
current_block = []
hits = []
all_selects_tables = {}  # tabulka -> pocet vyskytu

def get_from_tables(text):
    """Extrahuj tabulky z FROM klauzule."""
    tables = re.findall(r'\bFROM\s+([A-Z_][A-Z0-9_]*)', text, re.IGNORECASE)
    tables += re.findall(r'\bJOIN\s+([A-Z_][A-Z0-9_]*)', text, re.IGNORECASE)
    return [t.upper() for t in tables]

def process_block(ts, lines):
    text = '\n'.join(lines)
    if not re.search(r'\bSELECT\b', text, re.IGNORECASE):
        return

    text_up = text.upper()

    # Spocitej tabulky
    tables = get_from_tables(text_up)
    for t in tables:
        all_selects_tables[t] = all_selects_tables.get(t, 0) + 1

    # Filtr - hledej klicova slova
    for kw in KEYWORDS:
        if kw in text_up:
            hits.append((ts, kw, text))
            return

    # Alternativne: dotazy na PORTAL nebo FAK s datumem
    if ('PORTAL' in text_up or 'FAKDAV' in text_up) and 'SELECT' in text_up:
        # Vynech jednoduche dotazy
        if len(text) > 100:
            hits.append((ts, 'PORTAL/FAKDAV', text))

try:
    with open(TRACE_FILE, 'r', encoding='utf-8', errors='replace') as f:
        for line in f:
            line = line.rstrip('\n')
            m = ts_re.match(line)
            if m:
                if current_block and in_window:
                    process_block(current_ts, current_block)
                current_ts = line[:19]
                current_block = [line]
                in_window = (TIME_FROM <= current_ts <= TIME_TO)
            else:
                if in_window:
                    current_block.append(line)
        if current_block and in_window:
            process_block(current_ts, current_block)
except FileNotFoundError:
    print(f"CHYBA: soubor nenalezen: {TRACE_FILE}")
    sys.exit(1)

print(f"\nNejčastější tabulky v SELECT dotazech v okně (top 40):")
sorted_tables = sorted(all_selects_tables.items(), key=lambda x: -x[1])
for t, cnt in sorted_tables[:40]:
    print(f"  {t:<40} {cnt}x")

print()
print(f"{'='*70}")
print(f"Nalezeno relevantních bloků (klíčová slova): {len(hits)}")
print()

for i, (ts, kw, text) in enumerate(hits):
    print(f"{'='*70}")
    print(f"[{i+1}] {ts}  klíčové slovo: {kw}")
    print(text[:4000])
    print()

if not hits:
    print("Žádné přesné shody. Zkusíme PREPARE_STATEMENT bloky s FROM tabulkami:")
    print("(viz seznam tabulek nahoře - neobvyklé názvy mohou být klíč)")