GTI_GERENCIADOR_FISCAL/scripts/bloco2_pis_cofins_cross_db.py

"""Missão 2 / Bloco 2.1 — evidência cruzada PIS/COFINS cadastro-vs-emissão, somente leitura, nos 6 bancos.
Nunca escreve. Não decide regra fiscal: só conta e cruza fatos.
"""
import json
import os
import sys
from pathlib import Path
from collections import Counter

ROOT = Path(__file__).resolve().parents[1]
sys.path.insert(0, str(ROOT)); sys.path.insert(0, str(ROOT / "scripts")); sys.path.insert(0, str(ROOT / "agent" / "src"))
import fdb  # noqa: E402
import run_fiscal_decision_v1 as base  # noqa: E402
from gti_fiscal_agent import company_context as cc  # noqa: E402

DBS = {
    "GTI": (r"D:\PROJETOS\GTISISTEMAS\Banco\GTI.FDB", {}),
    "TESTESOFICIAIS": (str(ROOT / "TESTESOFICIAIS.FDB"), {}),
    "CONSTRUCAO": (str(ROOT / "Banco" / "CONSTRUCAO.FDB"), {"company_code": "000001"}),
    "MMMODAS": (str(ROOT / "Banco" / "MMMODAS.FDB"), {"company_code": "000001", "crt_source": "EMISSOR"}),
    "RINCAO": (str(ROOT / "Banco" / "RINCAO.FDB"), {"company_code": "000001", "crt_source": "EMISSOR"}),
    "UNIAO": (str(ROOT / "Banco" / "UNIAO.FDB"), {"company_code": "000001", "crt_source": "EMISSOR"}),
}


def connect(path):
    return fdb.connect(database=path, user=os.environ["FB_USER"], password=os.environ["FB_PASSWORD"],
                        fb_library_name=os.environ.get("FIREBIRD_CLIENT_LIBRARY"), isolation_level=fdb.ISOLATION_LEVEL_READ_COMMITED_RO)


def analyze_one(name, path, cfg):
    conn = connect(path); cur = conn.cursor()
    company = cc.resolve(cc.collect_evidence(cur), cc.config_from_env({("COMPANY_CODE" if k == "company_code" else "COMPANY_CRT_SOURCE"): v for k, v in cfg.items()}))
    out = {"name": name, "company_status": company["status"], "crt": company.get("company_crt"), "uf": company.get("company_uf")}
    try:
        cur.execute("SELECT COUNT(*), COUNT(NULLIF(TRIM(CST_PIS_COFINS),'')) FROM NFCE_ITENS")
        tot, nonblank = cur.fetchone()
        out["items_total"] = tot; out["items_cst_pis_cofins_filled"] = nonblank
        out["pct_cst_filled"] = round(100.0 * nonblank / tot, 2) if tot else None
    except Exception as e:  # noqa: BLE001
        out["items_error"] = str(e)[:150]; conn.close(); return out
    cur.execute("SELECT TRIM(CST_PIS_COFINS), COUNT(*) FROM NFCE_ITENS WHERE CST_PIS_COFINS IS NOT NULL GROUP BY 1 ORDER BY 2 DESC")
    out["cst_distribution"] = {(r[0] or "(blank)"): r[1] for r in cur.fetchall()}
    try:
        cur.execute("SELECT EXTRACT(YEAR FROM DATA), COUNT(*), COUNT(NULLIF(TRIM(CST_PIS_COFINS),'')) FROM NFCE_ITENS GROUP BY 1 ORDER BY 1")
        out["by_year"] = {int(y): {"items": t, "cst_filled": f} for y, t, f in cur.fetchall() if y is not None}
    except Exception as e:  # noqa: BLE001
        out["by_year_error"] = str(e)[:150]
    try:
        cur.execute("SELECT ROUND(ALIQUOTA_PIS,4), COUNT(*) FROM NFCE_ITENS WHERE ALIQUOTA_PIS IS NOT NULL GROUP BY 1 ORDER BY 2 DESC")
        out["aliquota_pis_distribution"] = {str(r[0]): r[1] for r in cur.fetchmany(10)}
        cur.execute("SELECT ROUND(ALIQUOTA_COFINS,4), COUNT(*) FROM NFCE_ITENS WHERE ALIQUOTA_COFINS IS NOT NULL GROUP BY 1 ORDER BY 2 DESC")
        out["aliquota_cofins_distribution"] = {str(r[0]): r[1] for r in cur.fetchmany(10)}
    except Exception as e:  # noqa: BLE001
        out["aliquota_error"] = str(e)[:150]
    # cadastro vs emissão, apenas nos itens cujo NCM emitido == NCM cadastrado (join identity-consistent, evita o viés de 2022-12 já documentado)
    try:
        cur.execute("""
            SELECT COUNT(*),
                   SUM(CASE WHEN TRIM(COALESCE(p.CLASSIFICACAO_FISCAL,'')) != '' AND p.CLASSIFICACAO_FISCAL = i.NCM THEN 1 ELSE 0 END)
            FROM NFCE_ITENS i JOIN C000025 p ON p.CODIGO = i.CODPRODUTO
        """)
        joined_tot, identity_consistent = cur.fetchone()
        out["identity_consistent_items"] = identity_consistent
        cur.execute("""
            SELECT COUNT(*)
            FROM NFCE_ITENS i JOIN C000025 p ON p.CODIGO = i.CODPRODUTO
            WHERE p.CLASSIFICACAO_FISCAL = i.NCM
              AND COALESCE(TRIM(p.CST_ENTRADA),'') = '' -- produto sem CST/PIS cadastral (a maioria)
              AND COALESCE(TRIM(i.CST_PIS_COFINS),'') != ''
        """)
        out["emitted_cst_present_when_cadastro_blank"] = cur.fetchone()[0]
    except Exception as e:  # noqa: BLE001
        out["cadastro_vs_emissao_error"] = str(e)[:150]
    conn.close()
    return out


def main():
    base.load_dotenv(ROOT / ".env")
    results = {}
    for name, (path, cfg) in DBS.items():
        print(f"== {name}", flush=True)
        try:
            results[name] = analyze_one(name, path, cfg)
        except Exception as e:  # noqa: BLE001
            results[name] = {"name": name, "error": str(e)[:300]}
        print(json.dumps(results[name], ensure_ascii=False, indent=1, default=str)[:800], flush=True)
    out_path = ROOT / "docs" / "runs" / "BLOCO2_PIS_COFINS_CROSS_DB_V1.json"
    out_path.write_text(json.dumps(results, ensure_ascii=False, indent=1, default=str), encoding="utf-8")
    print(f"-> {out_path}")


if __name__ == "__main__":
    main()