feat: add autonomous fiscal completion v1

scripts/analyze_inbound_observations.py

@@ -0,0 +1,82 @@
+"""Read-only extraction of populated supplier-item fiscal observations."""
+
+from __future__ import annotations
+
+import hashlib
+import json
+import os
+import sys
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[1]
+sys.path.insert(0, str(ROOT))
+
+from central.app.evidence_aggregation import Observation, aggregate_observations  # noqa: E402
+
+
+def _hash(value: object) -> str:
+    return hashlib.sha256(str(value).encode("utf-8")).hexdigest()[:16]
+
+
+def summarize_rows(rows: list[tuple[object, ...]]) -> dict:
+    observations = []
+    for product, supplier, observed_at, ncm, cfop, cst in rows:
+        identity = _hash(product)
+        supplier_id = _hash(supplier)
+        document = _hash(f"{supplier}:{product}:{observed_at}")
+        for attribute, value in (("ncm", ncm), ("cfop", cfop), ("cst", cst)):
+            if value not in (None, ""):
+                observations.append(Observation(identity, attribute, str(value).strip(), supplier_id, document, str(observed_at), "SUPPLIER_OBSERVATION"))
+    aggregate = aggregate_observations(observations)
+    return {
+        "observation_count": len(observations),
+        "aggregate_count": len(aggregate),
+        "sources": ["SUPPLIER_OBSERVATION"] if observations else [],
+        "strong_evidence_count": sum(1 for row in aggregate.values() if row["strong_evidence"]),
+        "conflict_count": sum(row["conflict_count"] for row in aggregate.values()),
+    }
+
+
+def _env() -> None:
+    path = ROOT / ".env"
+    if path.exists():
+        for line in path.read_text(encoding="utf-8").splitlines():
+            if "=" in line and not line.lstrip().startswith("#"):
+                key, _, value = line.partition("=")
+                os.environ.setdefault(key.strip(), value.strip())
+
+
+def analyze_database(path: Path) -> dict:
+    import fdb
+
+    conn = fdb.connect(database=str(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)
+    try:
+        cur = conn.cursor()
+        cur.execute("SELECT TRIM(RDB$FIELD_NAME) FROM RDB$RELATION_FIELDS WHERE RDB$RELATION_NAME = 'C000088'")
+        fields = {row[0] for row in cur.fetchall()}
+        required = {"CODPRODUTO", "CODFORNECEDOR", "DATA", "CLASSIFICACAO_FISCAL", "CFOP", "CST"}
+        if not required <= fields:
+            return {"database": path.name, "status": "NO_COMPLETE_INBOUND_ITEM_SHAPE", "fields_available": sorted(fields)}
+        cur.execute("SELECT CODPRODUTO, CODFORNECEDOR, DATA, CLASSIFICACAO_FISCAL, CFOP, CST FROM C000088")
+        rows = cur.fetchall()
+        summary = summarize_rows(rows)
+        return {"database": path.name, "status": "ANALYZED", "items_analyzed": len(rows), **summary, "origin_field_available": False, "origin_resolution": "ORIGIN_NEEDS_EVIDENCE"}
+    finally:
+        conn.close()
+
+
+def main() -> None:
+    _env()
+    paths = [ROOT / "TESTESOFICIAIS.FDB"] + sorted((ROOT / "Banco").glob("*.FDB"))
+    result = {"artifact": "INBOUND_SUPPLIER_OBSERVATIONS_V1", "mode": "READ_ONLY", "firebird_writes": 0, "databases": []}
+    for path in paths:
+        try:
+            result["databases"].append(analyze_database(path))
+        except Exception as exc:
+            result["databases"].append({"database": path.name, "status": "UNAVAILABLE", "error_type": type(exc).__name__})
+    (ROOT / "docs/INBOUND_SUPPLIER_OBSERVATIONS_V1.json").write_text(json.dumps(result, ensure_ascii=False, indent=2) + "\n", encoding="utf-8")
+    print(json.dumps({"artifact": result["artifact"], "databases": len(result["databases"]), "firebird_writes": 0}))
+
+
+if __name__ == "__main__":
+    main()