23 lines
1.4 KiB
Python
23 lines
1.4 KiB
Python
"""Read-only: per-column stats for a table. usage: inspect_erp_table.py TABLE [regex-of-cols]"""
|
|
import os, re, sys
|
|
from pathlib import Path
|
|
ROOT = Path(__file__).resolve().parents[1]
|
|
sys.path.insert(0, str(ROOT / "agent" / "src"))
|
|
from gti_fiscal_agent.firebird_reader import _connect
|
|
for l in (ROOT / ".env").read_text(encoding="utf-8").splitlines():
|
|
if "=" in l and not l.startswith("#"):
|
|
k, _, v = l.partition("="); os.environ.setdefault(k.strip(), v.strip())
|
|
t = sys.argv[1]; rx = re.compile(sys.argv[2], re.I) if len(sys.argv) > 2 else None
|
|
conn = _connect(os.environ); cur = conn.cursor()
|
|
cur.execute("select trim(rf.RDB$FIELD_NAME), f.RDB$FIELD_TYPE, coalesce(f.RDB$CHARACTER_LENGTH,f.RDB$FIELD_LENGTH) from RDB$RELATION_FIELDS rf join RDB$FIELDS f on f.RDB$FIELD_NAME=rf.RDB$FIELD_SOURCE where rf.RDB$RELATION_NAME=? order by rf.RDB$FIELD_POSITION", (t,))
|
|
for name, ty, ln in cur.fetchall():
|
|
if rx and not rx.search(name): continue
|
|
if ty in (261,): # blob
|
|
print(f"{name} BLOB"); continue
|
|
cur.execute(f'select "{name}", count(*) from "{t}" group by "{name}" order by 2 desc rows 1 to 400')
|
|
rows = cur.fetchall()
|
|
nn = sum(c for v, c in rows if v is not None)
|
|
top = [(str(v).strip()[:24] if v is not None else None, c) for v, c in rows[:5]]
|
|
print(f"{name} t{ty}/{ln} nn={nn} distinct{'>=' if len(rows)>=400 else '='}{len(rows)} {top}")
|
|
conn.close()
|