File size: 473 Bytes
e7543dc
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
import json
import sys
from pathlib import Path

p = Path(sys.argv[1])
rows = [json.loads(x) for x in p.read_text(encoding='utf-8').splitlines() if x.strip()]
ents = sum(len(r.get('entities', [])) for r in rows)
docs = len(rows)
print('%s docs=%d entities=%d' % (p.name, docs, ents))
if len(sys.argv) > 2:
    target = float(sys.argv[2])
    lo, hi = int(target * 0.5), int(target * 1.5)
    print('band=[%d,%d] %s' % (lo, hi, 'OK' if lo <= ents <= hi else 'OUT_OF_BAND'))