Fichiers
mesure-audience/aggregate.py
T

50 lignes
2.0 KiB
Python

"""Agrégation quotidienne : SQLite -> bloc JSON pour data.json (stats.gastinel.me).
Purge : lignes brutes > RAW_DAYS jours, sels des mois révolus (les identifiants
des mois passés deviennent alors définitivement non reliables).
Les agrégats (des compteurs, sans identifiant) peuvent être gardés 25 mois.
"""
import argparse, json, sqlite3, time
from collector import open_db
RAW_DAYS = 62 # couvre le mois en cours + le précédent (visiteurs uniques mensuels)
def purge(db, now=None):
now = time.time() if now is None else now
cutoff = time.strftime("%Y-%m-%d", time.gmtime(now - RAW_DAYS * 86400))
db.execute("DELETE FROM hits WHERE day < ?", (cutoff,))
t = time.gmtime(now)
prev = f"{t.tm_year - (t.tm_mon == 1)}-{(t.tm_mon - 2) % 12 + 1:02d}"
db.execute("DELETE FROM salts WHERE month < ?", (prev,))
def aggregate(db):
q = lambda sql: db.execute(sql).fetchall()
return {
"methode": "uid-localstorage-v1",
"jours": {d: {"visiteurs": v, "pages_vues": p} for d, v, p in
q("SELECT day, COUNT(DISTINCT vid), COUNT(*) FROM hits GROUP BY day ORDER BY day")},
"mois": {m: {"visiteurs": v, "pages_vues": p} for m, v, p in
q("SELECT substr(day,1,7) m, COUNT(DISTINCT vid), COUNT(*) FROM hits GROUP BY m ORDER BY m")},
"sources": dict(q("SELECT src, COUNT(*) FROM hits GROUP BY src ORDER BY 2 DESC")),
"pages": dict(q("SELECT path, COUNT(*) FROM hits GROUP BY path ORDER BY 2 DESC LIMIT 20")),
"appareils": dict(q("SELECT dev, COUNT(DISTINCT vid) FROM hits GROUP BY dev")),
"navigateurs": dict(q("SELECT nav, COUNT(DISTINCT vid) FROM hits GROUP BY nav")),
}
def main():
ap = argparse.ArgumentParser()
ap.add_argument("--db", default="audience.sqlite3")
ap.add_argument("--out", default="audience.json")
a = ap.parse_args()
db = open_db(a.db)
purge(db)
with open(a.out, "w", encoding="utf-8") as f:
json.dump(aggregate(db), f, ensure_ascii=False, indent=1)
if __name__ == "__main__":
main()