Add pure SQLite approach, with generate_series

bench.sh

@@ -2,6 +2,14 @@
 # runs each of the scripts one after another, prints the measurements to stdout
 export TZ=":Asia/Kolkata"
 
+# benching sqlite cli
+rm -rf sqlite3.db
+echo "$(date)" "[SQLite] running sqlite3 (100_000_000) inserts"
+time sqlite3 sqlite3.db '.read load.sql'
+if [[ $(sqlite3 sqlite3.db  "select count(*) from user";) != 100000000 ]]; then
+  echo "data verification failed"
+fi
+
 # benching naive version
 rm -rf naive.db naive.db-shm naive.db-wal
 echo "$(date)" "[PYTHON] running naive.py (10_000_000) inserts"

load.sql

@@ -0,0 +1,13 @@
+create table IF NOT EXISTS user (
+  id INTEGER not null primary key,
+  area CHAR(6),
+  age INTEGER not null,
+  active INTEGER not null
+);
+
+insert into user (area, age, active)
+select 
+  abs(random() % 99999) as area, 
+  (abs(random() % 3) + 1) * 5 as age, 
+  abs(random() % 1) as active
+from generate_series(1, 100000000);