Oracle 11g Release 2

By using db<>fiddle, you agree to license everything you submit by Creative Commons CC0.
create table leg (datainizio) as
          select date '2018-03-23' from dual
union all select date '2013-03-15' from dual;
2 rows affected

create table par (datainizio) as
          select date '2018-04-02' from dual
union all select date '2017-03-04' from dual
union all select date '2016-03-17' from dual;
3 rows affected

select datainizio as date_from,
  coalesce(lead(datainizio) over (order by datainizio), sysdate) as date_to
from leg;


DATE_FROM
DATE_TO


15-MAR-13
23-MAR-18

23-MAR-18
28-JUN-18


with ranges as (
  select datainizio as date_from,
    coalesce(lead(datainizio) over (order by datainizio), sysdate) as date_to
  from leg
)
select r.date_from, r.date_to - 1/86400 as date_to, count(p.datainizio)
from ranges r
left join par p
on p.datainizio >= r.date_from
and p.datainizio < r.date_to
group by r.date_from, r.date_to
order by r.date_from;


DATE_FROM
DATE_TO
COUNT(P.DATAINIZIO)


15-MAR-13
22-MAR-18
2

23-MAR-18
28-JUN-18
1


with ranges as (
  select datainizio as date_from,
    lead(datainizio) over (order by datainizio) as date_to
  from leg
)
select to_char(r.date_from, 'DD-MON-RR') || ' - ' ||
  case when date_to is null then 'sysdate'
       else to_char(date_to - 1/86400, 'DD-MON-RR')
  end as range,
  count(p.datainizio)
from ranges r
left join par p
on p.datainizio >= r.date_from
and p.datainizio < coalesce(r.date_to, sysdate)
group by r.date_from, r.date_to
order by r.date_from;


RANGE
COUNT(P.DATAINIZIO)


15-MAR-13 - 22-MAR-18
2

23-MAR-18 - sysdate
1