with file_nm as (
select 'test_csv_2013-11-30_0800_region_1_080012678.csv' test_fl 
 from dual),
inl_query as
(select 'test_csv_2014-05-20_0800_region_1.csv' a
   from dual
  union all
 select 'test_csv_2014-05-20_0800_region_2.csv' a
   from dual
  union all
 select 'test_txt_2014-05-20_0800-region_2.csv' 
   from dual)
select a from inl_query
where <regexp conditions>

SQL> with inl_query as
  2  (select 'test_csv_2014-05-20_0800_region_1.csv' a
  3     from dual
  4    union all
  5   select 'test_csv_2014-05-20_0800_region_2.csv' a
  6     from dual
  7    union all
  8   select 'test_txt_2014-05-20_0800-region_2.csv' 
  9     from dual),
 10    inl_query_ranked as (
 11      select a,
 12             rank() over (
 13               order by 
 14                 case when a like 'test_csv%' then 1
 15                      when a like 'test_txt%' then 2
 16                      else 3
 17                 end,
 18                 case when a like '%region_1.csv' then 1
 19                      when a like '%region_2.csv' then 2
 20                      else 3
 21                 end
 22             ) rk
 23      from inl_query
 24    )
 25  select a from inl_query_ranked where rk = 1
 26  /
A
-------------------------------------
test_csv_2014-05-20_0800_region_1.csv

1 row selected.