What is the Oracle equivalent of the Microsoft Access FIRST function?

Question

Using: Oracle 10gR2 RAC on SUSE Linux 9 (10.2.0.3)
In the process of converting a Microsoft Access database to Oracle, an Access query is using the FIRST function.
What is the Oracle equivalent of the Microsoft Access FIRST function?
In the attempt to convert, the Oracle FIRST_VALUE function was used. However, the same results was not achieved.
Thanks,
(BLL)
Query:
h2. ACCESS:
SELECT
     TRE.GCUSNO,
     UCASE([DCUSNO]) AS DCUSNO_STD,
     *FIRST(UCASE([DNAME])) AS DNAME_STD*,
     *FIRST(UCASE([DADDR])) AS DADDR_STD*,
     *FIRST(UCASE([DCITY])) AS DCITY_STD*,
     TRE.DSTATE,
     FIRST(TRE.DZIP) AS DZIP,
     TRE.DREGN,
     TRE.DDIST,
     TRE.DSLSMN,
     TRE.DCHAIN,
     TRE.MARKET,
     TRE.MKTPGM,
     TRE.EUMKT
FROM
     TRE
GROUP BY
     TRE.GCUSNO,
     UCASE([DCUSNO]),
     TRE.DSTATE,
     TRE.DREGN,
     TRE.DDIST,
     TRE.DSLSMN,
     TRE.DCHAIN,
     TRE.MARKET,
     TRE.MKTPGM,
     TRE.EUMKT;
h2. ORACLE:
SELECT DISTINCT
TRE.GCUSNO,
UPPER(TRIM(TRE.DCUSNO)) AS DCUSNO_STD,
UPPER(TRIM(TRE.DNAME)) AS DNAME_STD,
UPPER(TRIM(TRE.DADDR)) AS DADDR_STD,
     FIRST_VALUE(UPPER(TRIM(TRE.DNAME)) IGNORE NULLS) OVER (ORDER BY TRE.GCUSNO) AS DNAME_STD,
     FIRST_VALUE(UPPER(TRIM(TRE.DADDR)) IGNORE NULLS) OVER (ORDER BY TRE.GCUSNO) AS DADDR_STD,
     FIRST_VALUE(UPPER(TRIM(TRE.DCITY)) IGNORE NULLS) OVER (ORDER BY TRE.GCUSNO) AS DCITY_STD,
TRE.DSTATE,
TRE.DZIP,
FIRST_VALUE(UPPER(TRIM(TRE.DZIP)) IGNORE NULLS) OVER (ORDER BY TRE.DZIP ASC) AS DZIP,
TRE.DREGN,
TRE.DDIST,
TRE.DSLSMN,
TRE.DCHAIN,
TRE.MARKET,
TRE.MKTPGM,
TRE.EUMKT
FROM CRM.TREUP100R TRE
GROUP BY
TRE.GCUSNO,
UPPER(TRIM(TRE.DCUSNO)),
TRE.DNAME,
TRE.DADDR,
TRE.DCITY,
TRE.DSTATE,
TRE.DZIP,
TRE.DREGN,
TRE.DDIST,
TRE.DSLSMN,
TRE.DCHAIN,
TRE.MARKET,
TRE.MKTPGM,
TRE.EUMKT;

Answer 1

A slight correction to odie's post. I think you want min not max to replicate the Access first function, but see below to be sure. So:
min(upper(trim(tre.dname))) keep (dense_rank first order by tre.gcusno) as dname_std
user10860953 wrote:How does one ignore null values?The min and max functions will ignore nulls automatically, so if there is a null value in tre.dname, it will not be be returned, unless all of the values are null. For example:
SQL> WITH t AS (
  2     SELECT 65 id, 'ABCD' col FROM dual UNION ALL
  3     SELECT 37, 'DEFG' FROM dual UNION ALL
  4     SELECT 65, 'DEFG' FROM dual UNION ALL
  5     SELECT 65, null FROM dual UNION ALL
  6     SELECT 70, null FROM dual UNION ALL
  7     SELECT 70, null FROM dual UNION ALL
  8     SELECT 37, 'ABC' from dual)
  9  SELECT id,
10         MIN(col) keep (DENSE_RANK FIRST ORDER BY id) min_dname_std,
11         MAX(col) keep (DENSE_RANK FIRST ORDER BY id) max_dname_std
12  FROM t
13  GROUP BY id;
        ID MIN_ MAX_
        37 ABC  DEFG
        65 ABCD DEFG
        70John