Oracle SQL REGEX_LIKE - sql

Question

SELECT first_name, last_name
FROM employees
WHERE REGEXP_LIKE (first_name, '^Ste(v|ph)en$');
The following query returns the first and last names for those employees with a first name of Steven or Stephen (where first_name begins with Ste and ends with en and in between is either v or ph)
is there a call that is opposite where the query will return everything that would not have (v or ph) between Ste and en?
so that it would return things like:
Stezen
Stellen
is it as simple as putting NOT in front of REGEXP_LIKE?

Answer 1

How about MINUS
SELECT *
FROM employees
WHERE REGEXP_LIKE( first_name , '^Ste([[:alpha:]])+en$')
MINUS
SELECT *
FROM employees
WHERE REGEXP_LIKE( first_name , '^Ste(v|ph)en$');
and this too:
WITH t AS
( SELECT 'Stezen' first_name FROM dual
UNION ALL
SELECT 'Steven' FROM dual
UNION ALL
SELECT 'Stephen' FROM dual
)
SELECT *
FROM t
WHERE REGEXP_LIKE( first_name , '^Ste([[:alpha:]])+en$')
AND NOT REGEXP_LIKE( first_name , '^Ste(v|ph)en$');

Answer 2

You need something like this:
SELECT 'Match'
FROM dual
WHERE REGEXP_LIKE ('Steden', '^Ste[^(v|ph)]en$');
EDIT
This will exclude any two (or more) letter combinations but still allow "v" :
SELECT 'Match'
FROM dual
WHERE REGEXP_LIKE ('Stephen', '^Ste[[:alpha:]]en$');
Since Oracle does not support look-ahead functionality, I will have to agree with others that we will have to deal with "v" explicitly, either by excluding the entire name(word) or at least specifying its exact position.
SELECT name
FROM WhateverTable
WHERE REGEXP_LIKE (name, '^Ste[[:alpha:]]en$') AND SUBSTR(name, 4, 1) <> 'v';

Answer 3

Two options:
The first query uses two REGEXP_LIKE tests: one regular expression to generically match; and one for excluding the invalid matches.
The second query uses REGEXP_SUBSTR to testfor a generic match and extract the sub-group of the match and then tests to see whether it should be exluded.
The third query then looks at how you can extend the query by having another table containing the match criteria and allows you to build and test multiple name variants.
SQL Fiddle
Oracle 11g R2 Schema Setup:
CREATE TABLE tbl ( str ) AS
SELECT 'Stephen' FROM DUAL
UNION ALL SELECT 'Steven' FROM DUAL
UNION ALL SELECT 'Stepen' FROM DUAL
UNION ALL SELECT 'Steephen' FROM DUAL
UNION ALL SELECT 'Steeven' FROM DUAL
UNION ALL SELECT 'Steeven' FROM DUAL
UNION ALL SELECT 'Smith' FROM DUAL
UNION ALL SELECT 'Smithe' FROM DUAL
UNION ALL SELECT 'Smythe' FROM DUAL
UNION ALL SELECT 'Smythee' FROM DUAL;
CREATE TABLE exclusions ( prefix, exclusion, suffix ) AS
SELECT 'Ste', 'v|ph', 'en' FROM DUAL
UNION ALL SELECT 'Sm', 'ithe?|ythe', '' FROM DUAL;
Query 1:
SELECT str
FROM tbl
WHERE REGEXP_LIKE( str, '^Ste(\w+)en$' )
AND NOT REGEXP_LIKE( str, '^Ste(v|ph)en$' )
Results:
| STR |
|----------|
| Stepen |
| Steephen |
| Steeven |
| Steeven |
Query 2:
SELECT str
FROM (SELECT str,
REGEXP_SUBSTR( str, '^Ste(\w+)en$', 1, 1, NULL, 1 ) AS match
FROM tbl)
WHERE match IS NOT NULL
AND NOT REGEXP_LIKE( match, '^(v|ph)$' )
Results:
| STR |
|----------|
| Stepen |
| Steephen |
| Steeven |
| Steeven |
Query 3:
SELECT str
FROM tbl t
WHERE EXISTS ( SELECT 1
FROM exclusions e
WHERE REGEXP_LIKE( t.str, '^' || e.prefix || '(\w+)' || e.suffix || '$' )
AND NOT REGEXP_LIKE( t.str, '^' || e.prefix || '(' || e.exclusion || ')' || e.suffix || '$' )
)
Results:
| STR |
|----------|
| Stepen |
| Steephen |
| Steeven |
| Steeven |
| Smythee |