regexp_substr help, please

Thread view:

Enable EMail Alerts

Start New Thread

Thread rating:

Doug Miller - 29 Apr 2008 16:34 GMT

I need to be able to pull just the last name out of a string consisting of
lastname and firstname, separated by a comma, or space, or comma and space.
Complicating matters somewhat is the fact that lastname might be something
like "Mc Kay" or "St. Louis" so simply grabbing everything before the first
space isn't sufficient.

The closest I've come so far is
select regexp_substr ('St. Louis, Ted', '.{4}[A-Z]+') from dual;
but this returns only
St. L

I thought regular expression matching was supposed to be "greedy", and take as
many characters as would satisfy the pattern ("St. Louis" in this case).

What am I doing wrong?

Reply to this Message

md - 29 Apr 2008 17:33 GMT

> I need to be able to pull just the last name out of a string consisting of
> lastname and firstname, separated by a comma, or space, or comma and space.
[quoted text clipped - 11 lines]
>
> What am I doing wrong?

These are you combos?

ln,fn|ln, fn|ln fn
l n,fn|l n, fn|l n fn

I'd start by nuking the fn. It's the last solid string and it has an
nice anchor fn$

Reply to this Message

Malcolm Dew-Jones - 29 Apr 2008 17:46 GMT

: I need to be able to pull just the last name out of a string consisting of
: lastname and firstname, separated by a comma, or space, or comma and space.
: Complicating matters somewhat is the fact that lastname might be something
: like "Mc Kay" or "St. Louis" so simply grabbing everything before the first
: space isn't sufficient.

: The closest I've come so far is
: select regexp_substr ('St. Louis, Ted', '.{4}[A-Z]+') from dual;
: but this returns only
: St. L

[A-Z] doesn't match o

Reply to this Message

md - 29 Apr 2008 17:52 GMT

On Apr 29, 12:43 pm, yf...@vtn1.victoria.tc.ca (Malcolm Dew-Jones)
wrote:

> : I need to be able to pull just the last name out of a string consisting of
> : lastname and firstname, separated by a comma, or space, or comma and space.
[quoted text clipped - 8 lines]
>
> [A-Z] doesn't match o

Here a perl test I did to check out a maybe.
@ is an array of your combos (did I get them all?)
The loop goes thru each. $1 will contain the "last name".

@s = ("mc winter, first",
"mc. winter, first",
"winter, first",
"mc winter,first",
"mc. winter,first",
"winter,first",
"mc winter first",
"mc. winter first",
"winter first",
);

foreach $x (@s) {
print qq/look at "$x"\n/;
{
$x =~ /(.*)?[, ]{1}[a-zA-Z]*$/;
print $1 . "\n";
}
}

Reply to this Message

Doug Miller - 29 Apr 2008 17:56 GMT

>: I need to be able to pull just the last name out of a string consisting of
>: lastname and firstname, separated by a comma, or space, or comma and space.
[quoted text clipped - 8 lines]
>
>[A-Z] doesn't match o

Yeah, I just got back from lunch and realized the same thing. Knew it had to
be something stupid like that. Thanks.

Reply to this Message

Peter Nilsson - 30 Apr 2008 04:17 GMT

> I need to be able to pull just the last name out of a string
> consisting of lastname and firstname, separated by a
[quoted text clipped - 7 lines]
> but this returns only
> St. L

Start with a 'student' aproach...

with
names as
(
select 'mc winter, first' nme from dual union all
select 'mc. winter, first' nme from dual union all
select 'winter, first second' nme from dual union all
select 'mc winter,first' nme from dual union all
select 'mc. winter,first' nme from dual union all
select 'winter,first' nme from dual union all
select 'mc winter first second' nme from dual union all
select 'mc. winter first' nme from dual union all
select 'winter first' nme from dual union all
select 'macwilliams' nme from dual
)
select
nme,
regexp_replace
(
trim(nme),
'^'
|| '(' -- with comma
|| '([^,]*)' -- surname
|| ', *'
|| '(.*)' -- given name(s)
|| ')'
|| '|'
|| '(' -- without comma
|| '(' -- surname
|| '((mc|st)\.?)? *' -- optional prefix
|| '[^ ]+'
|| ')'
|| ' *'
|| '(.*)' -- given name(s)
|| ')'
|| '$',
'"\2\5", "\3\8"', -- "surname", "given"
1, 0, 'i' -- case insensitive
)
from
names

--
Peter

Reply to this Message