Board index » delphi » Last name, First name versus First name Last name

Last name, First name versus First name Last name

Quote
AlanGLLoyd <alangll...@aol.com> wrote in message

news:19991114073246.15736.00000700@ngol07.aol.com...

Quote
> In article <942539134.17624.0.nnrp-07.c2de6...@news.demon.co.uk>, "Jeremy
> Collins" <jer...@kansai.co.uk> writes:

> >Maybe you know five "Jeremy's", and two "Collins's". If you want
> >to look me up, you scan down to the "C's", find "Collins", and hey
> >presto, from a choice of two you find me pretty quickly. It would
> >take you approximately twice as long to find me in the
> >  "forename(s), surname"
> >system.

> Depends entirely on the popularity of the surname and forename - I would
> suggest that Collins is actually more popular than Jeremy - but I don't know.

You might be right - I used my name in case I accidentally
offended anyone else. In the general sense, I'm 100% sure
that, in the Western world at least, there are more surnames
than forenames.

Quote
> Searching for a name (for final choice by selection) is better done by
locating
> or indexing on the soundex detivative of the name (remove vowels, vowel-likes,
> and repeat consonants).

I totally agree. The compilers of telephone books recognise this
to some extent, (for example upon looking "Smith" you might
be prompted to try "Smyth" too), but people who develop
databases rarely do the same.

The last time I checked, it was *way* expensive to license those
kinds of clever search routines - do you know if this is still the
case?

--
Jeremy Collins
Kansai Business Systems
http://www.kansai.co.uk/

 

Re:Last name, First name versus First name Last name


Jeremy Collins <jer...@kansai.co.uk> skrev i en
nyhedsmeddelelse:942587544.17561.0.nnrp-11.c2de6...@news.demon.co.uk...

Quote

> Finn Tolderlund <XnospamYfinn.tolderlu...@Ymobilixnet.dkXnospamY> wrote in
> message news:sDwX3.403$d3.5266@news0.mobilixnet.dk...

> > Jeremy Collins <jer...@kansai.co.uk> skrev i en
> > nyhedsmeddelelse:942539134.17624.0.nnrp-07.c2de6...@news.demon.co.uk...

> > > Finn Tolderlund <XnospamYfinn.tolderlu...@Ymobilixnet.dkXnospamY>
wrote in
> > > message news:JhjX3.352$d3.4665@news0.mobilixnet.dk...
> > > > I have always wondered why names sometimes are stored as "Last name,
> > First
> > > > name".

> > > Indexing.

> > I think you missed my point.
> > My point was the last lines in my post.

> Them I'm confused. I don't see how the name is *displayed*
> can affect the ability to search the database? Do we agree that
> the way data is displayed is a very separate issue to how it is
> stored?

Yes we agree on that.
And you are still missing my point.

Finn Tolderlund
(not Tolderlund, Finn)

Re:Last name, First name versus First name Last name


Jeremy Collins <jer...@kansai.co.uk> skrev i en
nyhedsmeddelelse:942591422.2126.0.nnrp-03.c2de6...@news.demon.co.uk...

Quote
> > Searching for a name (for final choice by selection) is better done by
> locating
> > or indexing on the soundex detivative of the name (remove vowels,
vowel-likes,
> > and repeat consonants).

> I totally agree. The compilers of telephone books recognise this
> to some extent, (for example upon looking "Smith" you might
> be prompted to try "Smyth" too), but people who develop
> databases rarely do the same.

> The last time I checked, it was *way* expensive to license those
> kinds of clever search routines - do you know if this is still the
> case?

You can just make your own routine.
Look at
http://www.delphimag.com/features/1998/03/di199803rs_f/di199803rs_d.asp
for examples.

Finn Tolderlund

Re:Last name, First name versus First name Last name


JRS:  In article <7vtX3.9317$1C4.91...@news1.online.no> of Sun, 14 Nov
1999 08:33:40 in news:comp.lang.pascal.delphi.misc, Kai Inge Buseth

Quote
<kibus...@online.no> wrote:
>in my application I use a separate field for First, Middle and Last name for
>the user input.
>The problem is the table the user had before. His table has the whole string
>in one field, with no separator but a space, in the previous mentioned
>order. Since over four thousand names has to be split up, it has to be done
>by code. As you say, it is impossible to know if the middle parts of the
>string is a double surname or double/ triple  given name. The closest I get,
>is to assume that at least the rightmost word is a surename.

You'll have problems with Queen Elizabeth II, and with both Prince
Philip and the Duke of Edinburgh; and, IIRC, with the entire populations
of China, Japan?, ?, and Hungary.

Family Name and Given Name are fairly safe, and it's probably best to
store the full formal name (Lt-Gen Sir Frederick Smith, OM, PC), and the
form to be used at the beginning of a letter (Sir Frederick, Mr
President, ...)

--
? John Stockton, Surrey, UK.  j...@merlyn.demon.co.uk   Turnpike v4.00   MIME. ?
 Web <URL: http://www.merlyn.demon.co.uk/> - FAQish topics, acronyms, & links.
 Proper 4-line sig. separator is as above, a line exactly "-- " (SonOfRFC1036)
 Do not Mail News to me. Before a reply, quote with ">" or "> " (SonOfRFC1036)

Re:Last name, First name versus First name Last name


In article <942591422.2126.0.nnrp-03.c2de6...@news.demon.co.uk>, "Jeremy

Quote
Collins" <jer...@kansai.co.uk> writes:
>The last time I checked, it was *way* expensive to license those
>kinds of clever search routines - do you know if this is still the
>case?

Soundex was included in dBase 4 and Foxbase, so its pretty old hat. I don't
remember the detailed algorithm but it was more or less as I said, so that
names with a similar sound came together. One might have to vary the algorithm
with diffeent languages (for example "w" is a "v" in German), but English is
fairly straightforward. I think the "funnies" were treating "y" as a vowel,
"ck"  and "c" surrounded by vowels as a "k", "z" as an "s" and similar things.

I also have a faint memory of seeing the algorithm described in an old Delphi
magazine, but which one has gone into my "write-only" memory <g>

Alan Lloyd
alangll...@aol.com

Re:Last name, First name versus First name Last name


On Sat, 13 Nov 1999 23:20:51 +0100, "Guido Festraets"

Quote
<gu...@festra.com> wrote:
>[...]
>(although I don't see any reason to index on _first_ name
>in a business application...?)
>[...]

Have you ever looked in an Icelandic telephone directory?
I suppose they would like their software sort in the same way (on
first name).

------------------------------------
Anders Isaksson, Sweden

BlockCAD: http://user.tninet.se/~hbh828t/proglego.htm
Gallery:  http://user.tninet.se/~hbh828t/gallery.htm

Re:Last name, First name versus First name Last name


AlanGLLoyd <alangll...@aol.com> skrev i en
nyhedsmeddelelse:19991115082539.28555.00000...@ngol02.aol.com...

Quote
> In article <942591422.2126.0.nnrp-03.c2de6...@news.demon.co.uk>, "Jeremy
> Collins" <jer...@kansai.co.uk> writes:

> >The last time I checked, it was *way* expensive to license those
> >kinds of clever search routines - do you know if this is still the
> >case?

> Soundex was included in dBase 4 and Foxbase, so its pretty old hat. I
don't
> remember the detailed algorithm but it was more or less as I said, so that
> names with a similar sound came together. One might have to vary the
algorithm
> with diffeent languages (for example "w" is a "v" in German), but English
is
> fairly straightforward. I think the "funnies" were treating "y" as a
vowel,
> "ck"  and "c" surrounded by vowels as a "k", "z" as an "s" and similar
things.

> I also have a faint memory of seeing the algorithm described in an old
Delphi
> magazine, but which one has gone into my "write-only" memory <g>

Look at
http://www.delphimag.com/features/1998/03/di199803rs_f/di199803rs_d.asp
for examples.

Finn Tolderlund

Re:Last name, First name versus First name Last name


Finn Tolderlund <XnospamYfinn.tolderlu...@Ymobilixnet.dkXnospamY> wrote in
message news:JhjX3.352$d3.4665@news0.mobilixnet.dk...

Quote

> And by the way:
> My name is not Tolderlund, Finn.
> My name is Finn Tolderlund

Not in China

Re:Last name, First name versus First name Last name


There are several problems with breaking a name up into different fields.
Among them is that assumptions have to be made as to the structure of the
names to be stored. Names are very cultural centric and any assumptions that
someone makes is bound to be restrictive in other cultural contexts. The
other significant problem is that you end up allocating a lot of wasted
space and/or creating a situation in which some names cannot be easily or
fully stored.

Several years ago I adopted the practice of having a single, (long), name
field and having users enter western names as surname\given names. (Actually
I support multiple \ so that one can create entries like Smith\John
Edward\Mr. Its very simple to reformat the name for output.) I have not
encountered any resistance from users to this scheme and it offers several
advantages:

1. For the same amount of space as 2/3 separate fields you can accept longer
names, i.e. hyphenated names, multiple middle names, unusually long given
names. In fact the set of unaltered names that can be accepted is very much
larger.
2. Company names can be intermixed with individual names with no special
handling. And, company names that start with a preposition can take
advantage of the scheme, e.g. Great Big Insurance Company\The.
3. Users in parts of the world that normally use a surname given name
convention can use the software without having to reverse the names (which
is what they have to do with the multi field approach).
4. Surnames and given names can be easily extracted if required.
5. With multiple \ support, titles can be handled without any special
processing.

The approach isn't perfect. Some people like to have awards listed after
their name, e.g. 'John Smith, B.A, M.Sc.', which means an entry like 'Smith,
B.A. M.Sc.\John'. Also people that have numerals following their surname
don't get listed immediately after their parent, e.g. William Gates, William
Gates II, and William Gates III do not necessarily appear one after the
other.

Kai Inge Buseth <kibus...@online.no> wrote in message
news:7vtX3.9317$1C4.91404@news1.online.no...

Quote
> Hello Rod,

> in my application I use a separate field for First, Middle and Last name
for
> the user input.
> The problem is the table the user had before. His table has the whole
string
> in one field, with no separator but a space, in the previous mentioned
> order. Since over four thousand names has to be split up, it has to be
done
> by code. As you say, it is impossible to know if the middle parts of the
> string is a double surname or double/ triple  given name. The closest I
get,
> is to assume that at least the rightmost word is a surename.

> --
> Kind regards
> Kai Inge

> {The limit of my Delphi
> is my lack of knowledge}
> Rod Allen wrote in message <382E168B.520AE...@san.rr.com>...
> >Hi,

> >Isolating the rightmost word is not sufficient; what about people with
> surnames
> >such as 'St. John'.  You really need to keep the surname and the given
> name(s)
> >separate internally in your application.  How you present them externally
> to the
> >user is a different matter.

> >Rod

Re:Last name, First name versus First name Last name


Very smart ! It may actually handle names like:

Elisabeth/Queen
Harald/King
Pius 2

in a sensible way..that is very useful...;-)

--
Bjoerge Saether
Consultant / Developer
Asker, Norway
bsaether.removet...@online.no (remove the obvious)

Bruce Roberts skrev i meldingen ...

Quote
>There are several problems with breaking a name up into different
fields.
>Among them is that assumptions have to be made as to the structure of
the
>names to be stored. Names are very cultural centric and any assumptions
that
>someone makes is bound to be restrictive in other cultural contexts.
The
>other significant problem is that you end up allocating a lot of wasted
>space and/or creating a situation in which some names cannot be easily
or
>fully stored.

>Several years ago I adopted the practice of having a single, (long),
name
>field and having users enter western names as surname\given names.
(Actually
>I support multiple \ so that one can create entries like Smith\John
>Edward\Mr. Its very simple to reformat the name for output.) I have not
>encountered any resistance from users to this scheme and it offers
several
>advantages:

>1. For the same amount of space as 2/3 separate fields you can accept
longer
>names, i.e. hyphenated names, multiple middle names, unusually long
given
>names. In fact the set of unaltered names that can be accepted is very
much
>larger.
>2. Company names can be intermixed with individual names with no
special
>handling. And, company names that start with a preposition can take
>advantage of the scheme, e.g. Great Big Insurance Company\The.
>3. Users in parts of the world that normally use a surname given name
>convention can use the software without having to reverse the names
(which
>is what they have to do with the multi field approach).
>4. Surnames and given names can be easily extracted if required.
>5. With multiple \ support, titles can be handled without any special
>processing.

>The approach isn't perfect. Some people like to have awards listed
after
>their name, e.g. 'John Smith, B.A, M.Sc.', which means an entry like
'Smith,
>B.A. M.Sc.\John'. Also people that have numerals following their
surname
>don't get listed immediately after their parent, e.g. William Gates,
William
>Gates II, and William Gates III do not necessarily appear one after the
>other.

>Kai Inge Buseth <kibus...@online.no> wrote in message
>news:7vtX3.9317$1C4.91404@news1.online.no...
>> Hello Rod,

>> in my application I use a separate field for First, Middle and Last
name
>for
>> the user input.
>> The problem is the table the user had before. His table has the whole
>string
>> in one field, with no separator but a space, in the previous
mentioned
>> order. Since over four thousand names has to be split up, it has to
be
>done
>> by code. As you say, it is impossible to know if the middle parts of
the
>> string is a double surname or double/ triple  given name. The closest
I
>get,
>> is to assume that at least the rightmost word is a surename.

>> --
>> Kind regards
>> Kai Inge

>> {The limit of my Delphi
>> is my lack of knowledge}
>> Rod Allen wrote in message <382E168B.520AE...@san.rr.com>...
>> >Hi,

>> >Isolating the rightmost word is not sufficient; what about people
with
>> surnames
>> >such as 'St. John'.  You really need to keep the surname and the
given
>> name(s)
>> >separate internally in your application.  How you present them
externally
>> to the
>> >user is a different matter.

>> >Rod

Go to page: [1] [2]

Other Threads