Database design - Multiple "Contact information" for different tables

Question

I have a database with tables "person", "company", "shop", etc. Many of these tables must have "contact information". The possibility to design this was asked in Database design - Similar Contact Information for multiple entities Now, in my database I leave a possibility to have a multiple addresses, multiple phones and multiple emails to each contact data. This is my database schema:

So, I make an intermediate table "contact" as a simplest way to link a "contact information" to each table.
My question: is it a good practice to do this and to have a table with only one row?

Answer 1

This is how I would design your database:

address_types
    id              unsigned int(P)
    description     varchar(10) // Mailing, Physical, etc.

addresses
    id              unsigned int(P)
    line1           varchar(50) // 123 Main Street, etc.
    line2           varchar(50) // Default NULL
    city_id         unsigned int(F cities.id)
    zip             varchar(6) // 12345, A1A 1A1, etc.
    zip4            char(4) // Default NULL
    lat             decimal(10,8) // 13.12345678, etc.
    lon             decimal(11,8) // 110.12345678, etc.

cities
    id              unsigned int(P)
    state_id        unsigned int(F states.id)
    name            varchar(50) // Omaha, Detroit, Tampa, etc.

companies
    id              unsigned int(P)
    name            varchar(75) // IBM, Microsoft, RedHat, etc.
    ...

companies_addresses
    id                  unsigned int(P)
    company_id          unsigned int(F companies.id)
    address_id          unsigned int(F addresses.id)
    address_type_id     unsigned int(F address_types.id)

companies_contacts
    id                  unsigned int(P)
    company_id          unsigned int(F companies.id)
    contact_id          unsigned int(F contacts.id)
    contact_type_id     unsigned int(F contact_types.id)

companies_emails
    id                  unsigned int(P)
    company_id          unsigned int(F companies.id)
    email_id            unsigned int(F emails.id)
    email_type_id       unsigned int(F email_types.id)

contact_types
    id              unsigned int(P)
    description     varchar(10) // Home phone, Mobile phone, FAX, etc.

In North America phone numbers look like this: CC-AAA-EEE-SSSS-XXXXXXX where CC is the country code, AAA is the area code, EEE is the exchange, SSSS is the station and XXXXX is the extension.

contacts
    id              unsigned int(P)
    country_code    varchar(3)
    area_code       varchar(3)
    exchange        varchar(3)
    station         varchar(4)
    extension       varchar(10) // Default NULL

See ISO 3166-1.

countries
    id              char(2) // ca, mx, us, etc.
    iso3            char(3) // can, mex, usa, etc.
    iso_num         char(3)
    name            varchar(44) // Canada, Mexico, United States, etc.

email_types
    id              unsigned int(P)
    description     varchar(10) // Personal, Work, etc.

emails
    id              unsigned int(P)
    address         varchar(255) // [email protected], etc.

shops
    id              unsigned int(P)
    name            varchar(45) // Shop A, Shop B, etc.
    ...

shops_addresses
    id                  unsigned int(P)
    shop_id             unsigned int(F shops.id)
    address_id          unsigned int(F addresses.id)
    address_type_id     unsigned int(F address_types.id)

shops_contacts
    id                  unsigned int(P)
    shop_id             unsigned int(F shops.id)
    contact_id          unsigned int(F contacts.id)
    contact_type_id     unsigned int(F contact_types.id)

shops_emails
    id                      unsigned int(P)
    shop_id                 unsigned int(F shops.id)
    email_id                unsigned int(F emails.id)
    email_type_id           unsigned int(F email_types.id)

See ISO 3166-2.

states
    id              unsigned int(P)
    country_id      char(2)(F countries.id)
    code            varchar(2) // AL, NF, NL, etc.
    name            varchar(50) // Alabama, Newfoundland, Nuevo León, etc.

Answer 2

My question: is it a good practice to do this and to have a table with only one row?

Not really. Looking at your diagram, I have to ask : can a contact really be linked to an arbitrary number of persons ? If not, you should use 'person' as your parent table and make the other tables link to it.

Answer 3

Personaly, i don't like realy this method because we need to create any Entity for each table and a lot of associative tables. I propose one table for contact informations and another one for addresses.

informations
- id INT
- name VARCHAR
- value VARCHAR
- type ENUM(phone, email, url, custom)
- idcompany INT NULL
- idcontact INT NULL

And

addresses
- id
- address1
- address2
- district
- postcode
- idcity
- idcompany
- idcontact