Parse and extract a Canadian Address and Phone# from a Text Varchar2 field

Question

Happy Remembrance/Veterans Day everyone.
I have a "data migration" issue in my hand and there is a Text field that now needs to be organized nicely into proper address fields.
So far, I have been able to parse the Postal Code correctly from the text field which is storing the address.
SELECT code, address, regexp_substr(address,'[A-Z]\d[A-Z] *\d[A-Z]\d') postal_code
FROM student_tb
WHERE REGEXP_LIKE(address,'[V]\d[A-Z] *\d[A-Z]\d');
I still have to extract the Address and sometimes a Phone number which is all stored in one field. Here are 4 examples and acceptable formats:
#2115-101 Adelaide Street East Toronto ON M5A 1N1 [phone]
Testing ABC #2101-212 King Street West, Toronto, ON, M5A 4T1
#3114-212 Bloor Street E      Toronto    ON   M3C1S1  651-1213
HOTEL PANAORAMA      Waterloo    ON   M1A 2A1
Script for table creations:
create table student_tb
    code     varchar2(4),
    address varchar2(500)
create table student_address_tb
  code           varchar2(4),
  address_type   varchar2(1),
  hotel_name     varchar2(50),
  street         varchar2(50),
  city           varchar2(50),
  province       varchar2(2),
  postal_code    varchar2(10),
  phone          varchar2(10)
Inserting records into table student_tb:
insert into student_tb values('1001','#2115-101 Adelaide Street East Toronto ON M5A 1N1 [phone]');
insert into student_tb values('1002','Testing ABC #2101-212 King Street West, Toronto, ON, M5A 4T1');
insert into student_tb values('1003','#3114-212 Bloor Street E      Toronto    ON   M3C1S1  651-1213');
insert into student_tb values('1004','#HOTEL PANAORAMA
Toronto
ON   M1A 2A1');
create table city_gis(city_name varchar2(50));
insert into city_gis values('Toronto');
insert into city_gis values('Waterloo');
insert into city_gis values('Barrie');
The above 4 records from student_tb need to be transferred over to the student_address_tb.
Assumptions:
1. Address can be detected when it starts with a # sign.
2. City will be picked up based on the table city_gis. So, the street will not contain the City.
3. Province will be detected as 'ON'
4. Phone numbers will be starting with numbers or a bracket followed by number and should not contain any letters.
For phone numbers, I attempted with this but I am not sure I am picking it up correctly. In fact, its picking up the address of the Apartment.
select address, regexp_substr(address,'[[:digit:]]+( ?)(-?)( ?)[[:digit:]]+( ?)(-?)( ?)[[:digit:]]+')
from   student_tb;
Any help in this regard would be great.
Thanks!

Answer 1

FrankKulash wrote:
Aren't Canadian phone numbers always 10 digits (where the first 3 digits are soemtimes implied), and separators, when used, only occur after the 3rd and 6th digits?
Being Canadian myself, I can answer this   Should be always 10 (sometimes 11 if somebody insists on tacking the "assumed" 1 on the front). Seperators *should* occur after the 3rd and 6th, however, I've seen all kinds of crazy ..
Parsing phone #'s has to be one of the harder things business keep asking us to do O_o (since everyone insists on entering them in their own way) *sigh*
On another note, I did get the following working, however, as Frank says, not sure what it'll do on a large data set
with w_data as (
         select '#2115-101 Adelaide Street East Toronto ON M5A 1N1 [phone]  ' data from dual union all
         select 'Testing ABC #2101-212 King Street West, Toronto, ON, M5A 4T1      ' data from dual union all
         select '#3114-212 Bloor Street E      Toronto    ON   M3C1S1  651-1213    ' data from dual union all
         select 'HOTEL PANAORAMA      Waterloo    ON   M1A 2A1                     ' data from dual
      w_data1 as (
         select data,
         regexp_replace(data, '^(.*?)[ ,]+([A-Za-z]+)[ ,]+([A-Z]{2})[ ,]+([A-Z]\d[A-Z] *\d[A-Z]\d)[ ,]+(.*?)$', '\1::\2::\3::\4::\5')  data_parsed
         from w_data
Select regexp_substr(data_parsed, '[^:]+', 1, 1) address,
       regexp_substr(data_parsed, '[^:]+', 1, 2) city,
       regexp_substr(data_parsed, '[^:]+', 1, 3) province,
       regexp_substr(data_parsed, '[^:]+', 1, 4) postal,
       regexp_substr(data_parsed, '[^:]+', 1, 5) phone
   from w_data1
ADDRESS                                  CITY                 PROVINCE   POSTAL               PHONE
#2115-101 Adelaide Street East           Toronto              ON         M5A 1N1              [phone]
Testing ABC #2101-212 King Street West   Toronto              ON         M5A 4T1
#3114-212 Bloor Street E                 Toronto              ON         M3C1S1               651-1213
HOTEL PANAORAMA                          Waterloo             ON         M1A 2A1