Advance sql session - strings

Advance SQL
String Manipulation in SQL
Eyal Trabelsi

Agenda
○ Gotha’s
○ Regex
○ String Functions
○ Sentence Splits
○ Aggregations
○ NLP

String Gotchas #1
Users
Id FirstName LastName Email
1 NULL Trabelsi eya
invalidId Eyal Trabelsi eyal@gmail.com

String Gotchas #1
Users
Calculate and retrieve full name (Concatenation of first name and last name from T2)
1 NULL Trabelsi eya

String Gotchas #1
Users
SELECT first_name + last_name
FROM users
1 NULL Trabelsi eya

String Gotchas #1
Users
FROM users
Is it correct ?
1 NULL Trabelsi eya

String Gotchas #1
Users
FROM users
No it will return null for the first row
1 NULL Trabelsi eya

String Gotchas #1
Users
SELECT ISNULL(first_name,””) + ISNULL(last_name,””)
FROM users
1 NULL Trabelsi eya

String Gotchas #1
Users
SELECT ISNULL(first_name,””) + ISNULL(last_name,””)
FROM users
Is it correct ?
1 NULL Trabelsi eya

String Gotchas #2
Users
1 NULL Trabelsi eya

String Gotchas #2
Users
Retrieve all the rows with last_name Trabelsi
1 NULL Trabelsi eya

String Gotchas #2
Users
SELECT *
FROM T1
WHERE last_name =“trabelsi”
1 NULL Trabelsi eya

String Gotchas #2
Users
SELECT *
FROM Users
Is it correct ?
1 NULL Trabelsi eya

String Gotchas #2
Users
SELECT *
FROM T1
No SQL Strings are case sensitive
1 NULL Trabelsi eya

String Gotchas #2
Users
SELECT *
FROM Users
WHERE LOWER(last_name) =“trabelsi”
Is it correct ?
1 NULL Trabelsi eya

String Regex Example #1
Users
1 NULL Trabelsi eya

Users
Retrieve all the rows with valid Id
1 NULL Trabelsi eya

Users
SELECT *
FROM Users
WHERE id ~ ‘^[0-9]*$’
1 NULL Trabelsi eya

Users
SELECT *
FROM Users
WHERE id ~ ‘^[0-9]*$’
Is it correct ?
1 NULL Trabelsi eya

Users
1 NULL Trabelsi eya

Users
Retrieve all the rows with valid email
1 NULL Trabelsi eya

Users
SELECT *
FROM Users
WHERE email ~ ^[a-z0-9._%-]+@[a-z0-9.-]+.[a-z]{2,4}$
1 NULL Trabelsi eya

Users
SELECT *
FROM Users
WHERE email ~ ^[a-z0-9._%-]+@[a-z0-9.-]+.[a-z]{2,4}$
Is it correct ?
1 NULL Trabelsi eya

String Regex – Additional Resources

● Regex Curated List

● Regex Curated List
● Redshift additional functionality

Supported String functions
• String function are vendor ‘ dependent!!!!

• Most vendors support the all the basic window functions
like lengh, replace, reverse etc.

like lengh, replace, reverse etc.
• These are the vendor specific supported window
functions:

like length, replace, reverse etc.
• These are the vendor specific supported window
functions:

String Function Example #1
Tickets
Id Days_for_resolution Description
1 5 Bad product
2 5 Not working
3 1 The button doesn’t click on IE
4 1 My user is locked after I installed it

Tickets
Checking whether ticket description length correlate to days for resolution
1 5 Bad product
2 5 Not working

Tickets
1 5 Bad product
2 5 Not working
SELECT AVG(length(description)), days_for_resolution
FROM tickets
GROUP BY days_for_resolution

Tickets
Is it correct ?
1 5 Bad product
2 5 Not working
SELECT AVG(length(description)), days_for_resolution
FROM tickets
GROUP BY days_for_resolution

Sentence Splitting
https://blog-holistics-io.cdn.ampproject.org/c/s/blog.holistics.io/splitting-
array-string-into-rows-in-amazon-redshift-or-mysql/amp/

String Splitting Example #1
Tickets
Id description
1 The beer is very nice
2 The trash cans have small holes

Tickets
Id description
Split the ticket descriptions into words related to ticket

Tickets
Id description
WITH NS AS ( SELECT GENERATE_SERIES(0,10000
AS n)
SELECT SPLIT_PART(description,’ ’, NS.n)
FROM ticket
LEFT JOIN NS
ON NS.n < REGEXP_COUNT((description,’ ’)+1

Tickets
Id description
AS n)
FROM ticket
LEFT JOIN NS
4

Tickets
Id description
AS n)
FROM ticket
LEFT JOIN NS
4
NS.n =1: beer

Tickets
Id description
AS n)
FROM ticket
LEFT JOIN NS
4
NS.n =2: is

Tickets
Id description
AS n)
FROM ticket
LEFT JOIN NS
4
NS.n =2: very

Tickets
Id description
AS n)
FROM ticket
LEFT JOIN NS
4
NS.n =4: nice

Supported String aggregation functions

• String aggregation function are vendor ‘ dependent!!!!

• Each vendor support different functions and call them in
different way

• Each vendor support different functions and call them in
different way
• These are the vendor specific supported window functions:

String Aggregation Example #1
Accounts
Id Name
1 microsoft
2 amazon
Id Account_id Phone
1 1 050 0000000
2 1 050 0000001
3 2 050 0000002
Contacts

Accounts
Getting all the phone numbers of an account
Id Name
1 microsoft
2 amazon
Id Account_id Phone
1 1 050 0000000
2 1 050 0000001
3 2 050 0000002
Contacts

Accounts
Id Name
1 microsoft
2 amazon
SELECT LISTAGG(contacts.phone), account.id
FROM accounts
INNER JOIN contacts ON accounts.id = contacts.account_id
GROUP BY account.id
Id Account_id Phone
1 1 050 0000000
2 1 050 0000001
3 2 050 0000002
Contacts

Accounts
Is it correct ?
Id Name
1 microsoft
2 amazon
SELECT LISTAGG(contacts.phone), account.id
FROM accounts
INNER JOIN contacts ON accounts.id = contacts.account_id
GROUP BY account.id
Id Account_id Phone
1 1 050 0000000
2 1 050 0000001
3 2 050 0000002
Contacts

When we know exactly the number of strings we want to concatenate

When we know exactly the number of strings we want to concatenate
https://www.periscopedata.com/blog/aggregating-into-strings-without-string-
agg-or-array-agg-in-amazon-redshift

NLP- NPS Calculation
http://www.silota.com/docs/recipes/sql-nps-customer-loyalty-survey-analysis.html

Advance sql session - strings

More Related Content

What's hot (20)

Similar to Advance sql session - strings (20)

More from Eyal Trabelsi (6)

Recently uploaded (20)

Advance sql session - strings

Editor's Notes