matching names that are close by not always exact [duplicate] - sql

This question already has an answer here:
SQL Fuzzy Matching
(1 answer)
Closed 6 years ago.
attempting to match a list of names that are similar in one very long column to another that are close but often vary do to missing letters and punctuation? is there a simple solution via a macro and/or sql?

using Levenstein functions can help
check the function and algorithm here: Levenshtein distance in T-SQL
after you create a function - compare the distance, for example:
select ..... from....
where dbo.Levenstein(str1,str2)>0.9 --means, the match between str1 and str2 is 90%

Related

SQL Server: dynamic columns based on row values (Date) [duplicate]

This question already has answers here:
T-SQL dynamic pivot
(5 answers)
Why is processing a sorted array faster than processing an unsorted array?
(26 answers)
Closed 4 years ago.
I have spent an hour already on this problem.
I want to dynamically generate columns based on the values from the column AttendanceDate.
I have found some similar questions, but unfortunately the examples were too complicated for me to comprehend.
Data:
Expected output:
This can be done with the stuff method as mention in comments or with a while exists implementation:
http://rextester.com/FPU47008

Identify missing numbers in an Oracle Table [duplicate]

This question already has answers here:
How to check any missing number from a series of numbers?
(11 answers)
SQL - Find missing int values in mostly ordered sequential series
(6 answers)
Closed 5 years ago.
I have a number sequence field in a table that has some gaps/skipped numbers in it. I need to identify the skipped numbers. The only solution I can think of is to use iterative/cursor based loops and I suspect that will be fairly slow. Is there a faster method?

SQL: Splitting first, last and middle name into their own columns? [duplicate]

This question already has answers here:
Splitting the full name and writing it to another table in SQL Server 2008
(2 answers)
Closed 9 years ago.
How would I go about splitting one column that has the first, last and middle name. To there own separate columns in a SQL Server 2008 query?
The column is called NAME
NAME(char(25),null)
Mctasrren ,David Max
Cressler ,Patti L
Basil ,Vessen Eddie
Chapplestait ,Victoy
this is what i've used so far my main issue is the middle name. or if someone has a better way to shorten the first name code.
--last name code
left([NAME],charindex(' ,',[NAME]))
-first name code
substring([NAME],charindex(',',[NAME])+1,charindex(' ',substring([NAME],charindex(',',[NAME])+1,25-charindex(',',[NAME])+1)))
Do you want to split it into columns as part of a result set, create computed columns on the table, or actually update the schema to have the data split in the source?
In any case the basic nuts and bolts can be done by either:
Use a combination of CHARINDEX, SUBSTRING, and LEFT or RIGHT to find commas or spaces and split around that. If you sure you data will always be 'L_NAME ,F_NAME M_NAME_OR_INITIAL' that will pretty easy. I am actually I surprised I didn't find an similar question here near the top of a google search, but there is an example of similar from SQLServerCentral.
Use a RegEx via the CLR, which can be more robust if there is any variety in the data. If you are familiar with RegEx this should be a straight forward parse. Again, a simplified example can found on MSDN.
Whatever you choose, you'll probably quickly run into names that don't easily follow that format. In that case you want to build more logic into a function handle different types of names.

SQL Server parsing function? [duplicate]

This question already has answers here:
Closed 10 years ago.
Possible Duplicate:
Split Function equivalent in tsql?
I have a column that contains data in the form:
CustomerZip::12345||AccountId::1111111||s_Is_Advertiser::True||ManagedBy::3000||CustomerID::5555555||
Does SQL have any sort of built in function to easily parse out this data, or will I have to build my own complicated mess of patindex/substring functions to pull each value into its own field?
I don't believe there is anything built in. Look at the comments posted against your original question.
If this is something you're going to need on a regular basis, consider writing a view.

Situation with SQL query [duplicate]

This question already has answers here:
How to search for a comma separated value
(3 answers)
Closed 8 years ago.
I have data in table in below format
id brand_ids
--------------
2 77,2
3 77
6 3,77,5
8 2,45,77
--------------
(Note the brand ids will be stored like comma separated values, this is common for values in this field)
Now i am trying to get a query which is capable of querying out only rows which have '77'
in it..
I know i can use LIKE command in three formats LIKE '77,%' OR LIKE '%,77,%' OR LIKE '%,77' with or condition to achieve it. But i hope this will increase the load time of the sql.
is there any straight forward method to achieve this? if so please suggest.
Thanks,
Balan
A strict answer to your question would be: no. Your suggestion of using LIKE is your best option with this data model. However, as mentioned, it is highly suggested that you use a more normalized model.