比较不同表中的两列

Question

I know that this question has been asked several times and I did read them all. However, my situation is little different I can still can't get the right results with my SQL statement below.

I have two tables as shown below:

TABLE A
######################################################################
   |   ID   | IP_Address    | Username  | Comments
----------------------------------------------------------------------  
    1       128.        abc     travel to US
    2       127.        dzd     author
    3       127.        abc     It's not redundant at all. Not offering a single function
    4       124.        deb     I just lost laptop.  How do I report it?
-----------------------------------------------------------------------



TABLE B
######################################################################
   |   keywords     |
----------------------------------------------------------------------  
    author
    How do I report
-----------------------------------------------------------------------

My intention is to pull out all records in TABLE A when the Comments column partially matches one of the keywords in TABLE B.

Here is the query:

SELECT ID, IP_Address, UserName, Comments FROM TABLEA
FULL JOIN TABLEB
ON TABLEA.Comments LIKE TABLEB.Keywords
WHERE TABLEA.Comments IS NOT NULL AND TABLEB.Keywords IS NOT NULL

It works but it only pulls out the EXACT match, not a part of the comments.
In my sample, it pulls out "author" record but not "I just lost laptop. How do I report it?" record.

Is there a way to twist my query to meet my requirements?

Thanks

JPL

Answer 1

Add % to the LIKE clause

SELECT ID, IP_Address, UserName, Comments FROM TABLEA
FULL JOIN TABLEB
ON TABLEA.Comments LIKE '%' + TABLEB.Keywords + '%'
WHERE TABLEA.Comments IS NOT NULL AND TABLEB.Keywords IS NOT NULL

Answer 2

This will do what you need.

SELECT ID, IP_Address, UserName
FROM TABLE1 T1
CROSS APPLY (SELECT Keywords FROM Table2) T2
WHERE CHARINDEX(T2.Keywords,T1.Comments) >0

The CROSS APPLY could be replaced with the full join, but cross apply works better if there are more columns in the 2nd table.