Please help me in writing this query

Question

Hi all,
How to do a token match of a string with the token of another mathc
EX:
LET THE FIRST STRINGS WITH A COMMA separated like below
"firsttoken,second,third,are,the,parts,of,a,string"
LET THE SECOND STRING WITH A COMMA separated like below
"this,is,the,search,tokens"
Now I have to match any token of the SECOND string with any TOKEN OF THE FIRST string
If atleast one match is exist then its treated as success..
I need only sql, not pl/sql
Preferrable if analytical functions not used.
Thanks in advance,
Khaleel

Answer 1

Khaleel,
I understand the limitation to use PL/SQL. I have similar kind of limitations at my end, I am not allowed
to leave any custom procedure/function definitions in the db with out taking the permissions for them.
Here is the SQL solution for you for variable number of TOKENS. Only requirement is UPPER BOUND
should be known. You could have variable number of tokens in either of the strings as long as you code it for the maximum number of tokens in each string. Here I coded it for 4 tokens in each string, hence I could use any number of tokens <= 4 in each string. Couple of tests here, -- Strings with different number of tokens but with a matching word
SQL> SELECT DECODE(count(*), 0, 'MATCH NOT FOUND', 'MATCH FOUND') yes_no FROM DUAL,
  2  (SELECT DECODE(str1_comma1, 0, str1, SUBSTR(str1, 1, str1_comma1 - 1)) str1_token1,
  3          DECODE(str1_comma1, 0, NULL, decode(str1_comma2, 0, SUBSTR(str1, str1_comma1 + 1),
  4                    SUBSTR(str1, str1_comma1 + 1, str1_comma2 - str1_comma1 - 1))) str1_token2,
  5          DECODE(str1_comma2, 0, NULL, decode(str1_comma3, 0,  SUBSTR(str1, str1_comma2 + 1 ),
  6                    SUBSTR(str1, str1_comma2 + 1, str1_comma3 - str1_comma2 - 1))) str1_token3,
  7          DECODE(str1_comma3, 0, NULL, SUBSTR(str1, str1_comma3 + 1)) str1_token4,
  8          -- string2 tokens
  9          DECODE(str2_comma1, 0, str1, SUBSTR(str2, 1, str2_comma1 - 1)) str2_token1,
10          DECODE(str2_comma1, 0, NULL, decode(str2_comma2, 0, SUBSTR(str2, str2_comma1 + 1),
11                    SUBSTR(str2, str2_comma1 + 1, str2_comma2 - str2_comma1 - 1))) str2_token2,
12          DECODE(str2_comma2, 0, NULL, decode(str2_comma3, 0,  SUBSTR(str2, str2_comma2 + 1 ),
13                    SUBSTR(str2, str2_comma2 + 1, str2_comma3 - str2_comma2 - 1))) str2_token3,
14          DECODE(str2_comma3, 0, NULL, SUBSTR(str2, str2_comma3 + 1)) str2_token4
15  FROM
16  (
17  Select A.str1, Instr(A.str1, ',', 1, 1) str1_comma1,
18                 Instr(A.str1, ',', 1, 2) str1_comma2,
19            Instr(A.str1, ',', 1, 3) str1_comma3,
20         B.str2,  Instr(B.str2, ',', 1, 1) str2_comma1,
21             Instr(B.str2, ',', 1, 2) str2_comma2,
22             Instr(B.str2, ',', 1, 3) str2_comma3
23  From
24  (Select 'Shaik,Khaleel,Java,Solution' str1 from dual) A,
25  (Select 'SriDHAR,SQL,Solution' str2 from dual) B
26  )
27  ) X
28  WHERE X.str2_token1 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
29  OR    X.str2_token2 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
30  OR    X.str2_token3 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
31  OR    X.str2_token4 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
32  /
YES_NO
MATCH FOUND
-- Strings with same numer of tokens, with no matching word
SQL> SELECT DECODE(count(*), 0, 'MATCH NOT FOUND', 'MATCH FOUND') yes_no FROM DUAL,
  2  (SELECT DECODE(str1_comma1, 0, str1, SUBSTR(str1, 1, str1_comma1 - 1)) str1_token1,
  3          DECODE(str1_comma1, 0, NULL, decode(str1_comma2, 0, SUBSTR(str1, str1_comma1 + 1),
  4                    SUBSTR(str1, str1_comma1 + 1, str1_comma2 - str1_comma1 - 1))) str1_token2,
  5          DECODE(str1_comma2, 0, NULL, decode(str1_comma3, 0,  SUBSTR(str1, str1_comma2 + 1 ),
  6                    SUBSTR(str1, str1_comma2 + 1, str1_comma3 - str1_comma2 - 1))) str1_token3,
  7          DECODE(str1_comma3, 0, NULL, SUBSTR(str1, str1_comma3 + 1)) str1_token4,
  8          -- string2 tokens
  9          DECODE(str2_comma1, 0, str1, SUBSTR(str2, 1, str2_comma1 - 1)) str2_token1,
10          DECODE(str2_comma1, 0, NULL, decode(str2_comma2, 0, SUBSTR(str2, str2_comma1 + 1),
11                    SUBSTR(str2, str2_comma1 + 1, str2_comma2 - str2_comma1 - 1))) str2_token2,
12          DECODE(str2_comma2, 0, NULL, decode(str2_comma3, 0,  SUBSTR(str2, str2_comma2 + 1 ),
13                    SUBSTR(str2, str2_comma2 + 1, str2_comma3 - str2_comma2 - 1))) str2_token3,
14          DECODE(str2_comma3, 0, NULL, SUBSTR(str2, str2_comma3 + 1)) str2_token4
15  FROM
16  (
17  Select A.str1, Instr(A.str1, ',', 1, 1) str1_comma1,
18                 Instr(A.str1, ',', 1, 2) str1_comma2,
19            Instr(A.str1, ',', 1, 3) str1_comma3,
20         B.str2,  Instr(B.str2, ',', 1, 1) str2_comma1,
21             Instr(B.str2, ',', 1, 2) str2_comma2,
22             Instr(B.str2, ',', 1, 3) str2_comma3
23  From
24  (Select 'Shaik,Khaleel,Java,Program' str1 from dual) A,
25  (Select 'SriDHAR,SQL,Solution,Possible' str2 from dual) B
26  )
27  ) X
28  WHERE X.str2_token1 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
29  OR    X.str2_token2 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
30  OR    X.str2_token3 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
31  OR    X.str2_token4 IN (X.str1_token1, X.str1_token2, X.str1_token3, X.str1_token4)
32  /
YES_NO
MATCH NOT FOUNDLooks cumbersome, but works. You should be able to take it from here to extend the functionality also,
like if you want to know which words are matching OR which words are not matching etc.., you should be
able to tweak the SQL accordingly.
Good luck,
Thx,
SriDHAR