拆分字符串Teradata SQL

Question

I'm looking to split a string in Teradata.

The table might look something like this.

column1
hello:goodbye:afternoon

I'm trying to use SUBSTRING and INSTR to extract specific words. So, say I want to select "goodbye". I'm trying the following query.

SELECT SUBSTRING(a.column1 from index(a.column1,':')+1 for INSTR(a.column1,':',0,2))
FROM db.table as a

I get the following error.

SELECT Failed. [3707] Syntax error, expected something like ')' between the word 'INSTR' and '('

I'm not sure why I'm getting that error. It lets me use INDEX to deduce a number in place of INSTR, so I'm not sure why it is acting this way when I use INSTR.

Answer 1

If this was TD14 you wouldn't need INSTR , there's a STRTOK function :-)

STRTOK(column1,':',2),

For earlier releases it's

CASE 
   WHEN column1 LIKE '%:%:%'
   THEN SUBSTRING(column1 FROM POSITION(':' IN column1) + 1 FOR POSITION(':' IN 
        SUBSTRING(column1 FROM POSITION(':' IN column1) + 1)) - 1)
   WHEN column1 LIKE '%:%'
   THEN SUBSTRING(column1 FROM POSITION(':' IN column1) + 1)
END

The CASE LIKE ist just to prevent an " string subscript out of bound " error when there no colon.