Problem with special chars in BLOB datatype using contains keyword

Question

Facing problem, when part searching with special chars in BLOB datatype. It is considering the non alpha-numeric chars as a separtor in a provided string
EX:
SELECT *
FROM RESUME_TEST P,grst_candidate d
WHERE d.candidate_id = p.candidate_id
AND CONTAINS(P.CAND_RESUME,'%VB.NET%',1) > 0
Strings: , VB.NET , PL/SQL AS/400 , C etc..
Followed the below approaches
1) created a table:
Syntax: create table resume_Test(cand_id number(10),cand_resume blob);
2) inserted the values into this table upto 60,000
3) created a context index
3.1 created preferences
Syntax:
BEGIN
ctx_ddl.create_preference('try_lexer3','BASIC_LEXER');
ctx_ddl.set_attribute('try_lexer3','printjoins','-_~!@#$%^&*(){}[],=?\;|><.+');
END;
3.2 created context index
Syntax:
CREATE INDEX CANDRESUME_CTX_IDX ON resume_test (cand_resume)
INDEXTYPE IS CTXSYS.CONTEXT
PARAMETERS ('LEXER try_lexer3 memory 500M');
4) while executing this index, it is taking much time approx 6 hrs(plz explain why it is taking time)
5) Problems:
5.1 when searching with string(VB.NET , PL/SQL AS/400 , C etc..) it is considering the special char as a separator
5.2 used escape char (\) also, but no effect
5.3 when searching with single char, it is giving error (ORA-29902,ORA-20000,DRG-51030)
5.4 getting the above error with wild card chars (& ,_, (),{},[])
So, please explain the clear scenarios, why am getting this error , and how to get the proper results.

Answer 1

Have you tried adding the / char to the printjoin characters?
Indexing can take a lot of time, depending on the amount of data and your machine's power. You could try to parallelize the index creation and / or assign more memory
CREATE INDEX CANDRESUME_CTX_IDX ON resume_test (cand_resume)
INDEXTYPE IS CTXSYS.CONTEXT
PARAMETERS ('LEXER try_lexer3 memory 2000M') PARALLEL 8;