Single quote in basic lexer preference

Question

Hi Gurus,
Using basic_lexer preference I wanted to consider columns having single quotes *'* also as white space. Please help me to achieve it.
Original Code
    ctx_ddl.create_preference ('test_lex1', 'basic_lexer');
    ctx_ddl.set_attribute ('test_lex1', 'whitespace', '/\|-_+&');
I tried using q operator
SQL>  begin
ctx_ddl.create_preference ('test_lex1', 'basic_lexer');
ctx_ddl.set_attribute ('test_lex1', 'whitespace', '/\|\'-_+&q'');\
end;  2    3    4
  5  /
Enter value for q:
I also tried
SQL>
SQL>  begin
ctx_ddl.create_preference ('test_lex1', 'basic_lexer');
ctx_ddl.set_attribute ('test_lex1', 'whitespace', '/\|-_+&\'');
end;  2    3    4
  5  /
ERROR:
ORA-01756: quoted string not properly terminated
I tried this but it is not indexing as intended.
SQL>  begin
ctx_ddl.create_preference ('test_lex1', 'basic_lexer');
ctx_ddl.set_attribute ('test_lex1', 'whitespace', '/\|-_+&'''); --translating special characters as white space using lexer
end;  2    3    4
  5  /
PL/SQL procedure successfully completedplease help me here.

Answer 1

Thanks a lot Barabara for quick reply.
Let me paste my code here.
create table test_sh4 (text1 clob,text2 clob,text3 clob);
insert  into test_sh4  values ('ORACLE''s','It isn''t Oracle''s fault.','y');
begin
   ctx_ddl.create_preference ('test_mcd', 'multi_column_datastore'); -- utilizing the same index for multiple columns
    ctx_ddl.set_attribute
      ('test_mcd',
       'columns',
       'replace (text1, '' '', '''') nd1, -- virtual column to eliminate white spaces in text1 column
        text1 text1, 
        replace (text2, '' '', '''') nd2,  -- virtual column to eliminate white spaces in text2 column
        text2 text2');
    ctx_ddl.create_preference ('test_lex1', 'basic_lexer');
    ctx_ddl.set_attribute ('test_lex1', 'whitespace', '/\|-_+&'''); --translating special characters as white space using lexer
    ctx_ddl.create_section_group ('test_sg', 'basic_section_group'); -- creating section group to search within sections.
    ctx_ddl.add_field_section ('test_sg', 'text1', 'text1', true);
    ctx_ddl.add_field_section ('test_sg', 'nd1', 'nd1', true);
    ctx_ddl.add_field_section ('test_sg', 'text2', 'text2', true);
    ctx_ddl.add_field_section ('test_sg', 'nd2', 'nd2', true);
     ctx_ddl.create_preference('SUBSTRING_PREF', 'BASIC_WORDLIST');
    ctx_ddl.set_attribute('SUBSTRING_PREF','SUBSTRING_INDEX','TRUE');
     ctx_ddl.set_attribute('SUBSTRING_PREF','PREFIX_INDEX','TRUE');
     ctx_ddl.set_attribute('SUBSTRING_PREF','PREFIX_MIN_LENGTH', '3');
     ctx_ddl.set_attribute('SUBSTRING_PREF','PREFIX_MAX_LENGTH', '8');
     ctx_ddl.create_preference('textstore', 'BASIC_STORAGE');
    ctx_ddl.set_attribute('textstore', 'I_TABLE_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
    ctx_ddl.set_attribute('textstore', 'K_TABLE_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
    ctx_ddl.set_attribute('textstore', 'R_TABLE_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
    ctx_ddl.set_attribute('textstore', 'N_TABLE_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
    ctx_ddl.set_attribute('textstore', 'I_INDEX_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
    ctx_ddl.set_attribute('textstore', 'P_TABLE_CLAUSE','tablespace TEST_INDEX storage (initial 64K)');
end;
SQL>  create index IX_test_sh4  on test_sh4 (text3)
  2  indextype is ctxsys.context
  3  parameters
  4      ('datastore test_mcd
  5        lexer          test_lex1
  6        section group  test_sg
  7       wordlist SUBSTRING_PREF MEMORY 50M
  8       SYNC ( ON COMMIT)
  9       storage textstore'
10       );
SQL> select count(*) from (
SELECT SCORE(1) score,t.* FROM test_sh4 t
WHERE CONTAINS (text3, 'ORACLE S' ,1) > 1
order by score desc
)  2    3    4    5
  6  /
  COUNT(*)
         1
I was thinking that combination of lexer preference and virtual column nd1 when indexed will take care of following  part. But it is not
SQL> select count(*) from (
SELECT SCORE(1) score,t.* FROM test_sh4 t
WHERE CONTAINS (text3, 'ORACLES' ,1) > 1
order by score desc
)  2    3    4    5
  6  /
  COUNT(*)
         0
I had to correct the code to
    ctx_ddl.set_attribute
      ('test_mcd',
       'columns',
       'regexp_replace(text1,''[[:punct:]*|[[:space:]*]'') nd1,
        text1 text1, 
        regexp_replace(text2,''[[:punct:]*|[[:space:]*]'') nd2, 
        text2 text2');
Now it is returning rows as expected.The requirement was to consider any special characters as white spaces, Can we do that in Lexer preference as the logic we are performing iin virtual columns nd1 and nd2.