BROWSE_WORDS

This procedure enables you to browse words in an Oracle Text index. You specify a seed word and BROWSE_WORDS returns the words around it in the index, and a rough count of the number of documents that contain each word.

This feature is useful for refining queries. You can identify the following:

• unselective words (words that have low document count)

• misspelled words in the document set

Syntax 1: To Store Results in Table

ctx_query.browse_words( 

index_name  IN   VARCHAR2, 
seed        IN   VARCHAR2, 
restab      IN   VARCHAR2, 
browse_id   IN   NUMBER   DEFAULT 0, 
numwords    IN   NUMBER   DEFAULT 10, 
direction   IN   VARCHAR2 DEFAULT BROWSE_AROUND,
part_name   IN   VARCHAR2 DEFAULT NULL

); 

Syntax 2: To Store Results in Memory

ctx_query.browse_words( 

index_name  IN      VARCHAR2, 
seed        IN      VARCHAR2, 
resarr      IN OUT  BROWSE_TAB, 
numwords    IN      NUMBER   DEFAULT 10, 
direction   IN      VARCHAR2 DEFAULT BROWSE_AROUND,
part_name   IN      VARCHAR2 DEFAULT NULL 

); 

index

Specify the name of the index. You can specify schema.name. Must be a local index.

seed

Specify the seed word. This word is lexed before browse expansion. The word need not exist in the token table. seed must be a single word. Using multiple words as the seed will result in an error.

restab

Specify the name of the result table. You can enter restab as schema.name. The table must exist before you call this procedure, and you must have INSERT permissions on the table. This table must have the following schema.

Column	Datatype
browse_id	number
word	varchar2(64)
doc_count	number

Existing rows in restab are not deleted before BROWSE_WORDS is called.

resarr

Specify the name of the result array. resarr is of type ctx_query.browse_tab.

type browse_rec is record (
   word varchar2(64),
   doc_count number
);
type browse_tab is table of browse_rec index by binary_integer;

browse_id

Specify a numeric identifier between 0 and 2³². The rows produced for this browse have a value of in the browse_id column in restab. When you do not specify browse_id, it defaults to 0.

numwords

Specify the number of words returned.

direction

Specify the direction for the browse. You can specify one of:

value	behavior
BEFORE	Browse seed word and words alphabetically before the seed.
AROUND	Browse seed word and words alphabetically before and after the seed.
AFTER	Browse seed word and words alphabetically after the seed.

Symbols CTX_QUERY.BROWSE_BEFORE, CTX_QUERY.BROWSE_AROUND, and CTX_QUERY.BROWSE_AFTER are defined for these literal values as well.

part_name

Specify the name of the index partition to browse.

Example

Browsing Words with Result Table

begin
ctx_query.browse_words('myindex','dog','myres',numwords=>5,direction=>'AROUND');
end;

select word, doc_count from myres order by word;

WORD       DOC_COUNT
--------   ----------
CZAR       15
DARLING    5
DOC        73
DUNK       100
EAR        3

Browsing Words with Result Array

set serveroutput on;
declare
  resarr ctx_query.browse_tab;
begin
ctx_query.browse_words('myindex','dog',resarr,5,CTX_QUERY.BROWSE_AROUND);
for i in 1..resarr.count loop
  dbms_output.put_line(resarr(i).word || ':' || resarr(i).doc_count);
end loop;
end;