This table compares WD and TEXT indexes in terms of syntax and capability.
Feature |
Supported by WD index? |
Supported by TEXT index? |
|---|---|---|
Conjunction of terms |
Yes, expressed in the form:
tbl.col
CONTAINS('great','white'
,'whale')
|
Yes, expressed in the form: CONTAINS(tbl.col,'great white whale') |
General boolean expressions |
Yes, expressed in the form:
tbl.col CONTAINS
('great') AND ( tbl.col
CONTAINS('white) OR
tbl.col CONTAINS('whale')
AND NOT tbl.col
CONTAINS('ship'))
|
Yes, expressed in the form: CONTAINS(tbl.col, 'great AND ( white OR whale AND NOT ship )') |
Search for terms matching prefix |
No |
Yes, for example: CONTAINS (tbl.col,'whale*') |
Acceleration of LIKE predicates |
Yes, for example: tbl.col LIKE 'whale%' |
No |
Searches for terms in proximity |
No |
Yes, for example: CONTAINS(tbl.col, 'white BEFORE whale') CONTAINS(tbl.col, 'whale NEAR white')
CONTAINS(tbl.col, ' "white whale" ') |
Ordering of results based on search scoring |
No |
Yes |
In TEXT index, searching for terms matching a prefix and searching for a LIKE expression have different semantics and may return very different results depending on the text configuration. The specification of minimum length, maximum length and a stoplist will govern the prefix processing but does not affect LIKE semantics.
Meaning of boolean expressions will differ between WD index
and TEXT index when term dropping occurs because,
the effect of dropped terms in TEXT index processing
has no equivalent in the WD index.
