Databases Reference
In-Depth Information
Value
Description
HA_READ_AFTER_KEY
In this mode,
index_read()
should find a row
with the key that follows the
key
in the index, in
other words, the smallest key that is larger than the
key
argument.
HA_READ_KEY_OR_NEXT
Here it should find a smallest key that is larger
than or equal to the
key
. That is, if the row with
the key of
key
exists, this mode behaves exactly as
HA_READ_KEY_EXACT
, but if such a row does not
exist it behaves as
HA_READ_AFTER_KEY
.
HA_READ_PREFIX
In this mode, the
key
should match the prefix of
the key of the found row. Although in all search
modes MySQL can search for a prefix of a key
(for example, for a specific value of the
a
ield
in a
KEY (a,b)
index), only in this mode can it
search for a prefix of a field (first 5 characters in the
20-character string key). This search mode is never
used in MySQL 5.1.
HA_READ_KEY_OR_PREV
This is similar to
HA_READ_KEY_OR_NEXT
—
MySQL wants a row with the key, which is smaller
than or equal to
key
. In other words, when looking
for a key, if we find it, we return it at once. If we do
not find it, we take the previous key in the index,
the key that had been directly preceding to our
key if it would have been in the index. Only the
HANDLER
statement may use this search mode.
HA_READ_BEFORE_KEY
Similar to
HA_READ_AFTER_KEY
, this means a
request for a largest key which is less than the
value of the
key
argument.
HA_READ_PREFIX_LAST
This means a search for a last (in the index order)
key that starts from the
key
.
HA_READ_PREFIX_LAST_OR_PREV
This means, not surprisingly, a search for a last
key that starts from the
key
or a previous key if a
matching key was not found.
These modes are only used for spatial indexes and
we will not discuss them in this topic.
HA_READ_MBR_CONTAIN
HA_READ_MBR_INTERSECT
HA_READ_MBR_WITHIN
HA_READ_MBR_DISJOINT
HA_READ_MBR_EQUAL
