Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
Overview
| Comment: | Update an r-tree extension test case to account for recent changes to the query planner. Also fix a comment in rtree.c. |
|---|---|
| Downloads: | Tarball | ZIP archive | SQL archive |
| Timelines: | family | ancestors | descendants | both | trunk |
| Files: | files | file ages | folders |
| SHA1: |
eaaca669a4afc83906806b22365b010c |
| User & Date: | dan 2010-08-11 12:26:46 |
Context
|
2010-08-11
| ||
| 18:56 | Improve coverage of pager.c. (check-in: 2fa05d01 user: dan tags: trunk) | |
| 12:26 | Update an r-tree extension test case to account for recent changes to the query planner. Also fix a comment in rtree.c. (check-in: eaaca669 user: dan tags: trunk) | |
| 11:59 | Use sqlite3DbFree() instead of sqlite3_free() to free any error message reported by the parser layer while attempting to parse a CREATE TABLE statement passed to sqlite3_declare_vtab(). (check-in: 7c674aab user: dan tags: trunk) | |
Changes
Changes to ext/rtree/rtree.c.
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 |
** Rtree virtual table module xBestIndex method. There are three ** table scan strategies to choose from (in order from most to ** least desirable): ** ** idxNum idxStr Strategy ** ------------------------------------------------ ** 1 Unused Direct lookup by rowid. ** 2 See below R-tree query. ** 3 Unused Full table scan. ** ------------------------------------------------ ** ** If strategy 1 or 3 is used, then idxStr is not meaningful. If strategy ** 2 is used, idxStr is formatted to contain 2 bytes for each ** constraint used. The first two bytes of idxStr correspond to ** the constraint in sqlite3_index_info.aConstraintUsage[] with ** (argvIndex==1) etc. ** ** The first of each pair of bytes in idxStr identifies the constraint ** operator as follows: |
| < | |
1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 |
** Rtree virtual table module xBestIndex method. There are three
** table scan strategies to choose from (in order from most to
** least desirable):
**
** idxNum idxStr Strategy
** ------------------------------------------------
** 1 Unused Direct lookup by rowid.
** 2 See below R-tree query or full-table scan.
** ------------------------------------------------
**
** If strategy 1 is used, then idxStr is not meaningful. If strategy
** 2 is used, idxStr is formatted to contain 2 bytes for each
** constraint used. The first two bytes of idxStr correspond to
** the constraint in sqlite3_index_info.aConstraintUsage[] with
** (argvIndex==1) etc.
**
** The first of each pair of bytes in idxStr identifies the constraint
** operator as follows:
|
Changes to ext/rtree/rtree6.test.
84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 |
{TABLE t1 VIRTUAL TABLE INDEX 2:Ca} \
{TABLE t2 USING PRIMARY KEY} \
]
do_test rtree6.2.3 {
query_plan {SELECT * FROM t1,t2 WHERE k=ii}
} [list \
{TABLE t2} \
{TABLE t1 VIRTUAL TABLE INDEX 1:} \
]
do_test rtree6.2.4 {
query_plan {SELECT * FROM t1,t2 WHERE v=10 and x1<10 and x2>10}
} [list \
{TABLE t1 VIRTUAL TABLE INDEX 2:CaEb} \
{TABLE t2} \
]
do_test rtree6.2.5 {
query_plan {SELECT * FROM t1,t2 WHERE k=ii AND x1<v}
} [list \
{TABLE t2} \
{TABLE t1 VIRTUAL TABLE INDEX 1:} \
]
finish_test
|
| | | | |
84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 |
{TABLE t1 VIRTUAL TABLE INDEX 2:Ca} \
{TABLE t2 USING PRIMARY KEY} \
]
do_test rtree6.2.3 {
query_plan {SELECT * FROM t1,t2 WHERE k=ii}
} [list \
{TABLE t1 VIRTUAL TABLE INDEX 2:} \
{TABLE t2 USING PRIMARY KEY} \
]
do_test rtree6.2.4 {
query_plan {SELECT * FROM t1,t2 WHERE v=10 and x1<10 and x2>10}
} [list \
{TABLE t1 VIRTUAL TABLE INDEX 2:CaEb} \
{TABLE t2} \
]
do_test rtree6.2.5 {
query_plan {SELECT * FROM t1,t2 WHERE k=ii AND x1<v}
} [list \
{TABLE t1 VIRTUAL TABLE INDEX 2:} \
{TABLE t2 USING PRIMARY KEY} \
]
finish_test
|