-C Fix\sproblems\sin\sestimating\sthe\snumber\sof\srows\svisited\sby\sa\srange\squery\susing\ssqlite_stat4\sdata\swhen\sthe\scolumn\ssubject\sto\sthe\srange\squery\sis\snot\sthe\sleftmost\sof\sthe\sindex.
-D 2013-08-08T19:38:40.828
+C Update\sthe\sheader\scomment\son\sanalyze.c\sto\sdescribe\sthe\ssqlite_stat4\stable\nformat.
+D 2013-08-09T14:07:55.264
F Makefile.arm-wince-mingw32ce-gcc d6df77f1f48d690bd73162294bbba7f59507c72f
F Makefile.in 5e41da95d92656a5004b03d3576e8b226858a28e
F Makefile.linux-gcc 91d710bdc4998cb015f39edf3cb314ec4f4d7e23
F sqlite3.1 6be1ad09113570e1fc8dcaff84c9b0b337db5ffc
F sqlite3.pc.in ae6f59a76e862f5c561eb32a380228a02afc3cad
F src/alter.c 2af0330bb1b601af7a7789bf7229675fd772a083
-F src/analyze.c 0f54320fb58eedf3690dd9609e6176fff3f8c157
+F src/analyze.c 9fced5c126f9655d8cea2be534da6caef93f203c
F src/attach.c 1816f5a9eea8d2010fc2b22b44f0f63eb3a62704
F src/auth.c 523da7fb4979469955d822ff9298352d6b31de34
F src/backup.c 43b348822db3e4cef48b2ae5a445fbeb6c73a165
F tool/warnings.sh fbc018d67fd7395f440c28f33ef0f94420226381
F tool/wherecosttest.c f407dc4c79786982a475261866a161cd007947ae
F tool/win/sqlite.vsix 97894c2790eda7b5bce3cc79cb2a8ec2fde9b3ac
-P e50dc30523210ba12324d5d8379503610f13aa34
-R b712a80bcc08bb0cfbaad8e54099473e
-U dan
-Z 6384d8e74fabf5ee0daae5c0c2206cf9
+P 9228aaf54dd2700c4f460f94f9c2309407578983
+R 8f8b5382a2066b3e84011ccbfba51880
+U drh
+Z c6ed6d4032cb2df263d6c33284a68607
** CREATE TABLE sqlite_stat1(tbl, idx, stat);
** CREATE TABLE sqlite_stat2(tbl, idx, sampleno, sample);
** CREATE TABLE sqlite_stat3(tbl, idx, nEq, nLt, nDLt, sample);
+** CREATE TABLE sqlite_stat4(tbl, idx, nEq, nLt, nDLt, sample);
**
** Additional tables might be added in future releases of SQLite.
** The sqlite_stat2 table is not created or used unless the SQLite version
** with SQLITE_ENABLE_STAT2. The sqlite_stat2 table is deprecated.
** The sqlite_stat2 table is superseded by sqlite_stat3, which is only
** created and used by SQLite versions 3.7.9 and later and with
-** SQLITE_ENABLE_STAT3 defined. The fucntionality of sqlite_stat3
-** is a superset of sqlite_stat2.
+** SQLITE_ENABLE_STAT3 defined. The functionality of sqlite_stat3
+** is a superset of sqlite_stat2. The sqlite_stat4 is an enhanced
+** version of sqlite_stat3 and is only available when compiled with
+** SQLITE_ENABLE_STAT4 and in SQLite versions 3.8.0 and later.
+**
+** For most applications, sqlite_stat1 provides all the statisics required
+** for the query planner to make good choices.
**
** Format of sqlite_stat1:
**
** name in the idx column. The tbl column is the name of the table to
** which the index belongs. In each such row, the stat column will be
** a string consisting of a list of integers. The first integer in this
-** list is the number of rows in the index and in the table. The second
+** list is the number of rows in the index. (This is the same as the
+** number of rows in the table, except for partial indices.) The second
** integer is the average number of rows in the index that have the same
** value in the first column of the index. The third integer is the average
** number of rows in the index that have the same value for the first two
**
** Format for sqlite_stat3:
**
-** The sqlite_stat3 is an enhancement to sqlite_stat2. A new name is
-** used to avoid compatibility problems.
+** The sqlite_stat3 format is a subset of sqlite_stat4. Hence, the
+** sqlite_stat4 format will be described first. Further information
+** about sqlite_stat3 follows the sqlite_stat4 description.
+**
+** Format for sqlite_stat4:
+**
+** As with sqlite_stat2, the sqlite_stat4 table contains histogram data
+** to aid the query planner in choosing good indices based on the values
+** that indexed columns are compared against in the WHERE clauses of
+** queries.
**
-** The format of the sqlite_stat3 table is similar to the format of
-** the sqlite_stat2 table. There are multiple entries for each index.
+** The sqlite_stat4 table contains multiple entries for each index.
** The idx column names the index and the tbl column is the table of the
** index. If the idx and tbl columns are the same, then the sample is
-** of the INTEGER PRIMARY KEY. The sample column is a value taken from
-** the left-most column of the index. The nEq column is the approximate
-** number of entires in the index whose left-most column exactly matches
-** the sample. nLt is the approximate number of entires whose left-most
-** column is less than the sample. The nDLt column is the approximate
-** number of distinct left-most entries in the index that are less than
-** the sample.
+** of the INTEGER PRIMARY KEY. The sample column is a blob which is the
+** binary encoding of a key from the index, with the trailing rowid
+** omitted. The nEq column is a list of integers. The first integer
+** is the approximate number of entires in the index whose left-most
+** column exactly matches the left-most column the sample. The second
+** integer in nEq is the approximate number of entires in the index where
+** the first two columns match the first two columns of the sample.
+** And so forth. nLt is another list of integer that show the approximate
+** number of entires that are strictly less than the sample. The first
+** integer in nLt contains the number of entries in the index where the
+** left-most column is less than the left-most column of the sample.
+** The K-th integer in the nLt entry is the number of index entries
+** where the first K columns are less than the first K columns of the
+** sample. The nDLt column is like nLt except that it contains the
+** number of distinct entries in the index that are less than the
+** sample.
**
-** Future versions of SQLite might change to store a string containing
-** multiple integers values in the nDLt column of sqlite_stat3. The first
-** integer will be the number of prior index entires that are distinct in
-** the left-most column. The second integer will be the number of prior index
-** entries that are distinct in the first two columns. The third integer
-** will be the number of prior index entries that are distinct in the first
-** three columns. And so forth. With that extension, the nDLt field is
-** similar in function to the sqlite_stat1.stat field.
-**
-** There can be an arbitrary number of sqlite_stat3 entries per index.
+** There can be an arbitrary number of sqlite_stat4 entries per index.
** The ANALYZE command will typically generate sqlite_stat3 tables
** that contain between 10 and 40 samples which are distributed across
** the key space, though not uniformly, and which include samples with
-** largest possible nEq values.
+** large nEq values.
+**
+** Format for sqlite_stat3 redux:
+**
+** The sqlite_stat3 table is like sqlite_stat4 except that it only
+** looks at the left-most column of the index. The sqlite_stat3.sample
+** column contains the actual value of the left-most column instead
+** of a blob encoding of the complete index key as is found in
+** sqlite_stat4.sample. The nEq, nLt, and nDLt entries of sqlite_stat3
+** all contain just a single integer which is the same as the first
+** integer in the equivalent columns in sqlite_stat4.
*/
#ifndef SQLITE_OMIT_ANALYZE
#include "sqliteInt.h"
projects / thirdparty / sqlite.git / commitdiff
