Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
Overview
Comment: | Update SQLite core library to the 3.23.0 release. |
---|---|
Downloads: | Tarball | ZIP archive |
Timelines: | family | ancestors | descendants | both | trunk |
Files: | files | file ages | folders |
SHA1: |
5540efea0c93d46e7c650b12799117d2 |
User & Date: | mistachkin 2018-04-05 03:42:54.486 |
Context
2018-04-05
| ||
03:45 | Pickup the SQLite core library 3.23.0 docs from upstream. check-in: 6bb5d25e27 user: mistachkin tags: trunk | |
03:42 | Update SQLite core library to the 3.23.0 release. check-in: 5540efea0c user: mistachkin tags: trunk | |
03:39 | Add some comments to a test file. check-in: 72990f698b user: mistachkin tags: trunk | |
Changes
Changes to Doc/Extra/Provider/version.html.
︙ | ︙ | |||
41 42 43 44 45 46 47 | </table> </div> <div id="mainSection"> <div id="mainBody"> <h1 class="heading">Version History</h1> <p><b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b></p> <ul> | | | 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 | </table> </div> <div id="mainSection"> <div id="mainBody"> <h1 class="heading">Version History</h1> <p><b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b></p> <ul> <li>Updated to <a href="https://www.sqlite.org/releaselog/3_23_0.html">SQLite 3.23.0</a>.</li> <li>Prevent GetSchemaTable from throwing InvalidCastException. Fix for <a href="https://system.data.sqlite.org/index.html/info/baf42ee135">[baf42ee135]</a>.</li> </ul> <p><b>1.0.108.0 - March 2, 2018</b></p> <ul> <li>Support extended result codes when messages are looked up without the SQLite core library.</li> <li>Override System.Object members for the SQLiteException class to improve its ToString return value. Pursuant to <a href="https://system.data.sqlite.org/index.html/info/53962f9eff">[53962f9eff]</a>.</li> <li>More database connection configuration options for the <a href="https://www.sqlite.org/c3ref/db_config.html">sqlite3_db_config()</a> interface. <b>** Potentially Incompatible Change **</b></li> |
︙ | ︙ |
Changes to SQLite.Interop/src/core/sqlite3.c.
︙ | ︙ | |||
209 210 211 212 213 214 215 | #if SQLITE_ENABLE_ATOMIC_WRITE "ENABLE_ATOMIC_WRITE", #endif #if SQLITE_ENABLE_BATCH_ATOMIC_WRITE "ENABLE_BATCH_ATOMIC_WRITE", #endif #if SQLITE_ENABLE_CEROD | | | 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 | #if SQLITE_ENABLE_ATOMIC_WRITE "ENABLE_ATOMIC_WRITE", #endif #if SQLITE_ENABLE_BATCH_ATOMIC_WRITE "ENABLE_BATCH_ATOMIC_WRITE", #endif #if SQLITE_ENABLE_CEROD "ENABLE_CEROD=" CTIMEOPT_VAL(SQLITE_ENABLE_CEROD), #endif #if SQLITE_ENABLE_COLUMN_METADATA "ENABLE_COLUMN_METADATA", #endif #if SQLITE_ENABLE_COLUMN_USED_MASK "ENABLE_COLUMN_USED_MASK", #endif |
︙ | ︙ | |||
1145 1146 1147 1148 1149 1150 1151 | ** ** See also: [sqlite3_libversion()], ** [sqlite3_libversion_number()], [sqlite3_sourceid()], ** [sqlite_version()] and [sqlite_source_id()]. */ #define SQLITE_VERSION "3.23.0" #define SQLITE_VERSION_NUMBER 3023000 | | | 1145 1146 1147 1148 1149 1150 1151 1152 1153 1154 1155 1156 1157 1158 1159 | ** ** See also: [sqlite3_libversion()], ** [sqlite3_libversion_number()], [sqlite3_sourceid()], ** [sqlite_version()] and [sqlite_source_id()]. */ #define SQLITE_VERSION "3.23.0" #define SQLITE_VERSION_NUMBER 3023000 #define SQLITE_SOURCE_ID "2018-04-02 11:04:16 736b53f57f70b23172c30880186dce7ad9baa3b74e3838cae5847cffb98f5cd2" /* ** CAPI3REF: Run-Time Library Version Numbers ** KEYWORDS: sqlite3_version sqlite3_sourceid ** ** These interfaces provide the same information as the [SQLITE_VERSION], ** [SQLITE_VERSION_NUMBER], and [SQLITE_SOURCE_ID] C preprocessor macros |
︙ | ︙ | |||
2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 | ** The [SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE] opcode causes all write ** operations since the previous successful call to ** [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE] to be rolled back. ** ^This file control takes the file descriptor out of batch write mode ** so that all subsequent write operations are independent. ** ^SQLite will never invoke SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE without ** a prior successful call to [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE]. ** </ul> */ #define SQLITE_FCNTL_LOCKSTATE 1 #define SQLITE_FCNTL_GET_LOCKPROXYFILE 2 #define SQLITE_FCNTL_SET_LOCKPROXYFILE 3 #define SQLITE_FCNTL_LAST_ERRNO 4 #define SQLITE_FCNTL_SIZE_HINT 5 | > > > > > > | 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 2101 2102 2103 | ** The [SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE] opcode causes all write ** operations since the previous successful call to ** [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE] to be rolled back. ** ^This file control takes the file descriptor out of batch write mode ** so that all subsequent write operations are independent. ** ^SQLite will never invoke SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE without ** a prior successful call to [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE]. ** ** <li>[[SQLITE_FCNTL_LOCK_TIMEOUT]] ** The [SQLITE_FCNTL_LOCK_TIMEOUT] opcode causes attempts to obtain ** a file lock using the xLock or xShmLock methods of the VFS to wait ** for up to M milliseconds before failing, where M is the single ** unsigned integer parameter. ** </ul> */ #define SQLITE_FCNTL_LOCKSTATE 1 #define SQLITE_FCNTL_GET_LOCKPROXYFILE 2 #define SQLITE_FCNTL_SET_LOCKPROXYFILE 3 #define SQLITE_FCNTL_LAST_ERRNO 4 #define SQLITE_FCNTL_SIZE_HINT 5 |
︙ | ︙ | |||
2118 2119 2120 2121 2122 2123 2124 2125 2126 2127 2128 2129 2130 2131 | #define SQLITE_FCNTL_VFS_POINTER 27 #define SQLITE_FCNTL_JOURNAL_POINTER 28 #define SQLITE_FCNTL_WIN32_GET_HANDLE 29 #define SQLITE_FCNTL_PDB 30 #define SQLITE_FCNTL_BEGIN_ATOMIC_WRITE 31 #define SQLITE_FCNTL_COMMIT_ATOMIC_WRITE 32 #define SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE 33 /* deprecated names */ #define SQLITE_GET_LOCKPROXYFILE SQLITE_FCNTL_GET_LOCKPROXYFILE #define SQLITE_SET_LOCKPROXYFILE SQLITE_FCNTL_SET_LOCKPROXYFILE #define SQLITE_LAST_ERRNO SQLITE_FCNTL_LAST_ERRNO | > | 2124 2125 2126 2127 2128 2129 2130 2131 2132 2133 2134 2135 2136 2137 2138 | #define SQLITE_FCNTL_VFS_POINTER 27 #define SQLITE_FCNTL_JOURNAL_POINTER 28 #define SQLITE_FCNTL_WIN32_GET_HANDLE 29 #define SQLITE_FCNTL_PDB 30 #define SQLITE_FCNTL_BEGIN_ATOMIC_WRITE 31 #define SQLITE_FCNTL_COMMIT_ATOMIC_WRITE 32 #define SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE 33 #define SQLITE_FCNTL_LOCK_TIMEOUT 34 /* deprecated names */ #define SQLITE_GET_LOCKPROXYFILE SQLITE_FCNTL_GET_LOCKPROXYFILE #define SQLITE_SET_LOCKPROXYFILE SQLITE_FCNTL_SET_LOCKPROXYFILE #define SQLITE_LAST_ERRNO SQLITE_FCNTL_LAST_ERRNO |
︙ | ︙ | |||
3074 3075 3076 3077 3078 3079 3080 | ** ** <dt>SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE</dt> ** <dd> Usually, when a database in wal mode is closed or detached from a ** database handle, SQLite checks if this will mean that there are now no ** connections at all to the database. If so, it performs a checkpoint ** operation before closing the connection. This option may be used to ** override this behaviour. The first parameter passed to this operation | | > | > > > > > > > | > | 3081 3082 3083 3084 3085 3086 3087 3088 3089 3090 3091 3092 3093 3094 3095 3096 3097 3098 3099 3100 3101 3102 3103 3104 3105 3106 3107 3108 3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 | ** ** <dt>SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE</dt> ** <dd> Usually, when a database in wal mode is closed or detached from a ** database handle, SQLite checks if this will mean that there are now no ** connections at all to the database. If so, it performs a checkpoint ** operation before closing the connection. This option may be used to ** override this behaviour. The first parameter passed to this operation ** is an integer - positive to disable checkpoints-on-close, or zero (the ** default) to enable them, and negative to leave the setting unchanged. ** The second parameter is a pointer to an integer ** into which is written 0 or 1 to indicate whether checkpoints-on-close ** have been disabled - 0 if they are not disabled, 1 if they are. ** </dd> ** ** <dt>SQLITE_DBCONFIG_ENABLE_QPSG</dt> ** <dd>^(The SQLITE_DBCONFIG_ENABLE_QPSG option activates or deactivates ** the [query planner stability guarantee] (QPSG). When the QPSG is active, ** a single SQL query statement will always use the same algorithm regardless ** of values of [bound parameters].)^ The QPSG disables some query optimizations ** that look at the values of bound parameters, which can make some queries ** slower. But the QPSG has the advantage of more predictable behavior. With ** the QPSG active, SQLite will always use the same query plan in the field as ** was used during testing in the lab. ** The first argument to this setting is an integer which is 0 to disable ** the QPSG, positive to enable QPSG, or negative to leave the setting ** unchanged. The second parameter is a pointer to an integer into which ** is written 0 or 1 to indicate whether the QPSG is disabled or enabled ** following this call. ** </dd> ** ** <dt>SQLITE_DBCONFIG_TRIGGER_EQP</dt> ** <dd> By default, the output of EXPLAIN QUERY PLAN commands does not ** include output for any operations performed by trigger programs. This ** option is used to set or clear (the default) a flag that governs this ** behavior. The first parameter passed to this operation is an integer - ** positive to enable output for trigger programs, or zero to disable it, ** or negative to leave the setting unchanged. ** The second parameter is a pointer to an integer into which is written ** 0 or 1 to indicate whether output-for-triggers has been disabled - 0 if ** it is not disabled, 1 if it is. ** </dd> ** </dl> */ #define SQLITE_DBCONFIG_MAINDBNAME 1000 /* const char* */ |
︙ | ︙ | |||
8249 8250 8251 8252 8253 8254 8255 8256 8257 8258 8259 8260 8261 8262 8263 8264 8265 8266 8267 8268 8269 8270 8271 8272 8273 8274 8275 | ** wal file in wal mode databases, or the number of pages written to the ** database file in rollback mode databases. Any pages written as part of ** transaction rollback or database recovery operations are not included. ** If an IO or other error occurs while writing a page to disk, the effect ** on subsequent SQLITE_DBSTATUS_CACHE_WRITE requests is undefined.)^ ^The ** highwater mark associated with SQLITE_DBSTATUS_CACHE_WRITE is always 0. ** </dd> ** ** [[SQLITE_DBSTATUS_DEFERRED_FKS]] ^(<dt>SQLITE_DBSTATUS_DEFERRED_FKS</dt> ** <dd>This parameter returns zero for the current value if and only if ** all foreign key constraints (deferred or immediate) have been ** resolved.)^ ^The highwater mark is always 0. ** </dd> ** </dl> */ #define SQLITE_DBSTATUS_LOOKASIDE_USED 0 #define SQLITE_DBSTATUS_CACHE_USED 1 #define SQLITE_DBSTATUS_SCHEMA_USED 2 #define SQLITE_DBSTATUS_STMT_USED 3 #define SQLITE_DBSTATUS_LOOKASIDE_HIT 4 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_SIZE 5 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_FULL 6 #define SQLITE_DBSTATUS_CACHE_HIT 7 #define SQLITE_DBSTATUS_CACHE_MISS 8 #define SQLITE_DBSTATUS_CACHE_WRITE 9 #define SQLITE_DBSTATUS_DEFERRED_FKS 10 #define SQLITE_DBSTATUS_CACHE_USED_SHARED 11 | > > > > > > > > > > | | 8265 8266 8267 8268 8269 8270 8271 8272 8273 8274 8275 8276 8277 8278 8279 8280 8281 8282 8283 8284 8285 8286 8287 8288 8289 8290 8291 8292 8293 8294 8295 8296 8297 8298 8299 8300 8301 8302 8303 8304 8305 8306 8307 8308 8309 | ** wal file in wal mode databases, or the number of pages written to the ** database file in rollback mode databases. Any pages written as part of ** transaction rollback or database recovery operations are not included. ** If an IO or other error occurs while writing a page to disk, the effect ** on subsequent SQLITE_DBSTATUS_CACHE_WRITE requests is undefined.)^ ^The ** highwater mark associated with SQLITE_DBSTATUS_CACHE_WRITE is always 0. ** </dd> ** ** [[SQLITE_DBSTATUS_CACHE_SPILL]] ^(<dt>SQLITE_DBSTATUS_CACHE_SPILL</dt> ** <dd>This parameter returns the number of dirty cache entries that have ** been written to disk in the middle of a transaction due to the page ** cache overflowing. Transactions are more efficient if they are written ** to disk all at once. When pages spill mid-transaction, that introduces ** additional overhead. This parameter can be used help identify ** inefficiencies that can be resolve by increasing the cache size. ** </dd> ** ** [[SQLITE_DBSTATUS_DEFERRED_FKS]] ^(<dt>SQLITE_DBSTATUS_DEFERRED_FKS</dt> ** <dd>This parameter returns zero for the current value if and only if ** all foreign key constraints (deferred or immediate) have been ** resolved.)^ ^The highwater mark is always 0. ** </dd> ** </dl> */ #define SQLITE_DBSTATUS_LOOKASIDE_USED 0 #define SQLITE_DBSTATUS_CACHE_USED 1 #define SQLITE_DBSTATUS_SCHEMA_USED 2 #define SQLITE_DBSTATUS_STMT_USED 3 #define SQLITE_DBSTATUS_LOOKASIDE_HIT 4 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_SIZE 5 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_FULL 6 #define SQLITE_DBSTATUS_CACHE_HIT 7 #define SQLITE_DBSTATUS_CACHE_MISS 8 #define SQLITE_DBSTATUS_CACHE_WRITE 9 #define SQLITE_DBSTATUS_DEFERRED_FKS 10 #define SQLITE_DBSTATUS_CACHE_USED_SHARED 11 #define SQLITE_DBSTATUS_CACHE_SPILL 12 #define SQLITE_DBSTATUS_MAX 12 /* Largest defined DBSTATUS */ /* ** CAPI3REF: Prepared Statement Status ** METHOD: sqlite3_stmt ** ** ^(Each prepared statement maintains various |
︙ | ︙ | |||
9812 9813 9814 9815 9816 9817 9818 | */ #define SQLITE_SERIALIZE_NOCOPY 0x001 /* Do no memory allocations */ /* ** CAPI3REF: Deserialize a database ** ** The sqlite3_deserialize(D,S,P,N,M,F) interface causes the | | | 9838 9839 9840 9841 9842 9843 9844 9845 9846 9847 9848 9849 9850 9851 9852 | */ #define SQLITE_SERIALIZE_NOCOPY 0x001 /* Do no memory allocations */ /* ** CAPI3REF: Deserialize a database ** ** The sqlite3_deserialize(D,S,P,N,M,F) interface causes the ** [database connection] D to disconnect from database S and then ** reopen S as an in-memory database based on the serialization contained ** in P. The serialized database P is N bytes in size. M is the size of ** the buffer P, which might be larger than N. If M is larger than N, and ** the SQLITE_DESERIALIZE_READONLY bit is not set in F, then SQLite is ** permitted to add content to the in-memory database as long as the total ** size does not exceed M bytes. ** |
︙ | ︙ | |||
9859 9860 9861 9862 9863 9864 9865 | ** The SQLITE_DESERIALIZE_FREEONCLOSE means that the database serialization ** in the P argument is held in memory obtained from [sqlite3_malloc64()] ** and that SQLite should take ownership of this memory and automatically ** free it when it has finished using it. Without this flag, the caller ** is resposible for freeing any dynamically allocated memory. ** ** The SQLITE_DESERIALIZE_RESIZEABLE flag means that SQLite is allowed to | | | 9885 9886 9887 9888 9889 9890 9891 9892 9893 9894 9895 9896 9897 9898 9899 | ** The SQLITE_DESERIALIZE_FREEONCLOSE means that the database serialization ** in the P argument is held in memory obtained from [sqlite3_malloc64()] ** and that SQLite should take ownership of this memory and automatically ** free it when it has finished using it. Without this flag, the caller ** is resposible for freeing any dynamically allocated memory. ** ** The SQLITE_DESERIALIZE_RESIZEABLE flag means that SQLite is allowed to ** grow the size of the database using calls to [sqlite3_realloc64()]. This ** flag should only be used if SQLITE_DESERIALIZE_FREEONCLOSE is also used. ** Without this flag, the deserialized database cannot increase in size beyond ** the number of bytes specified by the M parameter. ** ** The SQLITE_DESERIALIZE_READONLY flag means that the deserialized database ** should be treated as read-only. */ |
︙ | ︙ | |||
10953 10954 10955 10956 10957 10958 10959 | ** DESTRUCTOR: sqlite3_changegroup */ SQLITE_API void sqlite3changegroup_delete(sqlite3_changegroup*); /* ** CAPI3REF: Apply A Changeset To A Database ** | | | | | | | | | < | 10979 10980 10981 10982 10983 10984 10985 10986 10987 10988 10989 10990 10991 10992 10993 10994 10995 10996 10997 10998 10999 11000 11001 11002 11003 11004 | ** DESTRUCTOR: sqlite3_changegroup */ SQLITE_API void sqlite3changegroup_delete(sqlite3_changegroup*); /* ** CAPI3REF: Apply A Changeset To A Database ** ** Apply a changeset or patchset to a database. These functions attempt to ** update the "main" database attached to handle db with the changes found in ** the changeset passed via the second and third arguments. ** ** The fourth argument (xFilter) passed to these functions is the "filter ** callback". If it is not NULL, then for each table affected by at least one ** change in the changeset, the filter callback is invoked with ** the table name as the second argument, and a copy of the context pointer ** passed as the sixth argument as the first. If the "filter callback" ** returns zero, then no attempt is made to apply any changes to the table. ** Otherwise, if the return value is non-zero or the xFilter argument to ** is NULL, all changes related to the table are attempted. ** ** For each table that is not excluded by the filter callback, this function ** tests that the target database contains a compatible table. A table is ** considered compatible if all of the following are true: ** ** <ul> ** <li> The table has the same name as the name recorded in the |
︙ | ︙ | |||
11010 11011 11012 11013 11014 11015 11016 | ** actions are taken by sqlite3changeset_apply() depending on the value ** returned by each invocation of the conflict-handler function. Refer to ** the documentation for the three ** [SQLITE_CHANGESET_OMIT|available return values] for details. ** ** <dl> ** <dt>DELETE Changes<dd> | | | 11035 11036 11037 11038 11039 11040 11041 11042 11043 11044 11045 11046 11047 11048 11049 | ** actions are taken by sqlite3changeset_apply() depending on the value ** returned by each invocation of the conflict-handler function. Refer to ** the documentation for the three ** [SQLITE_CHANGESET_OMIT|available return values] for details. ** ** <dl> ** <dt>DELETE Changes<dd> ** For each DELETE change, the function checks if the target database ** contains a row with the same primary key value (or values) as the ** original row values stored in the changeset. If it does, and the values ** stored in all non-primary key columns also match the values stored in ** the changeset the row is deleted from the target database. ** ** If a row with matching primary key values is found, but one or more of ** the non-primary key fields contains a value different from the original |
︙ | ︙ | |||
11055 11056 11057 11058 11059 11060 11061 | ** violation (e.g. NOT NULL or UNIQUE), the conflict handler function is ** invoked with the second argument set to [SQLITE_CHANGESET_CONSTRAINT]. ** This includes the case where the INSERT operation is re-attempted because ** an earlier call to the conflict handler function returned ** [SQLITE_CHANGESET_REPLACE]. ** ** <dt>UPDATE Changes<dd> | | | 11080 11081 11082 11083 11084 11085 11086 11087 11088 11089 11090 11091 11092 11093 11094 | ** violation (e.g. NOT NULL or UNIQUE), the conflict handler function is ** invoked with the second argument set to [SQLITE_CHANGESET_CONSTRAINT]. ** This includes the case where the INSERT operation is re-attempted because ** an earlier call to the conflict handler function returned ** [SQLITE_CHANGESET_REPLACE]. ** ** <dt>UPDATE Changes<dd> ** For each UPDATE change, the function checks if the target database ** contains a row with the same primary key value (or values) as the ** original row values stored in the changeset. If it does, and the values ** stored in all modified non-primary key columns also match the values ** stored in the changeset the row is updated within the target database. ** ** If a row with matching primary key values is found, but one or more of ** the modified non-primary key fields contains a value different from an |
︙ | ︙ | |||
11086 11087 11088 11089 11090 11091 11092 | ** </dl> ** ** It is safe to execute SQL statements, including those that write to the ** table that the callback related to, from within the xConflict callback. ** This can be used to further customize the applications conflict ** resolution strategy. ** | | > > > > > > > > > > > > > > > > > > > > > > > > > > | 11111 11112 11113 11114 11115 11116 11117 11118 11119 11120 11121 11122 11123 11124 11125 11126 11127 11128 11129 11130 11131 11132 11133 11134 11135 11136 11137 11138 11139 11140 11141 11142 11143 11144 11145 11146 11147 11148 11149 11150 11151 11152 11153 11154 11155 11156 11157 11158 11159 11160 11161 11162 11163 11164 11165 11166 11167 11168 11169 11170 | ** </dl> ** ** It is safe to execute SQL statements, including those that write to the ** table that the callback related to, from within the xConflict callback. ** This can be used to further customize the applications conflict ** resolution strategy. ** ** All changes made by these functions are enclosed in a savepoint transaction. ** If any other error (aside from a constraint failure when attempting to ** write to the target database) occurs, then the savepoint transaction is ** rolled back, restoring the target database to its original state, and an ** SQLite error code returned. ** ** If the output parameters (ppRebase) and (pnRebase) are non-NULL and ** the input is a changeset (not a patchset), then sqlite3changeset_apply_v2() ** may set (*ppRebase) to point to a "rebase" that may be used with the ** sqlite3_rebaser APIs buffer before returning. In this case (*pnRebase) ** is set to the size of the buffer in bytes. It is the responsibility of the ** caller to eventually free any such buffer using sqlite3_free(). The buffer ** is only allocated and populated if one or more conflicts were encountered ** while applying the patchset. See comments surrounding the sqlite3_rebaser ** APIs for further details. */ SQLITE_API int sqlite3changeset_apply( sqlite3 *db, /* Apply change to "main" db of this handle */ int nChangeset, /* Size of changeset in bytes */ void *pChangeset, /* Changeset blob */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_apply_v2( sqlite3 *db, /* Apply change to "main" db of this handle */ int nChangeset, /* Size of changeset in bytes */ void *pChangeset, /* Changeset blob */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ); /* ** CAPI3REF: Constants Passed To The Conflict Handler ** ** Values that may be passed as the second argument to a conflict-handler. ** |
︙ | ︙ | |||
11204 11205 11206 11207 11208 11209 11210 11211 11212 11213 11214 11215 11216 11217 | ** and the call to sqlite3changeset_apply() returns SQLITE_ABORT. ** </dl> */ #define SQLITE_CHANGESET_OMIT 0 #define SQLITE_CHANGESET_REPLACE 1 #define SQLITE_CHANGESET_ABORT 2 /* ** CAPI3REF: Streaming Versions of API functions. ** ** The six streaming API xxx_strm() functions serve similar purposes to the ** corresponding non-streaming API functions: ** ** <table border=1 style="margin-left:8ex;margin-right:8ex"> | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 11255 11256 11257 11258 11259 11260 11261 11262 11263 11264 11265 11266 11267 11268 11269 11270 11271 11272 11273 11274 11275 11276 11277 11278 11279 11280 11281 11282 11283 11284 11285 11286 11287 11288 11289 11290 11291 11292 11293 11294 11295 11296 11297 11298 11299 11300 11301 11302 11303 11304 11305 11306 11307 11308 11309 11310 11311 11312 11313 11314 11315 11316 11317 11318 11319 11320 11321 11322 11323 11324 11325 11326 11327 11328 11329 11330 11331 11332 11333 11334 11335 11336 11337 11338 11339 11340 11341 11342 11343 11344 11345 11346 11347 11348 11349 11350 11351 11352 11353 11354 11355 11356 11357 11358 11359 11360 11361 11362 11363 11364 11365 11366 11367 11368 11369 11370 11371 11372 11373 11374 11375 11376 11377 11378 11379 11380 11381 11382 11383 11384 11385 11386 11387 11388 11389 11390 11391 11392 11393 11394 11395 11396 11397 11398 11399 11400 11401 11402 11403 11404 11405 11406 11407 11408 11409 11410 11411 11412 11413 11414 11415 11416 11417 11418 11419 11420 11421 11422 11423 | ** and the call to sqlite3changeset_apply() returns SQLITE_ABORT. ** </dl> */ #define SQLITE_CHANGESET_OMIT 0 #define SQLITE_CHANGESET_REPLACE 1 #define SQLITE_CHANGESET_ABORT 2 /* ** CAPI3REF: Rebasing changesets ** EXPERIMENTAL ** ** Suppose there is a site hosting a database in state S0. And that ** modifications are made that move that database to state S1 and a ** changeset recorded (the "local" changeset). Then, a changeset based ** on S0 is received from another site (the "remote" changeset) and ** applied to the database. The database is then in state ** (S1+"remote"), where the exact state depends on any conflict ** resolution decisions (OMIT or REPLACE) made while applying "remote". ** Rebasing a changeset is to update it to take those conflict ** resolution decisions into account, so that the same conflicts ** do not have to be resolved elsewhere in the network. ** ** For example, if both the local and remote changesets contain an ** INSERT of the same key on "CREATE TABLE t1(a PRIMARY KEY, b)": ** ** local: INSERT INTO t1 VALUES(1, 'v1'); ** remote: INSERT INTO t1 VALUES(1, 'v2'); ** ** and the conflict resolution is REPLACE, then the INSERT change is ** removed from the local changeset (it was overridden). Or, if the ** conflict resolution was "OMIT", then the local changeset is modified ** to instead contain: ** ** UPDATE t1 SET b = 'v2' WHERE a=1; ** ** Changes within the local changeset are rebased as follows: ** ** <dl> ** <dt>Local INSERT<dd> ** This may only conflict with a remote INSERT. If the conflict ** resolution was OMIT, then add an UPDATE change to the rebased ** changeset. Or, if the conflict resolution was REPLACE, add ** nothing to the rebased changeset. ** ** <dt>Local DELETE<dd> ** This may conflict with a remote UPDATE or DELETE. In both cases the ** only possible resolution is OMIT. If the remote operation was a ** DELETE, then add no change to the rebased changeset. If the remote ** operation was an UPDATE, then the old.* fields of change are updated ** to reflect the new.* values in the UPDATE. ** ** <dt>Local UPDATE<dd> ** This may conflict with a remote UPDATE or DELETE. If it conflicts ** with a DELETE, and the conflict resolution was OMIT, then the update ** is changed into an INSERT. Any undefined values in the new.* record ** from the update change are filled in using the old.* values from ** the conflicting DELETE. Or, if the conflict resolution was REPLACE, ** the UPDATE change is simply omitted from the rebased changeset. ** ** If conflict is with a remote UPDATE and the resolution is OMIT, then ** the old.* values are rebased using the new.* values in the remote ** change. Or, if the resolution is REPLACE, then the change is copied ** into the rebased changeset with updates to columns also updated by ** the conflicting remote UPDATE removed. If this means no columns would ** be updated, the change is omitted. ** </dl> ** ** A local change may be rebased against multiple remote changes ** simultaneously. If a single key is modified by multiple remote ** changesets, they are combined as follows before the local changeset ** is rebased: ** ** <ul> ** <li> If there has been one or more REPLACE resolutions on a ** key, it is rebased according to a REPLACE. ** ** <li> If there have been no REPLACE resolutions on a key, then ** the local changeset is rebased according to the most recent ** of the OMIT resolutions. ** </ul> ** ** Note that conflict resolutions from multiple remote changesets are ** combined on a per-field basis, not per-row. This means that in the ** case of multiple remote UPDATE operations, some fields of a single ** local change may be rebased for REPLACE while others are rebased for ** OMIT. ** ** In order to rebase a local changeset, the remote changeset must first ** be applied to the local database using sqlite3changeset_apply_v2() and ** the buffer of rebase information captured. Then: ** ** <ol> ** <li> An sqlite3_rebaser object is created by calling ** sqlite3rebaser_create(). ** <li> The new object is configured with the rebase buffer obtained from ** sqlite3changeset_apply_v2() by calling sqlite3rebaser_configure(). ** If the local changeset is to be rebased against multiple remote ** changesets, then sqlite3rebaser_configure() should be called ** multiple times, in the same order that the multiple ** sqlite3changeset_apply_v2() calls were made. ** <li> Each local changeset is rebased by calling sqlite3rebaser_rebase(). ** <li> The sqlite3_rebaser object is deleted by calling ** sqlite3rebaser_delete(). ** </ol> */ typedef struct sqlite3_rebaser sqlite3_rebaser; /* ** CAPI3REF: Create a changeset rebaser object. ** EXPERIMENTAL ** ** Allocate a new changeset rebaser object. If successful, set (*ppNew) to ** point to the new object and return SQLITE_OK. Otherwise, if an error ** occurs, return an SQLite error code (e.g. SQLITE_NOMEM) and set (*ppNew) ** to NULL. */ SQLITE_API int sqlite3rebaser_create(sqlite3_rebaser **ppNew); /* ** CAPI3REF: Configure a changeset rebaser object. ** EXPERIMENTAL ** ** Configure the changeset rebaser object to rebase changesets according ** to the conflict resolutions described by buffer pRebase (size nRebase ** bytes), which must have been obtained from a previous call to ** sqlite3changeset_apply_v2(). */ SQLITE_API int sqlite3rebaser_configure( sqlite3_rebaser*, int nRebase, const void *pRebase ); /* ** CAPI3REF: Rebase a changeset ** EXPERIMENTAL ** ** Argument pIn must point to a buffer containing a changeset nIn bytes ** in size. This function allocates and populates a buffer with a copy ** of the changeset rebased rebased according to the configuration of the ** rebaser object passed as the first argument. If successful, (*ppOut) ** is set to point to the new buffer containing the rebased changset and ** (*pnOut) to its size in bytes and SQLITE_OK returned. It is the ** responsibility of the caller to eventually free the new buffer using ** sqlite3_free(). Otherwise, if an error occurs, (*ppOut) and (*pnOut) ** are set to zero and an SQLite error code returned. */ SQLITE_API int sqlite3rebaser_rebase( sqlite3_rebaser*, int nIn, const void *pIn, int *pnOut, void **ppOut ); /* ** CAPI3REF: Delete a changeset rebaser object. ** EXPERIMENTAL ** ** Delete the changeset rebaser object and all associated resources. There ** should be one call to this function for each successful invocation ** of sqlite3rebaser_create(). */ SQLITE_API void sqlite3rebaser_delete(sqlite3_rebaser *p); /* ** CAPI3REF: Streaming Versions of API functions. ** ** The six streaming API xxx_strm() functions serve similar purposes to the ** corresponding non-streaming API functions: ** ** <table border=1 style="margin-left:8ex;margin-right:8ex"> |
︙ | ︙ | |||
11307 11308 11309 11310 11311 11312 11313 11314 11315 11316 11317 11318 11319 11320 | ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_concat_strm( int (*xInputA)(void *pIn, void *pData, int *pnData), void *pInA, int (*xInputB)(void *pIn, void *pData, int *pnData), void *pInB, int (*xOutput)(void *pOut, const void *pData, int nData), | > > > > > > > > > > > > > > > > | 11513 11514 11515 11516 11517 11518 11519 11520 11521 11522 11523 11524 11525 11526 11527 11528 11529 11530 11531 11532 11533 11534 11535 11536 11537 11538 11539 11540 11541 11542 | ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_apply_v2_strm( sqlite3 *db, /* Apply change to "main" db of this handle */ int (*xInput)(void *pIn, void *pData, int *pnData), /* Input function */ void *pIn, /* First arg for xInput */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ); SQLITE_API int sqlite3changeset_concat_strm( int (*xInputA)(void *pIn, void *pData, int *pnData), void *pInA, int (*xInputB)(void *pIn, void *pData, int *pnData), void *pInB, int (*xOutput)(void *pOut, const void *pData, int nData), |
︙ | ︙ | |||
11344 11345 11346 11347 11348 11349 11350 11351 11352 11353 11354 11355 11356 11357 | SQLITE_API int sqlite3changegroup_add_strm(sqlite3_changegroup*, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn ); SQLITE_API int sqlite3changegroup_output_strm(sqlite3_changegroup*, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); /* ** Make sure we can call this stuff from C++. */ #if 0 | > > > > > > > | 11566 11567 11568 11569 11570 11571 11572 11573 11574 11575 11576 11577 11578 11579 11580 11581 11582 11583 11584 11585 11586 | SQLITE_API int sqlite3changegroup_add_strm(sqlite3_changegroup*, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn ); SQLITE_API int sqlite3changegroup_output_strm(sqlite3_changegroup*, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); SQLITE_API int sqlite3rebaser_rebase_strm( sqlite3_rebaser *pRebaser, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); /* ** Make sure we can call this stuff from C++. */ #if 0 |
︙ | ︙ | |||
13224 13225 13226 13227 13228 13229 13230 | ** The sqlite.busyHandler member of the sqlite struct contains the busy ** callback for the database handle. Each pager opened via the sqlite ** handle is passed a pointer to sqlite.busyHandler. The busy-handler ** callback is currently invoked only from within pager.c. */ typedef struct BusyHandler BusyHandler; struct BusyHandler { | | | > | 13453 13454 13455 13456 13457 13458 13459 13460 13461 13462 13463 13464 13465 13466 13467 13468 13469 13470 | ** The sqlite.busyHandler member of the sqlite struct contains the busy ** callback for the database handle. Each pager opened via the sqlite ** handle is passed a pointer to sqlite.busyHandler. The busy-handler ** callback is currently invoked only from within pager.c. */ typedef struct BusyHandler BusyHandler; struct BusyHandler { int (*xBusyHandler)(void *,int); /* The busy callback */ void *pBusyArg; /* First arg to busy callback */ int nBusy; /* Incremented with each busy call */ u8 bExtraFileArg; /* Include sqlite3_file as callback arg */ }; /* ** Name of the master database table. The master database table ** is a special table that holds the names and attributes of all ** user tables and indices. */ |
︙ | ︙ | |||
14435 14436 14437 14438 14439 14440 14441 | int, void(*)(DbPage*) ); SQLITE_PRIVATE int sqlite3PagerClose(Pager *pPager, sqlite3*); SQLITE_PRIVATE int sqlite3PagerReadFileheader(Pager*, int, unsigned char*); /* Functions used to configure a Pager object. */ | | | 14665 14666 14667 14668 14669 14670 14671 14672 14673 14674 14675 14676 14677 14678 14679 | int, void(*)(DbPage*) ); SQLITE_PRIVATE int sqlite3PagerClose(Pager *pPager, sqlite3*); SQLITE_PRIVATE int sqlite3PagerReadFileheader(Pager*, int, unsigned char*); /* Functions used to configure a Pager object. */ SQLITE_PRIVATE void sqlite3PagerSetBusyHandler(Pager*, int(*)(void *), void *); SQLITE_PRIVATE int sqlite3PagerSetPagesize(Pager*, u32*, int); #ifdef SQLITE_HAS_CODEC SQLITE_PRIVATE void sqlite3PagerAlignReserve(Pager*,Pager*); #endif SQLITE_PRIVATE int sqlite3PagerMaxPageCount(Pager*, int); SQLITE_PRIVATE void sqlite3PagerSetCachesize(Pager*, int); SQLITE_PRIVATE int sqlite3PagerSetSpillsize(Pager*, int); |
︙ | ︙ | |||
14521 14522 14523 14524 14525 14526 14527 14528 14529 14530 14531 14532 14533 14534 | SQLITE_PRIVATE sqlite3_file *sqlite3PagerJrnlFile(Pager*); SQLITE_PRIVATE const char *sqlite3PagerJournalname(Pager*); SQLITE_PRIVATE void *sqlite3PagerTempSpace(Pager*); SQLITE_PRIVATE int sqlite3PagerIsMemdb(Pager*); SQLITE_PRIVATE void sqlite3PagerCacheStat(Pager *, int, int, int *); SQLITE_PRIVATE void sqlite3PagerClearCache(Pager*); SQLITE_PRIVATE int sqlite3SectorSize(sqlite3_file *); /* Functions used to truncate the database file. */ SQLITE_PRIVATE void sqlite3PagerTruncateImage(Pager*,Pgno); SQLITE_PRIVATE void sqlite3PagerRekey(DbPage*, Pgno, u16); #if defined(SQLITE_HAS_CODEC) && !defined(SQLITE_OMIT_WAL) | > > > > > | 14751 14752 14753 14754 14755 14756 14757 14758 14759 14760 14761 14762 14763 14764 14765 14766 14767 14768 14769 | SQLITE_PRIVATE sqlite3_file *sqlite3PagerJrnlFile(Pager*); SQLITE_PRIVATE const char *sqlite3PagerJournalname(Pager*); SQLITE_PRIVATE void *sqlite3PagerTempSpace(Pager*); SQLITE_PRIVATE int sqlite3PagerIsMemdb(Pager*); SQLITE_PRIVATE void sqlite3PagerCacheStat(Pager *, int, int, int *); SQLITE_PRIVATE void sqlite3PagerClearCache(Pager*); SQLITE_PRIVATE int sqlite3SectorSize(sqlite3_file *); #ifdef SQLITE_ENABLE_SETLK_TIMEOUT SQLITE_PRIVATE void sqlite3PagerResetLockTimeout(Pager *pPager); #else # define sqlite3PagerResetLockTimeout(X) #endif /* Functions used to truncate the database file. */ SQLITE_PRIVATE void sqlite3PagerTruncateImage(Pager*,Pgno); SQLITE_PRIVATE void sqlite3PagerRekey(DbPage*, Pgno, u16); #if defined(SQLITE_HAS_CODEC) && !defined(SQLITE_OMIT_WAL) |
︙ | ︙ | |||
15507 15508 15509 15510 15511 15512 15513 15514 15515 15516 15517 15518 15519 15520 | #define SQLITE_OrderByIdxJoin 0x0040 /* ORDER BY of joins via index */ #define SQLITE_Transitive 0x0080 /* Transitive constraints */ #define SQLITE_OmitNoopJoin 0x0100 /* Omit unused tables in joins */ #define SQLITE_CountOfView 0x0200 /* The count-of-view optimization */ #define SQLITE_CursorHints 0x0400 /* Add OP_CursorHint opcodes */ #define SQLITE_Stat34 0x0800 /* Use STAT3 or STAT4 data */ /* TH3 expects the Stat34 ^^^^^^ value to be 0x0800. Don't change it */ #define SQLITE_AllOpts 0xffff /* All optimizations */ /* ** Macros for testing whether or not optimizations are enabled or disabled. */ #define OptimizationDisabled(db, mask) (((db)->dbOptFlags&(mask))!=0) #define OptimizationEnabled(db, mask) (((db)->dbOptFlags&(mask))==0) | > > | 15742 15743 15744 15745 15746 15747 15748 15749 15750 15751 15752 15753 15754 15755 15756 15757 | #define SQLITE_OrderByIdxJoin 0x0040 /* ORDER BY of joins via index */ #define SQLITE_Transitive 0x0080 /* Transitive constraints */ #define SQLITE_OmitNoopJoin 0x0100 /* Omit unused tables in joins */ #define SQLITE_CountOfView 0x0200 /* The count-of-view optimization */ #define SQLITE_CursorHints 0x0400 /* Add OP_CursorHint opcodes */ #define SQLITE_Stat34 0x0800 /* Use STAT3 or STAT4 data */ /* TH3 expects the Stat34 ^^^^^^ value to be 0x0800. Don't change it */ #define SQLITE_PushDown 0x1000 /* The push-down optimization */ #define SQLITE_SimplifyJoin 0x2000 /* Convert LEFT JOIN to JOIN */ #define SQLITE_AllOpts 0xffff /* All optimizations */ /* ** Macros for testing whether or not optimizations are enabled or disabled. */ #define OptimizationDisabled(db, mask) (((db)->dbOptFlags&(mask))!=0) #define OptimizationEnabled(db, mask) (((db)->dbOptFlags&(mask))==0) |
︙ | ︙ | |||
16968 16969 16970 16971 16972 16973 16974 | yDbMask writeMask; /* Start a write transaction on these databases */ yDbMask cookieMask; /* Bitmask of schema verified databases */ int regRowid; /* Register holding rowid of CREATE TABLE entry */ int regRoot; /* Register holding root page number for new objects */ int nMaxArg; /* Max args passed to user function by sub-program */ #if SELECTTRACE_ENABLED int nSelect; /* Number of SELECT statements seen */ | < | 17205 17206 17207 17208 17209 17210 17211 17212 17213 17214 17215 17216 17217 17218 | yDbMask writeMask; /* Start a write transaction on these databases */ yDbMask cookieMask; /* Bitmask of schema verified databases */ int regRowid; /* Register holding rowid of CREATE TABLE entry */ int regRoot; /* Register holding root page number for new objects */ int nMaxArg; /* Max args passed to user function by sub-program */ #if SELECTTRACE_ENABLED int nSelect; /* Number of SELECT statements seen */ #endif #ifndef SQLITE_OMIT_SHARED_CACHE int nTableLock; /* Number of locks in aTableLock */ TableLock *aTableLock; /* Required table locks for shared-cache mode */ #endif AutoincInfo *pAinc; /* Information about AUTOINCREMENT counters */ Parse *pToplevel; /* Parse structure for main program (or NULL) */ |
︙ | ︙ | |||
17332 17333 17334 17335 17336 17337 17338 | int n; /* A counter */ int iCur; /* A cursor number */ SrcList *pSrcList; /* FROM clause */ struct SrcCount *pSrcCount; /* Counting column references */ struct CCurHint *pCCurHint; /* Used by codeCursorHint() */ int *aiCol; /* array of column indexes */ struct IdxCover *pIdxCover; /* Check for index coverage */ | | | | 17568 17569 17570 17571 17572 17573 17574 17575 17576 17577 17578 17579 17580 17581 17582 17583 17584 | int n; /* A counter */ int iCur; /* A cursor number */ SrcList *pSrcList; /* FROM clause */ struct SrcCount *pSrcCount; /* Counting column references */ struct CCurHint *pCCurHint; /* Used by codeCursorHint() */ int *aiCol; /* array of column indexes */ struct IdxCover *pIdxCover; /* Check for index coverage */ struct IdxExprTrans *pIdxTrans; /* Convert idxed expr to column */ ExprList *pGroupBy; /* GROUP BY clause */ Select *pSelect; /* HAVING to WHERE clause ctx */ } u; }; /* Forward declarations */ SQLITE_PRIVATE int sqlite3WalkExpr(Walker*, Expr*); SQLITE_PRIVATE int sqlite3WalkExprList(Walker*, ExprList*); SQLITE_PRIVATE int sqlite3WalkSelect(Walker*, Select*); |
︙ | ︙ | |||
17798 17799 17800 17801 17802 17803 17804 17805 17806 17807 17808 17809 17810 17811 | SQLITE_PRIVATE void sqlite3Vacuum(Parse*,Token*); SQLITE_PRIVATE int sqlite3RunVacuum(char**, sqlite3*, int); SQLITE_PRIVATE char *sqlite3NameFromToken(sqlite3*, Token*); SQLITE_PRIVATE int sqlite3ExprCompare(Parse*,Expr*, Expr*, int); SQLITE_PRIVATE int sqlite3ExprCompareSkip(Expr*, Expr*, int); SQLITE_PRIVATE int sqlite3ExprListCompare(ExprList*, ExprList*, int); SQLITE_PRIVATE int sqlite3ExprImpliesExpr(Parse*,Expr*, Expr*, int); SQLITE_PRIVATE void sqlite3ExprAnalyzeAggregates(NameContext*, Expr*); SQLITE_PRIVATE void sqlite3ExprAnalyzeAggList(NameContext*,ExprList*); SQLITE_PRIVATE int sqlite3ExprCoveredByIndex(Expr*, int iCur, Index *pIdx); SQLITE_PRIVATE int sqlite3FunctionUsesThisSrc(Expr*, SrcList*); SQLITE_PRIVATE Vdbe *sqlite3GetVdbe(Parse*); #ifndef SQLITE_UNTESTABLE SQLITE_PRIVATE void sqlite3PrngSaveState(void); | > | 18034 18035 18036 18037 18038 18039 18040 18041 18042 18043 18044 18045 18046 18047 18048 | SQLITE_PRIVATE void sqlite3Vacuum(Parse*,Token*); SQLITE_PRIVATE int sqlite3RunVacuum(char**, sqlite3*, int); SQLITE_PRIVATE char *sqlite3NameFromToken(sqlite3*, Token*); SQLITE_PRIVATE int sqlite3ExprCompare(Parse*,Expr*, Expr*, int); SQLITE_PRIVATE int sqlite3ExprCompareSkip(Expr*, Expr*, int); SQLITE_PRIVATE int sqlite3ExprListCompare(ExprList*, ExprList*, int); SQLITE_PRIVATE int sqlite3ExprImpliesExpr(Parse*,Expr*, Expr*, int); SQLITE_PRIVATE int sqlite3ExprImpliesNonNullRow(Expr*,int); SQLITE_PRIVATE void sqlite3ExprAnalyzeAggregates(NameContext*, Expr*); SQLITE_PRIVATE void sqlite3ExprAnalyzeAggList(NameContext*,ExprList*); SQLITE_PRIVATE int sqlite3ExprCoveredByIndex(Expr*, int iCur, Index *pIdx); SQLITE_PRIVATE int sqlite3FunctionUsesThisSrc(Expr*, SrcList*); SQLITE_PRIVATE Vdbe *sqlite3GetVdbe(Parse*); #ifndef SQLITE_UNTESTABLE SQLITE_PRIVATE void sqlite3PrngSaveState(void); |
︙ | ︙ | |||
18076 18077 18078 18079 18080 18081 18082 | SQLITE_PRIVATE int sqlite3ResolveOrderGroupBy(Parse*, Select*, ExprList*, const char*); SQLITE_PRIVATE void sqlite3ColumnDefault(Vdbe *, Table *, int, int); SQLITE_PRIVATE void sqlite3AlterFinishAddColumn(Parse *, Token *); SQLITE_PRIVATE void sqlite3AlterBeginAddColumn(Parse *, SrcList *); SQLITE_PRIVATE CollSeq *sqlite3GetCollSeq(Parse*, u8, CollSeq *, const char*); SQLITE_PRIVATE char sqlite3AffinityType(const char*, u8*); SQLITE_PRIVATE void sqlite3Analyze(Parse*, Token*, Token*); | | | 18313 18314 18315 18316 18317 18318 18319 18320 18321 18322 18323 18324 18325 18326 18327 | SQLITE_PRIVATE int sqlite3ResolveOrderGroupBy(Parse*, Select*, ExprList*, const char*); SQLITE_PRIVATE void sqlite3ColumnDefault(Vdbe *, Table *, int, int); SQLITE_PRIVATE void sqlite3AlterFinishAddColumn(Parse *, Token *); SQLITE_PRIVATE void sqlite3AlterBeginAddColumn(Parse *, SrcList *); SQLITE_PRIVATE CollSeq *sqlite3GetCollSeq(Parse*, u8, CollSeq *, const char*); SQLITE_PRIVATE char sqlite3AffinityType(const char*, u8*); SQLITE_PRIVATE void sqlite3Analyze(Parse*, Token*, Token*); SQLITE_PRIVATE int sqlite3InvokeBusyHandler(BusyHandler*, sqlite3_file*); SQLITE_PRIVATE int sqlite3FindDb(sqlite3*, Token*); SQLITE_PRIVATE int sqlite3FindDbName(sqlite3 *, const char *); SQLITE_PRIVATE int sqlite3AnalysisLoad(sqlite3*,int iDB); SQLITE_PRIVATE void sqlite3DeleteIndexSamples(sqlite3*,Index*); SQLITE_PRIVATE void sqlite3DefaultRowEst(Index*); SQLITE_PRIVATE void sqlite3RegisterLikeFunctions(sqlite3*, int); SQLITE_PRIVATE int sqlite3IsLikeFunction(sqlite3*,Expr*,int*,char*); |
︙ | ︙ | |||
19623 19624 19625 19626 19627 19628 19629 19630 19631 19632 19633 19634 19635 19636 | } /* ** Set *pCurrent to the total cache hits or misses encountered by all ** pagers the database handle is connected to. *pHighwater is always set ** to zero. */ case SQLITE_DBSTATUS_CACHE_HIT: case SQLITE_DBSTATUS_CACHE_MISS: case SQLITE_DBSTATUS_CACHE_WRITE:{ int i; int nRet = 0; assert( SQLITE_DBSTATUS_CACHE_MISS==SQLITE_DBSTATUS_CACHE_HIT+1 ); assert( SQLITE_DBSTATUS_CACHE_WRITE==SQLITE_DBSTATUS_CACHE_HIT+2 ); | > > > | 19860 19861 19862 19863 19864 19865 19866 19867 19868 19869 19870 19871 19872 19873 19874 19875 19876 | } /* ** Set *pCurrent to the total cache hits or misses encountered by all ** pagers the database handle is connected to. *pHighwater is always set ** to zero. */ case SQLITE_DBSTATUS_CACHE_SPILL: op = SQLITE_DBSTATUS_CACHE_WRITE+1; /* Fall through into the next case */ case SQLITE_DBSTATUS_CACHE_HIT: case SQLITE_DBSTATUS_CACHE_MISS: case SQLITE_DBSTATUS_CACHE_WRITE:{ int i; int nRet = 0; assert( SQLITE_DBSTATUS_CACHE_MISS==SQLITE_DBSTATUS_CACHE_HIT+1 ); assert( SQLITE_DBSTATUS_CACHE_WRITE==SQLITE_DBSTATUS_CACHE_HIT+2 ); |
︙ | ︙ | |||
21044 21045 21046 21047 21048 21049 21050 | ** when simply tossing information over the wall to the VFS and we do not ** really care if the VFS receives and understands the information since it ** is only a hint and can be safely ignored. The sqlite3OsFileControlHint() ** routine has no return value since the return value would be meaningless. */ SQLITE_PRIVATE int sqlite3OsFileControl(sqlite3_file *id, int op, void *pArg){ #ifdef SQLITE_TEST | | > > > | | 21284 21285 21286 21287 21288 21289 21290 21291 21292 21293 21294 21295 21296 21297 21298 21299 21300 21301 21302 21303 21304 21305 21306 21307 21308 21309 21310 21311 21312 21313 21314 21315 21316 21317 21318 | ** when simply tossing information over the wall to the VFS and we do not ** really care if the VFS receives and understands the information since it ** is only a hint and can be safely ignored. The sqlite3OsFileControlHint() ** routine has no return value since the return value would be meaningless. */ SQLITE_PRIVATE int sqlite3OsFileControl(sqlite3_file *id, int op, void *pArg){ #ifdef SQLITE_TEST if( op!=SQLITE_FCNTL_COMMIT_PHASETWO && op!=SQLITE_FCNTL_LOCK_TIMEOUT ){ /* Faults are not injected into COMMIT_PHASETWO because, assuming SQLite ** is using a regular VFS, it is called after the corresponding ** transaction has been committed. Injecting a fault at this point ** confuses the test scripts - the COMMIT comand returns SQLITE_NOMEM ** but the transaction is committed anyway. ** ** The core must call OsFileControl() though, not OsFileControlHint(), ** as if a custom VFS (e.g. zipvfs) returns an error here, it probably ** means the commit really has failed and an error should be returned ** to the user. */ DO_OS_MALLOC_TEST(id); } #endif if( id->pMethods==0 ) return SQLITE_NOTFOUND; return id->pMethods->xFileControl(id, op, pArg); } SQLITE_PRIVATE void sqlite3OsFileControlHint(sqlite3_file *id, int op, void *pArg){ if( id->pMethods ) (void)id->pMethods->xFileControl(id, op, pArg); } SQLITE_PRIVATE int sqlite3OsSectorSize(sqlite3_file *id){ int (*xSectorSize)(sqlite3_file*) = id->pMethods->xSectorSize; return (xSectorSize ? xSectorSize(id) : SQLITE_DEFAULT_SECTOR_SIZE); } SQLITE_PRIVATE int sqlite3OsDeviceCharacteristics(sqlite3_file *id){ |
︙ | ︙ | |||
27301 27302 27303 27304 27305 27306 27307 27308 27309 27310 27311 27312 27313 27314 27315 27316 27317 27318 27319 | pView = sqlite3TreeViewPush(pView, moreToFollow); if( p->pWith ){ sqlite3TreeViewWith(pView, p->pWith, 1); cnt = 1; sqlite3TreeViewPush(pView, 1); } do{ sqlite3TreeViewLine(pView, "SELECT%s%s (0x%p) selFlags=0x%x nSelectRow=%d", ((p->selFlags & SF_Distinct) ? " DISTINCT" : ""), ((p->selFlags & SF_Aggregate) ? " agg_flag" : ""), p, p->selFlags, (int)p->nSelectRow ); if( cnt++ ) sqlite3TreeViewPop(pView); if( p->pPrior ){ n = 1000; }else{ n = 0; if( p->pSrc && p->pSrc->nSrc ) n++; if( p->pWhere ) n++; | > > > > > > > > > > | 27544 27545 27546 27547 27548 27549 27550 27551 27552 27553 27554 27555 27556 27557 27558 27559 27560 27561 27562 27563 27564 27565 27566 27567 27568 27569 27570 27571 27572 | pView = sqlite3TreeViewPush(pView, moreToFollow); if( p->pWith ){ sqlite3TreeViewWith(pView, p->pWith, 1); cnt = 1; sqlite3TreeViewPush(pView, 1); } do{ #if SELECTTRACE_ENABLED sqlite3TreeViewLine(pView, "SELECT%s%s (%s/%p) selFlags=0x%x nSelectRow=%d", ((p->selFlags & SF_Distinct) ? " DISTINCT" : ""), ((p->selFlags & SF_Aggregate) ? " agg_flag" : ""), p->zSelName, p, p->selFlags, (int)p->nSelectRow ); #else sqlite3TreeViewLine(pView, "SELECT%s%s (0x%p) selFlags=0x%x nSelectRow=%d", ((p->selFlags & SF_Distinct) ? " DISTINCT" : ""), ((p->selFlags & SF_Aggregate) ? " agg_flag" : ""), p, p->selFlags, (int)p->nSelectRow ); #endif if( cnt++ ) sqlite3TreeViewPop(pView); if( p->pPrior ){ n = 1000; }else{ n = 0; if( p->pSrc && p->pSrc->nSrc ) n++; if( p->pWhere ) n++; |
︙ | ︙ | |||
30990 30991 30992 30993 30994 30995 30996 30997 30998 30999 31000 31001 31002 31003 | int deviceCharacteristics; /* Precomputed device characteristics */ #if SQLITE_ENABLE_LOCKING_STYLE int openFlags; /* The flags specified at open() */ #endif #if SQLITE_ENABLE_LOCKING_STYLE || defined(__APPLE__) unsigned fsFlags; /* cached details from statfs() */ #endif #if OS_VXWORKS struct vxworksFileId *pId; /* Unique file ID */ #endif #ifdef SQLITE_DEBUG /* The next group of variables are used to track whether or not the ** transaction counter in bytes 24-27 of database files are updated ** whenever any part of the database changes. An assertion fault will | > > > | 31243 31244 31245 31246 31247 31248 31249 31250 31251 31252 31253 31254 31255 31256 31257 31258 31259 | int deviceCharacteristics; /* Precomputed device characteristics */ #if SQLITE_ENABLE_LOCKING_STYLE int openFlags; /* The flags specified at open() */ #endif #if SQLITE_ENABLE_LOCKING_STYLE || defined(__APPLE__) unsigned fsFlags; /* cached details from statfs() */ #endif #ifdef SQLITE_ENABLE_SETLK_TIMEOUT unsigned iBusyTimeout; /* Wait this many millisec on locks */ #endif #if OS_VXWORKS struct vxworksFileId *pId; /* Unique file ID */ #endif #ifdef SQLITE_DEBUG /* The next group of variables are used to track whether or not the ** transaction counter in bytes 24-27 of database files are updated ** whenever any part of the database changes. An assertion fault will |
︙ | ︙ | |||
32423 32424 32425 32426 32427 32428 32429 32430 32431 32432 32433 32434 32435 32436 | unixLeaveMutex(); OSTRACE(("TEST WR-LOCK %d %d %d (unix)\n", pFile->h, rc, reserved)); *pResOut = reserved; return rc; } /* ** Attempt to set a system-lock on the file pFile. The lock is ** described by pLock. ** ** If the pFile was opened read/write from unix-excl, then the only lock ** ever obtained is an exclusive lock, and it is obtained exactly once | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 32679 32680 32681 32682 32683 32684 32685 32686 32687 32688 32689 32690 32691 32692 32693 32694 32695 32696 32697 32698 32699 32700 32701 32702 32703 32704 32705 32706 32707 32708 32709 32710 32711 32712 32713 32714 32715 32716 32717 32718 32719 32720 32721 32722 32723 32724 32725 32726 32727 32728 32729 | unixLeaveMutex(); OSTRACE(("TEST WR-LOCK %d %d %d (unix)\n", pFile->h, rc, reserved)); *pResOut = reserved; return rc; } /* ** Set a posix-advisory-lock. ** ** There are two versions of this routine. If compiled with ** SQLITE_ENABLE_SETLK_TIMEOUT then the routine has an extra parameter ** which is a pointer to a unixFile. If the unixFile->iBusyTimeout ** value is set, then it is the number of milliseconds to wait before ** failing the lock. The iBusyTimeout value is always reset back to ** zero on each call. ** ** If SQLITE_ENABLE_SETLK_TIMEOUT is not defined, then do a non-blocking ** attempt to set the lock. */ #ifndef SQLITE_ENABLE_SETLK_TIMEOUT # define osSetPosixAdvisoryLock(h,x,t) osFcntl(h,F_SETLK,x) #else static int osSetPosixAdvisoryLock( int h, /* The file descriptor on which to take the lock */ struct flock *pLock, /* The description of the lock */ unixFile *pFile /* Structure holding timeout value */ ){ int rc = osFcntl(h,F_SETLK,pLock); while( rc<0 && pFile->iBusyTimeout>0 ){ /* On systems that support some kind of blocking file lock with a timeout, ** make appropriate changes here to invoke that blocking file lock. On ** generic posix, however, there is no such API. So we simply try the ** lock once every millisecond until either the timeout expires, or until ** the lock is obtained. */ usleep(1000); rc = osFcntl(h,F_SETLK,pLock); pFile->iBusyTimeout--; } return rc; } #endif /* SQLITE_ENABLE_SETLK_TIMEOUT */ /* ** Attempt to set a system-lock on the file pFile. The lock is ** described by pLock. ** ** If the pFile was opened read/write from unix-excl, then the only lock ** ever obtained is an exclusive lock, and it is obtained exactly once |
︙ | ︙ | |||
32456 32457 32458 32459 32460 32461 32462 | if( pInode->bProcessLock==0 ){ struct flock lock; assert( pInode->nLock==0 ); lock.l_whence = SEEK_SET; lock.l_start = SHARED_FIRST; lock.l_len = SHARED_SIZE; lock.l_type = F_WRLCK; | | | | 32749 32750 32751 32752 32753 32754 32755 32756 32757 32758 32759 32760 32761 32762 32763 32764 32765 32766 32767 32768 32769 32770 32771 | if( pInode->bProcessLock==0 ){ struct flock lock; assert( pInode->nLock==0 ); lock.l_whence = SEEK_SET; lock.l_start = SHARED_FIRST; lock.l_len = SHARED_SIZE; lock.l_type = F_WRLCK; rc = osSetPosixAdvisoryLock(pFile->h, &lock, pFile); if( rc<0 ) return rc; pInode->bProcessLock = 1; pInode->nLock++; }else{ rc = 0; } }else{ rc = osSetPosixAdvisoryLock(pFile->h, pLock, pFile); } return rc; } /* ** Lock the file with the lock specified by parameter eFileLock - one ** of the following: |
︙ | ︙ | |||
34824 34825 34826 34827 34828 34829 34830 34831 34832 34833 34834 34835 34836 34837 | } return SQLITE_OK; } case SQLITE_FCNTL_HAS_MOVED: { *(int*)pArg = fileHasMoved(pFile); return SQLITE_OK; } #if SQLITE_MAX_MMAP_SIZE>0 case SQLITE_FCNTL_MMAP_SIZE: { i64 newLimit = *(i64*)pArg; int rc = SQLITE_OK; if( newLimit>sqlite3GlobalConfig.mxMmap ){ newLimit = sqlite3GlobalConfig.mxMmap; } | > > > > > > | 35117 35118 35119 35120 35121 35122 35123 35124 35125 35126 35127 35128 35129 35130 35131 35132 35133 35134 35135 35136 | } return SQLITE_OK; } case SQLITE_FCNTL_HAS_MOVED: { *(int*)pArg = fileHasMoved(pFile); return SQLITE_OK; } #ifdef SQLITE_ENABLE_SETLK_TIMEOUT case SQLITE_FCNTL_LOCK_TIMEOUT: { pFile->iBusyTimeout = *(int*)pArg; return SQLITE_OK; } #endif #if SQLITE_MAX_MMAP_SIZE>0 case SQLITE_FCNTL_MMAP_SIZE: { i64 newLimit = *(i64*)pArg; int rc = SQLITE_OK; if( newLimit>sqlite3GlobalConfig.mxMmap ){ newLimit = sqlite3GlobalConfig.mxMmap; } |
︙ | ︙ | |||
35143 35144 35145 35146 35147 35148 35149 | if( pShmNode->h>=0 ){ /* Initialize the locking parameters */ f.l_type = lockType; f.l_whence = SEEK_SET; f.l_start = ofst; f.l_len = n; | | | 35442 35443 35444 35445 35446 35447 35448 35449 35450 35451 35452 35453 35454 35455 35456 | if( pShmNode->h>=0 ){ /* Initialize the locking parameters */ f.l_type = lockType; f.l_whence = SEEK_SET; f.l_start = ofst; f.l_len = n; rc = osSetPosixAdvisoryLock(pShmNode->h, &f, pFile); rc = (rc!=(-1)) ? SQLITE_OK : SQLITE_BUSY; } /* Update the global lock state and do debug tracing */ #ifdef SQLITE_DEBUG { u16 mask; OSTRACE(("SHM-LOCK ")); |
︙ | ︙ | |||
45003 45004 45005 45006 45007 45008 45009 | ** ** May you do good and not evil. ** May you find forgiveness for yourself and forgive others. ** May you share freely, never taking more than you give. ** ****************************************************************************** ** | | | 45302 45303 45304 45305 45306 45307 45308 45309 45310 45311 45312 45313 45314 45315 45316 | ** ** May you do good and not evil. ** May you find forgiveness for yourself and forgive others. ** May you share freely, never taking more than you give. ** ****************************************************************************** ** ** This file implements an in-memory VFS. A database is held as a contiguous ** block of memory. ** ** This file also implements interface sqlite3_serialize() and ** sqlite3_deserialize(). */ #ifdef SQLITE_ENABLE_DESERIALIZE /* #include "sqliteInt.h" */ |
︙ | ︙ | |||
45442 45443 45444 45445 45446 45447 45448 45449 45450 45451 45452 45453 45454 45455 | Btree *pBt; sqlite3_int64 sz; int szPage = 0; sqlite3_stmt *pStmt = 0; unsigned char *pOut; char *zSql; int rc; if( zSchema==0 ) zSchema = db->aDb[0].zDbSName; p = memdbFromDbSchema(db, zSchema); iDb = sqlite3FindDbName(db, zSchema); if( piSize ) *piSize = -1; if( iDb<0 ) return 0; if( p ){ | > > > > > > > | 45741 45742 45743 45744 45745 45746 45747 45748 45749 45750 45751 45752 45753 45754 45755 45756 45757 45758 45759 45760 45761 | Btree *pBt; sqlite3_int64 sz; int szPage = 0; sqlite3_stmt *pStmt = 0; unsigned char *pOut; char *zSql; int rc; #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ){ (void)SQLITE_MISUSE_BKPT; return 0; } #endif if( zSchema==0 ) zSchema = db->aDb[0].zDbSName; p = memdbFromDbSchema(db, zSchema); iDb = sqlite3FindDbName(db, zSchema); if( piSize ) *piSize = -1; if( iDb<0 ) return 0; if( p ){ |
︙ | ︙ | |||
45512 45513 45514 45515 45516 45517 45518 45519 45520 45521 45522 45523 45524 45525 | unsigned mFlags /* Zero or more SQLITE_DESERIALIZE_* flags */ ){ MemFile *p; char *zSql; sqlite3_stmt *pStmt = 0; int rc; int iDb; sqlite3_mutex_enter(db->mutex); if( zSchema==0 ) zSchema = db->aDb[0].zDbSName; iDb = sqlite3FindDbName(db, zSchema); if( iDb<0 ){ rc = SQLITE_ERROR; goto end_deserialize; | > > > > > > > > | 45818 45819 45820 45821 45822 45823 45824 45825 45826 45827 45828 45829 45830 45831 45832 45833 45834 45835 45836 45837 45838 45839 | unsigned mFlags /* Zero or more SQLITE_DESERIALIZE_* flags */ ){ MemFile *p; char *zSql; sqlite3_stmt *pStmt = 0; int rc; int iDb; #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ){ return SQLITE_MISUSE_BKPT; } if( szDb<0 ) return SQLITE_MISUSE_BKPT; if( szBuf<0 ) return SQLITE_MISUSE_BKPT; #endif sqlite3_mutex_enter(db->mutex); if( zSchema==0 ) zSchema = db->aDb[0].zDbSName; iDb = sqlite3FindDbName(db, zSchema); if( iDb<0 ){ rc = SQLITE_ERROR; goto end_deserialize; |
︙ | ︙ | |||
46415 46416 46417 46418 46419 46420 46421 | } if( pPg ){ int rc; #ifdef SQLITE_LOG_CACHE_SPILL sqlite3_log(SQLITE_FULL, "spill page %d making room for %d - cache used: %d/%d", pPg->pgno, pgno, | | | 46729 46730 46731 46732 46733 46734 46735 46736 46737 46738 46739 46740 46741 46742 46743 | } if( pPg ){ int rc; #ifdef SQLITE_LOG_CACHE_SPILL sqlite3_log(SQLITE_FULL, "spill page %d making room for %d - cache used: %d/%d", pPg->pgno, pgno, sqlite3GlobalConfig.pcache2.xPagecount(pCache->pCache), numberOfCachePages(pCache)); #endif pcacheTrace(("%p.SPILL %d\n",pCache,pPg->pgno)); rc = pCache->xStress(pCache->pStress, pPg); pcacheDump(pCache); if( rc!=SQLITE_OK && rc!=SQLITE_BUSY ){ return rc; |
︙ | ︙ | |||
49471 49472 49473 49474 49475 49476 49477 | int pageSize; /* Number of bytes in a page */ Pgno mxPgno; /* Maximum allowed size of the database */ i64 journalSizeLimit; /* Size limit for persistent journal files */ char *zFilename; /* Name of the database file */ char *zJournal; /* Name of the journal file */ int (*xBusyHandler)(void*); /* Function to call when busy */ void *pBusyHandlerArg; /* Context argument for xBusyHandler */ | | | 49785 49786 49787 49788 49789 49790 49791 49792 49793 49794 49795 49796 49797 49798 49799 | int pageSize; /* Number of bytes in a page */ Pgno mxPgno; /* Maximum allowed size of the database */ i64 journalSizeLimit; /* Size limit for persistent journal files */ char *zFilename; /* Name of the database file */ char *zJournal; /* Name of the journal file */ int (*xBusyHandler)(void*); /* Function to call when busy */ void *pBusyHandlerArg; /* Context argument for xBusyHandler */ int aStat[4]; /* Total cache hits, misses, writes, spills */ #ifdef SQLITE_TEST int nRead; /* Database pages read */ #endif void (*xReiniter)(DbPage*); /* Call this routine when reloading pages */ int (*xGet)(Pager*,Pgno,DbPage**,int); /* Routine to fetch a patch */ #ifdef SQLITE_HAS_CODEC void *(*xCodec)(void*,void*,Pgno,int); /* Routine for en/decoding data */ |
︙ | ︙ | |||
49499 49500 49501 49502 49503 49504 49505 49506 49507 49508 49509 49510 49511 49512 | ** Indexes for use with Pager.aStat[]. The Pager.aStat[] array contains ** the values accessed by passing SQLITE_DBSTATUS_CACHE_HIT, CACHE_MISS ** or CACHE_WRITE to sqlite3_db_status(). */ #define PAGER_STAT_HIT 0 #define PAGER_STAT_MISS 1 #define PAGER_STAT_WRITE 2 /* ** The following global variables hold counters used for ** testing purposes only. These variables do not exist in ** a non-testing build. These variables are not thread-safe. */ #ifdef SQLITE_TEST | > | 49813 49814 49815 49816 49817 49818 49819 49820 49821 49822 49823 49824 49825 49826 49827 | ** Indexes for use with Pager.aStat[]. The Pager.aStat[] array contains ** the values accessed by passing SQLITE_DBSTATUS_CACHE_HIT, CACHE_MISS ** or CACHE_WRITE to sqlite3_db_status(). */ #define PAGER_STAT_HIT 0 #define PAGER_STAT_MISS 1 #define PAGER_STAT_WRITE 2 #define PAGER_STAT_SPILL 3 /* ** The following global variables hold counters used for ** testing purposes only. These variables do not exist in ** a non-testing build. These variables are not thread-safe. */ #ifdef SQLITE_TEST |
︙ | ︙ | |||
50901 50902 50903 50904 50905 50906 50907 | ** successfully committed, but the EXCLUSIVE lock is still held on the ** file. So it is safe to truncate the database file to its minimum ** required size. */ assert( pPager->eLock==EXCLUSIVE_LOCK ); rc = pager_truncate(pPager, pPager->dbSize); } | | | 51216 51217 51218 51219 51220 51221 51222 51223 51224 51225 51226 51227 51228 51229 51230 | ** successfully committed, but the EXCLUSIVE lock is still held on the ** file. So it is safe to truncate the database file to its minimum ** required size. */ assert( pPager->eLock==EXCLUSIVE_LOCK ); rc = pager_truncate(pPager, pPager->dbSize); } if( rc==SQLITE_OK && bCommit ){ rc = sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_COMMIT_PHASETWO, 0); if( rc==SQLITE_NOTFOUND ) rc = SQLITE_OK; } if( !pPager->exclusiveMode && (!pagerUseWal(pPager) || sqlite3WalExclusiveMode(pPager->pWal, 0)) ){ |
︙ | ︙ | |||
51720 51721 51722 51723 51724 51725 51726 | } /* Following a rollback, the database file should be back in its original ** state prior to the start of the transaction, so invoke the ** SQLITE_FCNTL_DB_UNCHANGED file-control method to disable the ** assertion that the transaction counter was modified. */ #ifdef SQLITE_DEBUG | < < | 52035 52036 52037 52038 52039 52040 52041 52042 52043 52044 52045 52046 52047 52048 52049 | } /* Following a rollback, the database file should be back in its original ** state prior to the start of the transaction, so invoke the ** SQLITE_FCNTL_DB_UNCHANGED file-control method to disable the ** assertion that the transaction counter was modified. */ #ifdef SQLITE_DEBUG sqlite3OsFileControlHint(pPager->fd,SQLITE_FCNTL_DB_UNCHANGED,0); #endif /* If this playback is happening automatically as a result of an IO or ** malloc error that occurred after the change-counter was updated but ** before the transaction was committed, then the change-counter ** modification may just have been reverted. If this happens in exclusive ** mode, then subsequent transactions performed by the connection will not |
︙ | ︙ | |||
52475 52476 52477 52478 52479 52480 52481 | ** SHARED_LOCK -> EXCLUSIVE_LOCK | No ** RESERVED_LOCK -> EXCLUSIVE_LOCK | Yes ** ** If the busy-handler callback returns non-zero, the lock is ** retried. If it returns zero, then the SQLITE_BUSY error is ** returned to the caller of the pager API function. */ | | > < < | < | 52788 52789 52790 52791 52792 52793 52794 52795 52796 52797 52798 52799 52800 52801 52802 52803 52804 52805 52806 52807 52808 52809 52810 52811 52812 52813 52814 | ** SHARED_LOCK -> EXCLUSIVE_LOCK | No ** RESERVED_LOCK -> EXCLUSIVE_LOCK | Yes ** ** If the busy-handler callback returns non-zero, the lock is ** retried. If it returns zero, then the SQLITE_BUSY error is ** returned to the caller of the pager API function. */ SQLITE_PRIVATE void sqlite3PagerSetBusyHandler( Pager *pPager, /* Pager object */ int (*xBusyHandler)(void *), /* Pointer to busy-handler function */ void *pBusyHandlerArg /* Argument to pass to xBusyHandler */ ){ void **ap; pPager->xBusyHandler = xBusyHandler; pPager->pBusyHandlerArg = pBusyHandlerArg; ap = (void **)&pPager->xBusyHandler; assert( ((int(*)(void *))(ap[0]))==xBusyHandler ); assert( ap[1]==pBusyHandlerArg ); sqlite3OsFileControlHint(pPager->fd, SQLITE_FCNTL_BUSYHANDLER, (void *)ap); } /* ** Change the page size used by the Pager object. The new page size ** is passed in *pPageSize. ** ** If the pager is in the error state when this function is called, it ** is a no-op. The value returned is the error state error code (i.e. |
︙ | ︙ | |||
53390 53391 53392 53393 53394 53395 53396 53397 53398 53399 53400 53401 53402 53403 | if( pPager->doNotSpill && ((pPager->doNotSpill & (SPILLFLAG_ROLLBACK|SPILLFLAG_OFF))!=0 || (pPg->flags & PGHDR_NEED_SYNC)!=0) ){ return SQLITE_OK; } pPg->pDirty = 0; if( pagerUseWal(pPager) ){ /* Write a single frame for this page to the log. */ rc = subjournalPageIfRequired(pPg); if( rc==SQLITE_OK ){ rc = pagerWalFrames(pPager, pPg, 0, 0); } | > | 53701 53702 53703 53704 53705 53706 53707 53708 53709 53710 53711 53712 53713 53714 53715 | if( pPager->doNotSpill && ((pPager->doNotSpill & (SPILLFLAG_ROLLBACK|SPILLFLAG_OFF))!=0 || (pPg->flags & PGHDR_NEED_SYNC)!=0) ){ return SQLITE_OK; } pPager->aStat[PAGER_STAT_SPILL]++; pPg->pDirty = 0; if( pagerUseWal(pPager) ){ /* Write a single frame for this page to the log. */ rc = subjournalPageIfRequired(pPg); if( rc==SQLITE_OK ){ rc = pagerWalFrames(pPager, pPg, 0, 0); } |
︙ | ︙ | |||
54463 54464 54465 54466 54467 54468 54469 54470 54471 54472 54473 54474 54475 54476 | } SQLITE_PRIVATE void sqlite3PagerUnrefPageOne(DbPage *pPg){ Pager *pPager; assert( pPg!=0 ); assert( pPg->pgno==1 ); assert( (pPg->flags & PGHDR_MMAP)==0 ); /* Page1 is never memory mapped */ pPager = pPg->pPager; sqlite3PcacheRelease(pPg); pagerUnlockIfUnused(pPager); } /* ** This function is called at the start of every write transaction. ** There must already be a RESERVED or EXCLUSIVE lock on the database | > | 54775 54776 54777 54778 54779 54780 54781 54782 54783 54784 54785 54786 54787 54788 54789 | } SQLITE_PRIVATE void sqlite3PagerUnrefPageOne(DbPage *pPg){ Pager *pPager; assert( pPg!=0 ); assert( pPg->pgno==1 ); assert( (pPg->flags & PGHDR_MMAP)==0 ); /* Page1 is never memory mapped */ pPager = pPg->pPager; sqlite3PagerResetLockTimeout(pPager); sqlite3PcacheRelease(pPg); pagerUnlockIfUnused(pPager); } /* ** This function is called at the start of every write transaction. ** There must already be a RESERVED or EXCLUSIVE lock on the database |
︙ | ︙ | |||
55058 55059 55060 55061 55062 55063 55064 | ** or pages with the Pager.noSync flag set. ** ** If successful, or if called on a pager for which it is a no-op, this ** function returns SQLITE_OK. Otherwise, an IO error code is returned. */ SQLITE_PRIVATE int sqlite3PagerSync(Pager *pPager, const char *zMaster){ int rc = SQLITE_OK; | < < < | 55371 55372 55373 55374 55375 55376 55377 55378 55379 55380 55381 55382 55383 55384 55385 55386 55387 | ** or pages with the Pager.noSync flag set. ** ** If successful, or if called on a pager for which it is a no-op, this ** function returns SQLITE_OK. Otherwise, an IO error code is returned. */ SQLITE_PRIVATE int sqlite3PagerSync(Pager *pPager, const char *zMaster){ int rc = SQLITE_OK; void *pArg = (void*)zMaster; rc = sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_SYNC, pArg); if( rc==SQLITE_NOTFOUND ) rc = SQLITE_OK; if( rc==SQLITE_OK && !pPager->noSync ){ assert( !MEMDB ); rc = sqlite3OsSync(pPager->fd, pPager->syncFlags); } return rc; } |
︙ | ︙ | |||
55510 55511 55512 55513 55514 55515 55516 | a[9] = pPager->nRead; a[10] = pPager->aStat[PAGER_STAT_WRITE]; return a; } #endif /* | | > > > > | > | > > | | | 55820 55821 55822 55823 55824 55825 55826 55827 55828 55829 55830 55831 55832 55833 55834 55835 55836 55837 55838 55839 55840 55841 55842 55843 55844 55845 55846 55847 55848 55849 55850 55851 55852 55853 55854 55855 55856 55857 55858 55859 55860 | a[9] = pPager->nRead; a[10] = pPager->aStat[PAGER_STAT_WRITE]; return a; } #endif /* ** Parameter eStat must be one of SQLITE_DBSTATUS_CACHE_HIT, _MISS, _WRITE, ** or _WRITE+1. The SQLITE_DBSTATUS_CACHE_WRITE+1 case is a translation ** of SQLITE_DBSTATUS_CACHE_SPILL. The _SPILL case is not contiguous because ** it was added later. ** ** Before returning, *pnVal is incremented by the ** current cache hit or miss count, according to the value of eStat. If the ** reset parameter is non-zero, the cache hit or miss count is zeroed before ** returning. */ SQLITE_PRIVATE void sqlite3PagerCacheStat(Pager *pPager, int eStat, int reset, int *pnVal){ assert( eStat==SQLITE_DBSTATUS_CACHE_HIT || eStat==SQLITE_DBSTATUS_CACHE_MISS || eStat==SQLITE_DBSTATUS_CACHE_WRITE || eStat==SQLITE_DBSTATUS_CACHE_WRITE+1 ); assert( SQLITE_DBSTATUS_CACHE_HIT+1==SQLITE_DBSTATUS_CACHE_MISS ); assert( SQLITE_DBSTATUS_CACHE_HIT+2==SQLITE_DBSTATUS_CACHE_WRITE ); assert( PAGER_STAT_HIT==0 && PAGER_STAT_MISS==1 && PAGER_STAT_WRITE==2 && PAGER_STAT_SPILL==3 ); eStat -= SQLITE_DBSTATUS_CACHE_HIT; *pnVal += pPager->aStat[eStat]; if( reset ){ pPager->aStat[eStat] = 0; } } /* ** Return true if this is an in-memory or temp-file backed pager. */ SQLITE_PRIVATE int sqlite3PagerIsMemdb(Pager *pPager){ |
︙ | ︙ | |||
55732 55733 55734 55735 55736 55737 55738 55739 55740 55741 55742 55743 55744 55745 | ** Return the file handle for the database file associated ** with the pager. This might return NULL if the file has ** not yet been opened. */ SQLITE_PRIVATE sqlite3_file *sqlite3PagerFile(Pager *pPager){ return pPager->fd; } /* ** Return the file handle for the journal file (if it exists). ** This will be either the rollback journal or the WAL file. */ SQLITE_PRIVATE sqlite3_file *sqlite3PagerJrnlFile(Pager *pPager){ #if SQLITE_OMIT_WAL | > > > > > > > > > > | 56049 56050 56051 56052 56053 56054 56055 56056 56057 56058 56059 56060 56061 56062 56063 56064 56065 56066 56067 56068 56069 56070 56071 56072 | ** Return the file handle for the database file associated ** with the pager. This might return NULL if the file has ** not yet been opened. */ SQLITE_PRIVATE sqlite3_file *sqlite3PagerFile(Pager *pPager){ return pPager->fd; } #ifdef SQLITE_ENABLE_SETLK_TIMEOUT /* ** Reset the lock timeout for pager. */ SQLITE_PRIVATE void sqlite3PagerResetLockTimeout(Pager *pPager){ int x = 0; sqlite3OsFileControl(pPager->fd, SQLITE_FCNTL_LOCK_TIMEOUT, &x); } #endif /* ** Return the file handle for the journal file (if it exists). ** This will be either the rollback journal or the WAL file. */ SQLITE_PRIVATE sqlite3_file *sqlite3PagerJrnlFile(Pager *pPager){ #if SQLITE_OMIT_WAL |
︙ | ︙ | |||
56193 56194 56195 56196 56197 56198 56199 56200 56201 56202 56203 56204 56205 56206 | if( pPager->pWal ){ rc = sqlite3WalCheckpoint(pPager->pWal, db, eMode, (eMode==SQLITE_CHECKPOINT_PASSIVE ? 0 : pPager->xBusyHandler), pPager->pBusyHandlerArg, pPager->walSyncFlags, pPager->pageSize, (u8 *)pPager->pTmpSpace, pnLog, pnCkpt ); } return rc; } SQLITE_PRIVATE int sqlite3PagerWalCallback(Pager *pPager){ return sqlite3WalCallback(pPager->pWal); } | > | 56520 56521 56522 56523 56524 56525 56526 56527 56528 56529 56530 56531 56532 56533 56534 | if( pPager->pWal ){ rc = sqlite3WalCheckpoint(pPager->pWal, db, eMode, (eMode==SQLITE_CHECKPOINT_PASSIVE ? 0 : pPager->xBusyHandler), pPager->pBusyHandlerArg, pPager->walSyncFlags, pPager->pageSize, (u8 *)pPager->pTmpSpace, pnLog, pnCkpt ); sqlite3PagerResetLockTimeout(pPager); } return rc; } SQLITE_PRIVATE int sqlite3PagerWalCallback(Pager *pPager){ return sqlite3WalCallback(pPager->pWal); } |
︙ | ︙ | |||
63469 63470 63471 63472 63473 63474 63475 | /* ** Invoke the busy handler for a btree. */ static int btreeInvokeBusyHandler(void *pArg){ BtShared *pBt = (BtShared*)pArg; assert( pBt->db ); assert( sqlite3_mutex_held(pBt->db->mutex) ); | | > | 63797 63798 63799 63800 63801 63802 63803 63804 63805 63806 63807 63808 63809 63810 63811 63812 | /* ** Invoke the busy handler for a btree. */ static int btreeInvokeBusyHandler(void *pArg){ BtShared *pBt = (BtShared*)pArg; assert( pBt->db ); assert( sqlite3_mutex_held(pBt->db->mutex) ); return sqlite3InvokeBusyHandler(&pBt->db->busyHandler, sqlite3PagerFile(pBt->pPager)); } /* ** Open a database file. ** ** zFilename is the name of the database file. If zFilename is NULL ** then an ephemeral database is created. The ephemeral database might |
︙ | ︙ | |||
63647 63648 63649 63650 63651 63652 63653 | rc = sqlite3PagerReadFileheader(pBt->pPager,sizeof(zDbHeader),zDbHeader); } if( rc!=SQLITE_OK ){ goto btree_open_out; } pBt->openFlags = (u8)flags; pBt->db = db; | | | 63976 63977 63978 63979 63980 63981 63982 63983 63984 63985 63986 63987 63988 63989 63990 | rc = sqlite3PagerReadFileheader(pBt->pPager,sizeof(zDbHeader),zDbHeader); } if( rc!=SQLITE_OK ){ goto btree_open_out; } pBt->openFlags = (u8)flags; pBt->db = db; sqlite3PagerSetBusyHandler(pBt->pPager, btreeInvokeBusyHandler, pBt); p->pBt = pBt; pBt->pCursor = 0; pBt->pPage1 = 0; if( sqlite3PagerIsreadonly(pBt->pPager) ) pBt->btsFlags |= BTS_READ_ONLY; #if defined(SQLITE_SECURE_DELETE) pBt->btsFlags |= BTS_SECURE_DELETE; |
︙ | ︙ | |||
64610 64611 64612 64613 64614 64615 64616 64617 64618 64619 64620 64621 64622 64623 | } if( rc!=SQLITE_OK ){ unlockBtreeIfUnused(pBt); } }while( (rc&0xFF)==SQLITE_BUSY && pBt->inTransaction==TRANS_NONE && btreeInvokeBusyHandler(pBt) ); if( rc==SQLITE_OK ){ if( p->inTrans==TRANS_NONE ){ pBt->nTransaction++; #ifndef SQLITE_OMIT_SHARED_CACHE if( p->sharable ){ assert( p->lock.pBtree==p && p->lock.iTable==1 ); | > | 64939 64940 64941 64942 64943 64944 64945 64946 64947 64948 64949 64950 64951 64952 64953 | } if( rc!=SQLITE_OK ){ unlockBtreeIfUnused(pBt); } }while( (rc&0xFF)==SQLITE_BUSY && pBt->inTransaction==TRANS_NONE && btreeInvokeBusyHandler(pBt) ); sqlite3PagerResetLockTimeout(pBt->pPager); if( rc==SQLITE_OK ){ if( p->inTrans==TRANS_NONE ){ pBt->nTransaction++; #ifndef SQLITE_OMIT_SHARED_CACHE if( p->sharable ){ assert( p->lock.pBtree==p && p->lock.iTable==1 ); |
︙ | ︙ | |||
80285 80286 80287 80288 80289 80290 80291 | /* ** Return the value of a status counter for a prepared statement */ SQLITE_API int sqlite3_stmt_status(sqlite3_stmt *pStmt, int op, int resetFlag){ Vdbe *pVdbe = (Vdbe*)pStmt; u32 v; #ifdef SQLITE_ENABLE_API_ARMOR | | > > | 80615 80616 80617 80618 80619 80620 80621 80622 80623 80624 80625 80626 80627 80628 80629 80630 80631 | /* ** Return the value of a status counter for a prepared statement */ SQLITE_API int sqlite3_stmt_status(sqlite3_stmt *pStmt, int op, int resetFlag){ Vdbe *pVdbe = (Vdbe*)pStmt; u32 v; #ifdef SQLITE_ENABLE_API_ARMOR if( !pStmt || (op!=SQLITE_STMTSTATUS_MEMUSED && (op<0||op>=ArraySize(pVdbe->aCounter))) ){ (void)SQLITE_MISUSE_BKPT; return 0; } #endif if( op==SQLITE_STMTSTATUS_MEMUSED ){ sqlite3 *db = pVdbe->db; sqlite3_mutex_enter(db->mutex); |
︙ | ︙ | |||
85171 85172 85173 85174 85175 85176 85177 85178 85179 85180 85181 85182 85183 85184 | VdbeFrame *pFrame; /* Root frame of VDBE */ v = 0; res = 0; pOut = out2Prerelease(p, pOp); assert( pOp->p1>=0 && pOp->p1<p->nCursor ); pC = p->apCsr[pOp->p1]; assert( pC!=0 ); assert( pC->eCurType==CURTYPE_BTREE ); assert( pC->uc.pCursor!=0 ); { /* The next rowid or record number (different terms for the same ** thing) is obtained in a two-step algorithm. ** | > > > > | 85503 85504 85505 85506 85507 85508 85509 85510 85511 85512 85513 85514 85515 85516 85517 85518 85519 85520 | VdbeFrame *pFrame; /* Root frame of VDBE */ v = 0; res = 0; pOut = out2Prerelease(p, pOp); assert( pOp->p1>=0 && pOp->p1<p->nCursor ); pC = p->apCsr[pOp->p1]; if( !pC->isTable ){ rc = SQLITE_CORRUPT_BKPT; goto abort_due_to_error; } assert( pC!=0 ); assert( pC->eCurType==CURTYPE_BTREE ); assert( pC->uc.pCursor!=0 ); { /* The next rowid or record number (different terms for the same ** thing) is obtained in a two-step algorithm. ** |
︙ | ︙ | |||
98642 98643 98644 98645 98646 98647 98648 98649 98650 98651 98652 98653 98654 98655 | if( pE2->op==TK_NOTNULL && pE1->op!=TK_ISNULL && pE1->op!=TK_IS ){ Expr *pX = sqlite3ExprSkipCollate(pE1->pLeft); testcase( pX!=pE1->pLeft ); if( sqlite3ExprCompare(pParse, pX, pE2->pLeft, iTab)==0 ) return 1; } return 0; } /* ** An instance of the following structure is used by the tree walker ** to determine if an expression can be evaluated by reference to the ** index only, without having to do a search for the corresponding ** table entry. The IdxCover.pIdx field is the index. IdxCover.iCur ** is the cursor for the table. | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 98978 98979 98980 98981 98982 98983 98984 98985 98986 98987 98988 98989 98990 98991 98992 98993 98994 98995 98996 98997 98998 98999 99000 99001 99002 99003 99004 99005 99006 99007 99008 99009 99010 99011 99012 99013 99014 99015 99016 99017 99018 99019 99020 99021 99022 99023 99024 99025 99026 99027 99028 99029 99030 99031 99032 99033 99034 99035 99036 99037 99038 99039 99040 99041 99042 99043 99044 99045 99046 99047 99048 99049 99050 99051 99052 99053 99054 99055 99056 99057 99058 99059 99060 99061 99062 99063 99064 99065 | if( pE2->op==TK_NOTNULL && pE1->op!=TK_ISNULL && pE1->op!=TK_IS ){ Expr *pX = sqlite3ExprSkipCollate(pE1->pLeft); testcase( pX!=pE1->pLeft ); if( sqlite3ExprCompare(pParse, pX, pE2->pLeft, iTab)==0 ) return 1; } return 0; } /* ** This is the Expr node callback for sqlite3ExprImpliesNotNullRow(). ** If the expression node requires that the table at pWalker->iCur ** have a non-NULL column, then set pWalker->eCode to 1 and abort. */ static int impliesNotNullRow(Walker *pWalker, Expr *pExpr){ /* This routine is only called for WHERE clause expressions and so it ** cannot have any TK_AGG_COLUMN entries because those are only found ** in HAVING clauses. We can get a TK_AGG_FUNCTION in a WHERE clause, ** but that is an illegal construct and the query will be rejected at ** a later stage of processing, so the TK_AGG_FUNCTION case does not ** need to be considered here. */ assert( pExpr->op!=TK_AGG_COLUMN ); testcase( pExpr->op==TK_AGG_FUNCTION ); if( ExprHasProperty(pExpr, EP_FromJoin) ) return WRC_Prune; switch( pExpr->op ){ case TK_ISNULL: case TK_IS: case TK_OR: case TK_CASE: case TK_IN: case TK_FUNCTION: testcase( pExpr->op==TK_ISNULL ); testcase( pExpr->op==TK_IS ); testcase( pExpr->op==TK_OR ); testcase( pExpr->op==TK_CASE ); testcase( pExpr->op==TK_IN ); testcase( pExpr->op==TK_FUNCTION ); return WRC_Prune; case TK_COLUMN: if( pWalker->u.iCur==pExpr->iTable ){ pWalker->eCode = 1; return WRC_Abort; } return WRC_Prune; default: return WRC_Continue; } } /* ** Return true (non-zero) if expression p can only be true if at least ** one column of table iTab is non-null. In other words, return true ** if expression p will always be NULL or false if every column of iTab ** is NULL. ** ** False negatives are acceptable. In other words, it is ok to return ** zero even if expression p will never be true of every column of iTab ** is NULL. A false negative is merely a missed optimization opportunity. ** ** False positives are not allowed, however. A false positive may result ** in an incorrect answer. ** ** Terms of p that are marked with EP_FromJoin (and hence that come from ** the ON or USING clauses of LEFT JOINS) are excluded from the analysis. ** ** This routine is used to check if a LEFT JOIN can be converted into ** an ordinary JOIN. The p argument is the WHERE clause. If the WHERE ** clause requires that some column of the right table of the LEFT JOIN ** be non-NULL, then the LEFT JOIN can be safely converted into an ** ordinary join. */ SQLITE_PRIVATE int sqlite3ExprImpliesNonNullRow(Expr *p, int iTab){ Walker w; w.xExprCallback = impliesNotNullRow; w.xSelectCallback = 0; w.xSelectCallback2 = 0; w.eCode = 0; w.u.iCur = iTab; sqlite3WalkExpr(&w, p); return w.eCode; } /* ** An instance of the following structure is used by the tree walker ** to determine if an expression can be evaluated by reference to the ** index only, without having to do a search for the corresponding ** table entry. The IdxCover.pIdx field is the index. IdxCover.iCur ** is the cursor for the table. |
︙ | ︙ | |||
100887 100888 100889 100890 100891 100892 100893 | if( v==0 || NEVER(pTab==0) ){ return; } if( pTab->tnum==0 ){ /* Do not gather statistics on views or virtual tables */ return; } | | | 101297 101298 101299 101300 101301 101302 101303 101304 101305 101306 101307 101308 101309 101310 101311 | if( v==0 || NEVER(pTab==0) ){ return; } if( pTab->tnum==0 ){ /* Do not gather statistics on views or virtual tables */ return; } if( sqlite3_strlike("sqlite\\_%", pTab->zName, '\\')==0 ){ /* Do not gather statistics on system tables */ return; } assert( sqlite3BtreeHoldsAllMutexes(db) ); iDb = sqlite3SchemaToIndex(db, pTab->pSchema); assert( iDb>=0 ); assert( sqlite3SchemaMutexHeld(db, iDb, 0) ); |
︙ | ︙ | |||
104566 104567 104568 104569 104570 104571 104572 | if( pEnd==0 && pSelect==0 ){ return; } assert( !db->mallocFailed ); p = pParse->pNewTable; if( p==0 ) return; | < < > > > > | 104976 104977 104978 104979 104980 104981 104982 104983 104984 104985 104986 104987 104988 104989 104990 104991 104992 104993 104994 104995 104996 104997 104998 104999 105000 105001 105002 105003 | if( pEnd==0 && pSelect==0 ){ return; } assert( !db->mallocFailed ); p = pParse->pNewTable; if( p==0 ) return; /* If the db->init.busy is 1 it means we are reading the SQL off the ** "sqlite_master" or "sqlite_temp_master" table on the disk. ** So do not write to the disk again. Extract the root page number ** for the table from the db->init.newTnum field. (The page number ** should have been put there by the sqliteOpenCb routine.) ** ** If the root page number is 1, that means this is the sqlite_master ** table itself. So mark it read-only. */ if( db->init.busy ){ if( pSelect ){ sqlite3ErrorMsg(pParse, ""); return; } p->tnum = db->init.newTnum; if( p->tnum==1 ) p->tabFlags |= TF_Readonly; } /* Special processing for WITHOUT ROWID Tables */ if( tabOpts & TF_WithoutRowid ){ if( (p->tabFlags & TF_Autoincrement) ){ |
︙ | ︙ | |||
112154 112155 112156 112157 112158 112159 112160 | ** ** There is at most one AutoincInfo structure per table even if the ** same table is autoincremented multiple times due to inserts within ** triggers. A new AutoincInfo structure is created if this is the ** first use of table pTab. On 2nd and subsequent uses, the original ** AutoincInfo structure is used. ** | | | | | > | 112566 112567 112568 112569 112570 112571 112572 112573 112574 112575 112576 112577 112578 112579 112580 112581 112582 112583 112584 112585 | ** ** There is at most one AutoincInfo structure per table even if the ** same table is autoincremented multiple times due to inserts within ** triggers. A new AutoincInfo structure is created if this is the ** first use of table pTab. On 2nd and subsequent uses, the original ** AutoincInfo structure is used. ** ** Four consecutive registers are allocated: ** ** (1) The name of the pTab table. ** (2) The maximum ROWID of pTab. ** (3) The rowid in sqlite_sequence of pTab ** (4) The original value of the max ROWID in pTab, or NULL if none ** ** The 2nd register is the one that is returned. That is all the ** insert routine needs to know about. */ static int autoIncBegin( Parse *pParse, /* Parsing context */ int iDb, /* Index of the database holding pTab */ |
︙ | ︙ | |||
112186 112187 112188 112189 112190 112191 112192 | if( pInfo==0 ) return 0; pInfo->pNext = pToplevel->pAinc; pToplevel->pAinc = pInfo; pInfo->pTab = pTab; pInfo->iDb = iDb; pToplevel->nMem++; /* Register to hold name of table */ pInfo->regCtr = ++pToplevel->nMem; /* Max rowid register */ | | | 112599 112600 112601 112602 112603 112604 112605 112606 112607 112608 112609 112610 112611 112612 112613 | if( pInfo==0 ) return 0; pInfo->pNext = pToplevel->pAinc; pToplevel->pAinc = pInfo; pInfo->pTab = pTab; pInfo->iDb = iDb; pToplevel->nMem++; /* Register to hold name of table */ pInfo->regCtr = ++pToplevel->nMem; /* Max rowid register */ pToplevel->nMem +=2; /* Rowid in sqlite_sequence + orig max val */ } memId = pInfo->regCtr; } return memId; } /* |
︙ | ︙ | |||
112214 112215 112216 112217 112218 112219 112220 | assert( sqlite3IsToplevel(pParse) ); assert( v ); /* We failed long ago if this is not so */ for(p = pParse->pAinc; p; p = p->pNext){ static const int iLn = VDBE_OFFSET_LINENO(2); static const VdbeOpList autoInc[] = { /* 0 */ {OP_Null, 0, 0, 0}, | | | > > | | | | | > > > | | 112627 112628 112629 112630 112631 112632 112633 112634 112635 112636 112637 112638 112639 112640 112641 112642 112643 112644 112645 112646 112647 112648 112649 112650 112651 112652 112653 112654 112655 112656 112657 112658 112659 112660 112661 112662 112663 112664 112665 112666 112667 112668 112669 112670 112671 112672 | assert( sqlite3IsToplevel(pParse) ); assert( v ); /* We failed long ago if this is not so */ for(p = pParse->pAinc; p; p = p->pNext){ static const int iLn = VDBE_OFFSET_LINENO(2); static const VdbeOpList autoInc[] = { /* 0 */ {OP_Null, 0, 0, 0}, /* 1 */ {OP_Rewind, 0, 10, 0}, /* 2 */ {OP_Column, 0, 0, 0}, /* 3 */ {OP_Ne, 0, 9, 0}, /* 4 */ {OP_Rowid, 0, 0, 0}, /* 5 */ {OP_Column, 0, 1, 0}, /* 6 */ {OP_AddImm, 0, 0, 0}, /* 7 */ {OP_Copy, 0, 0, 0}, /* 8 */ {OP_Goto, 0, 11, 0}, /* 9 */ {OP_Next, 0, 2, 0}, /* 10 */ {OP_Integer, 0, 0, 0}, /* 11 */ {OP_Close, 0, 0, 0} }; VdbeOp *aOp; pDb = &db->aDb[p->iDb]; memId = p->regCtr; assert( sqlite3SchemaMutexHeld(db, 0, pDb->pSchema) ); sqlite3OpenTable(pParse, 0, p->iDb, pDb->pSchema->pSeqTab, OP_OpenRead); sqlite3VdbeLoadString(v, memId-1, p->pTab->zName); aOp = sqlite3VdbeAddOpList(v, ArraySize(autoInc), autoInc, iLn); if( aOp==0 ) break; aOp[0].p2 = memId; aOp[0].p3 = memId+2; aOp[2].p3 = memId; aOp[3].p1 = memId-1; aOp[3].p3 = memId; aOp[3].p5 = SQLITE_JUMPIFNULL; aOp[4].p2 = memId+1; aOp[5].p3 = memId; aOp[6].p1 = memId; aOp[7].p2 = memId+2; aOp[7].p1 = memId; aOp[10].p2 = memId; } } /* ** Update the maximum rowid for an autoincrement calculation. ** ** This routine should be called when the regRowid register holds a |
︙ | ︙ | |||
112287 112288 112289 112290 112291 112292 112293 112294 112295 112296 112297 112298 112299 112300 | VdbeOp *aOp; Db *pDb = &db->aDb[p->iDb]; int iRec; int memId = p->regCtr; iRec = sqlite3GetTempReg(pParse); assert( sqlite3SchemaMutexHeld(db, 0, pDb->pSchema) ); sqlite3OpenTable(pParse, 0, p->iDb, pDb->pSchema->pSeqTab, OP_OpenWrite); aOp = sqlite3VdbeAddOpList(v, ArraySize(autoIncEnd), autoIncEnd, iLn); if( aOp==0 ) break; aOp[0].p1 = memId+1; aOp[1].p2 = memId+1; aOp[2].p1 = memId-1; aOp[2].p3 = iRec; | > > | 112705 112706 112707 112708 112709 112710 112711 112712 112713 112714 112715 112716 112717 112718 112719 112720 | VdbeOp *aOp; Db *pDb = &db->aDb[p->iDb]; int iRec; int memId = p->regCtr; iRec = sqlite3GetTempReg(pParse); assert( sqlite3SchemaMutexHeld(db, 0, pDb->pSchema) ); sqlite3VdbeAddOp3(v, OP_Le, memId+2, sqlite3VdbeCurrentAddr(v)+7, memId); VdbeCoverage(v); sqlite3OpenTable(pParse, 0, p->iDb, pDb->pSchema->pSeqTab, OP_OpenWrite); aOp = sqlite3VdbeAddOpList(v, ArraySize(autoIncEnd), autoIncEnd, iLn); if( aOp==0 ) break; aOp[0].p1 = memId+1; aOp[1].p2 = memId+1; aOp[2].p1 = memId-1; aOp[2].p3 = iRec; |
︙ | ︙ | |||
118920 118921 118922 118923 118924 118925 118926 | const char *zExtra /* Error information */ ){ sqlite3 *db = pData->db; if( !db->mallocFailed && (db->flags & SQLITE_WriteSchema)==0 ){ char *z; if( zObj==0 ) zObj = "?"; z = sqlite3MPrintf(db, "malformed database schema (%s)", zObj); | | | 119340 119341 119342 119343 119344 119345 119346 119347 119348 119349 119350 119351 119352 119353 119354 | const char *zExtra /* Error information */ ){ sqlite3 *db = pData->db; if( !db->mallocFailed && (db->flags & SQLITE_WriteSchema)==0 ){ char *z; if( zObj==0 ) zObj = "?"; z = sqlite3MPrintf(db, "malformed database schema (%s)", zObj); if( zExtra && zExtra[0] ) z = sqlite3MPrintf(db, "%z - %s", z, zExtra); sqlite3DbFree(db, *pData->pzErrMsg); *pData->pzErrMsg = z; } pData->rc = db->mallocFailed ? SQLITE_NOMEM_BKPT : SQLITE_CORRUPT_BKPT; } /* |
︙ | ︙ | |||
119814 119815 119816 119817 119818 119819 119820 | /* ** Trace output macros */ #if SELECTTRACE_ENABLED /***/ int sqlite3SelectTrace = 0; # define SELECTTRACE(K,P,S,X) \ if(sqlite3SelectTrace&(K)) \ | < | | 120234 120235 120236 120237 120238 120239 120240 120241 120242 120243 120244 120245 120246 120247 120248 | /* ** Trace output macros */ #if SELECTTRACE_ENABLED /***/ int sqlite3SelectTrace = 0; # define SELECTTRACE(K,P,S,X) \ if(sqlite3SelectTrace&(K)) \ sqlite3DebugPrintf("%s/%p: ",(S)->zSelName,(S)),\ sqlite3DebugPrintf X #else # define SELECTTRACE(K,P,S,X) #endif /* |
︙ | ︙ | |||
120175 120176 120177 120178 120179 120180 120181 120182 120183 120184 120185 120186 120187 120188 | setJoinExpr(p->x.pList->a[i].pExpr, iTable); } } setJoinExpr(p->pLeft, iTable); p = p->pRight; } } /* ** This routine processes the join information for a SELECT statement. ** ON and USING clauses are converted into extra terms of the WHERE clause. ** NATURAL joins also create extra WHERE clause terms. ** ** The terms of a FROM clause are contained in the Select.pSrc structure. | > > > > > > > > > > > > > > > > > > > > > > > | 120594 120595 120596 120597 120598 120599 120600 120601 120602 120603 120604 120605 120606 120607 120608 120609 120610 120611 120612 120613 120614 120615 120616 120617 120618 120619 120620 120621 120622 120623 120624 120625 120626 120627 120628 120629 120630 | setJoinExpr(p->x.pList->a[i].pExpr, iTable); } } setJoinExpr(p->pLeft, iTable); p = p->pRight; } } /* Undo the work of setJoinExpr(). In the expression tree p, convert every ** term that is marked with EP_FromJoin and iRightJoinTable==iTable into ** an ordinary term that omits the EP_FromJoin mark. ** ** This happens when a LEFT JOIN is simplified into an ordinary JOIN. */ static void unsetJoinExpr(Expr *p, int iTable){ while( p ){ if( ExprHasProperty(p, EP_FromJoin) && (iTable<0 || p->iRightJoinTable==iTable) ){ ExprClearProperty(p, EP_FromJoin); } if( p->op==TK_FUNCTION && p->x.pList ){ int i; for(i=0; i<p->x.pList->nExpr; i++){ unsetJoinExpr(p->x.pList->a[i].pExpr, iTable); } } unsetJoinExpr(p->pLeft, iTable); p = p->pRight; } } /* ** This routine processes the join information for a SELECT statement. ** ON and USING clauses are converted into extra terms of the WHERE clause. ** NATURAL joins also create extra WHERE clause terms. ** ** The terms of a FROM clause are contained in the Select.pSrc structure. |
︙ | ︙ | |||
123544 123545 123546 123547 123548 123549 123550 | ** (the only way this can happen is if the compound sub-query is ** currently part of pSub->pSrc). See ticket [d11a6e908f]. */ ExprList *pOrderBy = pSub->pOrderBy; for(i=0; i<pOrderBy->nExpr; i++){ pOrderBy->a[i].u.x.iOrderByCol = 0; } assert( pParent->pOrderBy==0 ); | < | 123986 123987 123988 123989 123990 123991 123992 123993 123994 123995 123996 123997 123998 123999 | ** (the only way this can happen is if the compound sub-query is ** currently part of pSub->pSrc). See ticket [d11a6e908f]. */ ExprList *pOrderBy = pSub->pOrderBy; for(i=0; i<pOrderBy->nExpr; i++){ pOrderBy->a[i].u.x.iOrderByCol = 0; } assert( pParent->pOrderBy==0 ); pParent->pOrderBy = pOrderBy; pSub->pOrderBy = 0; } pWhere = sqlite3ExprDup(db, pSub->pWhere, 0); if( isLeftJoin>0 ){ setJoinExpr(pWhere, iNewParent); } |
︙ | ︙ | |||
123628 123629 123630 123631 123632 123633 123634 | ** to suppress it. **) ** ** (2) The inner query is the recursive part of a common table expression. ** ** (3) The inner query has a LIMIT clause (since the changes to the WHERE ** close would change the meaning of the LIMIT). ** | | | | | > > > > > > > > > > | > | 124069 124070 124071 124072 124073 124074 124075 124076 124077 124078 124079 124080 124081 124082 124083 124084 124085 124086 124087 124088 124089 124090 124091 124092 124093 124094 124095 124096 124097 124098 124099 124100 124101 124102 124103 124104 124105 124106 124107 124108 | ** to suppress it. **) ** ** (2) The inner query is the recursive part of a common table expression. ** ** (3) The inner query has a LIMIT clause (since the changes to the WHERE ** close would change the meaning of the LIMIT). ** ** (4) The inner query is the right operand of a LEFT JOIN and the ** expression to be pushed down does not come from the ON clause ** on that LEFT JOIN. ** ** (5) The WHERE clause expression originates in the ON or USING clause ** of a LEFT JOIN where iCursor is not the right-hand table of that ** left join. An example: ** ** SELECT * ** FROM (SELECT 1 AS a1 UNION ALL SELECT 2) AS aa ** JOIN (SELECT 1 AS b2 UNION ALL SELECT 2) AS bb ON (a1=b2) ** LEFT JOIN (SELECT 8 AS c3 UNION ALL SELECT 9) AS cc ON (b2=2); ** ** The correct answer is three rows: (1,1,NULL),(2,2,8),(2,2,9). ** But if the (b2=2) term were to be pushed down into the bb subquery, ** then the (1,1,NULL) row would be suppressed. ** ** Return 0 if no changes are made and non-zero if one or more WHERE clause ** terms are duplicated into the subquery. */ static int pushDownWhereTerms( Parse *pParse, /* Parse context (for malloc() and error reporting) */ Select *pSubq, /* The subquery whose WHERE clause is to be augmented */ Expr *pWhere, /* The WHERE clause of the outer query */ int iCursor, /* Cursor number of the subquery */ int isLeftJoin /* True if pSubq is the right term of a LEFT JOIN */ ){ Expr *pNew; int nChng = 0; if( pWhere==0 ) return 0; if( pSubq->selFlags & SF_Recursive ) return 0; /* restriction (2) */ #ifdef SQLITE_DEBUG |
︙ | ︙ | |||
123666 123667 123668 123669 123670 123671 123672 | } #endif if( pSubq->pLimit!=0 ){ return 0; /* restriction (3) */ } while( pWhere->op==TK_AND ){ | | > > > > > > > | > > > | 124118 124119 124120 124121 124122 124123 124124 124125 124126 124127 124128 124129 124130 124131 124132 124133 124134 124135 124136 124137 124138 124139 124140 124141 124142 124143 124144 124145 124146 124147 124148 124149 124150 | } #endif if( pSubq->pLimit!=0 ){ return 0; /* restriction (3) */ } while( pWhere->op==TK_AND ){ nChng += pushDownWhereTerms(pParse, pSubq, pWhere->pRight, iCursor, isLeftJoin); pWhere = pWhere->pLeft; } if( isLeftJoin && (ExprHasProperty(pWhere,EP_FromJoin)==0 || pWhere->iRightJoinTable!=iCursor) ){ return 0; /* restriction (4) */ } if( ExprHasProperty(pWhere,EP_FromJoin) && pWhere->iRightJoinTable!=iCursor ){ return 0; /* restriction (5) */ } if( sqlite3ExprIsTableConstant(pWhere, iCursor) ){ nChng++; while( pSubq ){ SubstContext x; pNew = sqlite3ExprDup(pParse->db, pWhere, 0); unsetJoinExpr(pNew, -1); x.pParse = pParse; x.iTable = iCursor; x.iNewTable = iCursor; x.isLeftJoin = 0; x.pEList = pSubq->pEList; pNew = substExpr(&x, pNew); if( pSubq->selFlags & SF_Aggregate ){ |
︙ | ︙ | |||
124706 124707 124708 124709 124710 124711 124712 | ); } } #else # define explainSimpleCount(a,b,c) #endif | < < < < < < < < | | | | > | 125168 125169 125170 125171 125172 125173 125174 125175 125176 125177 125178 125179 125180 125181 125182 125183 125184 125185 125186 125187 125188 125189 125190 125191 125192 125193 125194 125195 125196 125197 125198 125199 125200 125201 125202 125203 125204 | ); } } #else # define explainSimpleCount(a,b,c) #endif /* ** sqlite3WalkExpr() callback used by havingToWhere(). ** ** If the node passed to the callback is a TK_AND node, return ** WRC_Continue to tell sqlite3WalkExpr() to iterate through child nodes. ** ** Otherwise, return WRC_Prune. In this case, also check if the ** sub-expression matches the criteria for being moved to the WHERE ** clause. If so, add it to the WHERE clause and replace the sub-expression ** within the HAVING expression with a constant "1". */ static int havingToWhereExprCb(Walker *pWalker, Expr *pExpr){ if( pExpr->op!=TK_AND ){ Select *pS = pWalker->u.pSelect; if( sqlite3ExprIsConstantOrGroupBy(pWalker->pParse, pExpr, pS->pGroupBy) ){ sqlite3 *db = pWalker->pParse->db; Expr *pNew = sqlite3ExprAlloc(db, TK_INTEGER, &sqlite3IntTokens[1], 0); if( pNew ){ Expr *pWhere = pS->pWhere; SWAP(Expr, *pNew, *pExpr); pNew = sqlite3ExprAnd(db, pWhere, pNew); pS->pWhere = pNew; pWalker->eCode = 1; } } return WRC_Prune; } return WRC_Continue; } |
︙ | ︙ | |||
124758 124759 124760 124761 124762 124763 124764 | ** ** SELECT * FROM <tables> WHERE a=? AND b=? GROUP BY b HAVING c=? ** ** A term of the HAVING expression is eligible for transfer if it consists ** entirely of constants and expressions that are also GROUP BY terms that ** use the "BINARY" collation sequence. */ | | < < < < < < < < < < | | > > > > > > | 125213 125214 125215 125216 125217 125218 125219 125220 125221 125222 125223 125224 125225 125226 125227 125228 125229 125230 125231 125232 125233 125234 125235 125236 125237 125238 125239 | ** ** SELECT * FROM <tables> WHERE a=? AND b=? GROUP BY b HAVING c=? ** ** A term of the HAVING expression is eligible for transfer if it consists ** entirely of constants and expressions that are also GROUP BY terms that ** use the "BINARY" collation sequence. */ static void havingToWhere(Parse *pParse, Select *p){ Walker sWalker; memset(&sWalker, 0, sizeof(sWalker)); sWalker.pParse = pParse; sWalker.xExprCallback = havingToWhereExprCb; sWalker.u.pSelect = p; sqlite3WalkExpr(&sWalker, p->pHaving); #if SELECTTRACE_ENABLED if( sWalker.eCode && (sqlite3SelectTrace & 0x100)!=0 ){ SELECTTRACE(0x100,pParse,p,("Move HAVING terms into WHERE:\n")); sqlite3TreeViewSelect(0, p, 0); } #endif } /* ** Check to see if the pThis entry of pTabList is a self-join of a prior view. ** If it is, then return the SrcList_item for the prior view. If it is not, ** then return 0. */ |
︙ | ︙ | |||
124935 124936 124937 124938 124939 124940 124941 | db = pParse->db; if( p==0 || db->mallocFailed || pParse->nErr ){ return 1; } if( sqlite3AuthCheck(pParse, SQLITE_SELECT, 0, 0, 0) ) return 1; memset(&sAggInfo, 0, sizeof(sAggInfo)); #if SELECTTRACE_ENABLED | < | 125386 125387 125388 125389 125390 125391 125392 125393 125394 125395 125396 125397 125398 125399 | db = pParse->db; if( p==0 || db->mallocFailed || pParse->nErr ){ return 1; } if( sqlite3AuthCheck(pParse, SQLITE_SELECT, 0, 0, 0) ) return 1; memset(&sAggInfo, 0, sizeof(sAggInfo)); #if SELECTTRACE_ENABLED SELECTTRACE(1,pParse,p, ("begin processing:\n")); if( sqlite3SelectTrace & 0x100 ){ sqlite3TreeViewSelect(0, p, 0); } #endif assert( p->pOrderBy==0 || pDest->eDest!=SRT_DistFifo ); |
︙ | ︙ | |||
124981 124982 124983 124984 124985 124986 124987 | ** does not already exist */ v = sqlite3GetVdbe(pParse); if( v==0 ) goto select_end; if( pDest->eDest==SRT_Output ){ generateColumnNames(pParse, p); } | > | > > > > > > > > > > > > > > > | 125431 125432 125433 125434 125435 125436 125437 125438 125439 125440 125441 125442 125443 125444 125445 125446 125447 125448 125449 125450 125451 125452 125453 125454 125455 125456 125457 125458 125459 125460 125461 125462 125463 125464 125465 125466 125467 | ** does not already exist */ v = sqlite3GetVdbe(pParse); if( v==0 ) goto select_end; if( pDest->eDest==SRT_Output ){ generateColumnNames(pParse, p); } /* Try to various optimizations (flattening subqueries, and strength ** reduction of join operators) in the FROM clause up into the main query */ #if !defined(SQLITE_OMIT_SUBQUERY) || !defined(SQLITE_OMIT_VIEW) for(i=0; !p->pPrior && i<pTabList->nSrc; i++){ struct SrcList_item *pItem = &pTabList->a[i]; Select *pSub = pItem->pSelect; Table *pTab = pItem->pTab; /* Convert LEFT JOIN into JOIN if there are terms of the right table ** of the LEFT JOIN used in the WHERE clause. */ if( (pItem->fg.jointype & JT_LEFT)!=0 && sqlite3ExprImpliesNonNullRow(p->pWhere, pItem->iCursor) && OptimizationEnabled(db, SQLITE_SimplifyJoin) ){ SELECTTRACE(0x100,pParse,p, ("LEFT-JOIN simplifies to JOIN on term %d\n",i)); pItem->fg.jointype &= ~(JT_LEFT|JT_OUTER); unsetJoinExpr(p->pWhere, pItem->iCursor); } /* No futher action if this term of the FROM clause is no a subquery */ if( pSub==0 ) continue; /* Catch mismatch in the declared columns of a view and the number of ** columns in the SELECT on the RHS */ if( pTab->nCol!=pSub->pEList->nExpr ){ sqlite3ErrorMsg(pParse, "expected %d columns for '%s' but got %d", pTab->nCol, pTab->zName, pSub->pEList->nExpr); |
︙ | ︙ | |||
125056 125057 125058 125059 125060 125061 125062 | ** procedure. */ if( p->pPrior ){ rc = multiSelect(pParse, p, pDest); explainSetInteger(pParse->iSelectId, iRestoreSelectId); #if SELECTTRACE_ENABLED SELECTTRACE(1,pParse,p,("end compound-select processing\n")); | < | 125522 125523 125524 125525 125526 125527 125528 125529 125530 125531 125532 125533 125534 125535 | ** procedure. */ if( p->pPrior ){ rc = multiSelect(pParse, p, pDest); explainSetInteger(pParse->iSelectId, iRestoreSelectId); #if SELECTTRACE_ENABLED SELECTTRACE(1,pParse,p,("end compound-select processing\n")); #endif return rc; } #endif /* For each term in the FROM clause, do two things: ** (1) Authorized unreferenced tables |
︙ | ︙ | |||
125129 125130 125131 125132 125133 125134 125135 | ** an exact limit. */ pParse->nHeight += sqlite3SelectExprHeight(p); /* Make copies of constant WHERE-clause terms in the outer query down ** inside the subquery. This can help the subquery to run more efficiently. */ | | | > > > | 125594 125595 125596 125597 125598 125599 125600 125601 125602 125603 125604 125605 125606 125607 125608 125609 125610 125611 125612 125613 125614 125615 125616 125617 125618 125619 | ** an exact limit. */ pParse->nHeight += sqlite3SelectExprHeight(p); /* Make copies of constant WHERE-clause terms in the outer query down ** inside the subquery. This can help the subquery to run more efficiently. */ if( OptimizationEnabled(db, SQLITE_PushDown) && pushDownWhereTerms(pParse, pSub, p->pWhere, pItem->iCursor, (pItem->fg.jointype & JT_OUTER)!=0) ){ #if SELECTTRACE_ENABLED if( sqlite3SelectTrace & 0x100 ){ SELECTTRACE(0x100,pParse,p,("After WHERE-clause push-down:\n")); sqlite3TreeViewSelect(0, p, 0); } #endif }else{ SELECTTRACE(0x100,pParse,p,("Push-down not possible\n")); } zSavedAuthContext = pParse->zAuthContext; pParse->zAuthContext = pItem->zName; /* Generate code to implement the subquery ** |
︙ | ︙ | |||
125340 125341 125342 125343 125344 125345 125346 125347 125348 125349 125350 125351 125352 125353 | if( !isAgg && pGroupBy==0 ){ /* No aggregate functions and no GROUP BY clause */ u16 wctrlFlags = (sDistinct.isTnct ? WHERE_WANT_DISTINCT : 0); assert( WHERE_USE_LIMIT==SF_FixedLimit ); wctrlFlags |= p->selFlags & SF_FixedLimit; /* Begin the database scan. */ pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, sSort.pOrderBy, p->pEList, wctrlFlags, p->nSelectRow); if( pWInfo==0 ) goto select_end; if( sqlite3WhereOutputRowCount(pWInfo) < p->nSelectRow ){ p->nSelectRow = sqlite3WhereOutputRowCount(pWInfo); } if( sDistinct.isTnct && sqlite3WhereIsDistinct(pWInfo) ){ | > | 125808 125809 125810 125811 125812 125813 125814 125815 125816 125817 125818 125819 125820 125821 125822 | if( !isAgg && pGroupBy==0 ){ /* No aggregate functions and no GROUP BY clause */ u16 wctrlFlags = (sDistinct.isTnct ? WHERE_WANT_DISTINCT : 0); assert( WHERE_USE_LIMIT==SF_FixedLimit ); wctrlFlags |= p->selFlags & SF_FixedLimit; /* Begin the database scan. */ SELECTTRACE(1,pParse,p,("WhereBegin\n")); pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, sSort.pOrderBy, p->pEList, wctrlFlags, p->nSelectRow); if( pWInfo==0 ) goto select_end; if( sqlite3WhereOutputRowCount(pWInfo) < p->nSelectRow ){ p->nSelectRow = sqlite3WhereOutputRowCount(pWInfo); } if( sDistinct.isTnct && sqlite3WhereIsDistinct(pWInfo) ){ |
︙ | ︙ | |||
125441 125442 125443 125444 125445 125446 125447 | sAggInfo.nSortingColumn = pGroupBy ? pGroupBy->nExpr : 0; sAggInfo.pGroupBy = pGroupBy; sqlite3ExprAnalyzeAggList(&sNC, pEList); sqlite3ExprAnalyzeAggList(&sNC, sSort.pOrderBy); if( pHaving ){ if( pGroupBy ){ assert( pWhere==p->pWhere ); | > > | | 125910 125911 125912 125913 125914 125915 125916 125917 125918 125919 125920 125921 125922 125923 125924 125925 125926 | sAggInfo.nSortingColumn = pGroupBy ? pGroupBy->nExpr : 0; sAggInfo.pGroupBy = pGroupBy; sqlite3ExprAnalyzeAggList(&sNC, pEList); sqlite3ExprAnalyzeAggList(&sNC, sSort.pOrderBy); if( pHaving ){ if( pGroupBy ){ assert( pWhere==p->pWhere ); assert( pHaving==p->pHaving ); assert( pGroupBy==p->pGroupBy ); havingToWhere(pParse, p); pWhere = p->pWhere; } sqlite3ExprAnalyzeAggregates(&sNC, pHaving); } sAggInfo.nAccumulator = sAggInfo.nColumn; if( p->pGroupBy==0 && p->pHaving==0 && sAggInfo.nFunc==1 ){ minMaxFlag = minMaxQuery(db, sAggInfo.aFunc[0].pExpr, &pMinMaxOrderBy); |
︙ | ︙ | |||
125528 125529 125530 125531 125532 125533 125534 125535 125536 125537 125538 125539 125540 125541 | /* Begin a loop that will extract all source rows in GROUP BY order. ** This might involve two separate loops with an OP_Sort in between, or ** it might be a single loop that uses an index to extract information ** in the right order to begin with. */ sqlite3VdbeAddOp2(v, OP_Gosub, regReset, addrReset); pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, pGroupBy, 0, WHERE_GROUPBY | (orderByGrp ? WHERE_SORTBYGROUP : 0), 0 ); if( pWInfo==0 ) goto select_end; if( sqlite3WhereIsOrdered(pWInfo)==pGroupBy->nExpr ){ /* The optimizer is able to deliver rows in group by order so ** we do not have to sort. The OP_OpenEphemeral table will be | > | 125999 126000 126001 126002 126003 126004 126005 126006 126007 126008 126009 126010 126011 126012 126013 | /* Begin a loop that will extract all source rows in GROUP BY order. ** This might involve two separate loops with an OP_Sort in between, or ** it might be a single loop that uses an index to extract information ** in the right order to begin with. */ sqlite3VdbeAddOp2(v, OP_Gosub, regReset, addrReset); SELECTTRACE(1,pParse,p,("WhereBegin\n")); pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, pGroupBy, 0, WHERE_GROUPBY | (orderByGrp ? WHERE_SORTBYGROUP : 0), 0 ); if( pWInfo==0 ) goto select_end; if( sqlite3WhereIsOrdered(pWInfo)==pGroupBy->nExpr ){ /* The optimizer is able to deliver rows in group by order so ** we do not have to sort. The OP_OpenEphemeral table will be |
︙ | ︙ | |||
125783 125784 125785 125786 125787 125788 125789 125790 125791 125792 125793 125794 125795 125796 | ** minMaxFlag will have been previously set to either ** WHERE_ORDERBY_MIN or WHERE_ORDERBY_MAX and pMinMaxOrderBy will ** be an appropriate ORDER BY expression for the optimization. */ assert( minMaxFlag==WHERE_ORDERBY_NORMAL || pMinMaxOrderBy!=0 ); assert( pMinMaxOrderBy==0 || pMinMaxOrderBy->nExpr==1 ); pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, pMinMaxOrderBy, 0, minMaxFlag, 0); if( pWInfo==0 ){ goto select_end; } updateAccumulator(pParse, &sAggInfo); if( sqlite3WhereIsOrdered(pWInfo)>0 ){ | > | 126255 126256 126257 126258 126259 126260 126261 126262 126263 126264 126265 126266 126267 126268 126269 | ** minMaxFlag will have been previously set to either ** WHERE_ORDERBY_MIN or WHERE_ORDERBY_MAX and pMinMaxOrderBy will ** be an appropriate ORDER BY expression for the optimization. */ assert( minMaxFlag==WHERE_ORDERBY_NORMAL || pMinMaxOrderBy!=0 ); assert( pMinMaxOrderBy==0 || pMinMaxOrderBy->nExpr==1 ); SELECTTRACE(1,pParse,p,("WhereBegin\n")); pWInfo = sqlite3WhereBegin(pParse, pTabList, pWhere, pMinMaxOrderBy, 0, minMaxFlag, 0); if( pWInfo==0 ){ goto select_end; } updateAccumulator(pParse, &sAggInfo); if( sqlite3WhereIsOrdered(pWInfo)>0 ){ |
︙ | ︙ | |||
125838 125839 125840 125841 125842 125843 125844 | select_end: explainSetInteger(pParse->iSelectId, iRestoreSelectId); sqlite3ExprListDelete(db, pMinMaxOrderBy); sqlite3DbFree(db, sAggInfo.aCol); sqlite3DbFree(db, sAggInfo.aFunc); #if SELECTTRACE_ENABLED SELECTTRACE(1,pParse,p,("end processing\n")); | < | 126311 126312 126313 126314 126315 126316 126317 126318 126319 126320 126321 126322 126323 126324 | select_end: explainSetInteger(pParse->iSelectId, iRestoreSelectId); sqlite3ExprListDelete(db, pMinMaxOrderBy); sqlite3DbFree(db, sAggInfo.aCol); sqlite3DbFree(db, sAggInfo.aFunc); #if SELECTTRACE_ENABLED SELECTTRACE(1,pParse,p,("end processing\n")); #endif return rc; } /************** End of select.c **********************************************/ /************** Begin file table.c *******************************************/ /* |
︙ | ︙ | |||
132405 132406 132407 132408 132409 132410 132411 | continue; } if( iLoop<3 && (pTerm->wtFlags & TERM_VARSELECT) ){ if( iNext==0 ) iNext = 3; continue; } | | | > | 132877 132878 132879 132880 132881 132882 132883 132884 132885 132886 132887 132888 132889 132890 132891 132892 132893 132894 132895 132896 132897 132898 132899 132900 132901 132902 132903 | continue; } if( iLoop<3 && (pTerm->wtFlags & TERM_VARSELECT) ){ if( iNext==0 ) iNext = 3; continue; } if( (pTerm->wtFlags & TERM_LIKECOND)!=0 ){ /* If the TERM_LIKECOND flag is set, that means that the range search ** is sufficient to guarantee that the LIKE operator is true, so we ** can skip the call to the like(A,B) function. But this only works ** for strings. So do not skip the call to the function on the pass ** that compares BLOBs. */ #ifdef SQLITE_LIKE_DOESNT_MATCH_BLOBS continue; #else u32 x = pLevel->iLikeRepCntr; if( x>0 ){ skipLikeAddr = sqlite3VdbeAddOp1(v, (x&1)?OP_IfNot:OP_If, (int)(x>>1)); } VdbeCoverage(v); #endif } #ifdef WHERETRACE_ENABLED /* 0xffff */ if( sqlite3WhereTrace ){ VdbeNoopComment((v, "WhereTerm[%d] (%p) priority=%d", pWC->nTerm-j, pTerm, iLoop)); |
︙ | ︙ | |||
142039 142040 142041 142042 142043 142044 142045 | ** then extract the first few alphanumeric characters from within that ** comment to be the zSelName value. Otherwise, the label is #N where ** is an integer that is incremented with each SELECT statement seen. */ if( yymsp[-8].minor.yy387!=0 ){ const char *z = s.z+6; int i; | | < | 142512 142513 142514 142515 142516 142517 142518 142519 142520 142521 142522 142523 142524 142525 142526 | ** then extract the first few alphanumeric characters from within that ** comment to be the zSelName value. Otherwise, the label is #N where ** is an integer that is incremented with each SELECT statement seen. */ if( yymsp[-8].minor.yy387!=0 ){ const char *z = s.z+6; int i; sqlite3_snprintf(sizeof(yymsp[-8].minor.yy387->zSelName), yymsp[-8].minor.yy387->zSelName,"#%d",++pParse->nSelect); while( z[0]==' ' ) z++; if( z[0]=='/' && z[1]=='*' ){ z += 2; while( z[0]==' ' ) z++; for(i=0; sqlite3Isalnum(z[i]); i++){} sqlite3_snprintf(sizeof(yymsp[-8].minor.yy387->zSelName), yymsp[-8].minor.yy387->zSelName, "%.*s", i, z); } |
︙ | ︙ | |||
145976 145977 145978 145979 145980 145981 145982 145983 145984 145985 | } /* ** This routine implements a busy callback that sleeps and tries ** again until a timeout value is reached. The timeout value is ** an integer number of milliseconds passed in as the first ** argument. */ static int sqliteDefaultBusyCallback( void *ptr, /* Database connection */ | > > > | > > > | > > > > > > > > > > > > > | | > > | > | | > > | | > > > > > > > > | > | 146448 146449 146450 146451 146452 146453 146454 146455 146456 146457 146458 146459 146460 146461 146462 146463 146464 146465 146466 146467 146468 146469 146470 146471 146472 146473 146474 146475 146476 146477 146478 146479 146480 146481 146482 146483 146484 146485 146486 146487 146488 146489 146490 146491 146492 146493 146494 146495 146496 146497 146498 146499 146500 146501 146502 146503 146504 146505 146506 146507 146508 146509 146510 146511 146512 146513 146514 146515 146516 146517 146518 146519 146520 146521 146522 146523 146524 146525 146526 146527 146528 146529 146530 146531 146532 146533 146534 146535 146536 146537 146538 146539 146540 146541 146542 146543 146544 146545 | } /* ** This routine implements a busy callback that sleeps and tries ** again until a timeout value is reached. The timeout value is ** an integer number of milliseconds passed in as the first ** argument. ** ** Return non-zero to retry the lock. Return zero to stop trying ** and cause SQLite to return SQLITE_BUSY. */ static int sqliteDefaultBusyCallback( void *ptr, /* Database connection */ int count, /* Number of times table has been busy */ sqlite3_file *pFile /* The file on which the lock occurred */ ){ #if SQLITE_OS_WIN || HAVE_USLEEP /* This case is for systems that have support for sleeping for fractions of ** a second. Examples: All windows systems, unix systems with usleep() */ static const u8 delays[] = { 1, 2, 5, 10, 15, 20, 25, 25, 25, 50, 50, 100 }; static const u8 totals[] = { 0, 1, 3, 8, 18, 33, 53, 78, 103, 128, 178, 228 }; # define NDELAY ArraySize(delays) sqlite3 *db = (sqlite3 *)ptr; int tmout = db->busyTimeout; int delay, prior; #ifdef SQLITE_ENABLE_SETLK_TIMEOUT if( sqlite3OsFileControl(pFile,SQLITE_FCNTL_LOCK_TIMEOUT,&tmout)==SQLITE_OK ){ if( count ){ tmout = 0; sqlite3OsFileControl(pFile, SQLITE_FCNTL_LOCK_TIMEOUT, &tmout); return 0; }else{ return 1; } } #else UNUSED_PARAMETER(pFile); #endif assert( count>=0 ); if( count < NDELAY ){ delay = delays[count]; prior = totals[count]; }else{ delay = delays[NDELAY-1]; prior = totals[NDELAY-1] + delay*(count-(NDELAY-1)); } if( prior + delay > tmout ){ delay = tmout - prior; if( delay<=0 ) return 0; } sqlite3OsSleep(db->pVfs, delay*1000); return 1; #else /* This case for unix systems that lack usleep() support. Sleeping ** must be done in increments of whole seconds */ sqlite3 *db = (sqlite3 *)ptr; int tmout = ((sqlite3 *)ptr)->busyTimeout; UNUSED_PARAMETER(pFile); if( (count+1)*1000 > tmout ){ return 0; } sqlite3OsSleep(db->pVfs, 1000000); return 1; #endif } /* ** Invoke the given busy handler. ** ** This routine is called when an operation failed to acquire a ** lock on VFS file pFile. ** ** If this routine returns non-zero, the lock is retried. If it ** returns 0, the operation aborts with an SQLITE_BUSY error. */ SQLITE_PRIVATE int sqlite3InvokeBusyHandler(BusyHandler *p, sqlite3_file *pFile){ int rc; if( p->xBusyHandler==0 || p->nBusy<0 ) return 0; if( p->bExtraFileArg ){ /* Add an extra parameter with the pFile pointer to the end of the ** callback argument list */ int (*xTra)(void*,int,sqlite3_file*); xTra = (int(*)(void*,int,sqlite3_file*))p->xBusyHandler; rc = xTra(p->pBusyArg, p->nBusy, pFile); }else{ /* Legacy style busy handler callback */ rc = p->xBusyHandler(p->pBusyArg, p->nBusy); } if( rc==0 ){ p->nBusy = -1; }else{ p->nBusy++; } return rc; } |
︙ | ︙ | |||
146048 146049 146050 146051 146052 146053 146054 | int (*xBusy)(void*,int), void *pArg ){ #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ) return SQLITE_MISUSE_BKPT; #endif sqlite3_mutex_enter(db->mutex); | | | > | 146553 146554 146555 146556 146557 146558 146559 146560 146561 146562 146563 146564 146565 146566 146567 146568 146569 146570 | int (*xBusy)(void*,int), void *pArg ){ #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ) return SQLITE_MISUSE_BKPT; #endif sqlite3_mutex_enter(db->mutex); db->busyHandler.xBusyHandler = xBusy; db->busyHandler.pBusyArg = pArg; db->busyHandler.nBusy = 0; db->busyHandler.bExtraFileArg = 0; db->busyTimeout = 0; sqlite3_mutex_leave(db->mutex); return SQLITE_OK; } #ifndef SQLITE_OMIT_PROGRESS_CALLBACK /* |
︙ | ︙ | |||
146098 146099 146100 146101 146102 146103 146104 | ** specified number of milliseconds before returning 0. */ SQLITE_API int sqlite3_busy_timeout(sqlite3 *db, int ms){ #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ) return SQLITE_MISUSE_BKPT; #endif if( ms>0 ){ | | > > | 146604 146605 146606 146607 146608 146609 146610 146611 146612 146613 146614 146615 146616 146617 146618 146619 146620 146621 | ** specified number of milliseconds before returning 0. */ SQLITE_API int sqlite3_busy_timeout(sqlite3 *db, int ms){ #ifdef SQLITE_ENABLE_API_ARMOR if( !sqlite3SafetyCheckOk(db) ) return SQLITE_MISUSE_BKPT; #endif if( ms>0 ){ sqlite3_busy_handler(db, (int(*)(void*,int))sqliteDefaultBusyCallback, (void*)db); db->busyTimeout = ms; db->busyHandler.bExtraFileArg = 1; }else{ sqlite3_busy_handler(db, 0, 0); } return SQLITE_OK; } /* |
︙ | ︙ | |||
148092 148093 148094 148095 148096 148097 148098 | rc = SQLITE_OK; }else if( op==SQLITE_FCNTL_VFS_POINTER ){ *(sqlite3_vfs**)pArg = sqlite3PagerVfs(pPager); rc = SQLITE_OK; }else if( op==SQLITE_FCNTL_JOURNAL_POINTER ){ *(sqlite3_file**)pArg = sqlite3PagerJrnlFile(pPager); rc = SQLITE_OK; | | < < | 148600 148601 148602 148603 148604 148605 148606 148607 148608 148609 148610 148611 148612 148613 148614 148615 | rc = SQLITE_OK; }else if( op==SQLITE_FCNTL_VFS_POINTER ){ *(sqlite3_vfs**)pArg = sqlite3PagerVfs(pPager); rc = SQLITE_OK; }else if( op==SQLITE_FCNTL_JOURNAL_POINTER ){ *(sqlite3_file**)pArg = sqlite3PagerJrnlFile(pPager); rc = SQLITE_OK; }else{ rc = sqlite3OsFileControl(fd, op, pArg); } sqlite3BtreeLeave(pBtree); } sqlite3_mutex_leave(db->mutex); return rc; } |
︙ | ︙ | |||
175671 175672 175673 175674 175675 175676 175677 | while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pXInfo) ){ int bKey = sqlite3_column_int(pXInfo, 5); if( bKey ){ int iCid = sqlite3_column_int(pXInfo, 1); int bDesc = sqlite3_column_int(pXInfo, 3); const char *zCollate = (const char*)sqlite3_column_text(pXInfo, 4); | | | 176177 176178 176179 176180 176181 176182 176183 176184 176185 176186 176187 176188 176189 176190 176191 | while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pXInfo) ){ int bKey = sqlite3_column_int(pXInfo, 5); if( bKey ){ int iCid = sqlite3_column_int(pXInfo, 1); int bDesc = sqlite3_column_int(pXInfo, 3); const char *zCollate = (const char*)sqlite3_column_text(pXInfo, 4); zCols = rbuMPrintf(p, "%z%sc%d %s COLLATE %Q", zCols, zComma, iCid, pIter->azTblType[iCid], zCollate ); zPk = rbuMPrintf(p, "%z%sc%d%s", zPk, zComma, iCid, bDesc?" DESC":""); zComma = ", "; } } zCols = rbuMPrintf(p, "%z, id INTEGER", zCols); |
︙ | ︙ | |||
175732 175733 175734 175735 175736 175737 175738 | ); if( pIter->eType==RBU_PK_IPK && pIter->abTblPk[iCol] ){ /* If the target table column is an "INTEGER PRIMARY KEY", add ** "PRIMARY KEY" to the imposter table column declaration. */ zPk = "PRIMARY KEY "; } | | | 176238 176239 176240 176241 176242 176243 176244 176245 176246 176247 176248 176249 176250 176251 176252 | ); if( pIter->eType==RBU_PK_IPK && pIter->abTblPk[iCol] ){ /* If the target table column is an "INTEGER PRIMARY KEY", add ** "PRIMARY KEY" to the imposter table column declaration. */ zPk = "PRIMARY KEY "; } zSql = rbuMPrintf(p, "%z%s\"%w\" %s %sCOLLATE %Q%s", zSql, zComma, zCol, pIter->azTblType[iCol], zPk, zColl, (pIter->abNotNull[iCol] ? " NOT NULL" : "") ); zComma = ", "; } if( pIter->eType==RBU_PK_WITHOUT_ROWID ){ |
︙ | ︙ | |||
179166 179167 179168 179169 179170 179171 179172 | pCsr->iOffset = (i64)pCsr->szPage * (pCsr->iPageno - 1); /* If connected to a ZIPVFS backend, override the page size and ** offset with actual values obtained from ZIPVFS. */ fd = sqlite3PagerFile(pPager); x[0] = pCsr->iPageno; | | | 179672 179673 179674 179675 179676 179677 179678 179679 179680 179681 179682 179683 179684 179685 179686 | pCsr->iOffset = (i64)pCsr->szPage * (pCsr->iPageno - 1); /* If connected to a ZIPVFS backend, override the page size and ** offset with actual values obtained from ZIPVFS. */ fd = sqlite3PagerFile(pPager); x[0] = pCsr->iPageno; if( sqlite3OsFileControl(fd, 230440, &x)==SQLITE_OK ){ pCsr->iOffset = x[0]; pCsr->szPage = (int)x[1]; } } /* ** Move a statvfs cursor to the next entry in the file. |
︙ | ︙ | |||
180096 180097 180098 180099 180100 180101 180102 | ** ** As in the changeset format, each field of the single record that is part ** of a patchset change is associated with the correspondingly positioned ** table column, counting from left to right within the CREATE TABLE ** statement. ** ** For a DELETE change, all fields within the record except those associated | | | | 180602 180603 180604 180605 180606 180607 180608 180609 180610 180611 180612 180613 180614 180615 180616 180617 | ** ** As in the changeset format, each field of the single record that is part ** of a patchset change is associated with the correspondingly positioned ** table column, counting from left to right within the CREATE TABLE ** statement. ** ** For a DELETE change, all fields within the record except those associated ** with PRIMARY KEY columns are omitted. The PRIMARY KEY fields contain the ** values identifying the row to delete. ** ** For an UPDATE change, all fields except those associated with PRIMARY KEY ** columns and columns that are modified by the UPDATE are set to "undefined". ** PRIMARY KEY fields contain the values identifying the table row to update, ** and fields associated with modified columns contain the new column values. ** ** The records associated with INSERT changes are in the same format as for |
︙ | ︙ | |||
180380 180381 180382 180383 180384 180385 180386 | ** The buffer that the argument points to contains a serialized SQL value. ** Return the number of bytes of space occupied by the value (including ** the type byte). */ static int sessionSerialLen(u8 *a){ int e = *a; int n; | | | 180886 180887 180888 180889 180890 180891 180892 180893 180894 180895 180896 180897 180898 180899 180900 | ** The buffer that the argument points to contains a serialized SQL value. ** Return the number of bytes of space occupied by the value (including ** the type byte). */ static int sessionSerialLen(u8 *a){ int e = *a; int n; if( e==0 || e==0xFF ) return 1; if( e==SQLITE_NULL ) return 1; if( e==SQLITE_INTEGER || e==SQLITE_FLOAT ) return 9; return sessionVarintGet(&a[1], &n) + 1 + n; } /* ** Based on the primary key values stored in change aRecord, calculate a |
︙ | ︙ | |||
180460 180461 180462 180463 180464 180465 180466 | int iCol; /* Used to iterate through table columns */ for(iCol=0; iCol<pTab->nCol; iCol++){ if( pTab->abPK[iCol] ){ int n1 = sessionSerialLen(a1); int n2 = sessionSerialLen(a2); | | | 180966 180967 180968 180969 180970 180971 180972 180973 180974 180975 180976 180977 180978 180979 180980 | int iCol; /* Used to iterate through table columns */ for(iCol=0; iCol<pTab->nCol; iCol++){ if( pTab->abPK[iCol] ){ int n1 = sessionSerialLen(a1); int n2 = sessionSerialLen(a2); if( n1!=n2 || memcmp(a1, a2, n1) ){ return 0; } a1 += n1; a2 += n2; }else{ if( bLeftPkOnly==0 ) a1 += sessionSerialLen(a1); if( bRightPkOnly==0 ) a2 += sessionSerialLen(a2); |
︙ | ︙ | |||
180703 180704 180705 180706 180707 180708 180709 | a += sessionVarintGet(a, &n); if( sqlite3_value_bytes(pVal)!=n ) return 0; if( eType==SQLITE_TEXT ){ z = sqlite3_value_text(pVal); }else{ z = sqlite3_value_blob(pVal); } | | | 181209 181210 181211 181212 181213 181214 181215 181216 181217 181218 181219 181220 181221 181222 181223 | a += sessionVarintGet(a, &n); if( sqlite3_value_bytes(pVal)!=n ) return 0; if( eType==SQLITE_TEXT ){ z = sqlite3_value_text(pVal); }else{ z = sqlite3_value_blob(pVal); } if( n>0 && memcmp(a, z, n) ) return 0; a += n; } } } return 1; } |
︙ | ︙ | |||
182047 182048 182049 182050 182051 182052 182053 182054 182055 182056 182057 182058 182059 182060 | int nSql = -1; if( 0==sqlite3_stricmp("sqlite_stat1", zTab) ){ zSql = sqlite3_mprintf( "SELECT tbl, ?2, stat FROM %Q.sqlite_stat1 WHERE tbl IS ?1 AND " "idx IS (CASE WHEN ?2=X'' THEN NULL ELSE ?2 END)", zDb ); }else{ int i; const char *zSep = ""; SessionBuffer buf = {0, 0, 0}; sessionAppendStr(&buf, "SELECT * FROM ", &rc); sessionAppendIdent(&buf, zDb, &rc); | > | 182553 182554 182555 182556 182557 182558 182559 182560 182561 182562 182563 182564 182565 182566 182567 | int nSql = -1; if( 0==sqlite3_stricmp("sqlite_stat1", zTab) ){ zSql = sqlite3_mprintf( "SELECT tbl, ?2, stat FROM %Q.sqlite_stat1 WHERE tbl IS ?1 AND " "idx IS (CASE WHEN ?2=X'' THEN NULL ELSE ?2 END)", zDb ); if( zSql==0 ) rc = SQLITE_NOMEM; }else{ int i; const char *zSep = ""; SessionBuffer buf = {0, 0, 0}; sessionAppendStr(&buf, "SELECT * FROM ", &rc); sessionAppendIdent(&buf, zDb, &rc); |
︙ | ︙ | |||
182782 182783 182784 182785 182786 182787 182788 | ** successfully advanced to the next change in the changeset, an SQLite ** error code if an error occurs, or SQLITE_DONE if there are no further ** changes in the changeset. */ static int sessionChangesetNext( sqlite3_changeset_iter *p, /* Changeset iterator */ u8 **paRec, /* If non-NULL, store record pointer here */ | | > | 183289 183290 183291 183292 183293 183294 183295 183296 183297 183298 183299 183300 183301 183302 183303 183304 | ** successfully advanced to the next change in the changeset, an SQLite ** error code if an error occurs, or SQLITE_DONE if there are no further ** changes in the changeset. */ static int sessionChangesetNext( sqlite3_changeset_iter *p, /* Changeset iterator */ u8 **paRec, /* If non-NULL, store record pointer here */ int *pnRec, /* If non-NULL, store size of record here */ int *pbNew /* If non-NULL, true if new table */ ){ int i; u8 op; assert( (paRec==0 && pnRec==0) || (paRec && pnRec) ); /* If the iterator is in the error-state, return immediately. */ |
︙ | ︙ | |||
182817 182818 182819 182820 182821 182822 182823 182824 182825 182826 182827 182828 182829 182830 | } sessionDiscardData(&p->in); p->in.iCurrent = p->in.iNext; op = p->in.aData[p->in.iNext++]; while( op=='T' || op=='P' ){ p->bPatchset = (op=='P'); if( sessionChangesetReadTblhdr(p) ) return p->rc; if( (p->rc = sessionInputBuffer(&p->in, 2)) ) return p->rc; p->in.iCurrent = p->in.iNext; if( p->in.iNext>=p->in.nData ) return SQLITE_DONE; op = p->in.aData[p->in.iNext++]; } | > | 183325 183326 183327 183328 183329 183330 183331 183332 183333 183334 183335 183336 183337 183338 183339 | } sessionDiscardData(&p->in); p->in.iCurrent = p->in.iNext; op = p->in.aData[p->in.iNext++]; while( op=='T' || op=='P' ){ if( pbNew ) *pbNew = 1; p->bPatchset = (op=='P'); if( sessionChangesetReadTblhdr(p) ) return p->rc; if( (p->rc = sessionInputBuffer(&p->in, 2)) ) return p->rc; p->in.iCurrent = p->in.iNext; if( p->in.iNext>=p->in.nData ) return SQLITE_DONE; op = p->in.aData[p->in.iNext++]; } |
︙ | ︙ | |||
182895 182896 182897 182898 182899 182900 182901 | ** change in the changeset. This function may return SQLITE_ROW, SQLITE_DONE ** or SQLITE_CORRUPT. ** ** This function may not be called on iterators passed to a conflict handler ** callback by changeset_apply(). */ SQLITE_API int sqlite3changeset_next(sqlite3_changeset_iter *p){ | | | 183404 183405 183406 183407 183408 183409 183410 183411 183412 183413 183414 183415 183416 183417 183418 | ** change in the changeset. This function may return SQLITE_ROW, SQLITE_DONE ** or SQLITE_CORRUPT. ** ** This function may not be called on iterators passed to a conflict handler ** callback by changeset_apply(). */ SQLITE_API int sqlite3changeset_next(sqlite3_changeset_iter *p){ return sessionChangesetNext(p, 0, 0, 0); } /* ** The following function extracts information on the current change ** from a changeset iterator. It may only be called after changeset_next() ** has returned SQLITE_ROW. */ |
︙ | ︙ | |||
183274 183275 183276 183277 183278 183279 183280 183281 183282 183283 183284 183285 183286 183287 | sqlite3_stmt *pSelect; /* SELECT statement */ int nCol; /* Size of azCol[] and abPK[] arrays */ const char **azCol; /* Array of column names */ u8 *abPK; /* Boolean array - true if column is in PK */ int bStat1; /* True if table is sqlite_stat1 */ int bDeferConstraints; /* True to defer constraints */ SessionBuffer constraints; /* Deferred constraints are stored here */ }; /* ** Formulate a statement to DELETE a row from database db. Assuming a table ** structure like this: ** ** CREATE TABLE x(a, b, c, d, PRIMARY KEY(a, c)); | > > | 183783 183784 183785 183786 183787 183788 183789 183790 183791 183792 183793 183794 183795 183796 183797 183798 | sqlite3_stmt *pSelect; /* SELECT statement */ int nCol; /* Size of azCol[] and abPK[] arrays */ const char **azCol; /* Array of column names */ u8 *abPK; /* Boolean array - true if column is in PK */ int bStat1; /* True if table is sqlite_stat1 */ int bDeferConstraints; /* True to defer constraints */ SessionBuffer constraints; /* Deferred constraints are stored here */ SessionBuffer rebase; /* Rebase information (if any) here */ int bRebaseStarted; /* If table header is already in rebase */ }; /* ** Formulate a statement to DELETE a row from database db. Assuming a table ** structure like this: ** ** CREATE TABLE x(a, b, c, d, PRIMARY KEY(a, c)); |
︙ | ︙ | |||
183540 183541 183542 183543 183544 183545 183546 | if( rc==SQLITE_OK ){ rc = sessionPrepare(db, &p->pDelete, "DELETE FROM main.sqlite_stat1 WHERE tbl=?1 AND idx IS " "CASE WHEN length(?2)=0 AND typeof(?2)='blob' THEN NULL ELSE ?2 END " "AND (?4 OR stat IS ?3)" ); } | < | 184051 184052 184053 184054 184055 184056 184057 184058 184059 184060 184061 184062 184063 184064 | if( rc==SQLITE_OK ){ rc = sessionPrepare(db, &p->pDelete, "DELETE FROM main.sqlite_stat1 WHERE tbl=?1 AND idx IS " "CASE WHEN length(?2)=0 AND typeof(?2)='blob' THEN NULL ELSE ?2 END " "AND (?4 OR stat IS ?3)" ); } return rc; } /* ** A wrapper around sqlite3_bind_value() that detects an extra problem. ** See comments in the body of this function for details. */ |
︙ | ︙ | |||
183654 183655 183656 183657 183658 183659 183660 183661 183662 183663 183664 183665 183666 183667 | if( rc==SQLITE_OK ){ rc = sqlite3_step(pSelect); if( rc!=SQLITE_ROW ) rc = sqlite3_reset(pSelect); } return rc; } /* ** Invoke the conflict handler for the change that the changeset iterator ** currently points to. ** ** Argument eType must be either CHANGESET_DATA or CHANGESET_CONFLICT. ** If argument pbReplace is NULL, then the type of conflict handler invoked | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 184164 184165 184166 184167 184168 184169 184170 184171 184172 184173 184174 184175 184176 184177 184178 184179 184180 184181 184182 184183 184184 184185 184186 184187 184188 184189 184190 184191 184192 184193 184194 184195 184196 184197 184198 184199 184200 184201 184202 184203 184204 184205 184206 184207 184208 184209 184210 184211 184212 184213 184214 184215 184216 184217 184218 184219 184220 184221 184222 184223 184224 184225 | if( rc==SQLITE_OK ){ rc = sqlite3_step(pSelect); if( rc!=SQLITE_ROW ) rc = sqlite3_reset(pSelect); } return rc; } /* ** This function is called from within sqlite3changset_apply_v2() when ** a conflict is encountered and resolved using conflict resolution ** mode eType (either SQLITE_CHANGESET_OMIT or SQLITE_CHANGESET_REPLACE).. ** It adds a conflict resolution record to the buffer in ** SessionApplyCtx.rebase, which will eventually be returned to the caller ** of apply_v2() as the "rebase" buffer. ** ** Return SQLITE_OK if successful, or an SQLite error code otherwise. */ static int sessionRebaseAdd( SessionApplyCtx *p, /* Apply context */ int eType, /* Conflict resolution (OMIT or REPLACE) */ sqlite3_changeset_iter *pIter /* Iterator pointing at current change */ ){ int rc = SQLITE_OK; int i; int eOp = pIter->op; if( p->bRebaseStarted==0 ){ /* Append a table-header to the rebase buffer */ const char *zTab = pIter->zTab; sessionAppendByte(&p->rebase, 'T', &rc); sessionAppendVarint(&p->rebase, p->nCol, &rc); sessionAppendBlob(&p->rebase, p->abPK, p->nCol, &rc); sessionAppendBlob(&p->rebase, (u8*)zTab, (int)strlen(zTab)+1, &rc); p->bRebaseStarted = 1; } assert( eType==SQLITE_CHANGESET_REPLACE||eType==SQLITE_CHANGESET_OMIT ); assert( eOp==SQLITE_DELETE || eOp==SQLITE_INSERT || eOp==SQLITE_UPDATE ); sessionAppendByte(&p->rebase, (eOp==SQLITE_DELETE ? SQLITE_DELETE : SQLITE_INSERT), &rc ); sessionAppendByte(&p->rebase, (eType==SQLITE_CHANGESET_REPLACE), &rc); for(i=0; i<p->nCol; i++){ sqlite3_value *pVal = 0; if( eOp==SQLITE_DELETE || (eOp==SQLITE_UPDATE && p->abPK[i]) ){ sqlite3changeset_old(pIter, i, &pVal); }else{ sqlite3changeset_new(pIter, i, &pVal); } sessionAppendValue(&p->rebase, pVal, &rc); } return rc; } /* ** Invoke the conflict handler for the change that the changeset iterator ** currently points to. ** ** Argument eType must be either CHANGESET_DATA or CHANGESET_CONFLICT. ** If argument pbReplace is NULL, then the type of conflict handler invoked |
︙ | ︙ | |||
183730 183731 183732 183733 183734 183735 183736 | }else if( rc==SQLITE_OK ){ if( p->bDeferConstraints && eType==SQLITE_CHANGESET_CONFLICT ){ /* Instead of invoking the conflict handler, append the change blob ** to the SessionApplyCtx.constraints buffer. */ u8 *aBlob = &pIter->in.aData[pIter->in.iCurrent]; int nBlob = pIter->in.iNext - pIter->in.iCurrent; sessionAppendBlob(&p->constraints, aBlob, nBlob, &rc); | | | 184288 184289 184290 184291 184292 184293 184294 184295 184296 184297 184298 184299 184300 184301 184302 | }else if( rc==SQLITE_OK ){ if( p->bDeferConstraints && eType==SQLITE_CHANGESET_CONFLICT ){ /* Instead of invoking the conflict handler, append the change blob ** to the SessionApplyCtx.constraints buffer. */ u8 *aBlob = &pIter->in.aData[pIter->in.iCurrent]; int nBlob = pIter->in.iNext - pIter->in.iCurrent; sessionAppendBlob(&p->constraints, aBlob, nBlob, &rc); return SQLITE_OK; }else{ /* No other row with the new.* primary key. */ res = xConflict(pCtx, eType+1, pIter); if( res==SQLITE_CHANGESET_REPLACE ) rc = SQLITE_MISUSE; } } |
︙ | ︙ | |||
183756 183757 183758 183759 183760 183761 183762 183763 183764 183765 183766 183767 183768 183769 | rc = SQLITE_ABORT; break; default: rc = SQLITE_MISUSE; break; } } return rc; } /* ** Attempt to apply the change that the iterator passed as the first argument | > > > | 184314 184315 184316 184317 184318 184319 184320 184321 184322 184323 184324 184325 184326 184327 184328 184329 184330 | rc = SQLITE_ABORT; break; default: rc = SQLITE_MISUSE; break; } if( rc==SQLITE_OK ){ rc = sessionRebaseAdd(p, res, pIter); } } return rc; } /* ** Attempt to apply the change that the iterator passed as the first argument |
︙ | ︙ | |||
183931 183932 183933 183934 183935 183936 183937 | void *pCtx /* First argument passed to xConflict */ ){ int bReplace = 0; int bRetry = 0; int rc; rc = sessionApplyOneOp(pIter, pApply, xConflict, pCtx, &bReplace, &bRetry); | | < | 184492 184493 184494 184495 184496 184497 184498 184499 184500 184501 184502 184503 184504 184505 184506 | void *pCtx /* First argument passed to xConflict */ ){ int bReplace = 0; int bRetry = 0; int rc; rc = sessionApplyOneOp(pIter, pApply, xConflict, pCtx, &bReplace, &bRetry); if( rc==SQLITE_OK ){ /* If the bRetry flag is set, the change has not been applied due to an ** SQLITE_CHANGESET_DATA problem (i.e. this is an UPDATE or DELETE and ** a row with the correct PK is present in the db, but one or more other ** fields do not contain the expected values) and the conflict handler ** returned SQLITE_CHANGESET_REPLACE. In this case retry the operation, ** but pass NULL as the final argument so that sessionApplyOneOp() ignores ** the SQLITE_CHANGESET_DATA problem. */ |
︙ | ︙ | |||
183969 183970 183971 183972 183973 183974 183975 183976 183977 183978 183979 183980 183981 183982 | if( rc==SQLITE_OK ){ rc = sessionApplyOneOp(pIter, pApply, xConflict, pCtx, 0, 0); } if( rc==SQLITE_OK ){ rc = sqlite3_exec(db, "RELEASE replace_op", 0, 0, 0); } } return rc; } /* ** Retry the changes accumulated in the pApply->constraints buffer. */ | > | 184529 184530 184531 184532 184533 184534 184535 184536 184537 184538 184539 184540 184541 184542 184543 | if( rc==SQLITE_OK ){ rc = sessionApplyOneOp(pIter, pApply, xConflict, pCtx, 0, 0); } if( rc==SQLITE_OK ){ rc = sqlite3_exec(db, "RELEASE replace_op", 0, 0, 0); } } } return rc; } /* ** Retry the changes accumulated in the pApply->constraints buffer. */ |
︙ | ︙ | |||
184042 184043 184044 184045 184046 184047 184048 | const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of fifth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), | | > | 184603 184604 184605 184606 184607 184608 184609 184610 184611 184612 184613 184614 184615 184616 184617 184618 | const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of fifth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase /* OUT: Rebase information */ ){ int schemaMismatch = 0; int rc; /* Return code */ const char *zTab = 0; /* Name of current table */ int nTab = 0; /* Result of sqlite3Strlen30(zTab) */ SessionApplyCtx sApply; /* changeset_apply() context object */ int bPatchset; |
︙ | ︙ | |||
184080 184081 184082 184083 184084 184085 184086 | if( rc!=SQLITE_OK ) break; sqlite3_free((char*)sApply.azCol); /* cast works around VC++ bug */ sqlite3_finalize(sApply.pDelete); sqlite3_finalize(sApply.pUpdate); sqlite3_finalize(sApply.pInsert); sqlite3_finalize(sApply.pSelect); | < > > > > > > > > > > | 184642 184643 184644 184645 184646 184647 184648 184649 184650 184651 184652 184653 184654 184655 184656 184657 184658 184659 184660 184661 184662 184663 184664 184665 184666 184667 | if( rc!=SQLITE_OK ) break; sqlite3_free((char*)sApply.azCol); /* cast works around VC++ bug */ sqlite3_finalize(sApply.pDelete); sqlite3_finalize(sApply.pUpdate); sqlite3_finalize(sApply.pInsert); sqlite3_finalize(sApply.pSelect); sApply.db = db; sApply.pDelete = 0; sApply.pUpdate = 0; sApply.pInsert = 0; sApply.pSelect = 0; sApply.nCol = 0; sApply.azCol = 0; sApply.abPK = 0; sApply.bStat1 = 0; sApply.bDeferConstraints = 1; sApply.bRebaseStarted = 0; memset(&sApply.constraints, 0, sizeof(SessionBuffer)); /* If an xFilter() callback was specified, invoke it now. If the ** xFilter callback returns zero, skip this table. If it returns ** non-zero, proceed. */ schemaMismatch = (xFilter && (0==xFilter(pCtx, zNew))); if( schemaMismatch ){ zTab = sqlite3_mprintf("%s", zNew); |
︙ | ︙ | |||
184192 184193 184194 184195 184196 184197 184198 184199 184200 184201 184202 184203 184204 184205 184206 184207 184208 184209 184210 184211 184212 184213 184214 | if( rc==SQLITE_OK ){ rc = sqlite3_exec(db, "RELEASE changeset_apply", 0, 0, 0); }else{ sqlite3_exec(db, "ROLLBACK TO changeset_apply", 0, 0, 0); sqlite3_exec(db, "RELEASE changeset_apply", 0, 0, 0); } sqlite3_finalize(sApply.pInsert); sqlite3_finalize(sApply.pDelete); sqlite3_finalize(sApply.pUpdate); sqlite3_finalize(sApply.pSelect); sqlite3_free((char*)sApply.azCol); /* cast works around VC++ bug */ sqlite3_free((char*)sApply.constraints.aBuf); sqlite3_mutex_leave(sqlite3_db_mutex(db)); return rc; } /* ** Apply the changeset passed via pChangeset/nChangeset to the main database ** attached to handle "db". Invoke the supplied conflict handler callback ** to resolve any conflicts encountered while applying the change. */ SQLITE_API int sqlite3changeset_apply( | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 184763 184764 184765 184766 184767 184768 184769 184770 184771 184772 184773 184774 184775 184776 184777 184778 184779 184780 184781 184782 184783 184784 184785 184786 184787 184788 184789 184790 184791 184792 184793 184794 184795 184796 184797 184798 184799 184800 184801 184802 184803 184804 184805 184806 184807 184808 184809 184810 184811 184812 184813 184814 184815 184816 184817 184818 184819 184820 184821 | if( rc==SQLITE_OK ){ rc = sqlite3_exec(db, "RELEASE changeset_apply", 0, 0, 0); }else{ sqlite3_exec(db, "ROLLBACK TO changeset_apply", 0, 0, 0); sqlite3_exec(db, "RELEASE changeset_apply", 0, 0, 0); } if( rc==SQLITE_OK && bPatchset==0 && ppRebase && pnRebase ){ *ppRebase = (void*)sApply.rebase.aBuf; *pnRebase = sApply.rebase.nBuf; sApply.rebase.aBuf = 0; } sqlite3_finalize(sApply.pInsert); sqlite3_finalize(sApply.pDelete); sqlite3_finalize(sApply.pUpdate); sqlite3_finalize(sApply.pSelect); sqlite3_free((char*)sApply.azCol); /* cast works around VC++ bug */ sqlite3_free((char*)sApply.constraints.aBuf); sqlite3_free((char*)sApply.rebase.aBuf); sqlite3_mutex_leave(sqlite3_db_mutex(db)); return rc; } /* ** Apply the changeset passed via pChangeset/nChangeset to the main ** database attached to handle "db". */ SQLITE_API int sqlite3changeset_apply_v2( sqlite3 *db, /* Apply change to "main" db of this handle */ int nChangeset, /* Size of changeset in bytes */ void *pChangeset, /* Changeset blob */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ){ sqlite3_changeset_iter *pIter; /* Iterator to skip through changeset */ int rc = sqlite3changeset_start(&pIter, nChangeset, pChangeset); if( rc==SQLITE_OK ){ rc = sessionChangesetApply( db, pIter, xFilter, xConflict, pCtx, ppRebase, pnRebase ); } return rc; } /* ** Apply the changeset passed via pChangeset/nChangeset to the main database ** attached to handle "db". Invoke the supplied conflict handler callback ** to resolve any conflicts encountered while applying the change. */ SQLITE_API int sqlite3changeset_apply( |
︙ | ︙ | |||
184222 184223 184224 184225 184226 184227 184228 | int(*xConflict)( void *pCtx, /* Copy of fifth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ){ | < | < | < < > > > > > > > > > > > > > > > > > > > > > > > > > > < | < | < < > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | | > > > > > > > > > > > > > > > > > > > | 184829 184830 184831 184832 184833 184834 184835 184836 184837 184838 184839 184840 184841 184842 184843 184844 184845 184846 184847 184848 184849 184850 184851 184852 184853 184854 184855 184856 184857 184858 184859 184860 184861 184862 184863 184864 184865 184866 184867 184868 184869 184870 184871 184872 184873 184874 184875 184876 184877 184878 184879 184880 184881 184882 184883 184884 184885 184886 184887 184888 184889 184890 184891 184892 184893 184894 184895 184896 184897 184898 184899 184900 184901 184902 184903 184904 184905 184906 184907 184908 184909 184910 184911 184912 184913 184914 184915 184916 184917 184918 184919 184920 184921 184922 184923 184924 184925 184926 184927 184928 184929 184930 184931 184932 184933 184934 184935 184936 184937 184938 184939 184940 184941 184942 184943 184944 184945 184946 184947 184948 184949 184950 184951 184952 184953 184954 184955 184956 184957 184958 184959 184960 184961 184962 184963 184964 184965 184966 184967 184968 184969 184970 184971 184972 184973 184974 184975 184976 184977 184978 184979 184980 184981 184982 184983 184984 184985 184986 184987 184988 184989 184990 184991 184992 184993 | int(*xConflict)( void *pCtx, /* Copy of fifth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ){ return sqlite3changeset_apply_v2( db, nChangeset, pChangeset, xFilter, xConflict, pCtx, 0, 0 ); } /* ** Apply the changeset passed via xInput/pIn to the main database ** attached to handle "db". Invoke the supplied conflict handler callback ** to resolve any conflicts encountered while applying the change. */ SQLITE_API int sqlite3changeset_apply_v2_strm( sqlite3 *db, /* Apply change to "main" db of this handle */ int (*xInput)(void *pIn, void *pData, int *pnData), /* Input function */ void *pIn, /* First arg for xInput */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ){ sqlite3_changeset_iter *pIter; /* Iterator to skip through changeset */ int rc = sqlite3changeset_start_strm(&pIter, xInput, pIn); if( rc==SQLITE_OK ){ rc = sessionChangesetApply( db, pIter, xFilter, xConflict, pCtx, ppRebase, pnRebase ); } return rc; } SQLITE_API int sqlite3changeset_apply_strm( sqlite3 *db, /* Apply change to "main" db of this handle */ int (*xInput)(void *pIn, void *pData, int *pnData), /* Input function */ void *pIn, /* First arg for xInput */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ){ return sqlite3changeset_apply_v2_strm( db, xInput, pIn, xFilter, xConflict, pCtx, 0, 0 ); } /* ** sqlite3_changegroup handle. */ struct sqlite3_changegroup { int rc; /* Error code */ int bPatch; /* True to accumulate patchsets */ SessionTable *pList; /* List of tables in current patch */ }; /* ** This function is called to merge two changes to the same row together as ** part of an sqlite3changeset_concat() operation. A new change object is ** allocated and a pointer to it stored in *ppNew. */ static int sessionChangeMerge( SessionTable *pTab, /* Table structure */ int bRebase, /* True for a rebase hash-table */ int bPatchset, /* True for patchsets */ SessionChange *pExist, /* Existing change */ int op2, /* Second change operation */ int bIndirect, /* True if second change is indirect */ u8 *aRec, /* Second change record */ int nRec, /* Number of bytes in aRec */ SessionChange **ppNew /* OUT: Merged change */ ){ SessionChange *pNew = 0; int rc = SQLITE_OK; if( !pExist ){ pNew = (SessionChange *)sqlite3_malloc(sizeof(SessionChange) + nRec); if( !pNew ){ return SQLITE_NOMEM; } memset(pNew, 0, sizeof(SessionChange)); pNew->op = op2; pNew->bIndirect = bIndirect; pNew->aRecord = (u8*)&pNew[1]; if( bIndirect==0 || bRebase==0 ){ pNew->nRecord = nRec; memcpy(pNew->aRecord, aRec, nRec); }else{ int i; u8 *pIn = aRec; u8 *pOut = pNew->aRecord; for(i=0; i<pTab->nCol; i++){ int nIn = sessionSerialLen(pIn); if( *pIn==0 ){ *pOut++ = 0; }else if( pTab->abPK[i]==0 ){ *pOut++ = 0xFF; }else{ memcpy(pOut, pIn, nIn); pOut += nIn; } pIn += nIn; } pNew->nRecord = pOut - pNew->aRecord; } }else if( bRebase ){ if( pExist->op==SQLITE_DELETE && pExist->bIndirect ){ *ppNew = pExist; }else{ int nByte = nRec + pExist->nRecord + sizeof(SessionChange); pNew = (SessionChange*)sqlite3_malloc(nByte); if( pNew==0 ){ rc = SQLITE_NOMEM; }else{ int i; u8 *a1 = pExist->aRecord; u8 *a2 = aRec; u8 *pOut; memset(pNew, 0, nByte); pNew->bIndirect = bIndirect || pExist->bIndirect; pNew->op = op2; pOut = pNew->aRecord = (u8*)&pNew[1]; for(i=0; i<pTab->nCol; i++){ int n1 = sessionSerialLen(a1); int n2 = sessionSerialLen(a2); if( *a1==0xFF || (pTab->abPK[i]==0 && bIndirect) ){ *pOut++ = 0xFF; }else if( *a2==0 ){ memcpy(pOut, a1, n1); pOut += n1; }else{ memcpy(pOut, a2, n2); pOut += n2; } a1 += n1; a2 += n2; } pNew->nRecord = pOut - pNew->aRecord; } sqlite3_free(pExist); } }else{ int op1 = pExist->op; /* ** op1=INSERT, op2=INSERT -> Unsupported. Discard op2. ** op1=INSERT, op2=UPDATE -> INSERT. ** op1=INSERT, op2=DELETE -> (none) |
︙ | ︙ | |||
184388 184389 184390 184391 184392 184393 184394 | pNew->nRecord = (int)(aCsr - pNew->aRecord); } sqlite3_free(pExist); } } *ppNew = pNew; | | | > < | | 185073 185074 185075 185076 185077 185078 185079 185080 185081 185082 185083 185084 185085 185086 185087 185088 185089 185090 185091 185092 185093 185094 185095 185096 185097 185098 185099 185100 185101 185102 185103 185104 | pNew->nRecord = (int)(aCsr - pNew->aRecord); } sqlite3_free(pExist); } } *ppNew = pNew; return rc; } /* ** Add all changes in the changeset traversed by the iterator passed as ** the first argument to the changegroup hash tables. */ static int sessionChangesetToHash( sqlite3_changeset_iter *pIter, /* Iterator to read from */ sqlite3_changegroup *pGrp, /* Changegroup object to add changeset to */ int bRebase /* True if hash table is for rebasing */ ){ u8 *aRec; int nRec; int rc = SQLITE_OK; SessionTable *pTab = 0; while( SQLITE_ROW==sessionChangesetNext(pIter, &aRec, &nRec, 0) ){ const char *zNew; int nCol; int op; int iHash; int bIndirect; SessionChange *pChange; SessionChange *pExist = 0; |
︙ | ︙ | |||
184485 184486 184487 184488 184489 184490 184491 | pExist = *pp; *pp = (*pp)->pNext; pTab->nEntry--; break; } } | | | 185170 185171 185172 185173 185174 185175 185176 185177 185178 185179 185180 185181 185182 185183 185184 | pExist = *pp; *pp = (*pp)->pNext; pTab->nEntry--; break; } } rc = sessionChangeMerge(pTab, bRebase, pIter->bPatchset, pExist, op, bIndirect, aRec, nRec, &pChange ); if( rc ) break; if( pChange ){ pChange->pNext = pTab->apChange[iHash]; pTab->apChange[iHash] = pChange; pTab->nEntry++; |
︙ | ︙ | |||
184593 184594 184595 184596 184597 184598 184599 | */ SQLITE_API int sqlite3changegroup_add(sqlite3_changegroup *pGrp, int nData, void *pData){ sqlite3_changeset_iter *pIter; /* Iterator opened on pData/nData */ int rc; /* Return code */ rc = sqlite3changeset_start(&pIter, nData, pData); if( rc==SQLITE_OK ){ | | | 185278 185279 185280 185281 185282 185283 185284 185285 185286 185287 185288 185289 185290 185291 185292 | */ SQLITE_API int sqlite3changegroup_add(sqlite3_changegroup *pGrp, int nData, void *pData){ sqlite3_changeset_iter *pIter; /* Iterator opened on pData/nData */ int rc; /* Return code */ rc = sqlite3changeset_start(&pIter, nData, pData); if( rc==SQLITE_OK ){ rc = sessionChangesetToHash(pIter, pGrp, 0); } sqlite3changeset_finalize(pIter); return rc; } /* ** Obtain a buffer containing a changeset representing the concatenation |
︙ | ︙ | |||
184624 184625 184626 184627 184628 184629 184630 | void *pIn ){ sqlite3_changeset_iter *pIter; /* Iterator opened on pData/nData */ int rc; /* Return code */ rc = sqlite3changeset_start_strm(&pIter, xInput, pIn); if( rc==SQLITE_OK ){ | | | 185309 185310 185311 185312 185313 185314 185315 185316 185317 185318 185319 185320 185321 185322 185323 | void *pIn ){ sqlite3_changeset_iter *pIter; /* Iterator opened on pData/nData */ int rc; /* Return code */ rc = sqlite3changeset_start_strm(&pIter, xInput, pIn); if( rc==SQLITE_OK ){ rc = sessionChangesetToHash(pIter, pGrp, 0); } sqlite3changeset_finalize(pIter); return rc; } /* ** Streaming versions of changegroup_output(). |
︙ | ︙ | |||
184708 184709 184710 184711 184712 184713 184714 184715 184716 184717 184718 184719 184720 184721 | if( rc==SQLITE_OK ){ rc = sqlite3changegroup_output_strm(pGrp, xOutput, pOut); } sqlite3changegroup_delete(pGrp); return rc; } #endif /* SQLITE_ENABLE_SESSION && SQLITE_ENABLE_PREUPDATE_HOOK */ /************** End of sqlite3session.c **************************************/ /************** Begin file json1.c *******************************************/ /* ** 2015-08-12 | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 185393 185394 185395 185396 185397 185398 185399 185400 185401 185402 185403 185404 185405 185406 185407 185408 185409 185410 185411 185412 185413 185414 185415 185416 185417 185418 185419 185420 185421 185422 185423 185424 185425 185426 185427 185428 185429 185430 185431 185432 185433 185434 185435 185436 185437 185438 185439 185440 185441 185442 185443 185444 185445 185446 185447 185448 185449 185450 185451 185452 185453 185454 185455 185456 185457 185458 185459 185460 185461 185462 185463 185464 185465 185466 185467 185468 185469 185470 185471 185472 185473 185474 185475 185476 185477 185478 185479 185480 185481 185482 185483 185484 185485 185486 185487 185488 185489 185490 185491 185492 185493 185494 185495 185496 185497 185498 185499 185500 185501 185502 185503 185504 185505 185506 185507 185508 185509 185510 185511 185512 185513 185514 185515 185516 185517 185518 185519 185520 185521 185522 185523 185524 185525 185526 185527 185528 185529 185530 185531 185532 185533 185534 185535 185536 185537 185538 185539 185540 185541 185542 185543 185544 185545 185546 185547 185548 185549 185550 185551 185552 185553 185554 185555 185556 185557 185558 185559 185560 185561 185562 185563 185564 185565 185566 185567 185568 185569 185570 185571 185572 185573 185574 185575 185576 185577 185578 185579 185580 185581 185582 185583 185584 185585 185586 185587 185588 185589 185590 185591 185592 185593 185594 185595 185596 185597 185598 185599 185600 185601 185602 185603 185604 185605 185606 185607 185608 185609 185610 185611 185612 185613 185614 185615 185616 185617 185618 185619 185620 185621 185622 185623 185624 185625 185626 185627 185628 185629 185630 185631 185632 185633 185634 185635 185636 185637 185638 185639 185640 185641 185642 185643 185644 185645 185646 185647 185648 185649 185650 185651 185652 185653 185654 185655 185656 185657 185658 185659 185660 185661 185662 185663 185664 185665 185666 185667 185668 185669 185670 185671 185672 185673 185674 185675 185676 185677 185678 185679 185680 185681 185682 185683 185684 185685 185686 185687 185688 185689 185690 185691 185692 185693 185694 185695 185696 185697 185698 185699 185700 185701 185702 185703 185704 185705 185706 185707 185708 185709 185710 185711 185712 185713 185714 185715 185716 185717 185718 185719 185720 185721 185722 185723 185724 185725 185726 185727 185728 185729 185730 185731 185732 185733 185734 185735 185736 185737 185738 185739 185740 185741 185742 185743 185744 185745 185746 185747 185748 185749 | if( rc==SQLITE_OK ){ rc = sqlite3changegroup_output_strm(pGrp, xOutput, pOut); } sqlite3changegroup_delete(pGrp); return rc; } /* ** Changeset rebaser handle. */ struct sqlite3_rebaser { sqlite3_changegroup grp; /* Hash table */ }; /* ** Buffers a1 and a2 must both contain a sessions module record nCol ** fields in size. This function appends an nCol sessions module ** record to buffer pBuf that is a copy of a1, except that for ** each field that is undefined in a1[], swap in the field from a2[]. */ static void sessionAppendRecordMerge( SessionBuffer *pBuf, /* Buffer to append to */ int nCol, /* Number of columns in each record */ u8 *a1, int n1, /* Record 1 */ u8 *a2, int n2, /* Record 2 */ int *pRc /* IN/OUT: error code */ ){ sessionBufferGrow(pBuf, n1+n2, pRc); if( *pRc==SQLITE_OK ){ int i; u8 *pOut = &pBuf->aBuf[pBuf->nBuf]; for(i=0; i<nCol; i++){ int nn1 = sessionSerialLen(a1); int nn2 = sessionSerialLen(a2); if( *a1==0 || *a1==0xFF ){ memcpy(pOut, a2, nn2); pOut += nn2; }else{ memcpy(pOut, a1, nn1); pOut += nn1; } a1 += nn1; a2 += nn2; } pBuf->nBuf = pOut-pBuf->aBuf; assert( pBuf->nBuf<=pBuf->nAlloc ); } } /* ** This function is called when rebasing a local UPDATE change against one ** or more remote UPDATE changes. The aRec/nRec buffer contains the current ** old.* and new.* records for the change. The rebase buffer (a single ** record) is in aChange/nChange. The rebased change is appended to buffer ** pBuf. ** ** Rebasing the UPDATE involves: ** ** * Removing any changes to fields for which the corresponding field ** in the rebase buffer is set to "replaced" (type 0xFF). If this ** means the UPDATE change updates no fields, nothing is appended ** to the output buffer. ** ** * For each field modified by the local change for which the ** corresponding field in the rebase buffer is not "undefined" (0x00) ** or "replaced" (0xFF), the old.* value is replaced by the value ** in the rebase buffer. */ static void sessionAppendPartialUpdate( SessionBuffer *pBuf, /* Append record here */ sqlite3_changeset_iter *pIter, /* Iterator pointed at local change */ u8 *aRec, int nRec, /* Local change */ u8 *aChange, int nChange, /* Record to rebase against */ int *pRc /* IN/OUT: Return Code */ ){ sessionBufferGrow(pBuf, 2+nRec+nChange, pRc); if( *pRc==SQLITE_OK ){ int bData = 0; u8 *pOut = &pBuf->aBuf[pBuf->nBuf]; int i; u8 *a1 = aRec; u8 *a2 = aChange; *pOut++ = SQLITE_UPDATE; *pOut++ = pIter->bIndirect; for(i=0; i<pIter->nCol; i++){ int n1 = sessionSerialLen(a1); int n2 = sessionSerialLen(a2); if( pIter->abPK[i] || a2[0]==0 ){ if( !pIter->abPK[i] ) bData = 1; memcpy(pOut, a1, n1); pOut += n1; }else if( a2[0]!=0xFF ){ bData = 1; memcpy(pOut, a2, n2); pOut += n2; }else{ *pOut++ = '\0'; } a1 += n1; a2 += n2; } if( bData ){ a2 = aChange; for(i=0; i<pIter->nCol; i++){ int n1 = sessionSerialLen(a1); int n2 = sessionSerialLen(a2); if( pIter->abPK[i] || a2[0]!=0xFF ){ memcpy(pOut, a1, n1); pOut += n1; }else{ *pOut++ = '\0'; } a1 += n1; a2 += n2; } pBuf->nBuf = (pOut - pBuf->aBuf); } } } /* ** pIter is configured to iterate through a changeset. This function rebases ** that changeset according to the current configuration of the rebaser ** object passed as the first argument. If no error occurs and argument xOutput ** is not NULL, then the changeset is returned to the caller by invoking ** xOutput zero or more times and SQLITE_OK returned. Or, if xOutput is NULL, ** then (*ppOut) is set to point to a buffer containing the rebased changeset ** before this function returns. In this case (*pnOut) is set to the size of ** the buffer in bytes. It is the responsibility of the caller to eventually ** free the (*ppOut) buffer using sqlite3_free(). ** ** If an error occurs, an SQLite error code is returned. If ppOut and ** pnOut are not NULL, then the two output parameters are set to 0 before ** returning. */ static int sessionRebase( sqlite3_rebaser *p, /* Rebaser hash table */ sqlite3_changeset_iter *pIter, /* Input data */ int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut, /* Context for xOutput callback */ int *pnOut, /* OUT: Number of bytes in output changeset */ void **ppOut /* OUT: Inverse of pChangeset */ ){ int rc = SQLITE_OK; u8 *aRec = 0; int nRec = 0; int bNew = 0; SessionTable *pTab = 0; SessionBuffer sOut = {0,0,0}; while( SQLITE_ROW==sessionChangesetNext(pIter, &aRec, &nRec, &bNew) ){ SessionChange *pChange = 0; int bDone = 0; if( bNew ){ const char *zTab = pIter->zTab; for(pTab=p->grp.pList; pTab; pTab=pTab->pNext){ if( 0==sqlite3_stricmp(pTab->zName, zTab) ) break; } bNew = 0; /* A patchset may not be rebased */ if( pIter->bPatchset ){ rc = SQLITE_ERROR; } /* Append a table header to the output for this new table */ sessionAppendByte(&sOut, pIter->bPatchset ? 'P' : 'T', &rc); sessionAppendVarint(&sOut, pIter->nCol, &rc); sessionAppendBlob(&sOut, pIter->abPK, pIter->nCol, &rc); sessionAppendBlob(&sOut,(u8*)pIter->zTab,(int)strlen(pIter->zTab)+1,&rc); } if( pTab && rc==SQLITE_OK ){ int iHash = sessionChangeHash(pTab, 0, aRec, pTab->nChange); for(pChange=pTab->apChange[iHash]; pChange; pChange=pChange->pNext){ if( sessionChangeEqual(pTab, 0, aRec, 0, pChange->aRecord) ){ break; } } } if( pChange ){ assert( pChange->op==SQLITE_DELETE || pChange->op==SQLITE_INSERT ); switch( pIter->op ){ case SQLITE_INSERT: if( pChange->op==SQLITE_INSERT ){ bDone = 1; if( pChange->bIndirect==0 ){ sessionAppendByte(&sOut, SQLITE_UPDATE, &rc); sessionAppendByte(&sOut, pIter->bIndirect, &rc); sessionAppendBlob(&sOut, pChange->aRecord, pChange->nRecord, &rc); sessionAppendBlob(&sOut, aRec, nRec, &rc); } } break; case SQLITE_UPDATE: bDone = 1; if( pChange->op==SQLITE_DELETE ){ if( pChange->bIndirect==0 ){ u8 *pCsr = aRec; sessionSkipRecord(&pCsr, pIter->nCol); sessionAppendByte(&sOut, SQLITE_INSERT, &rc); sessionAppendByte(&sOut, pIter->bIndirect, &rc); sessionAppendRecordMerge(&sOut, pIter->nCol, pCsr, nRec-(pCsr-aRec), pChange->aRecord, pChange->nRecord, &rc ); } }else{ sessionAppendPartialUpdate(&sOut, pIter, aRec, nRec, pChange->aRecord, pChange->nRecord, &rc ); } break; default: assert( pIter->op==SQLITE_DELETE ); bDone = 1; if( pChange->op==SQLITE_INSERT ){ sessionAppendByte(&sOut, SQLITE_DELETE, &rc); sessionAppendByte(&sOut, pIter->bIndirect, &rc); sessionAppendRecordMerge(&sOut, pIter->nCol, pChange->aRecord, pChange->nRecord, aRec, nRec, &rc ); } break; } } if( bDone==0 ){ sessionAppendByte(&sOut, pIter->op, &rc); sessionAppendByte(&sOut, pIter->bIndirect, &rc); sessionAppendBlob(&sOut, aRec, nRec, &rc); } if( rc==SQLITE_OK && xOutput && sOut.nBuf>SESSIONS_STRM_CHUNK_SIZE ){ rc = xOutput(pOut, sOut.aBuf, sOut.nBuf); sOut.nBuf = 0; } if( rc ) break; } if( rc!=SQLITE_OK ){ sqlite3_free(sOut.aBuf); memset(&sOut, 0, sizeof(sOut)); } if( rc==SQLITE_OK ){ if( xOutput ){ if( sOut.nBuf>0 ){ rc = xOutput(pOut, sOut.aBuf, sOut.nBuf); } }else{ *ppOut = (void*)sOut.aBuf; *pnOut = sOut.nBuf; sOut.aBuf = 0; } } sqlite3_free(sOut.aBuf); return rc; } /* ** Create a new rebaser object. */ SQLITE_API int sqlite3rebaser_create(sqlite3_rebaser **ppNew){ int rc = SQLITE_OK; sqlite3_rebaser *pNew; pNew = sqlite3_malloc(sizeof(sqlite3_rebaser)); if( pNew==0 ){ rc = SQLITE_NOMEM; }else{ memset(pNew, 0, sizeof(sqlite3_rebaser)); } *ppNew = pNew; return rc; } /* ** Call this one or more times to configure a rebaser. */ SQLITE_API int sqlite3rebaser_configure( sqlite3_rebaser *p, int nRebase, const void *pRebase ){ sqlite3_changeset_iter *pIter = 0; /* Iterator opened on pData/nData */ int rc; /* Return code */ rc = sqlite3changeset_start(&pIter, nRebase, (void*)pRebase); if( rc==SQLITE_OK ){ rc = sessionChangesetToHash(pIter, &p->grp, 1); } sqlite3changeset_finalize(pIter); return rc; } /* ** Rebase a changeset according to current rebaser configuration */ SQLITE_API int sqlite3rebaser_rebase( sqlite3_rebaser *p, int nIn, const void *pIn, int *pnOut, void **ppOut ){ sqlite3_changeset_iter *pIter = 0; /* Iterator to skip through input */ int rc = sqlite3changeset_start(&pIter, nIn, (void*)pIn); if( rc==SQLITE_OK ){ rc = sessionRebase(p, pIter, 0, 0, pnOut, ppOut); sqlite3changeset_finalize(pIter); } return rc; } /* ** Rebase a changeset according to current rebaser configuration */ SQLITE_API int sqlite3rebaser_rebase_strm( sqlite3_rebaser *p, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ){ sqlite3_changeset_iter *pIter = 0; /* Iterator to skip through input */ int rc = sqlite3changeset_start_strm(&pIter, xInput, pIn); if( rc==SQLITE_OK ){ rc = sessionRebase(p, pIter, xOutput, pOut, 0, 0); sqlite3changeset_finalize(pIter); } return rc; } /* ** Destroy a rebaser object */ SQLITE_API void sqlite3rebaser_delete(sqlite3_rebaser *p){ if( p ){ sessionDeleteTable(p->grp.pList); sqlite3_free(p); } } #endif /* SQLITE_ENABLE_SESSION && SQLITE_ENABLE_PREUPDATE_HOOK */ /************** End of sqlite3session.c **************************************/ /************** Begin file json1.c *******************************************/ /* ** 2015-08-12 |
︙ | ︙ | |||
204527 204528 204529 204530 204531 204532 204533 | static void fts5SourceIdFunc( sqlite3_context *pCtx, /* Function call context */ int nArg, /* Number of args */ sqlite3_value **apUnused /* Function arguments */ ){ assert( nArg==0 ); UNUSED_PARAM2(nArg, apUnused); | | | 205555 205556 205557 205558 205559 205560 205561 205562 205563 205564 205565 205566 205567 205568 205569 | static void fts5SourceIdFunc( sqlite3_context *pCtx, /* Function call context */ int nArg, /* Number of args */ sqlite3_value **apUnused /* Function arguments */ ){ assert( nArg==0 ); UNUSED_PARAM2(nArg, apUnused); sqlite3_result_text(pCtx, "fts5: 2018-04-02 11:04:16 736b53f57f70b23172c30880186dce7ad9baa3b74e3838cae5847cffb98f5cd2", -1, SQLITE_TRANSIENT); } static int fts5Init(sqlite3 *db){ static const sqlite3_module fts5Mod = { /* iVersion */ 2, /* xCreate */ fts5CreateMethod, /* xConnect */ fts5ConnectMethod, |
︙ | ︙ | |||
208797 208798 208799 208800 208801 208802 208803 | #endif return rc; } #endif /* SQLITE_CORE */ #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_STMTVTAB) */ /************** End of stmt.c ************************************************/ | | | | 209825 209826 209827 209828 209829 209830 209831 209832 209833 209834 209835 209836 209837 209838 | #endif return rc; } #endif /* SQLITE_CORE */ #endif /* !defined(SQLITE_CORE) || defined(SQLITE_ENABLE_STMTVTAB) */ /************** End of stmt.c ************************************************/ #if __LINE__!=209832 #undef SQLITE_SOURCE_ID #define SQLITE_SOURCE_ID "2018-04-02 11:04:16 736b53f57f70b23172c30880186dce7ad9baa3b74e3838cae5847cffb98falt2" #endif /* Return the source-id for this library */ SQLITE_API const char *sqlite3_sourceid(void){ return SQLITE_SOURCE_ID; } /************************** End of sqlite3.c ******************************/ |
Changes to SQLite.Interop/src/core/sqlite3.h.
︙ | ︙ | |||
121 122 123 124 125 126 127 | ** ** See also: [sqlite3_libversion()], ** [sqlite3_libversion_number()], [sqlite3_sourceid()], ** [sqlite_version()] and [sqlite_source_id()]. */ #define SQLITE_VERSION "3.23.0" #define SQLITE_VERSION_NUMBER 3023000 | | | 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 | ** ** See also: [sqlite3_libversion()], ** [sqlite3_libversion_number()], [sqlite3_sourceid()], ** [sqlite_version()] and [sqlite_source_id()]. */ #define SQLITE_VERSION "3.23.0" #define SQLITE_VERSION_NUMBER 3023000 #define SQLITE_SOURCE_ID "2018-04-02 11:04:16 736b53f57f70b23172c30880186dce7ad9baa3b74e3838cae5847cffb98f5cd2" /* ** CAPI3REF: Run-Time Library Version Numbers ** KEYWORDS: sqlite3_version sqlite3_sourceid ** ** These interfaces provide the same information as the [SQLITE_VERSION], ** [SQLITE_VERSION_NUMBER], and [SQLITE_SOURCE_ID] C preprocessor macros |
︙ | ︙ | |||
1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 | ** The [SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE] opcode causes all write ** operations since the previous successful call to ** [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE] to be rolled back. ** ^This file control takes the file descriptor out of batch write mode ** so that all subsequent write operations are independent. ** ^SQLite will never invoke SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE without ** a prior successful call to [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE]. ** </ul> */ #define SQLITE_FCNTL_LOCKSTATE 1 #define SQLITE_FCNTL_GET_LOCKPROXYFILE 2 #define SQLITE_FCNTL_SET_LOCKPROXYFILE 3 #define SQLITE_FCNTL_LAST_ERRNO 4 #define SQLITE_FCNTL_SIZE_HINT 5 | > > > > > > | 1060 1061 1062 1063 1064 1065 1066 1067 1068 1069 1070 1071 1072 1073 1074 1075 1076 1077 1078 1079 | ** The [SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE] opcode causes all write ** operations since the previous successful call to ** [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE] to be rolled back. ** ^This file control takes the file descriptor out of batch write mode ** so that all subsequent write operations are independent. ** ^SQLite will never invoke SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE without ** a prior successful call to [SQLITE_FCNTL_BEGIN_ATOMIC_WRITE]. ** ** <li>[[SQLITE_FCNTL_LOCK_TIMEOUT]] ** The [SQLITE_FCNTL_LOCK_TIMEOUT] opcode causes attempts to obtain ** a file lock using the xLock or xShmLock methods of the VFS to wait ** for up to M milliseconds before failing, where M is the single ** unsigned integer parameter. ** </ul> */ #define SQLITE_FCNTL_LOCKSTATE 1 #define SQLITE_FCNTL_GET_LOCKPROXYFILE 2 #define SQLITE_FCNTL_SET_LOCKPROXYFILE 3 #define SQLITE_FCNTL_LAST_ERRNO 4 #define SQLITE_FCNTL_SIZE_HINT 5 |
︙ | ︙ | |||
1094 1095 1096 1097 1098 1099 1100 1101 1102 1103 1104 1105 1106 1107 | #define SQLITE_FCNTL_VFS_POINTER 27 #define SQLITE_FCNTL_JOURNAL_POINTER 28 #define SQLITE_FCNTL_WIN32_GET_HANDLE 29 #define SQLITE_FCNTL_PDB 30 #define SQLITE_FCNTL_BEGIN_ATOMIC_WRITE 31 #define SQLITE_FCNTL_COMMIT_ATOMIC_WRITE 32 #define SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE 33 /* deprecated names */ #define SQLITE_GET_LOCKPROXYFILE SQLITE_FCNTL_GET_LOCKPROXYFILE #define SQLITE_SET_LOCKPROXYFILE SQLITE_FCNTL_SET_LOCKPROXYFILE #define SQLITE_LAST_ERRNO SQLITE_FCNTL_LAST_ERRNO | > | 1100 1101 1102 1103 1104 1105 1106 1107 1108 1109 1110 1111 1112 1113 1114 | #define SQLITE_FCNTL_VFS_POINTER 27 #define SQLITE_FCNTL_JOURNAL_POINTER 28 #define SQLITE_FCNTL_WIN32_GET_HANDLE 29 #define SQLITE_FCNTL_PDB 30 #define SQLITE_FCNTL_BEGIN_ATOMIC_WRITE 31 #define SQLITE_FCNTL_COMMIT_ATOMIC_WRITE 32 #define SQLITE_FCNTL_ROLLBACK_ATOMIC_WRITE 33 #define SQLITE_FCNTL_LOCK_TIMEOUT 34 /* deprecated names */ #define SQLITE_GET_LOCKPROXYFILE SQLITE_FCNTL_GET_LOCKPROXYFILE #define SQLITE_SET_LOCKPROXYFILE SQLITE_FCNTL_SET_LOCKPROXYFILE #define SQLITE_LAST_ERRNO SQLITE_FCNTL_LAST_ERRNO |
︙ | ︙ | |||
2050 2051 2052 2053 2054 2055 2056 | ** ** <dt>SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE</dt> ** <dd> Usually, when a database in wal mode is closed or detached from a ** database handle, SQLite checks if this will mean that there are now no ** connections at all to the database. If so, it performs a checkpoint ** operation before closing the connection. This option may be used to ** override this behaviour. The first parameter passed to this operation | | > | > > > > > > > | > | 2057 2058 2059 2060 2061 2062 2063 2064 2065 2066 2067 2068 2069 2070 2071 2072 2073 2074 2075 2076 2077 2078 2079 2080 2081 2082 2083 2084 2085 2086 2087 2088 2089 2090 2091 2092 2093 2094 2095 2096 2097 2098 2099 2100 | ** ** <dt>SQLITE_DBCONFIG_NO_CKPT_ON_CLOSE</dt> ** <dd> Usually, when a database in wal mode is closed or detached from a ** database handle, SQLite checks if this will mean that there are now no ** connections at all to the database. If so, it performs a checkpoint ** operation before closing the connection. This option may be used to ** override this behaviour. The first parameter passed to this operation ** is an integer - positive to disable checkpoints-on-close, or zero (the ** default) to enable them, and negative to leave the setting unchanged. ** The second parameter is a pointer to an integer ** into which is written 0 or 1 to indicate whether checkpoints-on-close ** have been disabled - 0 if they are not disabled, 1 if they are. ** </dd> ** ** <dt>SQLITE_DBCONFIG_ENABLE_QPSG</dt> ** <dd>^(The SQLITE_DBCONFIG_ENABLE_QPSG option activates or deactivates ** the [query planner stability guarantee] (QPSG). When the QPSG is active, ** a single SQL query statement will always use the same algorithm regardless ** of values of [bound parameters].)^ The QPSG disables some query optimizations ** that look at the values of bound parameters, which can make some queries ** slower. But the QPSG has the advantage of more predictable behavior. With ** the QPSG active, SQLite will always use the same query plan in the field as ** was used during testing in the lab. ** The first argument to this setting is an integer which is 0 to disable ** the QPSG, positive to enable QPSG, or negative to leave the setting ** unchanged. The second parameter is a pointer to an integer into which ** is written 0 or 1 to indicate whether the QPSG is disabled or enabled ** following this call. ** </dd> ** ** <dt>SQLITE_DBCONFIG_TRIGGER_EQP</dt> ** <dd> By default, the output of EXPLAIN QUERY PLAN commands does not ** include output for any operations performed by trigger programs. This ** option is used to set or clear (the default) a flag that governs this ** behavior. The first parameter passed to this operation is an integer - ** positive to enable output for trigger programs, or zero to disable it, ** or negative to leave the setting unchanged. ** The second parameter is a pointer to an integer into which is written ** 0 or 1 to indicate whether output-for-triggers has been disabled - 0 if ** it is not disabled, 1 if it is. ** </dd> ** </dl> */ #define SQLITE_DBCONFIG_MAINDBNAME 1000 /* const char* */ |
︙ | ︙ | |||
7225 7226 7227 7228 7229 7230 7231 7232 7233 7234 7235 7236 7237 7238 7239 7240 7241 7242 7243 7244 7245 7246 7247 7248 7249 7250 7251 | ** wal file in wal mode databases, or the number of pages written to the ** database file in rollback mode databases. Any pages written as part of ** transaction rollback or database recovery operations are not included. ** If an IO or other error occurs while writing a page to disk, the effect ** on subsequent SQLITE_DBSTATUS_CACHE_WRITE requests is undefined.)^ ^The ** highwater mark associated with SQLITE_DBSTATUS_CACHE_WRITE is always 0. ** </dd> ** ** [[SQLITE_DBSTATUS_DEFERRED_FKS]] ^(<dt>SQLITE_DBSTATUS_DEFERRED_FKS</dt> ** <dd>This parameter returns zero for the current value if and only if ** all foreign key constraints (deferred or immediate) have been ** resolved.)^ ^The highwater mark is always 0. ** </dd> ** </dl> */ #define SQLITE_DBSTATUS_LOOKASIDE_USED 0 #define SQLITE_DBSTATUS_CACHE_USED 1 #define SQLITE_DBSTATUS_SCHEMA_USED 2 #define SQLITE_DBSTATUS_STMT_USED 3 #define SQLITE_DBSTATUS_LOOKASIDE_HIT 4 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_SIZE 5 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_FULL 6 #define SQLITE_DBSTATUS_CACHE_HIT 7 #define SQLITE_DBSTATUS_CACHE_MISS 8 #define SQLITE_DBSTATUS_CACHE_WRITE 9 #define SQLITE_DBSTATUS_DEFERRED_FKS 10 #define SQLITE_DBSTATUS_CACHE_USED_SHARED 11 | > > > > > > > > > > | | 7241 7242 7243 7244 7245 7246 7247 7248 7249 7250 7251 7252 7253 7254 7255 7256 7257 7258 7259 7260 7261 7262 7263 7264 7265 7266 7267 7268 7269 7270 7271 7272 7273 7274 7275 7276 7277 7278 7279 7280 7281 7282 7283 7284 7285 | ** wal file in wal mode databases, or the number of pages written to the ** database file in rollback mode databases. Any pages written as part of ** transaction rollback or database recovery operations are not included. ** If an IO or other error occurs while writing a page to disk, the effect ** on subsequent SQLITE_DBSTATUS_CACHE_WRITE requests is undefined.)^ ^The ** highwater mark associated with SQLITE_DBSTATUS_CACHE_WRITE is always 0. ** </dd> ** ** [[SQLITE_DBSTATUS_CACHE_SPILL]] ^(<dt>SQLITE_DBSTATUS_CACHE_SPILL</dt> ** <dd>This parameter returns the number of dirty cache entries that have ** been written to disk in the middle of a transaction due to the page ** cache overflowing. Transactions are more efficient if they are written ** to disk all at once. When pages spill mid-transaction, that introduces ** additional overhead. This parameter can be used help identify ** inefficiencies that can be resolve by increasing the cache size. ** </dd> ** ** [[SQLITE_DBSTATUS_DEFERRED_FKS]] ^(<dt>SQLITE_DBSTATUS_DEFERRED_FKS</dt> ** <dd>This parameter returns zero for the current value if and only if ** all foreign key constraints (deferred or immediate) have been ** resolved.)^ ^The highwater mark is always 0. ** </dd> ** </dl> */ #define SQLITE_DBSTATUS_LOOKASIDE_USED 0 #define SQLITE_DBSTATUS_CACHE_USED 1 #define SQLITE_DBSTATUS_SCHEMA_USED 2 #define SQLITE_DBSTATUS_STMT_USED 3 #define SQLITE_DBSTATUS_LOOKASIDE_HIT 4 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_SIZE 5 #define SQLITE_DBSTATUS_LOOKASIDE_MISS_FULL 6 #define SQLITE_DBSTATUS_CACHE_HIT 7 #define SQLITE_DBSTATUS_CACHE_MISS 8 #define SQLITE_DBSTATUS_CACHE_WRITE 9 #define SQLITE_DBSTATUS_DEFERRED_FKS 10 #define SQLITE_DBSTATUS_CACHE_USED_SHARED 11 #define SQLITE_DBSTATUS_CACHE_SPILL 12 #define SQLITE_DBSTATUS_MAX 12 /* Largest defined DBSTATUS */ /* ** CAPI3REF: Prepared Statement Status ** METHOD: sqlite3_stmt ** ** ^(Each prepared statement maintains various |
︙ | ︙ | |||
8788 8789 8790 8791 8792 8793 8794 | */ #define SQLITE_SERIALIZE_NOCOPY 0x001 /* Do no memory allocations */ /* ** CAPI3REF: Deserialize a database ** ** The sqlite3_deserialize(D,S,P,N,M,F) interface causes the | | | 8814 8815 8816 8817 8818 8819 8820 8821 8822 8823 8824 8825 8826 8827 8828 | */ #define SQLITE_SERIALIZE_NOCOPY 0x001 /* Do no memory allocations */ /* ** CAPI3REF: Deserialize a database ** ** The sqlite3_deserialize(D,S,P,N,M,F) interface causes the ** [database connection] D to disconnect from database S and then ** reopen S as an in-memory database based on the serialization contained ** in P. The serialized database P is N bytes in size. M is the size of ** the buffer P, which might be larger than N. If M is larger than N, and ** the SQLITE_DESERIALIZE_READONLY bit is not set in F, then SQLite is ** permitted to add content to the in-memory database as long as the total ** size does not exceed M bytes. ** |
︙ | ︙ | |||
8835 8836 8837 8838 8839 8840 8841 | ** The SQLITE_DESERIALIZE_FREEONCLOSE means that the database serialization ** in the P argument is held in memory obtained from [sqlite3_malloc64()] ** and that SQLite should take ownership of this memory and automatically ** free it when it has finished using it. Without this flag, the caller ** is resposible for freeing any dynamically allocated memory. ** ** The SQLITE_DESERIALIZE_RESIZEABLE flag means that SQLite is allowed to | | | 8861 8862 8863 8864 8865 8866 8867 8868 8869 8870 8871 8872 8873 8874 8875 | ** The SQLITE_DESERIALIZE_FREEONCLOSE means that the database serialization ** in the P argument is held in memory obtained from [sqlite3_malloc64()] ** and that SQLite should take ownership of this memory and automatically ** free it when it has finished using it. Without this flag, the caller ** is resposible for freeing any dynamically allocated memory. ** ** The SQLITE_DESERIALIZE_RESIZEABLE flag means that SQLite is allowed to ** grow the size of the database using calls to [sqlite3_realloc64()]. This ** flag should only be used if SQLITE_DESERIALIZE_FREEONCLOSE is also used. ** Without this flag, the deserialized database cannot increase in size beyond ** the number of bytes specified by the M parameter. ** ** The SQLITE_DESERIALIZE_READONLY flag means that the deserialized database ** should be treated as read-only. */ |
︙ | ︙ | |||
9929 9930 9931 9932 9933 9934 9935 | ** DESTRUCTOR: sqlite3_changegroup */ SQLITE_API void sqlite3changegroup_delete(sqlite3_changegroup*); /* ** CAPI3REF: Apply A Changeset To A Database ** | | | | | | | | | < | 9955 9956 9957 9958 9959 9960 9961 9962 9963 9964 9965 9966 9967 9968 9969 9970 9971 9972 9973 9974 9975 9976 9977 9978 9979 9980 | ** DESTRUCTOR: sqlite3_changegroup */ SQLITE_API void sqlite3changegroup_delete(sqlite3_changegroup*); /* ** CAPI3REF: Apply A Changeset To A Database ** ** Apply a changeset or patchset to a database. These functions attempt to ** update the "main" database attached to handle db with the changes found in ** the changeset passed via the second and third arguments. ** ** The fourth argument (xFilter) passed to these functions is the "filter ** callback". If it is not NULL, then for each table affected by at least one ** change in the changeset, the filter callback is invoked with ** the table name as the second argument, and a copy of the context pointer ** passed as the sixth argument as the first. If the "filter callback" ** returns zero, then no attempt is made to apply any changes to the table. ** Otherwise, if the return value is non-zero or the xFilter argument to ** is NULL, all changes related to the table are attempted. ** ** For each table that is not excluded by the filter callback, this function ** tests that the target database contains a compatible table. A table is ** considered compatible if all of the following are true: ** ** <ul> ** <li> The table has the same name as the name recorded in the |
︙ | ︙ | |||
9986 9987 9988 9989 9990 9991 9992 | ** actions are taken by sqlite3changeset_apply() depending on the value ** returned by each invocation of the conflict-handler function. Refer to ** the documentation for the three ** [SQLITE_CHANGESET_OMIT|available return values] for details. ** ** <dl> ** <dt>DELETE Changes<dd> | | | 10011 10012 10013 10014 10015 10016 10017 10018 10019 10020 10021 10022 10023 10024 10025 | ** actions are taken by sqlite3changeset_apply() depending on the value ** returned by each invocation of the conflict-handler function. Refer to ** the documentation for the three ** [SQLITE_CHANGESET_OMIT|available return values] for details. ** ** <dl> ** <dt>DELETE Changes<dd> ** For each DELETE change, the function checks if the target database ** contains a row with the same primary key value (or values) as the ** original row values stored in the changeset. If it does, and the values ** stored in all non-primary key columns also match the values stored in ** the changeset the row is deleted from the target database. ** ** If a row with matching primary key values is found, but one or more of ** the non-primary key fields contains a value different from the original |
︙ | ︙ | |||
10031 10032 10033 10034 10035 10036 10037 | ** violation (e.g. NOT NULL or UNIQUE), the conflict handler function is ** invoked with the second argument set to [SQLITE_CHANGESET_CONSTRAINT]. ** This includes the case where the INSERT operation is re-attempted because ** an earlier call to the conflict handler function returned ** [SQLITE_CHANGESET_REPLACE]. ** ** <dt>UPDATE Changes<dd> | | | 10056 10057 10058 10059 10060 10061 10062 10063 10064 10065 10066 10067 10068 10069 10070 | ** violation (e.g. NOT NULL or UNIQUE), the conflict handler function is ** invoked with the second argument set to [SQLITE_CHANGESET_CONSTRAINT]. ** This includes the case where the INSERT operation is re-attempted because ** an earlier call to the conflict handler function returned ** [SQLITE_CHANGESET_REPLACE]. ** ** <dt>UPDATE Changes<dd> ** For each UPDATE change, the function checks if the target database ** contains a row with the same primary key value (or values) as the ** original row values stored in the changeset. If it does, and the values ** stored in all modified non-primary key columns also match the values ** stored in the changeset the row is updated within the target database. ** ** If a row with matching primary key values is found, but one or more of ** the modified non-primary key fields contains a value different from an |
︙ | ︙ | |||
10062 10063 10064 10065 10066 10067 10068 | ** </dl> ** ** It is safe to execute SQL statements, including those that write to the ** table that the callback related to, from within the xConflict callback. ** This can be used to further customize the applications conflict ** resolution strategy. ** | | > > > > > > > > > > > > > > > > > > > > > > > > > > | 10087 10088 10089 10090 10091 10092 10093 10094 10095 10096 10097 10098 10099 10100 10101 10102 10103 10104 10105 10106 10107 10108 10109 10110 10111 10112 10113 10114 10115 10116 10117 10118 10119 10120 10121 10122 10123 10124 10125 10126 10127 10128 10129 10130 10131 10132 10133 10134 10135 10136 10137 10138 10139 10140 10141 10142 10143 10144 10145 10146 | ** </dl> ** ** It is safe to execute SQL statements, including those that write to the ** table that the callback related to, from within the xConflict callback. ** This can be used to further customize the applications conflict ** resolution strategy. ** ** All changes made by these functions are enclosed in a savepoint transaction. ** If any other error (aside from a constraint failure when attempting to ** write to the target database) occurs, then the savepoint transaction is ** rolled back, restoring the target database to its original state, and an ** SQLite error code returned. ** ** If the output parameters (ppRebase) and (pnRebase) are non-NULL and ** the input is a changeset (not a patchset), then sqlite3changeset_apply_v2() ** may set (*ppRebase) to point to a "rebase" that may be used with the ** sqlite3_rebaser APIs buffer before returning. In this case (*pnRebase) ** is set to the size of the buffer in bytes. It is the responsibility of the ** caller to eventually free any such buffer using sqlite3_free(). The buffer ** is only allocated and populated if one or more conflicts were encountered ** while applying the patchset. See comments surrounding the sqlite3_rebaser ** APIs for further details. */ SQLITE_API int sqlite3changeset_apply( sqlite3 *db, /* Apply change to "main" db of this handle */ int nChangeset, /* Size of changeset in bytes */ void *pChangeset, /* Changeset blob */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_apply_v2( sqlite3 *db, /* Apply change to "main" db of this handle */ int nChangeset, /* Size of changeset in bytes */ void *pChangeset, /* Changeset blob */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ); /* ** CAPI3REF: Constants Passed To The Conflict Handler ** ** Values that may be passed as the second argument to a conflict-handler. ** |
︙ | ︙ | |||
10180 10181 10182 10183 10184 10185 10186 10187 10188 10189 10190 10191 10192 10193 | ** and the call to sqlite3changeset_apply() returns SQLITE_ABORT. ** </dl> */ #define SQLITE_CHANGESET_OMIT 0 #define SQLITE_CHANGESET_REPLACE 1 #define SQLITE_CHANGESET_ABORT 2 /* ** CAPI3REF: Streaming Versions of API functions. ** ** The six streaming API xxx_strm() functions serve similar purposes to the ** corresponding non-streaming API functions: ** ** <table border=1 style="margin-left:8ex;margin-right:8ex"> | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | 10231 10232 10233 10234 10235 10236 10237 10238 10239 10240 10241 10242 10243 10244 10245 10246 10247 10248 10249 10250 10251 10252 10253 10254 10255 10256 10257 10258 10259 10260 10261 10262 10263 10264 10265 10266 10267 10268 10269 10270 10271 10272 10273 10274 10275 10276 10277 10278 10279 10280 10281 10282 10283 10284 10285 10286 10287 10288 10289 10290 10291 10292 10293 10294 10295 10296 10297 10298 10299 10300 10301 10302 10303 10304 10305 10306 10307 10308 10309 10310 10311 10312 10313 10314 10315 10316 10317 10318 10319 10320 10321 10322 10323 10324 10325 10326 10327 10328 10329 10330 10331 10332 10333 10334 10335 10336 10337 10338 10339 10340 10341 10342 10343 10344 10345 10346 10347 10348 10349 10350 10351 10352 10353 10354 10355 10356 10357 10358 10359 10360 10361 10362 10363 10364 10365 10366 10367 10368 10369 10370 10371 10372 10373 10374 10375 10376 10377 10378 10379 10380 10381 10382 10383 10384 10385 10386 10387 10388 10389 10390 10391 10392 10393 10394 10395 10396 10397 10398 10399 | ** and the call to sqlite3changeset_apply() returns SQLITE_ABORT. ** </dl> */ #define SQLITE_CHANGESET_OMIT 0 #define SQLITE_CHANGESET_REPLACE 1 #define SQLITE_CHANGESET_ABORT 2 /* ** CAPI3REF: Rebasing changesets ** EXPERIMENTAL ** ** Suppose there is a site hosting a database in state S0. And that ** modifications are made that move that database to state S1 and a ** changeset recorded (the "local" changeset). Then, a changeset based ** on S0 is received from another site (the "remote" changeset) and ** applied to the database. The database is then in state ** (S1+"remote"), where the exact state depends on any conflict ** resolution decisions (OMIT or REPLACE) made while applying "remote". ** Rebasing a changeset is to update it to take those conflict ** resolution decisions into account, so that the same conflicts ** do not have to be resolved elsewhere in the network. ** ** For example, if both the local and remote changesets contain an ** INSERT of the same key on "CREATE TABLE t1(a PRIMARY KEY, b)": ** ** local: INSERT INTO t1 VALUES(1, 'v1'); ** remote: INSERT INTO t1 VALUES(1, 'v2'); ** ** and the conflict resolution is REPLACE, then the INSERT change is ** removed from the local changeset (it was overridden). Or, if the ** conflict resolution was "OMIT", then the local changeset is modified ** to instead contain: ** ** UPDATE t1 SET b = 'v2' WHERE a=1; ** ** Changes within the local changeset are rebased as follows: ** ** <dl> ** <dt>Local INSERT<dd> ** This may only conflict with a remote INSERT. If the conflict ** resolution was OMIT, then add an UPDATE change to the rebased ** changeset. Or, if the conflict resolution was REPLACE, add ** nothing to the rebased changeset. ** ** <dt>Local DELETE<dd> ** This may conflict with a remote UPDATE or DELETE. In both cases the ** only possible resolution is OMIT. If the remote operation was a ** DELETE, then add no change to the rebased changeset. If the remote ** operation was an UPDATE, then the old.* fields of change are updated ** to reflect the new.* values in the UPDATE. ** ** <dt>Local UPDATE<dd> ** This may conflict with a remote UPDATE or DELETE. If it conflicts ** with a DELETE, and the conflict resolution was OMIT, then the update ** is changed into an INSERT. Any undefined values in the new.* record ** from the update change are filled in using the old.* values from ** the conflicting DELETE. Or, if the conflict resolution was REPLACE, ** the UPDATE change is simply omitted from the rebased changeset. ** ** If conflict is with a remote UPDATE and the resolution is OMIT, then ** the old.* values are rebased using the new.* values in the remote ** change. Or, if the resolution is REPLACE, then the change is copied ** into the rebased changeset with updates to columns also updated by ** the conflicting remote UPDATE removed. If this means no columns would ** be updated, the change is omitted. ** </dl> ** ** A local change may be rebased against multiple remote changes ** simultaneously. If a single key is modified by multiple remote ** changesets, they are combined as follows before the local changeset ** is rebased: ** ** <ul> ** <li> If there has been one or more REPLACE resolutions on a ** key, it is rebased according to a REPLACE. ** ** <li> If there have been no REPLACE resolutions on a key, then ** the local changeset is rebased according to the most recent ** of the OMIT resolutions. ** </ul> ** ** Note that conflict resolutions from multiple remote changesets are ** combined on a per-field basis, not per-row. This means that in the ** case of multiple remote UPDATE operations, some fields of a single ** local change may be rebased for REPLACE while others are rebased for ** OMIT. ** ** In order to rebase a local changeset, the remote changeset must first ** be applied to the local database using sqlite3changeset_apply_v2() and ** the buffer of rebase information captured. Then: ** ** <ol> ** <li> An sqlite3_rebaser object is created by calling ** sqlite3rebaser_create(). ** <li> The new object is configured with the rebase buffer obtained from ** sqlite3changeset_apply_v2() by calling sqlite3rebaser_configure(). ** If the local changeset is to be rebased against multiple remote ** changesets, then sqlite3rebaser_configure() should be called ** multiple times, in the same order that the multiple ** sqlite3changeset_apply_v2() calls were made. ** <li> Each local changeset is rebased by calling sqlite3rebaser_rebase(). ** <li> The sqlite3_rebaser object is deleted by calling ** sqlite3rebaser_delete(). ** </ol> */ typedef struct sqlite3_rebaser sqlite3_rebaser; /* ** CAPI3REF: Create a changeset rebaser object. ** EXPERIMENTAL ** ** Allocate a new changeset rebaser object. If successful, set (*ppNew) to ** point to the new object and return SQLITE_OK. Otherwise, if an error ** occurs, return an SQLite error code (e.g. SQLITE_NOMEM) and set (*ppNew) ** to NULL. */ SQLITE_API int sqlite3rebaser_create(sqlite3_rebaser **ppNew); /* ** CAPI3REF: Configure a changeset rebaser object. ** EXPERIMENTAL ** ** Configure the changeset rebaser object to rebase changesets according ** to the conflict resolutions described by buffer pRebase (size nRebase ** bytes), which must have been obtained from a previous call to ** sqlite3changeset_apply_v2(). */ SQLITE_API int sqlite3rebaser_configure( sqlite3_rebaser*, int nRebase, const void *pRebase ); /* ** CAPI3REF: Rebase a changeset ** EXPERIMENTAL ** ** Argument pIn must point to a buffer containing a changeset nIn bytes ** in size. This function allocates and populates a buffer with a copy ** of the changeset rebased rebased according to the configuration of the ** rebaser object passed as the first argument. If successful, (*ppOut) ** is set to point to the new buffer containing the rebased changset and ** (*pnOut) to its size in bytes and SQLITE_OK returned. It is the ** responsibility of the caller to eventually free the new buffer using ** sqlite3_free(). Otherwise, if an error occurs, (*ppOut) and (*pnOut) ** are set to zero and an SQLite error code returned. */ SQLITE_API int sqlite3rebaser_rebase( sqlite3_rebaser*, int nIn, const void *pIn, int *pnOut, void **ppOut ); /* ** CAPI3REF: Delete a changeset rebaser object. ** EXPERIMENTAL ** ** Delete the changeset rebaser object and all associated resources. There ** should be one call to this function for each successful invocation ** of sqlite3rebaser_create(). */ SQLITE_API void sqlite3rebaser_delete(sqlite3_rebaser *p); /* ** CAPI3REF: Streaming Versions of API functions. ** ** The six streaming API xxx_strm() functions serve similar purposes to the ** corresponding non-streaming API functions: ** ** <table border=1 style="margin-left:8ex;margin-right:8ex"> |
︙ | ︙ | |||
10283 10284 10285 10286 10287 10288 10289 10290 10291 10292 10293 10294 10295 10296 | ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_concat_strm( int (*xInputA)(void *pIn, void *pData, int *pnData), void *pInA, int (*xInputB)(void *pIn, void *pData, int *pnData), void *pInB, int (*xOutput)(void *pOut, const void *pData, int nData), | > > > > > > > > > > > > > > > > | 10489 10490 10491 10492 10493 10494 10495 10496 10497 10498 10499 10500 10501 10502 10503 10504 10505 10506 10507 10508 10509 10510 10511 10512 10513 10514 10515 10516 10517 10518 | ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx /* First argument passed to xConflict */ ); SQLITE_API int sqlite3changeset_apply_v2_strm( sqlite3 *db, /* Apply change to "main" db of this handle */ int (*xInput)(void *pIn, void *pData, int *pnData), /* Input function */ void *pIn, /* First arg for xInput */ int(*xFilter)( void *pCtx, /* Copy of sixth arg to _apply() */ const char *zTab /* Table name */ ), int(*xConflict)( void *pCtx, /* Copy of sixth arg to _apply() */ int eConflict, /* DATA, MISSING, CONFLICT, CONSTRAINT */ sqlite3_changeset_iter *p /* Handle describing change and conflict */ ), void *pCtx, /* First argument passed to xConflict */ void **ppRebase, int *pnRebase ); SQLITE_API int sqlite3changeset_concat_strm( int (*xInputA)(void *pIn, void *pData, int *pnData), void *pInA, int (*xInputB)(void *pIn, void *pData, int *pnData), void *pInB, int (*xOutput)(void *pOut, const void *pData, int nData), |
︙ | ︙ | |||
10320 10321 10322 10323 10324 10325 10326 10327 10328 10329 10330 10331 10332 10333 | SQLITE_API int sqlite3changegroup_add_strm(sqlite3_changegroup*, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn ); SQLITE_API int sqlite3changegroup_output_strm(sqlite3_changegroup*, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); /* ** Make sure we can call this stuff from C++. */ #ifdef __cplusplus | > > > > > > > | 10542 10543 10544 10545 10546 10547 10548 10549 10550 10551 10552 10553 10554 10555 10556 10557 10558 10559 10560 10561 10562 | SQLITE_API int sqlite3changegroup_add_strm(sqlite3_changegroup*, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn ); SQLITE_API int sqlite3changegroup_output_strm(sqlite3_changegroup*, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); SQLITE_API int sqlite3rebaser_rebase_strm( sqlite3_rebaser *pRebaser, int (*xInput)(void *pIn, void *pData, int *pnData), void *pIn, int (*xOutput)(void *pOut, const void *pData, int nData), void *pOut ); /* ** Make sure we can call this stuff from C++. */ #ifdef __cplusplus |
︙ | ︙ |
Changes to SQLite.Interop/src/ext/fts5.c.
︙ | ︙ | |||
17459 17460 17461 17462 17463 17464 17465 | static void fts5SourceIdFunc( sqlite3_context *pCtx, /* Function call context */ int nArg, /* Number of args */ sqlite3_value **apUnused /* Function arguments */ ){ assert( nArg==0 ); UNUSED_PARAM2(nArg, apUnused); | | | 17459 17460 17461 17462 17463 17464 17465 17466 17467 17468 17469 17470 17471 17472 17473 | static void fts5SourceIdFunc( sqlite3_context *pCtx, /* Function call context */ int nArg, /* Number of args */ sqlite3_value **apUnused /* Function arguments */ ){ assert( nArg==0 ); UNUSED_PARAM2(nArg, apUnused); sqlite3_result_text(pCtx, "fts5: 2018-04-02 11:04:16 736b53f57f70b23172c30880186dce7ad9baa3b74e3838cae5847cffb98f5cd2", -1, SQLITE_TRANSIENT); } static int fts5Init(sqlite3 *db){ static const sqlite3_module fts5Mod = { /* iVersion */ 2, /* xCreate */ fts5CreateMethod, /* xConnect */ fts5ConnectMethod, |
︙ | ︙ |
Changes to readme.htm.
1 2 3 4 5 6 7 8 | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <title></title> </head> <body> ADO.NET SQLite Data Provider<br /> Version 1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font><br /> | | | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 | <!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN"> <html> <head> <title></title> </head> <body> ADO.NET SQLite Data Provider<br /> Version 1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font><br /> Using <a href="https://www.sqlite.org/releaselog/3_23_0.html">SQLite 3.23.0</a><br />Originally written by Robert Simpson<br /> Released to the public domain, use at your own risk!<br /> Official provider website: <a href="https://system.data.sqlite.org/">https://system.data.sqlite.org/</a><br /> Legacy versions: <a href="https://sourceforge.net/projects/sqlite-dotnet2/">https://sourceforge.net/projects/sqlite-dotnet2/</a><br /> <br /> The current development version can be downloaded from <a href="https://system.data.sqlite.org/index.html/timeline?y=ci"> https://system.data.sqlite.org/index.html/timeline?y=ci</a> <br /> |
︙ | ︙ | |||
207 208 209 210 211 212 213 | <h2><b>Version History</b></h2> <p> <b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b> </p> <ul> | | | 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 | <h2><b>Version History</b></h2> <p> <b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b> </p> <ul> <li>Updated to <a href="https://www.sqlite.org/releaselog/3_23_0.html">SQLite 3.23.0</a>.</li> <li>Prevent GetSchemaTable from throwing InvalidCastException. Fix for [baf42ee135].</li> </ul> <p> <b>1.0.XXX.0 - March 2, 2018</b> </p> <ul> <li>Support extended result codes when messages are looked up without the SQLite core library.</li> |
︙ | ︙ |
Changes to www/news.wiki.
︙ | ︙ | |||
44 45 46 47 48 49 50 | <div align="center"><h2><b>Version History</b></h2></div> <p> <b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b> </p> <ul> | | | 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 | <div align="center"><h2><b>Version History</b></h2></div> <p> <b>1.0.109.0 - May XX, 2018 <font color="red">(release scheduled)</font></b> </p> <ul> <li>Updated to [https://www.sqlite.org/releaselog/3_23_0.html|SQLite 3.23.0].</li> <li>Prevent GetSchemaTable from throwing InvalidCastException. Fix for [baf42ee135].</li> </ul> <p> <b>1.0.108.0 - March 2, 2018</b> </p> <ul> <li>Support extended result codes when messages are looked up without the SQLite core library.</li> |
︙ | ︙ |