Merge pull request #1601 from penberg/merge-sqlite-3.45.1-new

Merge upstream SQLite 3.45.1

Author: MarinPostma

libsql-ffi/bundled/src/sqlite3.c

@@ -149,9 +149,9 @@ extern "C" {
 ** [sqlite3_libversion_number()], [sqlite3_sourceid()],
 ** [sqlite_version()] and [sqlite_source_id()].
 */
-#define SQLITE_VERSION        "3.44.0"
-#define SQLITE_VERSION_NUMBER 3044000
-#define SQLITE_SOURCE_ID      "2023-11-01 11:23:50 17129ba1ff7f0daf37100ee82d507aef7827cf38de1866e2633096ae6ad8alt1"
+#define SQLITE_VERSION        "3.45.1"
+#define SQLITE_VERSION_NUMBER 3045001
+#define SQLITE_SOURCE_ID      "2024-01-30 16:01:20 e876e51a0ed5c5b3126f52e532044363a014bc594cfefa87ffb5b82257ccalt1"
 
 #define LIBSQL_VERSION        "0.2.3"
 
@@ -4015,15 +4015,17 @@ SQLITE_API void sqlite3_free_filename(sqlite3_filename);
 ** </ul>
 **
 ** ^The sqlite3_errmsg() and sqlite3_errmsg16() return English-language
-** text that describes the error, as either UTF-8 or UTF-16 respectively.
+** text that describes the error, as either UTF-8 or UTF-16 respectively,
+** or NULL if no error message is available.
 ** (See how SQLite handles [invalid UTF] for exceptions to this rule.)
 ** ^(Memory to hold the error message string is managed internally.
 ** The application does not need to worry about freeing the result.
 ** However, the error string might be overwritten or deallocated by
 ** subsequent calls to other SQLite interface functions.)^
 **
-** ^The sqlite3_errstr() interface returns the English-language text
-** that describes the [result code], as UTF-8.
+** ^The sqlite3_errstr(E) interface returns the English-language text
+** that describes the [result code] E, as UTF-8, or NULL if E is not an
+** result code for which a text error message is available.
 ** ^(Memory to hold the error message string is managed internally
 ** and must not be freed by the application)^.
 **
@@ -5634,20 +5636,35 @@ SQLITE_API int sqlite3_create_window_function(
 ** </dd>
 **
 ** [[SQLITE_SUBTYPE]] <dt>SQLITE_SUBTYPE</dt><dd>
-** The SQLITE_SUBTYPE flag indicates to SQLite that a function may call
+** The SQLITE_SUBTYPE flag indicates to SQLite that a function might call
 ** [sqlite3_value_subtype()] to inspect the sub-types of its arguments.
-** Specifying this flag makes no difference for scalar or aggregate user
-** functions. However, if it is not specified for a user-defined window
-** function, then any sub-types belonging to arguments passed to the window
-** function may be discarded before the window function is called (i.e.
-** sqlite3_value_subtype() will always return 0).
+** This flag instructs SQLite to omit some corner-case optimizations that
+** might disrupt the operation of the [sqlite3_value_subtype()] function,
+** causing it to return zero rather than the correct subtype().
+** SQL functions that invokes [sqlite3_value_subtype()] should have this
+** property.  If the SQLITE_SUBTYPE property is omitted, then the return
+** value from [sqlite3_value_subtype()] might sometimes be zero even though
+** a non-zero subtype was specified by the function argument expression.
+**
+** [[SQLITE_RESULT_SUBTYPE]] <dt>SQLITE_RESULT_SUBTYPE</dt><dd>
+** The SQLITE_RESULT_SUBTYPE flag indicates to SQLite that a function might call
+** [sqlite3_result_subtype()] to cause a sub-type to be associated with its
+** result.
+** Every function that invokes [sqlite3_result_subtype()] should have this
+** property.  If it does not, then the call to [sqlite3_result_subtype()]
+** might become a no-op if the function is used as term in an
+** [expression index].  On the other hand, SQL functions that never invoke
+** [sqlite3_result_subtype()] should avoid setting this property, as the
+** purpose of this property is to disable certain optimizations that are
+** incompatible with subtypes.
 ** </dd>
 ** </dl>
 */
 #define SQLITE_DETERMINISTIC    0x000000800
 #define SQLITE_DIRECTONLY       0x000080000
 #define SQLITE_SUBTYPE          0x000100000
 #define SQLITE_INNOCUOUS        0x000200000
+#define SQLITE_RESULT_SUBTYPE   0x001000000
 
 /*
 ** CAPI3REF: Deprecated Functions
@@ -5844,6 +5861,12 @@ SQLITE_API int sqlite3_value_encoding(sqlite3_value*);
 ** information can be used to pass a limited amount of context from
 ** one SQL function to another.  Use the [sqlite3_result_subtype()]
 ** routine to set the subtype for the return value of an SQL function.
+**
+** Every [application-defined SQL function] that invoke this interface
+** should include the [SQLITE_SUBTYPE] property in the text
+** encoding argument when the function is [sqlite3_create_function|registered].
+** If the [SQLITE_SUBTYPE] property is omitted, then sqlite3_value_subtype()
+** might return zero instead of the upstream subtype in some corner cases.
 */
 SQLITE_API unsigned int sqlite3_value_subtype(sqlite3_value*);
 
@@ -5974,14 +5997,22 @@ SQLITE_API sqlite3 *sqlite3_context_db_handle(sqlite3_context*);
 ** <li> ^(when sqlite3_set_auxdata() is invoked again on the same
 **       parameter)^, or
 ** <li> ^(during the original sqlite3_set_auxdata() call when a memory
-**      allocation error occurs.)^ </ul>
+**      allocation error occurs.)^
+** <li> ^(during the original sqlite3_set_auxdata() call if the function
+**      is evaluated during query planning instead of during query execution,
+**      as sometimes happens with [SQLITE_ENABLE_STAT4].)^ </ul>
 **
-** Note the last bullet in particular.  The destructor X in
+** Note the last two bullets in particular.  The destructor X in
 ** sqlite3_set_auxdata(C,N,P,X) might be called immediately, before the
 ** sqlite3_set_auxdata() interface even returns.  Hence sqlite3_set_auxdata()
 ** should be called near the end of the function implementation and the
 ** function implementation should not make any use of P after
-** sqlite3_set_auxdata() has been called.
+** sqlite3_set_auxdata() has been called.  Furthermore, a call to
+** sqlite3_get_auxdata() that occurs immediately after a corresponding call
+** to sqlite3_set_auxdata() might still return NULL if an out-of-memory
+** condition occurred during the sqlite3_set_auxdata() call or if the
+** function is being evaluated during query planning rather than during
+** query execution.
 **
 ** ^(In practice, auxiliary data is preserved between function calls for
 ** function parameters that are compile-time constants, including literal
@@ -6255,6 +6286,20 @@ SQLITE_API int sqlite3_result_zeroblob64(sqlite3_context*, sqlite3_uint64 n);
 ** higher order bits are discarded.
 ** The number of subtype bytes preserved by SQLite might increase
 ** in future releases of SQLite.
+**
+** Every [application-defined SQL function] that invokes this interface
+** should include the [SQLITE_RESULT_SUBTYPE] property in its
+** text encoding argument when the SQL function is
+** [sqlite3_create_function|registered].  If the [SQLITE_RESULT_SUBTYPE]
+** property is omitted from the function that invokes sqlite3_result_subtype(),
+** then in some cases the sqlite3_result_subtype() might fail to set
+** the result subtype.
+**
+** If SQLite is compiled with -DSQLITE_STRICT_SUBTYPE=1, then any
+** SQL function that invokes the sqlite3_result_subtype() interface
+** and that does not have the SQLITE_RESULT_SUBTYPE property will raise
+** an error.  Future versions of SQLite might enable -DSQLITE_STRICT_SUBTYPE=1
+** by default.
 */
 SQLITE_API void sqlite3_result_subtype(sqlite3_context*,unsigned int);
 
@@ -8082,9 +8127,11 @@ typedef struct libsql_wal_methods libsql_wal_methods;
 **
 ** ^(Some systems (for example, Windows 95) do not support the operation
 ** implemented by sqlite3_mutex_try().  On those systems, sqlite3_mutex_try()
-** will always return SQLITE_BUSY. The SQLite core only ever uses
-** sqlite3_mutex_try() as an optimization so this is acceptable
-** behavior.)^
+** will always return SQLITE_BUSY. In most cases the SQLite core only uses
+** sqlite3_mutex_try() as an optimization, so this is acceptable
+** behavior. The exceptions are unix builds that set the
+** SQLITE_ENABLE_SETLK_TIMEOUT build option. In that case a working
+** sqlite3_mutex_try() is required.)^
 **
 ** ^The sqlite3_mutex_leave() routine exits a mutex that was
 ** previously entered by the same thread.   The behavior
@@ -8343,6 +8390,7 @@ SQLITE_API int sqlite3_test_control(int op, ...);
 #define SQLITE_TESTCTRL_ASSERT                  12
 #define SQLITE_TESTCTRL_ALWAYS                  13
 #define SQLITE_TESTCTRL_RESERVE                 14  /* NOT USED */
+#define SQLITE_TESTCTRL_JSON_SELFCHECK          14
 #define SQLITE_TESTCTRL_OPTIMIZATIONS           15
 #define SQLITE_TESTCTRL_ISKEYWORD               16  /* NOT USED */
 #define SQLITE_TESTCTRL_SCRATCHMALLOC           17  /* NOT USED */
@@ -12889,8 +12937,11 @@ struct Fts5PhraseIter {
 **   created with the "columnsize=0" option.
 **
 ** xColumnText:
-**   This function attempts to retrieve the text of column iCol of the
-**   current document. If successful, (*pz) is set to point to a buffer
+**   If parameter iCol is less than zero, or greater than or equal to the
+**   number of columns in the table, SQLITE_RANGE is returned.
+**
+**   Otherwise, this function attempts to retrieve the text of column iCol of
+**   the current document. If successful, (*pz) is set to point to a buffer
 **   containing the text in utf-8 encoding, (*pn) is set to the size in bytes
 **   (not characters) of the buffer and SQLITE_OK is returned. Otherwise,
 **   if an error occurs, an SQLite error code is returned and the final values
@@ -12900,8 +12951,10 @@ struct Fts5PhraseIter {
 **   Returns the number of phrases in the current query expression.
 **
 ** xPhraseSize:
-**   Returns the number of tokens in phrase iPhrase of the query. Phrases
-**   are numbered starting from zero.
+**   If parameter iCol is less than zero, or greater than or equal to the
+**   number of phrases in the current query, as returned by xPhraseCount,
+**   0 is returned. Otherwise, this function returns the number of tokens in
+**   phrase iPhrase of the query. Phrases are numbered starting from zero.
 **
 ** xInstCount:
 **   Set *pnInst to the total number of occurrences of all phrases within
@@ -12917,12 +12970,13 @@ struct Fts5PhraseIter {
 **   Query for the details of phrase match iIdx within the current row.
 **   Phrase matches are numbered starting from zero, so the iIdx argument
 **   should be greater than or equal to zero and smaller than the value
-**   output by xInstCount().
+**   output by xInstCount(). If iIdx is less than zero or greater than
+**   or equal to the value returned by xInstCount(), SQLITE_RANGE is returned.
 **
-**   Usually, output parameter *piPhrase is set to the phrase number, *piCol
+**   Otherwise, output parameter *piPhrase is set to the phrase number, *piCol
 **   to the column in which it occurs and *piOff the token offset of the
-**   first token of the phrase. Returns SQLITE_OK if successful, or an error
-**   code (i.e. SQLITE_NOMEM) if an error occurs.
+**   first token of the phrase. SQLITE_OK is returned if successful, or an
+**   error code (i.e. SQLITE_NOMEM) if an error occurs.
 **
 **   This API can be quite slow if used with an FTS5 table created with the
 **   "detail=none" or "detail=column" option.
@@ -12948,6 +13002,10 @@ struct Fts5PhraseIter {
 **   Invoking Api.xUserData() returns a copy of the pointer passed as
 **   the third argument to pUserData.
 **
+**   If parameter iPhrase is less than zero, or greater than or equal to
+**   the number of phrases in the query, as returned by xPhraseCount(),
+**   this function returns SQLITE_RANGE.
+**
 **   If the callback function returns any value other than SQLITE_OK, the
 **   query is abandoned and the xQueryPhrase function returns immediately.
 **   If the returned value is SQLITE_DONE, xQueryPhrase returns SQLITE_OK.
@@ -13062,9 +13120,42 @@ struct Fts5PhraseIter {
 **
 ** xPhraseNextColumn()
 **   See xPhraseFirstColumn above.
+**
+** xQueryToken(pFts5, iPhrase, iToken, ppToken, pnToken)
+**   This is used to access token iToken of phrase iPhrase of the current
+**   query. Before returning, output parameter *ppToken is set to point
+**   to a buffer containing the requested token, and *pnToken to the
+**   size of this buffer in bytes.
+**
+**   If iPhrase or iToken are less than zero, or if iPhrase is greater than
+**   or equal to the number of phrases in the query as reported by
+**   xPhraseCount(), or if iToken is equal to or greater than the number of
+**   tokens in the phrase, SQLITE_RANGE is returned and *ppToken and *pnToken
+     are both zeroed.
+**
+**   The output text is not a copy of the query text that specified the
+**   token. It is the output of the tokenizer module. For tokendata=1
+**   tables, this includes any embedded 0x00 and trailing data.
+**
+** xInstToken(pFts5, iIdx, iToken, ppToken, pnToken)
+**   This is used to access token iToken of phrase hit iIdx within the
+**   current row. If iIdx is less than zero or greater than or equal to the
+**   value returned by xInstCount(), SQLITE_RANGE is returned.  Otherwise,
+**   output variable (*ppToken) is set to point to a buffer containing the
+**   matching document token, and (*pnToken) to the size of that buffer in
+**   bytes. This API is not available if the specified token matches a
+**   prefix query term. In that case both output variables are always set
+**   to 0.
+**
+**   The output text is not a copy of the document text that was tokenized.
+**   It is the output of the tokenizer module. For tokendata=1 tables, this
+**   includes any embedded 0x00 and trailing data.
+**
+**   This API can be quite slow if used with an FTS5 table created with the
+**   "detail=none" or "detail=column" option.
 */
 struct Fts5ExtensionApi {
-  int iVersion;                   /* Currently always set to 2 */
+  int iVersion;                   /* Currently always set to 3 */
 
   void *(*xUserData)(Fts5Context*);
 
@@ -13099,6 +13190,13 @@ struct Fts5ExtensionApi {
 
   int (*xPhraseFirstColumn)(Fts5Context*, int iPhrase, Fts5PhraseIter*, int*);
   void (*xPhraseNextColumn)(Fts5Context*, Fts5PhraseIter*, int *piCol);
+
+  /* Below this point are iVersion>=3 only */
+  int (*xQueryToken)(Fts5Context*,
+      int iPhrase, int iToken,
+      const char **ppToken, int *pnToken
+  );
+  int (*xInstToken)(Fts5Context*, int iIdx, int iToken, const char**, int*);
 };
 
 /*

libsql-ffi/bundled/src/sqlite3.h

@@ -493,6 +493,7 @@ TESTSRC += \
   $(TOP)/ext/misc/percentile.c \
   $(TOP)/ext/misc/prefixes.c \
   $(TOP)/ext/misc/qpvtab.c \
+  $(TOP)/ext/misc/randomjson.c \
   $(TOP)/ext/misc/regexp.c \
   $(TOP)/ext/misc/remember.c \
   $(TOP)/ext/misc/series.c \
@@ -649,6 +650,7 @@ SHELL_OPT += -DSQLITE_ENABLE_DBPAGE_VTAB
 SHELL_OPT += -DSQLITE_ENABLE_DBSTAT_VTAB
 SHELL_OPT += -DSQLITE_ENABLE_BYTECODE_VTAB
 SHELL_OPT += -DSQLITE_ENABLE_OFFSET_SQL_FUNC
+SHELL_OPT += -DSQLITE_STRICT_SUBTYPE=1
 FUZZERSHELL_OPT =
 FUZZCHECK_OPT += -I$(TOP)/test
 FUZZCHECK_OPT += -I$(TOP)/ext/recover
@@ -679,14 +681,17 @@ FUZZCHECK_OPT += \
   -DSQLITE_MAX_MMAP_SIZE=0 \
   -DSQLITE_OMIT_LOAD_EXTENSION \
   -DSQLITE_PRINTF_PRECISION_LIMIT=1000 \
-  -DSQLITE_PRIVATE=""
+  -DSQLITE_PRIVATE="" \
+  -DSQLITE_STRICT_SUBTYPE=1 \
+  -DSQLITE_STATIC_RANDOMJSON
 
 FUZZCHECK_SRC += $(TOP)/test/fuzzcheck.c
 FUZZCHECK_SRC += $(TOP)/test/ossfuzz.c
 FUZZCHECK_SRC += $(TOP)/test/fuzzinvariants.c
 FUZZCHECK_SRC += $(TOP)/ext/recover/dbdata.c
 FUZZCHECK_SRC += $(TOP)/ext/recover/sqlite3recover.c
 FUZZCHECK_SRC += $(TOP)/test/vt02.c
+FUZZCHECK_SRC += $(TOP)/ext/misc/randomjson.c
 DBFUZZ_OPT =
 ST_OPT = -DSQLITE_OS_KV_OPTIONAL
 
@@ -769,6 +774,21 @@ fuzzcheck-asan$(TEXE):	$(FUZZCHECK_SRC) sqlite3.c sqlite3.h $(FUZZCHECK_DEP)
 fuzzcheck-ubsan$(TEXE):	$(FUZZCHECK_SRC) sqlite3.c sqlite3.h $(FUZZCHECK_DEP)
 	$(LTLINK) -o $@ -fsanitize=undefined $(FUZZCHECK_OPT) $(FUZZCHECK_SRC) sqlite3.c $(TLIBS)
 
+# Usage:    FUZZDB=filename make run-fuzzcheck
+#
+# Where filename is a fuzzcheck database, this target builds and runs
+# fuzzcheck, fuzzcheck-asan, and fuzzcheck-ubsan on that database.
+#
+# FUZZDB can be a glob pattern of two or more databases. Example:
+#
+#     FUZZDB=test/fuzzdata*.db make run-fuzzcheck
+#
+run-fuzzcheck:	fuzzcheck$(TEXE) fuzzcheck-asan$(TEXE) fuzzcheck-ubsan$(TEXE)
+	@if test "$(FUZZDB)" = ""; then echo 'ERROR: No FUZZDB specified. Rerun with FUZZDB=filename'; exit 1; fi
+	./fuzzcheck$(TEXE) --spinner $(FUZZDB)
+	./fuzzcheck-asan$(TEXE) --spinner $(FUZZDB)
+	./fuzzcheck-ubsan$(TEXE) --spinner $(FUZZDB)
+
 ossshell$(TEXE):	$(TOP)/test/ossfuzz.c $(TOP)/test/ossshell.c sqlite3.c sqlite3.h
 	$(LTLINK) -o $@ $(FUZZCHECK_OPT) $(TOP)/test/ossshell.c \
              $(TOP)/test/ossfuzz.c sqlite3.c $(TLIBS) $(OPT_STATIC_LIBLIBSQL_WASM)
@@ -1223,19 +1243,21 @@ keywordhash.h:	$(TOP)/tool/mkkeywordhash.c
 # Source files that go into making shell.c
 SHELL_SRC = \
 	$(TOP)/src/shell.c.in \
-        $(TOP)/ext/misc/appendvfs.c \
+	$(TOP)/ext/consio/console_io.c \
+	$(TOP)/ext/consio/console_io.h \
+	$(TOP)/ext/misc/appendvfs.c \
 	$(TOP)/ext/misc/completion.c \
-        $(TOP)/ext/misc/decimal.c \
-        $(TOP)/ext/misc/basexx.c \
-        $(TOP)/ext/misc/base64.c \
-        $(TOP)/ext/misc/base85.c \
+	$(TOP)/ext/misc/decimal.c \
+	$(TOP)/ext/misc/basexx.c \
+	$(TOP)/ext/misc/base64.c \
+	$(TOP)/ext/misc/base85.c \
 	$(TOP)/ext/misc/fileio.c \
-        $(TOP)/ext/misc/ieee754.c \
-        $(TOP)/ext/misc/regexp.c \
-        $(TOP)/ext/misc/series.c \
+	$(TOP)/ext/misc/ieee754.c \
+	$(TOP)/ext/misc/regexp.c \
+	$(TOP)/ext/misc/series.c \
 	$(TOP)/ext/misc/shathree.c \
 	$(TOP)/ext/misc/sqlar.c \
-        $(TOP)/ext/misc/uint.c \
+	$(TOP)/ext/misc/uint.c \
 	$(TOP)/ext/expert/sqlite3expert.c \
 	$(TOP)/ext/expert/sqlite3expert.h \
 	$(TOP)/ext/misc/zipfile.c \
@@ -1244,7 +1266,7 @@ SHELL_SRC = \
 	$(TOP)/ext/recover/dbdata.c \
 	$(TOP)/ext/recover/sqlite3recover.c \
 	$(TOP)/ext/recover/sqlite3recover.h \
-        $(TOP)/src/test_windirent.c
+	$(TOP)/src/test_windirent.c
 
 shell.c:	$(SHELL_SRC) $(TOP)/tool/mkshellc.tcl has_tclsh84
 	$(TCLSH_CMD) $(TOP)/tool/mkshellc.tcl >shell.c
@@ -1365,6 +1387,8 @@ TESTFIXTURE_FLAGS += -DSQLITE_ENABLE_STMTVTAB
 TESTFIXTURE_FLAGS += -DSQLITE_ENABLE_DBPAGE_VTAB
 TESTFIXTURE_FLAGS += -DSQLITE_ENABLE_BYTECODE_VTAB
 TESTFIXTURE_FLAGS += -DSQLITE_CKSUMVFS_STATIC
+TESTFIXTURE_FLAGS += -DSQLITE_STATIC_RANDOMJSON
+TESTFIXTURE_FLAGS += -DSQLITE_STRICT_SUBTYPE=1
 
 TESTFIXTURE_SRC0 = $(TESTSRC2) libsqlite3.la
 TESTFIXTURE_SRC1 = sqlite3.c