SQLite: Check-in [063f1b37]

Many hyperlinks are disabled.
Use anonymous login to enable hyperlinks.

Overview

Comment:	Merge the sqlite3_system_errno() change from trunk.
Downloads:	Tarball \| ZIP archive \| SQL archive
Timelines:	family \| ancestors \| descendants \| both \| apple-osx
Files:	files \| file ages \| folders
SHA1:	063f1b3705625b2f115af8e909d86705b892b529
User & Date:	drh 2016-03-21 14:52:04

Context

2016-03-29
10:30		Merge 3.12.0 changes. (check-in: eef55744 user: drh tags: apple-osx)
2016-03-21
14:52		Merge the sqlite3_system_errno() change from trunk. (check-in: 063f1b37 user: drh tags: apple-osx)
14:46		Add the sqlite3_system_errno() interface. (check-in: 4bd12b57 user: drh tags: trunk)
2016-03-19
00:40		Fix a VDBE register allocation bug exposed by recent enhancements to PRAGMA integrity_check. (check-in: c73b5b9c user: drh tags: apple-osx)

Changes

Hide Diffs Unified Diffs Ignore Whitespace Patch

Changes to doc/lemon.html.

<html>
<head>
<title>The Lemon Parser Generator</title>
</head>
<body bgcolor=white>
<h1 align=center>The Lemon Parser Generator</h1>

<p>Lemon is an LALR(1) parser generator for C or C++.  
It does the same job as ``bison'' and ``yacc''.
But lemon is not another bison or yacc clone.  It
uses a different grammar syntax which is designed to
reduce the number of coding errors.  Lemon also uses a more
sophisticated parsing engine that is faster than yacc and
bison and which is both reentrant and thread-safe.



Furthermore, Lemon implements features that can be used
to eliminate resource leaks, making is suitable for use
in long-running programs such as graphical user interfaces
or embedded controllers.</p>

<p>This document is an introduction to the Lemon
parser generator.</p>

................................................................................
<ul>
<li>C code to implement the parser.
<li>A header file defining an integer ID for each terminal symbol.
<li>An information file that describes the states of the generated parser
    automaton.
</ul>
By default, all three of these output files are generated.
The header file is suppressed if the ``-m'' command-line option is
used and the report file is omitted when ``-q'' is selected.</p>

<p>The grammar specification file uses a ``.y'' suffix, by convention.
In the examples used in this document, we'll assume the name of the
grammar file is ``gram.y''.  A typical use of Lemon would be the
following command:
<pre>
   lemon gram.y
</pre>
This command will generate three output files named ``gram.c'',
``gram.h'' and ``gram.out''.
The first is C code to implement the parser.  The second
is the header file that defines numerical values for all
terminal symbols, and the last is the report that explains
the states used by the parser automaton.</p>

<h3>Command Line Options</h3>

................................................................................
You can obtain a list of the available command-line options together
with a brief explanation of what each does by typing
<pre>
   lemon -?
</pre>
As of this writing, the following command-line options are supported:
<ul>
<li><tt>-b</tt>
<li><tt>-c</tt>
<li><tt>-g</tt>
<li><tt>-m</tt>
<li><tt>-q</tt>
<li><tt>-s</tt>
<li><tt>-x</tt>
</ul>
The ``-b'' option reduces the amount of text in the report file by
printing only the basis of each parser state, rather than the full
configuration.
The ``-c'' option suppresses action table compression.  Using -c
will make the parser a little larger and slower but it will detect
syntax errors sooner.
The ``-g'' option causes no output files to be generated at all.
Instead, the input grammar file is printed on standard output but
with all comments, actions and other extraneous text deleted.  This
is a useful way to get a quick summary of a grammar.
The ``-m'' option causes the output C source file to be compatible
with the ``makeheaders'' program.
Makeheaders is a program that automatically generates header files
from C source code.  When the ``-m'' option is used, the header
file is not output since the makeheaders program will take care
of generated all header files automatically.
The ``-q'' option suppresses the report file.
Using ``-s'' causes a brief summary of parser statistics to be
printed.  Like this:
<pre>
   Parser statistics: 74 terminals, 70 nonterminals, 179 rules
                      340 states, 2026 parser table entries, 0 conflicts
</pre>
Finally, the ``-x'' option causes Lemon to print its version number
and then stops without attempting to read the grammar or generate a parser.</p>

<h3>The Parser Interface</h3>

<p>Lemon doesn't generate a complete, working program.  It only generates
a few subroutines that implement a parser.  This section describes
the interface to those subroutines.  It is up to the programmer to
call these subroutines in an appropriate way in order to produce a
................................................................................
must first create the parser.
A new parser is created as follows:
<pre>
   void *pParser = ParseAlloc( malloc );
</pre>
The ParseAlloc() routine allocates and initializes a new parser and
returns a pointer to it.
The actual data structure used to represent a parser is opaque --
its internal structure is not visible or usable by the calling routine.
For this reason, the ParseAlloc() routine returns a pointer to void
rather than a pointer to some particular structure.
The sole argument to the ParseAlloc() routine is a pointer to the
subroutine used to allocate memory.  Typically this means ``malloc()''.</p>

<p>After a program is finished using a parser, it can reclaim all
memory allocated by that parser by calling
<pre>
   ParseFree(pParser, free);
</pre>
The first argument is the same pointer returned by ParseAlloc().  The
................................................................................
The first argument to the Parse() routine is the pointer returned by
ParseAlloc().
The second argument is a small positive integer that tells the parse the
type of the next token in the data stream.
There is one token type for each terminal symbol in the grammar.
The gram.h file generated by Lemon contains #define statements that
map symbolic terminal symbol names into appropriate integer values.
(A value of 0 for the second argument is a special flag to the
parser to indicate that the end of input has been reached.)
The third argument is the value of the given token.  By default,
the type of the third argument is integer, but the grammar will
usually redefine this type to be some kind of structure.
Typically the second argument will be a broad category of tokens
such as ``identifier'' or ``number'' and the third argument will
be the name of the identifier or the value of the number.</p>

<p>The Parse() function may have either three or four arguments,
depending on the grammar.  If the grammar specification file requests
it (via the <a href='#extraarg'><tt>extra_argument</tt> directive</a>),
the Parse() function will have a fourth parameter that can be
of any type chosen by the programmer.  The parser doesn't do anything
................................................................................
   15    ParseFree(pParser, free );
   16    TokenizerFree(pTokenizer);
   17    return sState.treeRoot;
   18 }
</pre>
This example shows a user-written routine that parses a file of
text and returns a pointer to the parse tree.
(We've omitted all error-handling from this example to keep it
simple.)
We assume the existence of some kind of tokenizer which is created
using TokenizerCreate() on line 8 and deleted by TokenizerFree()
on line 16.  The GetNextToken() function on line 11 retrieves the
next token from the input file and puts its type in the 
integer variable hTokenId.  The sToken variable is assumed to be
some kind of structure that contains details about each token,
................................................................................
declaration can occur at any point in the file.
Lemon ignores whitespace (except where it is needed to separate
tokens) and it honors the same commenting conventions as C and C++.</p>

<h3>Terminals and Nonterminals</h3>

<p>A terminal symbol (token) is any string of alphanumeric
and underscore characters
that begins with an upper case letter.
A terminal can contain lowercase letters after the first character,
but the usual convention is to make terminals all upper case.
A nonterminal, on the other hand, is any string of alphanumeric
and underscore characters than begins with a lower case letter.
Again, the usual convention is to make nonterminals use all lower
case letters.</p>
................................................................................
must have alphanumeric names.</p>

<h3>Grammar Rules</h3>

<p>The main component of a Lemon grammar file is a sequence of grammar
rules.
Each grammar rule consists of a nonterminal symbol followed by
the special symbol ``::='' and then a list of terminals and/or nonterminals.
The rule is terminated by a period.
The list of terminals and nonterminals on the right-hand side of the
rule can be empty.
Rules can occur in any order, except that the left-hand side of the
first rule is assumed to be the start symbol for the grammar (unless
specified otherwise using the <tt>%start</tt> directive described below.)
A typical sequence of grammar rules might look something like this:
................................................................................
  expr ::= expr PLUS expr.
  expr ::= expr TIMES expr.
  expr ::= LPAREN expr RPAREN.
  expr ::= VALUE.
</pre>
</p>

<p>There is one non-terminal in this example, ``expr'', and five
terminal symbols or tokens: ``PLUS'', ``TIMES'', ``LPAREN'',
``RPAREN'' and ``VALUE''.</p>

<p>Like yacc and bison, Lemon allows the grammar to specify a block
of C code that will be executed whenever a grammar rule is reduced
by the parser.
In Lemon, this action is specified by putting the C code (contained
within curly braces <tt>{...}</tt>) immediately after the
period that closes the rule.
................................................................................
<pre>
  expr ::= expr PLUS expr.   { printf("Doing an addition...\n"); }
</pre>
</p>

<p>In order to be useful, grammar actions must normally be linked to
their associated grammar rules.
In yacc and bison, this is accomplished by embedding a ``$$'' in the
action to stand for the value of the left-hand side of the rule and
symbols ``$1'', ``$2'', and so forth to stand for the value of
the terminal or nonterminal at position 1, 2 and so forth on the
right-hand side of the rule.
This idea is very powerful, but it is also very error-prone.  The
single most common source of errors in a yacc or bison grammar is
to miscount the number of symbols on the right-hand side of a grammar
rule and say ``$7'' when you really mean ``$8''.</p>

<p>Lemon avoids the need to count grammar symbols by assigning symbolic
names to each symbol in a grammar rule and then using those symbolic
names in the action.
In yacc or bison, one would write this:
<pre>
  expr -> expr PLUS expr  { $$ = $1 + $3; };
................................................................................
includes a linking symbol in parentheses but that linking symbol
is not actually used in the reduce action, then an error message
is generated.
For example, the rule
<pre>
  expr(A) ::= expr(B) PLUS expr(C).  { A = B; }
</pre>
will generate an error because the linking symbol ``C'' is used
in the grammar rule but not in the reduce action.</p>

<p>The Lemon notation for linking grammar rules to reduce actions
also facilitates the use of destructors for reclaiming memory
allocated by the values of terminals and nonterminals on the
right-hand side of a rule.</p>


<h3>Precedence Rules</h3>

<p>Lemon resolves parsing ambiguities in exactly the same way as
yacc and bison.  A shift-reduce conflict is resolved in favor
of the shift, and a reduce-reduce conflict is resolved by reducing
whichever rule comes first in the grammar file.</p>

<p>Just like in
yacc and bison, Lemon allows a measure of control 
over the resolution of paring conflicts using precedence rules.
A precedence value can be assigned to any terminal symbol



using the %left, %right or %nonassoc directives.  Terminal symbols
mentioned in earlier directives have a lower precedence that
terminal symbols mentioned in later directives.  For example:</p>

<p><pre>
   %left AND.
   %left OR.
   %nonassoc EQ NE GT GE LT LE.
................................................................................

<p>Lemon supports the following special directives:
<ul>
<li><tt>%code</tt>
<li><tt>%default_destructor</tt>
<li><tt>%default_type</tt>
<li><tt>%destructor</tt>

<li><tt>%extra_argument</tt>



<li><tt>%include</tt>
<li><tt>%left</tt>
<li><tt>%name</tt>
<li><tt>%nonassoc</tt>
<li><tt>%parse_accept</tt>
<li><tt>%parse_failure </tt>
<li><tt>%right</tt>
<li><tt>%stack_overflow</tt>
<li><tt>%stack_size</tt>
<li><tt>%start_symbol</tt>
<li><tt>%syntax_error</tt>

<li><tt>%token_destructor</tt>
<li><tt>%token_prefix</tt>
<li><tt>%token_type</tt>
<li><tt>%type</tt>

</ul>
Each of these directives will be described separately in the
following sections:</p>


<h4>The <tt>%code</tt> directive</h4>

<p>The %code directive is used to specify addition C/C++ code that
is added to the end of the main output file.  This is similar to
the %include directive except that %include is inserted at the

beginning of the main output file.</p>

<p>%code is typically used to include some action routines or perhaps

a tokenizer as part of the output file.</p>


<h4>The <tt>%default_destructor</tt> directive</h4>

<p>The %default_destructor directive specifies a destructor to 
use for non-terminals that do not have their own destructor
specified by a separate %destructor directive.  See the documentation

on the %destructor directive below for additional information.</p>

<p>In some grammers, many different non-terminal symbols have the
same datatype and hence the same destructor.  This directive is
a convenience way to specify the same destructor for all those
non-terminals using a single statement.</p>


<h4>The <tt>%default_type</tt> directive</h4>

<p>The %default_type directive specifies the datatype of non-terminal
symbols that do no have their own datatype defined using a separate
%type directive.  See the documentation on %type below for addition
information.</p>


<h4>The <tt>%destructor</tt> directive</h4>

<p>The %destructor directive is used to specify a destructor for
a non-terminal symbol.
(See also the %token_destructor directive which is used to
specify a destructor for terminal symbols.)</p>

<p>A non-terminal's destructor is called to dispose of the
non-terminal's value whenever the non-terminal is popped from
the stack.  This includes all of the following circumstances:
<ul>
<li> When a rule reduces and the value of a non-terminal on
     the right-hand side is not linked to C code.
................................................................................
<pre>
   %type nt {void*}
   %destructor nt { free($$); }
   nt(A) ::= ID NUM.   { A = malloc( 100 ); }
</pre>
This example is a bit contrived but it serves to illustrate how
destructors work.  The example shows a non-terminal named
``nt'' that holds values of type ``void*''.  When the rule for
an ``nt'' reduces, it sets the value of the non-terminal to
space obtained from malloc().  Later, when the nt non-terminal
is popped from the stack, the destructor will fire and call
free() on this malloced space, thus avoiding a memory leak.
(Note that the symbol ``$$'' in the destructor code is replaced
by the value of the non-terminal.)</p>

<p>It is important to note that the value of a non-terminal is passed
to the destructor whenever the non-terminal is removed from the
stack, unless the non-terminal is used in a C-code action.  If
the non-terminal is used by C-code, then it is assumed that the
C-code will take care of destroying it if it should really
be destroyed.  More commonly, the value is used to build some
larger structure and we don't want to destroy it, which is why
the destructor is not called in this circumstance.</p>

<p>By appropriate use of destructors, it is possible to
build a parser using Lemon that can be used within a long-running
program, such as a GUI, that will not leak memory or other resources.
To do the same using yacc or bison is much more difficult.</p>

<a name="extraarg"></a>
<h4>The <tt>%extra_argument</tt> directive</h4>

The %extra_argument directive instructs Lemon to add a 4th parameter
to the parameter list of the Parse() function it generates.  Lemon
................................................................................
and so forth.  For example, if the grammar file contains:</p>

<p><pre>
    %extra_argument { MyStruct *pAbc }
</pre></p>

<p>Then the Parse() function generated will have an 4th parameter
of type ``MyStruct*'' and all action routines will have access to
a variable named ``pAbc'' that is the value of the 4th parameter
in the most recent call to Parse().</p>

















































<h4>The <tt>%include</tt> directive</h4>

<p>The %include directive specifies C code that is included at the
top of the generated parser.  You can include any text you want --
the Lemon parser generator copies it blindly.  If you have multiple
%include directives in your grammar file the value of the last
%include directive overwrites all the others.</p.



<p>The %include directive is very handy for getting some extra #include
preprocessor statements at the beginning of the generated parser.
For example:</p>

<p><pre>
   %include {#include &lt;unistd.h&gt;}
</pre></p>

<p>This might be needed, for example, if some of the C actions in the
grammar call functions that are prototyed in unistd.h.</p>


<h4>The <tt>%left</tt> directive</h4>

The %left directive is used (along with the %right and
%nonassoc directives) to declare precedences of terminal
symbols.  Every terminal symbol whose name appears after
a %left directive but before the next period (``.'') is
given the same left-associative precedence value.  Subsequent
%left directives have higher precedence.  For example:</p>

<p><pre>
   %left AND.
   %left OR.
   %nonassoc EQ NE GT GE LT LE.
................................................................................
directive.</p>

<p>LALR(1) grammars can get into a situation where they require
a large amount of stack space if you make heavy use or right-associative
operators.  For this reason, it is recommended that you use %left
rather than %right whenever possible.</p>


<h4>The <tt>%name</tt> directive</h4>

<p>By default, the functions generated by Lemon all begin with the
five-character string ``Parse''.  You can change this string to something
different using the %name directive.  For instance:</p>

<p><pre>
   %name Abcde
</pre></p>

<p>Putting this directive in the grammar file will cause Lemon to generate
................................................................................
<li> AbcdeTrace(), and
<li> Abcde().
</ul>
The %name directive allows you to generator two or more different
parsers and link them all into the same executable.
</p>


<h4>The <tt>%nonassoc</tt> directive</h4>

<p>This directive is used to assign non-associative precedence to
one or more terminal symbols.  See the section on precedence rules

or on the %left directive for additional information.</p>


<h4>The <tt>%parse_accept</tt> directive</h4>

<p>The %parse_accept directive specifies a block of C code that is
executed whenever the parser accepts its input string.  To ``accept''
an input string means that the parser was able to process all tokens
without error.</p>

<p>For example:</p>

<p><pre>
   %parse_accept {
      printf("parsing complete!\n");
   }
</pre></p>


<h4>The <tt>%parse_failure</tt> directive</h4>

<p>The %parse_failure directive specifies a block of C code that
is executed whenever the parser fails complete.  This code is not
executed until the parser has tried and failed to resolve an input
error using is usual error recovery strategy.  The routine is
only invoked when parsing is unable to continue.</p>
................................................................................

<p><pre>
   %parse_failure {
     fprintf(stderr,"Giving up.  Parser is hopelessly lost...\n");
   }
</pre></p>


<h4>The <tt>%right</tt> directive</h4>

<p>This directive is used to assign right-associative precedence to
one or more terminal symbols.  See the section on precedence rules

or on the %left directive for additional information.</p>


<h4>The <tt>%stack_overflow</tt> directive</h4>

<p>The %stack_overflow directive specifies a block of C code that
is executed if the parser's internal stack ever overflows.  Typically
this just prints an error message.  After a stack overflow, the parser
will be unable to continue and must be reset.</p>

................................................................................
</pre>
Not like this:
<pre>
   list ::= element list.      // right-recursion.  Bad!
   list ::= .
</pre>


<h4>The <tt>%stack_size</tt> directive</h4>

<p>If stack overflow is a problem and you can't resolve the trouble
by using left-recursion, then you might want to increase the size
of the parser's stack using this directive.  Put an positive integer
after the %stack_size directive and Lemon will generate a parse
with a stack of the requested size.  The default value is 100.</p>

<p><pre>
   %stack_size 2000
</pre></p>


<h4>The <tt>%start_symbol</tt> directive</h4>

<p>By default, the start-symbol for the grammar that Lemon generates
is the first non-terminal that appears in the grammar file.  But you
can choose a different start-symbol using the %start_symbol directive.</p>

<p><pre>
   %start_symbol  prog
</pre></p>


<h4>The <tt>%token_destructor</tt> directive</h4>

<p>The %destructor directive assigns a destructor to a non-terminal
symbol.  (See the description of the %destructor directive above.)
This directive does the same thing for all terminal symbols.</p>

<p>Unlike non-terminal symbols which may each have a different data type
for their values, terminals all use the same data type (defined by
the %token_type directive) and so they use a common destructor.  Other
than that, the token destructor works just like the non-terminal
destructors.</p>


<h4>The <tt>%token_prefix</tt> directive</h4>

<p>Lemon generates #defines that assign small integer constants
to each terminal symbol in the grammar.  If desired, Lemon will
add a prefix specified by this directive
to each of the #defines it generates.
So if the default output of Lemon looked like this:
................................................................................
<pre>
    #define TOKEN_AND        1
    #define TOKEN_MINUS      2
    #define TOKEN_OR         3
    #define TOKEN_PLUS       4
</pre>


<h4>The <tt>%token_type</tt> and <tt>%type</tt> directives</h4>

<p>These directives are used to specify the data types for values
on the parser's stack associated with terminal and non-terminal
symbols.  The values of all terminal symbols must be of the same
type.  This turns out to be the same data type as the 3rd parameter
to the Parse() function generated by Lemon.  Typically, you will
................................................................................
token structure.  Like this:</p>

<p><pre>
   %token_type    {Token*}
</pre></p>

<p>If the data type of terminals is not specified, the default value
is ``int''.</p>

<p>Non-terminal symbols can each have their own data types.  Typically
the data type  of a non-terminal is a pointer to the root of a parse-tree
structure that contains all information about that non-terminal.
For example:</p>

<p><pre>
................................................................................
on what the corresponding non-terminal or terminal symbol is.  But
the grammar designer should keep in mind that the size of the union
will be the size of its largest element.  So if you have a single
non-terminal whose data type requires 1K of storage, then your 100
entry parser stack will require 100K of heap space.  If you are willing
and able to pay that price, fine.  You just need to know.</p>












<h3>Error Processing</h3>

<p>After extensive experimentation over several years, it has been
discovered that the error recovery strategy used by yacc is about
as good as it gets.  And so that is what Lemon uses.</p>

<p>When a Lemon-generated parser encounters a syntax error, it
first invokes the code specified by the %syntax_error directive, if
any.  It then enters its error recovery strategy.  The error recovery
strategy is to begin popping the parsers stack until it enters a
state where it is permitted to shift a special non-terminal symbol
named ``error''.  It then shifts this non-terminal and continues
parsing.  But the %syntax_error routine will not be called again
until at least three new tokens have been successfully shifted.</p>

<p>If the parser pops its stack until the stack is empty, and it still
is unable to shift the error symbol, then the %parse_failed routine
is invoked and the parser resets itself to its start state, ready
to begin parsing a new file.  This is what will happen at the very
first syntax error, of course, if there are no instances of the 
``error'' non-terminal in your grammar.</p>

</body>
</html>








|
|
|

|
|
|
>
>
>
|







 







|
|

|

|




|
|







 







|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
<
<
<
<







 







|




|







 







|
|




|







 







|







 







|







 







|







 







|
|
|







 







|

|





|







 







|







>











>
>
>
|







 







>

>
>
>











>




>




>


|

<
>
|


>
|

>





>
|






>




|
|

>




|
|







 







|
|



|






|
|



|
|
<







 







|
|


>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>





|
<
>
>












>


|
|
|
|







 







>



|







 







>



|
>
|

>



|











|







 







>



|
>
|

>







 







>












>










>












>







 







>







 







|







 







>
>
>
>
>
>
>
>
>
>
>











|








|

<html>
<head>
<title>The Lemon Parser Generator</title>
</head>
<body bgcolor=white>
<h1 align=center>The Lemon Parser Generator</h1>

<p>Lemon is an LALR(1) parser generator for C.
It does the same job as "bison" and "yacc".
But lemon is not a bison or yacc clone.  Lemon
uses a different grammar syntax which is designed to
reduce the number of coding errors.  Lemon also uses a
parsing engine that is faster than yacc and
bison and which is both reentrant and threadsafe.
(Update: Since the previous sentence was written, bison
has also been updated so that it too can generate a
reentrant and threadsafe parser.)
Lemon also implements features that can be used
to eliminate resource leaks, making is suitable for use
in long-running programs such as graphical user interfaces
or embedded controllers.</p>

<p>This document is an introduction to the Lemon
parser generator.</p>

................................................................................
<ul>
<li>C code to implement the parser.
<li>A header file defining an integer ID for each terminal symbol.
<li>An information file that describes the states of the generated parser
    automaton.
</ul>
By default, all three of these output files are generated.
The header file is suppressed if the "-m" command-line option is
used and the report file is omitted when "-q" is selected.</p>

<p>The grammar specification file uses a ".y" suffix, by convention.
In the examples used in this document, we'll assume the name of the
grammar file is "gram.y".  A typical use of Lemon would be the
following command:
<pre>
   lemon gram.y
</pre>
This command will generate three output files named "gram.c",
"gram.h" and "gram.out".
The first is C code to implement the parser.  The second
is the header file that defines numerical values for all
terminal symbols, and the last is the report that explains
the states used by the parser automaton.</p>

<h3>Command Line Options</h3>

................................................................................
You can obtain a list of the available command-line options together
with a brief explanation of what each does by typing
<pre>
   lemon -?
</pre>
As of this writing, the following command-line options are supported:
<ul>
<li><b>-b</b>
Show only the basis for each parser state in the report file.
<li><b>-c</b>
Do not compress the generated action tables.
<li><b>-D<i>name</i></b>
Define C preprocessor macro <i>name</i>.  This macro is useable by
"%ifdef" lines in the grammar file.
<li><b>-g</b>
Do not generate a parser.  Instead write the input grammar to standard
output with all comments, actions, and other extraneous text removed.
<li><b>-l</b>
Omit "#line" directives int the generated parser C code.
<li><b>-m</b>
Cause the output C source code to be compatible with the "makeheaders"
program. 
<li><b>-p</b>
Display all conflicts that are resolved by 
<a href='#precrules'>precedence rules</a>.
<li><b>-q</b>
Suppress generation of the report file.
<li><b>-r</b>
Do not sort or renumber the parser states as part of optimization.
<li><b>-s</b>
Show parser statistics before existing.
<li><b>-T<i>file</i></b>
Use <i>file</i> as the template for the generated C-code parser implementation.
<li><b>-x</b>
Print the Lemon version number.
</ul>





<h3>The Parser Interface</h3>

<p>Lemon doesn't generate a complete, working program.  It only generates
a few subroutines that implement a parser.  This section describes
the interface to those subroutines.  It is up to the programmer to
call these subroutines in an appropriate way in order to produce a
................................................................................
must first create the parser.
A new parser is created as follows:
<pre>
   void *pParser = ParseAlloc( malloc );
</pre>
The ParseAlloc() routine allocates and initializes a new parser and
returns a pointer to it.
The actual data structure used to represent a parser is opaque &mdash;
its internal structure is not visible or usable by the calling routine.
For this reason, the ParseAlloc() routine returns a pointer to void
rather than a pointer to some particular structure.
The sole argument to the ParseAlloc() routine is a pointer to the
subroutine used to allocate memory.  Typically this means malloc().</p>

<p>After a program is finished using a parser, it can reclaim all
memory allocated by that parser by calling
<pre>
   ParseFree(pParser, free);
</pre>
The first argument is the same pointer returned by ParseAlloc().  The
................................................................................
The first argument to the Parse() routine is the pointer returned by
ParseAlloc().
The second argument is a small positive integer that tells the parse the
type of the next token in the data stream.
There is one token type for each terminal symbol in the grammar.
The gram.h file generated by Lemon contains #define statements that
map symbolic terminal symbol names into appropriate integer values.
A value of 0 for the second argument is a special flag to the
parser to indicate that the end of input has been reached.
The third argument is the value of the given token.  By default,
the type of the third argument is integer, but the grammar will
usually redefine this type to be some kind of structure.
Typically the second argument will be a broad category of tokens
such as "identifier" or "number" and the third argument will
be the name of the identifier or the value of the number.</p>

<p>The Parse() function may have either three or four arguments,
depending on the grammar.  If the grammar specification file requests
it (via the <a href='#extraarg'><tt>extra_argument</tt> directive</a>),
the Parse() function will have a fourth parameter that can be
of any type chosen by the programmer.  The parser doesn't do anything
................................................................................
   15    ParseFree(pParser, free );
   16    TokenizerFree(pTokenizer);
   17    return sState.treeRoot;
   18 }
</pre>
This example shows a user-written routine that parses a file of
text and returns a pointer to the parse tree.
(All error-handling code is omitted from this example to keep it
simple.)
We assume the existence of some kind of tokenizer which is created
using TokenizerCreate() on line 8 and deleted by TokenizerFree()
on line 16.  The GetNextToken() function on line 11 retrieves the
next token from the input file and puts its type in the 
integer variable hTokenId.  The sToken variable is assumed to be
some kind of structure that contains details about each token,
................................................................................
declaration can occur at any point in the file.
Lemon ignores whitespace (except where it is needed to separate
tokens) and it honors the same commenting conventions as C and C++.</p>

<h3>Terminals and Nonterminals</h3>

<p>A terminal symbol (token) is any string of alphanumeric
and/or underscore characters
that begins with an upper case letter.
A terminal can contain lowercase letters after the first character,
but the usual convention is to make terminals all upper case.
A nonterminal, on the other hand, is any string of alphanumeric
and underscore characters than begins with a lower case letter.
Again, the usual convention is to make nonterminals use all lower
case letters.</p>
................................................................................
must have alphanumeric names.</p>

<h3>Grammar Rules</h3>

<p>The main component of a Lemon grammar file is a sequence of grammar
rules.
Each grammar rule consists of a nonterminal symbol followed by
the special symbol "::=" and then a list of terminals and/or nonterminals.
The rule is terminated by a period.
The list of terminals and nonterminals on the right-hand side of the
rule can be empty.
Rules can occur in any order, except that the left-hand side of the
first rule is assumed to be the start symbol for the grammar (unless
specified otherwise using the <tt>%start</tt> directive described below.)
A typical sequence of grammar rules might look something like this:
................................................................................
  expr ::= expr PLUS expr.
  expr ::= expr TIMES expr.
  expr ::= LPAREN expr RPAREN.
  expr ::= VALUE.
</pre>
</p>

<p>There is one non-terminal in this example, "expr", and five
terminal symbols or tokens: "PLUS", "TIMES", "LPAREN",
"RPAREN" and "VALUE".</p>

<p>Like yacc and bison, Lemon allows the grammar to specify a block
of C code that will be executed whenever a grammar rule is reduced
by the parser.
In Lemon, this action is specified by putting the C code (contained
within curly braces <tt>{...}</tt>) immediately after the
period that closes the rule.
................................................................................
<pre>
  expr ::= expr PLUS expr.   { printf("Doing an addition...\n"); }
</pre>
</p>

<p>In order to be useful, grammar actions must normally be linked to
their associated grammar rules.
In yacc and bison, this is accomplished by embedding a "$$" in the
action to stand for the value of the left-hand side of the rule and
symbols "$1", "$2", and so forth to stand for the value of
the terminal or nonterminal at position 1, 2 and so forth on the
right-hand side of the rule.
This idea is very powerful, but it is also very error-prone.  The
single most common source of errors in a yacc or bison grammar is
to miscount the number of symbols on the right-hand side of a grammar
rule and say "$7" when you really mean "$8".</p>

<p>Lemon avoids the need to count grammar symbols by assigning symbolic
names to each symbol in a grammar rule and then using those symbolic
names in the action.
In yacc or bison, one would write this:
<pre>
  expr -> expr PLUS expr  { $$ = $1 + $3; };
................................................................................
includes a linking symbol in parentheses but that linking symbol
is not actually used in the reduce action, then an error message
is generated.
For example, the rule
<pre>
  expr(A) ::= expr(B) PLUS expr(C).  { A = B; }
</pre>
will generate an error because the linking symbol "C" is used
in the grammar rule but not in the reduce action.</p>

<p>The Lemon notation for linking grammar rules to reduce actions
also facilitates the use of destructors for reclaiming memory
allocated by the values of terminals and nonterminals on the
right-hand side of a rule.</p>

<a name='precrules'></a>
<h3>Precedence Rules</h3>

<p>Lemon resolves parsing ambiguities in exactly the same way as
yacc and bison.  A shift-reduce conflict is resolved in favor
of the shift, and a reduce-reduce conflict is resolved by reducing
whichever rule comes first in the grammar file.</p>

<p>Just like in
yacc and bison, Lemon allows a measure of control 
over the resolution of paring conflicts using precedence rules.
A precedence value can be assigned to any terminal symbol
using the 
<a href='#pleft'>%left</a>,
<a href='#pright'>%right</a> or
<a href='#pnonassoc'>%nonassoc</a> directives.  Terminal symbols
mentioned in earlier directives have a lower precedence that
terminal symbols mentioned in later directives.  For example:</p>

<p><pre>
   %left AND.
   %left OR.
   %nonassoc EQ NE GT GE LT LE.
................................................................................

<p>Lemon supports the following special directives:
<ul>
<li><tt>%code</tt>
<li><tt>%default_destructor</tt>
<li><tt>%default_type</tt>
<li><tt>%destructor</tt>
<li><tt>%endif</tt>
<li><tt>%extra_argument</tt>
<li><tt>%fallback</tt>
<li><tt>%ifdef</tt>
<li><tt>%ifndef</tt>
<li><tt>%include</tt>
<li><tt>%left</tt>
<li><tt>%name</tt>
<li><tt>%nonassoc</tt>
<li><tt>%parse_accept</tt>
<li><tt>%parse_failure </tt>
<li><tt>%right</tt>
<li><tt>%stack_overflow</tt>
<li><tt>%stack_size</tt>
<li><tt>%start_symbol</tt>
<li><tt>%syntax_error</tt>
<li><tt>%token_class</tt>
<li><tt>%token_destructor</tt>
<li><tt>%token_prefix</tt>
<li><tt>%token_type</tt>
<li><tt>%type</tt>
<li><tt>%wildcard</tt>
</ul>
Each of these directives will be described separately in the
following sections:</p>

<a name='pcode'></a>
<h4>The <tt>%code</tt> directive</h4>

<p>The %code directive is used to specify addition C code that
is added to the end of the main output file.  This is similar to

the <a href='#pinclude'>%include</a> directive except that %include
is inserted at the beginning of the main output file.</p>

<p>%code is typically used to include some action routines or perhaps
a tokenizer or even the "main()" function 
as part of the output file.</p>

<a name='default_destructor'></a>
<h4>The <tt>%default_destructor</tt> directive</h4>

<p>The %default_destructor directive specifies a destructor to 
use for non-terminals that do not have their own destructor
specified by a separate %destructor directive.  See the documentation
on the <a name='#destructor'>%destructor</a> directive below for
additional information.</p>

<p>In some grammers, many different non-terminal symbols have the
same datatype and hence the same destructor.  This directive is
a convenience way to specify the same destructor for all those
non-terminals using a single statement.</p>

<a name='default_type'></a>
<h4>The <tt>%default_type</tt> directive</h4>

<p>The %default_type directive specifies the datatype of non-terminal
symbols that do no have their own datatype defined using a separate
<a href='#ptype'>%type</a> directive.  
</p>

<a name='destructor'></a>
<h4>The <tt>%destructor</tt> directive</h4>

<p>The %destructor directive is used to specify a destructor for
a non-terminal symbol.
(See also the <a href='#token_destructor'>%token_destructor</a>
directive which is used to specify a destructor for terminal symbols.)</p>

<p>A non-terminal's destructor is called to dispose of the
non-terminal's value whenever the non-terminal is popped from
the stack.  This includes all of the following circumstances:
<ul>
<li> When a rule reduces and the value of a non-terminal on
     the right-hand side is not linked to C code.
................................................................................
<pre>
   %type nt {void*}
   %destructor nt { free($$); }
   nt(A) ::= ID NUM.   { A = malloc( 100 ); }
</pre>
This example is a bit contrived but it serves to illustrate how
destructors work.  The example shows a non-terminal named
"nt" that holds values of type "void*".  When the rule for
an "nt" reduces, it sets the value of the non-terminal to
space obtained from malloc().  Later, when the nt non-terminal
is popped from the stack, the destructor will fire and call
free() on this malloced space, thus avoiding a memory leak.
(Note that the symbol "$$" in the destructor code is replaced
by the value of the non-terminal.)</p>

<p>It is important to note that the value of a non-terminal is passed
to the destructor whenever the non-terminal is removed from the
stack, unless the non-terminal is used in a C-code action.  If
the non-terminal is used by C-code, then it is assumed that the
C-code will take care of destroying it.
More commonly, the value is used to build some
larger structure and we don't want to destroy it, which is why
the destructor is not called in this circumstance.</p>

<p>Destructors help avoid memory leaks by automatically freeing
allocated objects when they go out of scope.

To do the same using yacc or bison is much more difficult.</p>

<a name="extraarg"></a>
<h4>The <tt>%extra_argument</tt> directive</h4>

The %extra_argument directive instructs Lemon to add a 4th parameter
to the parameter list of the Parse() function it generates.  Lemon
................................................................................
and so forth.  For example, if the grammar file contains:</p>

<p><pre>
    %extra_argument { MyStruct *pAbc }
</pre></p>

<p>Then the Parse() function generated will have an 4th parameter
of type "MyStruct*" and all action routines will have access to
a variable named "pAbc" that is the value of the 4th parameter
in the most recent call to Parse().</p>

<a name='pfallback'></a>
<h4>The <tt>%fallback</tt> directive</h4>

<p>The %fallback directive specifies an alternative meaning for one
or more tokens.  The alternative meaning is tried if the original token
would have generated a syntax error.

<p>The %fallback directive was added to support robust parsing of SQL
syntax in <a href="https://www.sqlite.org/">SQLite</a>.
The SQL language contains a large assortment of keywords, each of which
appears as a different token to the language parser.  SQL contains so
many keywords, that it can be difficult for programmers to keep up with
them all.  Programmers will, therefore, sometimes mistakenly use an
obscure language keyword for an identifier.  The %fallback directive
provides a mechanism to tell the parser:  "If you are unable to parse
this keyword, try treating it as an identifier instead."

<p>The syntax of %fallback is as follows:

<blockquote>
<tt>%fallback</tt>  <i>ID</i> <i>TOKEN...</i> <b>.</b>
</blockquote>

<p>In words, the %fallback directive is followed by a list of token names
terminated by a period.  The first token name is the fallback token - the
token to which all the other tokens fall back to.  The second and subsequent
arguments are tokens which fall back to the token identified by the first
argument.

<a name='pifdef'></a>
<h4>The <tt>%ifdef</tt>, <tt>%ifndef</tt>, and <tt>%endif</tt> directives.</h4>

<p>The %ifdef, %ifndef, and %endif directives are similar to
#ifdef, #ifndef, and #endif in the C-preprocessor, just not as general.
Each of these directives must begin at the left margin.  No whitespace
is allowed between the "%" and the directive name.

<p>Grammar text in between "%ifdef MACRO" and the next nested "%endif" is
ignored unless the "-DMACRO" command-line option is used.  Grammar text
betwen "%ifndef MACRO" and the next nested "%endif" is included except when
the "-DMACRO" command-line option is used.

<p>Note that the argument to %ifdef and %ifndef must be a single 
preprocessor symbol name, not a general expression.  There is no "%else"
directive.


<a name='pinclude'></a>
<h4>The <tt>%include</tt> directive</h4>

<p>The %include directive specifies C code that is included at the
top of the generated parser.  You can include any text you want --
the Lemon parser generator copies it blindly.  If you have multiple
%include directives in your grammar file, their values are concatenated

so that all %include code ultimately appears near the top of the
generated parser, in the same order as it appeared in the grammer.</p>

<p>The %include directive is very handy for getting some extra #include
preprocessor statements at the beginning of the generated parser.
For example:</p>

<p><pre>
   %include {#include &lt;unistd.h&gt;}
</pre></p>

<p>This might be needed, for example, if some of the C actions in the
grammar call functions that are prototyed in unistd.h.</p>

<a name='pleft'></a>
<h4>The <tt>%left</tt> directive</h4>

The %left directive is used (along with the <a href='#pright'>%right</a> and
<a href='#pnonassoc'>%nonassoc</a> directives) to declare precedences of 
terminal symbols.  Every terminal symbol whose name appears after
a %left directive but before the next period (".") is
given the same left-associative precedence value.  Subsequent
%left directives have higher precedence.  For example:</p>

<p><pre>
   %left AND.
   %left OR.
   %nonassoc EQ NE GT GE LT LE.
................................................................................
directive.</p>

<p>LALR(1) grammars can get into a situation where they require
a large amount of stack space if you make heavy use or right-associative
operators.  For this reason, it is recommended that you use %left
rather than %right whenever possible.</p>

<a name='pname'></a>
<h4>The <tt>%name</tt> directive</h4>

<p>By default, the functions generated by Lemon all begin with the
five-character string "Parse".  You can change this string to something
different using the %name directive.  For instance:</p>

<p><pre>
   %name Abcde
</pre></p>

<p>Putting this directive in the grammar file will cause Lemon to generate
................................................................................
<li> AbcdeTrace(), and
<li> Abcde().
</ul>
The %name directive allows you to generator two or more different
parsers and link them all into the same executable.
</p>

<a name='pnonassoc'></a>
<h4>The <tt>%nonassoc</tt> directive</h4>

<p>This directive is used to assign non-associative precedence to
one or more terminal symbols.  See the section on 
<a href='#precrules'>precedence rules</a>
or on the <a href='#pleft'>%left</a> directive for additional information.</p>

<a name='parse_accept'></a>
<h4>The <tt>%parse_accept</tt> directive</h4>

<p>The %parse_accept directive specifies a block of C code that is
executed whenever the parser accepts its input string.  To "accept"
an input string means that the parser was able to process all tokens
without error.</p>

<p>For example:</p>

<p><pre>
   %parse_accept {
      printf("parsing complete!\n");
   }
</pre></p>

<a name='parse_failure'></a>
<h4>The <tt>%parse_failure</tt> directive</h4>

<p>The %parse_failure directive specifies a block of C code that
is executed whenever the parser fails complete.  This code is not
executed until the parser has tried and failed to resolve an input
error using is usual error recovery strategy.  The routine is
only invoked when parsing is unable to continue.</p>
................................................................................

<p><pre>
   %parse_failure {
     fprintf(stderr,"Giving up.  Parser is hopelessly lost...\n");
   }
</pre></p>

<a name='pright'></a>
<h4>The <tt>%right</tt> directive</h4>

<p>This directive is used to assign right-associative precedence to
one or more terminal symbols.  See the section on 
<a href='#precrules'>precedence rules</a>
or on the <a href='#pleft'>%left</a> directive for additional information.</p>

<a name='stack_overflow'></a>
<h4>The <tt>%stack_overflow</tt> directive</h4>

<p>The %stack_overflow directive specifies a block of C code that
is executed if the parser's internal stack ever overflows.  Typically
this just prints an error message.  After a stack overflow, the parser
will be unable to continue and must be reset.</p>

................................................................................
</pre>
Not like this:
<pre>
   list ::= element list.      // right-recursion.  Bad!
   list ::= .
</pre>

<a name='stack_size'></a>
<h4>The <tt>%stack_size</tt> directive</h4>

<p>If stack overflow is a problem and you can't resolve the trouble
by using left-recursion, then you might want to increase the size
of the parser's stack using this directive.  Put an positive integer
after the %stack_size directive and Lemon will generate a parse
with a stack of the requested size.  The default value is 100.</p>

<p><pre>
   %stack_size 2000
</pre></p>

<a name='start_symbol'></a>
<h4>The <tt>%start_symbol</tt> directive</h4>

<p>By default, the start-symbol for the grammar that Lemon generates
is the first non-terminal that appears in the grammar file.  But you
can choose a different start-symbol using the %start_symbol directive.</p>

<p><pre>
   %start_symbol  prog
</pre></p>

<a name='token_destructor'></a>
<h4>The <tt>%token_destructor</tt> directive</h4>

<p>The %destructor directive assigns a destructor to a non-terminal
symbol.  (See the description of the %destructor directive above.)
This directive does the same thing for all terminal symbols.</p>

<p>Unlike non-terminal symbols which may each have a different data type
for their values, terminals all use the same data type (defined by
the %token_type directive) and so they use a common destructor.  Other
than that, the token destructor works just like the non-terminal
destructors.</p>

<a name='token_prefix'></a>
<h4>The <tt>%token_prefix</tt> directive</h4>

<p>Lemon generates #defines that assign small integer constants
to each terminal symbol in the grammar.  If desired, Lemon will
add a prefix specified by this directive
to each of the #defines it generates.
So if the default output of Lemon looked like this:
................................................................................
<pre>
    #define TOKEN_AND        1
    #define TOKEN_MINUS      2
    #define TOKEN_OR         3
    #define TOKEN_PLUS       4
</pre>

<a name='token_type'></a><a name='ptype'></a>
<h4>The <tt>%token_type</tt> and <tt>%type</tt> directives</h4>

<p>These directives are used to specify the data types for values
on the parser's stack associated with terminal and non-terminal
symbols.  The values of all terminal symbols must be of the same
type.  This turns out to be the same data type as the 3rd parameter
to the Parse() function generated by Lemon.  Typically, you will
................................................................................
token structure.  Like this:</p>

<p><pre>
   %token_type    {Token*}
</pre></p>

<p>If the data type of terminals is not specified, the default value
is "int".</p>

<p>Non-terminal symbols can each have their own data types.  Typically
the data type  of a non-terminal is a pointer to the root of a parse-tree
structure that contains all information about that non-terminal.
For example:</p>

<p><pre>
................................................................................
on what the corresponding non-terminal or terminal symbol is.  But
the grammar designer should keep in mind that the size of the union
will be the size of its largest element.  So if you have a single
non-terminal whose data type requires 1K of storage, then your 100
entry parser stack will require 100K of heap space.  If you are willing
and able to pay that price, fine.  You just need to know.</p>

<a name='pwildcard'></a>
<h4>The <tt>%wildcard</tt> directive</h4>

<p>The %wildcard directive is followed by a single token name and a
period.  This directive specifies that the identified token should 
match any input token.

<p>When the generated parser has the choice of matching an input against
the wildcard token and some other token, the other token is always used.
The wildcard token is only matched if there are no other alternatives.

<h3>Error Processing</h3>

<p>After extensive experimentation over several years, it has been
discovered that the error recovery strategy used by yacc is about
as good as it gets.  And so that is what Lemon uses.</p>

<p>When a Lemon-generated parser encounters a syntax error, it
first invokes the code specified by the %syntax_error directive, if
any.  It then enters its error recovery strategy.  The error recovery
strategy is to begin popping the parsers stack until it enters a
state where it is permitted to shift a special non-terminal symbol
named "error".  It then shifts this non-terminal and continues
parsing.  But the %syntax_error routine will not be called again
until at least three new tokens have been successfully shifted.</p>

<p>If the parser pops its stack until the stack is empty, and it still
is unable to shift the error symbol, then the %parse_failed routine
is invoked and the parser resets itself to its start state, ready
to begin parsing a new file.  This is what will happen at the very
first syntax error, of course, if there are no instances of the 
"error" non-terminal in your grammar.</p>

</body>
</html>

Changes to ext/fts5/fts5_index.c.

static int fts5AllocateSegid(Fts5Index *p, Fts5Structure *pStruct){
  int iSegid = 0;

  if( p->rc==SQLITE_OK ){
    if( pStruct->nSegment>=FTS5_MAX_SEGMENT ){
      p->rc = SQLITE_FULL;
    }else{
      while( iSegid==0 ){


        int iLvl, iSeg;
        sqlite3_randomness(sizeof(u32), (void*)&iSegid);
        iSegid = iSegid & ((1 << FTS5_DATA_ID_B)-1);



        for(iLvl=0; iLvl<pStruct->nLevel; iLvl++){
          for(iSeg=0; iSeg<pStruct->aLevel[iLvl].nSeg; iSeg++){
            if( iSegid==pStruct->aLevel[iLvl].aSeg[iSeg].iSegid ){
              iSegid = 0;


            }
          }
        }
      }













    }
  }

  return iSegid;
}

/*
................................................................................
  Fts5PageWriter *pLeaf = &pWriter->writer;
  if( p->rc==SQLITE_OK ){
    assert( pLeaf->pgno>=1 );
    if( pLeaf->buf.n>4 ){
      fts5WriteFlushLeaf(p, pWriter);
    }
    *pnLeaf = pLeaf->pgno-1;

    fts5WriteFlushBtree(p, pWriter);

  }
  fts5BufferFree(&pLeaf->term);
  fts5BufferFree(&pLeaf->buf);
  fts5BufferFree(&pLeaf->pgidx);
  fts5BufferFree(&pWriter->btterm);

  for(i=0; i<pWriter->nDlidx; i++){








|
>
>
|
<
<
>
>
>
|
|
|
<
>
>
|
|
|
|
>
>
>
>
>
>
>
>
>
>
>
>
>







 







>
|
>

static int fts5AllocateSegid(Fts5Index *p, Fts5Structure *pStruct){
  int iSegid = 0;

  if( p->rc==SQLITE_OK ){
    if( pStruct->nSegment>=FTS5_MAX_SEGMENT ){
      p->rc = SQLITE_FULL;
    }else{
      /* FTS5_MAX_SEGMENT is currently defined as 2000. So the following
      ** array is 63 elements, or 252 bytes, in size.  */
      u32 aUsed[(FTS5_MAX_SEGMENT+31) / 32];
      int iLvl, iSeg;


      int i;
      u32 mask;
      memset(aUsed, 0, sizeof(aUsed));
      for(iLvl=0; iLvl<pStruct->nLevel; iLvl++){
        for(iSeg=0; iSeg<pStruct->aLevel[iLvl].nSeg; iSeg++){
          int iId = pStruct->aLevel[iLvl].aSeg[iSeg].iSegid;

          if( iId<=FTS5_MAX_SEGMENT ){
            aUsed[(iId-1) / 32] |= 1 << ((iId-1) % 32);
          }
        }
      }

      for(i=0; aUsed[i]==0xFFFFFFFF; i++);
      mask = aUsed[i];
      for(iSegid=0; mask & (1 << iSegid); iSegid++);
      iSegid += 1 + i*32;

#ifdef SQLITE_DEBUG
      for(iLvl=0; iLvl<pStruct->nLevel; iLvl++){
        for(iSeg=0; iSeg<pStruct->aLevel[iLvl].nSeg; iSeg++){
          assert( iSegid!=pStruct->aLevel[iLvl].aSeg[iSeg].iSegid );
        }
      }
      assert( iSegid>0 && iSegid<=FTS5_MAX_SEGMENT );
#endif
    }
  }

  return iSegid;
}

/*
................................................................................
  Fts5PageWriter *pLeaf = &pWriter->writer;
  if( p->rc==SQLITE_OK ){
    assert( pLeaf->pgno>=1 );
    if( pLeaf->buf.n>4 ){
      fts5WriteFlushLeaf(p, pWriter);
    }
    *pnLeaf = pLeaf->pgno-1;
    if( pLeaf->pgno>1 ){
      fts5WriteFlushBtree(p, pWriter);
    }
  }
  fts5BufferFree(&pLeaf->term);
  fts5BufferFree(&pLeaf->buf);
  fts5BufferFree(&pLeaf->pgidx);
  fts5BufferFree(&pWriter->btterm);

  for(i=0; i<pWriter->nDlidx; i++){

Changes to ext/fts5/tool/fts5txt2db.tcl.

    {fts5                 "use fts5 (this is the default)"}
    {fts4                 "use fts4"}
    {colsize   "10 10 10" "list of column sizes"}
    {tblname   "t1"       "table name to create"}
    {detail    "full"     "Fts5 detail mode to use"}
    {repeat    1          "Load each file this many times"}
    {prefix    ""         "Fts prefix= option"}

    database
    file...
  } {
  This script is designed to create fts4/5 tables with more than one column.
  The -colsize option should be set to a Tcl list of integer values, one for
  each column in the table. Each value is the number of tokens that will be
  inserted into the column value for each row. For example, setting the -colsize
................................................................................
set cols [create_table]
set sql "INSERT INTO $A(tblname) VALUES(\$R([lindex $cols 0])"
foreach c [lrange $cols 1 end] {
  append sql ", \$R($c)"
}
append sql ")"

db eval BEGIN
  while {$i < $N} {
    foreach c $cols s $A(colsize) {
      set R($c) [lrange $tokens $i [expr $i+$s-1]]
      incr i $s
    }
    db eval $sql
  }
db eval COMMIT

    {fts5                 "use fts5 (this is the default)"}
    {fts4                 "use fts4"}
    {colsize   "10 10 10" "list of column sizes"}
    {tblname   "t1"       "table name to create"}
    {detail    "full"     "Fts5 detail mode to use"}
    {repeat    1          "Load each file this many times"}
    {prefix    ""         "Fts prefix= option"}
    {trans     1          "True to use a transaction"}
    database
    file...
  } {
  This script is designed to create fts4/5 tables with more than one column.
  The -colsize option should be set to a Tcl list of integer values, one for
  each column in the table. Each value is the number of tokens that will be
  inserted into the column value for each row. For example, setting the -colsize
................................................................................
set cols [create_table]
set sql "INSERT INTO $A(tblname) VALUES(\$R([lindex $cols 0])"
foreach c [lrange $cols 1 end] {
  append sql ", \$R($c)"
}
append sql ")"

if {$A(trans)} { db eval BEGIN }
  while {$i < $N} {
    foreach c $cols s $A(colsize) {
      set R($c) [lrange $tokens $i [expr $i+$s-1]]
      incr i $s
    }
    db eval $sql
  }
if {$A(trans)} { db eval COMMIT }

Changes to ext/rbu/rbu1.test.

         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 4);
       } {SQLITE_ERROR - invalid rbu_control value}

       9 {
         CREATE TABLE t1(a, b PRIMARY KEY) WITHOUT ROWID;
         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 2);
       } {SQLITE_ERROR - invalid rbu_control value}

       10 {
         CREATE TABLE t2(a, b);
         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 2);
       } {SQLITE_ERROR - no such table: t1}

         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 4);
       } {SQLITE_ERROR - invalid rbu_control value}

       9 {
         CREATE TABLE t1(a, b PRIMARY KEY) WITHOUT ROWID;
         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 3);
       } {SQLITE_ERROR - invalid rbu_control value}

       10 {
         CREATE TABLE t2(a, b);
         CREATE TABLE rbu.data_t1(a, b, rbu_control);
         INSERT INTO rbu.data_t1 VALUES(1, 2, 2);
       } {SQLITE_ERROR - no such table: t1}

Changes to ext/rbu/rbudiff.test.

#
if {![info exists testdir]} {
  set testdir [file join [file dirname [info script]] .. .. test]
}
source $testdir/tester.tcl
set testprefix rbudiff

if {$tcl_platform(platform)=="windows"} {
  set PROG "sqldiff.exe"
} else {
  set PROG "./sqldiff"
}
if {![file exe $PROG]} {
  puts "rbudiff.test cannot run because $PROG is not available"
  finish_test
  return
}
db close

proc get_rbudiff_sql {db1 db2} {
  exec $::PROG --rbu $db1 $db2
}

proc step_rbu {target rbu} {
................................................................................
    rbu close
    if {$rc != "SQLITE_OK"} break
  }
  set rc
}

proc apply_rbudiff {sql target} {

  forcedelete rbu.db
  sqlite3 rbudb rbu.db
  rbudb eval $sql
  rbudb close
  step_rbu $target rbu.db
}






















proc rbudiff_cksum {db1} {
  set txt ""

  sqlite3 dbtmp $db1
  foreach tbl [dbtmp eval {SELECT name FROM sqlite_master WHERE type='table'}] {
    set cols [list]








<
<
<
|
<
<
<
<
<
<







 







>






>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

#
if {![info exists testdir]} {
  set testdir [file join [file dirname [info script]] .. .. test]
}
source $testdir/tester.tcl
set testprefix rbudiff




set PROG [test_find_sqldiff]






db close

proc get_rbudiff_sql {db1 db2} {
  exec $::PROG --rbu $db1 $db2
}

proc step_rbu {target rbu} {
................................................................................
    rbu close
    if {$rc != "SQLITE_OK"} break
  }
  set rc
}

proc apply_rbudiff {sql target} {
  test_rbucount $sql
  forcedelete rbu.db
  sqlite3 rbudb rbu.db
  rbudb eval $sql
  rbudb close
  step_rbu $target rbu.db
}

# The only argument is the output of an [sqldiff -rbu] run. This command
# tests that the contents of the rbu_count table is correct. An exception
# is thrown if it is not.
#
proc test_rbucount {sql} {
  sqlite3 tmpdb ""
  tmpdb eval $sql
  tmpdb eval {
    SELECT name FROM sqlite_master WHERE name LIKE 'data%' AND type='table'
  } {
    set a [tmpdb eval "SELECT count(*) FROM $name"]
    set b [tmpdb eval {SELECT cnt FROM rbu_count WHERE tbl = $name}]
    if {$a != $b} { 
      tmpdb close
      error "rbu_count error - tbl = $name" 
    }
  }
  tmpdb close
  return ""
}

proc rbudiff_cksum {db1} {
  set txt ""

  sqlite3 dbtmp $db1
  foreach tbl [dbtmp eval {SELECT name FROM sqlite_master WHERE type='table'}] {
    set cols [list]

Added ext/rbu/rbuprogress.test.

>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

# 2016 March 18
#
# The author disclaims copyright to this source code.  In place of
# a legal notice, here is a blessing:
#
#    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.
#
#***********************************************************************
#

source [file join [file dirname [info script]] rbu_common.tcl]
set ::testprefix rbuprogress


proc create_db_file {filename sql} {
  forcedelete $filename
  sqlite3 tmpdb $filename  
  tmpdb eval $sql
  tmpdb close
}

# Create a simple RBU database. That expects to write to a table:
#
#   CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
#
proc create_rbu1 {filename} {
  create_db_file $filename {
    CREATE TABLE data_t1(a, b, c, rbu_control);
    INSERT INTO data_t1 VALUES(1, 2, 3, 0);
    INSERT INTO data_t1 VALUES(2, 'two', 'three', 0);
    INSERT INTO data_t1 VALUES(3, NULL, 8.2, 0);

    CREATE TABLE rbu_count(tbl, cnt);
    INSERT INTO rbu_count VALUES('data_t1', 3);
  }
  return $filename
}


do_execsql_test 1.0 {
  CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
}

do_test 1.1 {
  create_rbu1 rbu.db
  sqlite3rbu rbu test.db rbu.db
  rbu bp_progress
} {0 0}
do_test 1.2 { rbu step ; rbu bp_progress } {3333 0}
do_test 1.3 { rbu step ; rbu bp_progress } {6666 0}
do_test 1.4 { rbu step ; rbu bp_progress } {10000 0}
do_test 1.5 { rbu step ; rbu bp_progress } {10000 0}
do_test 1.6 { rbu step ; rbu bp_progress } {10000 0}
do_test 1.7 { rbu step ; rbu bp_progress } {10000 5000}
do_test 1.8 { rbu step ; rbu bp_progress } {10000 10000}
do_test 1.9 { rbu step ; rbu bp_progress } {10000 10000}

do_test 1.10 {
  rbu close
} {SQLITE_DONE}

#-------------------------------------------------------------------------
#
proc do_sp_test {tn bReopen target rbu reslist} {
  uplevel [list do_test $tn [subst -nocommands {
    if {$bReopen==0} { sqlite3rbu rbu $target $rbu }
    set res [list]
    while 1 {
      if {$bReopen} { sqlite3rbu rbu $target $rbu }
      set rc [rbu step]
      if {[set rc] != "SQLITE_OK"} { rbu close ; error "error 1" }
      lappend res [lindex [rbu bp_progress] 0]
      if {[lindex [set res] end]==10000} break
      if {$bReopen} { rbu close }
    }
    if {[set res] != [list $reslist]} {
      rbu close
      error "1. reslist incorrect (expect=$reslist got=[set res])"
    }

    # One step to clean up the temporary tables used to update the only
    # target table in the rbu database. And one more to move the *-oal 
    # file to *-wal. After each of these steps, the progress remains
    # at "10000 0".
    #
    if {[lindex [list $reslist] 0]!=-1} {
      rbu step
      set res [rbu bp_progress]
      if {[set res] != [list 10000 0]} {
        rbu close
        error "2. reslist incorrect (expect=10000 0 got=[set res])"
      }
    }

    rbu step
    set res [rbu bp_progress]
    if {[set res] != [list 10000 0]} {
      rbu close
      error "3. reslist incorrect (expect=10000 0 got=[set res])"
    }

    # Do the checkpoint.
    while {[rbu step]=="SQLITE_OK"} { 
      foreach {a b} [rbu bp_progress] {}
      if {[set a]!=10000 || [set b]<=0 || [set b]>10000} {
        rbu close
        error "4. reslist incorrect (expect=10000 1..10000 got=[set a] [set b])"
      }
    }

    set res [rbu bp_progress]
    if {[set res] != [list 10000 10000]} {
      rbu close
      error "5. reslist is incorrect (expect=10000 10000 got=[set res])"
    }

    rbu close
  }] {SQLITE_DONE}]
}

foreach {bReopen} { 0 1 } {
  reset_db
  do_test 2.$bReopen.1.0 {
    execsql {
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, c, rbu_control);
      INSERT INTO data_t1 VALUES(4, 4, 4, 0);
      INSERT INTO data_t1 VALUES(5, 5, 5, 0);
  
      CREATE TABLE rbu_count(tbl, cnt);
      INSERT INTO rbu_count VALUES('data_t1', 2);
    }
  } {}
  do_sp_test 2.$bReopen.1.1 $bReopen test.db rbu.db {5000 10000}
  
  reset_db
  do_test 2.$bReopen.2.0 {
    execsql { CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c) }
    create_rbu1 rbu.db
  } {rbu.db}
  do_sp_test 2.$bReopen.2.1 $bReopen test.db rbu.db {3333 6666 10000}
  
  reset_db
  do_test 2.$bReopen.3.0 {
    execsql { 
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
      CREATE INDEX i1 ON t1(b);
      INSERT INTO t1 VALUES(1, 1, 1);
      INSERT INTO t1 VALUES(2, 2, 2);
      INSERT INTO t1 VALUES(3, 3, 3);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, c, rbu_control);
      INSERT INTO data_t1 VALUES(4, 4, 4, 0);
      INSERT INTO data_t1 VALUES(2, NULL, NULL, 1);
      INSERT INTO data_t1 VALUES(5, NULL, NULL, 1);
  
      CREATE TABLE rbu_count(tbl, cnt);
      INSERT INTO rbu_count VALUES('data_t1', 3);
    }
  } {}
  do_sp_test 2.$bReopen.3.1 $bReopen test.db rbu.db {1666 3333 6000 8000 10000}
  
  reset_db
  do_test 2.$bReopen.4.0 {
    execsql { 
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
      CREATE INDEX i1 ON t1(b);
      INSERT INTO t1 VALUES(1, 1, 1);
      INSERT INTO t1 VALUES(2, 2, 2);
      INSERT INTO t1 VALUES(3, 3, 3);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, c, rbu_control);
      INSERT INTO data_t1 VALUES(2, 4, 4, '.xx');
  
      CREATE TABLE rbu_count(tbl, cnt);
      INSERT INTO rbu_count VALUES('data_t1', 1);
    }
  } {}
  do_sp_test 2.$bReopen.4.1 $bReopen test.db rbu.db {3333 6666 10000}
  
  reset_db
  do_test 2.$bReopen.5.0 {
    execsql { 
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
      CREATE INDEX i1 ON t1(b);
      INSERT INTO t1 VALUES(1, 1, 1);
      INSERT INTO t1 VALUES(2, 2, 2);
      INSERT INTO t1 VALUES(3, 3, 3);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, c, rbu_control);
      INSERT INTO data_t1 VALUES(4, NULL, 4, '.xx');
  
      CREATE TABLE rbu_count(tbl, cnt);
      INSERT INTO rbu_count VALUES('data_t1', 1);
    }
  } {}
  do_sp_test 2.$bReopen.5.1 $bReopen test.db rbu.db {10000}

  reset_db
  do_test 2.$bReopen.6.0 {
    execsql { 
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c);
      CREATE INDEX i1 ON t1(b);
      INSERT INTO t1 VALUES(1, 1, 1);
      INSERT INTO t1 VALUES(2, 2, 2);
      INSERT INTO t1 VALUES(3, 3, 3);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, c, rbu_control);
      INSERT INTO data_t1 VALUES(4, 4, 4, 0);
      INSERT INTO data_t1 VALUES(2, NULL, NULL, 1);
      INSERT INTO data_t1 VALUES(5, NULL, NULL, 1);
    }
  } {}
  do_sp_test 2.$bReopen.6.1 $bReopen test.db rbu.db {-1 -1 -1 -1 -1 10000}
}

#-------------------------------------------------------------------------
# The following tests verify that the API works when resuming an update
# during the incremental checkpoint stage.
#
proc do_phase2_test {tn bReopen target rbu nStep} {
  uplevel [list do_test $tn [subst -nocommands {

    # Build the OAL/WAL file:
    sqlite3rbu rbu $target $rbu
    while {[lindex [rbu bp_progress] 0]<10000} { 
      set rc [rbu step]
      if {"SQLITE_OK" != [set rc]} { rbu close }
    }

    # Clean up the temp tables and move the *-oal file to *-wal.
    rbu step
    rbu step

    for {set i 0} {[set i] < $nStep} {incr i} {
      if {$bReopen} {
        rbu close
        sqlite3rbu rbu $target $rbu
      }
      rbu step
      set res [rbu bp_progress]
      set expect [expr (1 + [set i]) * 10000 / $nStep]
      if {[lindex [set res] 1] != [set expect]} {
        error "Have [set res], expected 10000 [set expect]"
      }
    }

    set rc [rbu step]
    if {[set rc] != "SQLITE_DONE"} {
      error "Have [set rc], expected SQLITE_DONE" 
    }

    rbu close
  }] {SQLITE_DONE}]
}

foreach bReopen {0 1} {
  do_test 3.$bReopen.1.0 {
    reset_db
    execsql {
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b);
      CREATE TABLE t2(a INTEGER PRIMARY KEY, b);
      CREATE TABLE t3(a INTEGER PRIMARY KEY, b);
      CREATE TABLE t4(a INTEGER PRIMARY KEY, b);
    }
    create_db_file rbu.db {
      CREATE TABLE data_t1(a, b, rbu_control);
      CREATE TABLE data_t2(a, b, rbu_control);
      CREATE TABLE data_t3(a, b, rbu_control);
      CREATE TABLE data_t4(a, b, rbu_control);
      INSERT INTO data_t1 VALUES(1, 2, 0);
      INSERT INTO data_t2 VALUES(1, 2, 0);
      INSERT INTO data_t3 VALUES(1, 2, 0);
      INSERT INTO data_t4 VALUES(1, 2, 0);
  
      CREATE TABLE rbu_count(tbl, cnt);
      INSERT INTO rbu_count VALUES('data_t1', 1);
      INSERT INTO rbu_count VALUES('data_t2', 1);
      INSERT INTO rbu_count VALUES('data_t3', 1);
      INSERT INTO rbu_count VALUES('data_t4', 1);
    }
  } {}
  do_phase2_test 3.$bReopen.1.1 $bReopen test.db rbu.db 5
}


foreach {bReopen} { 0 1 } {
  foreach {tn tbl} {
    ipk { CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c) }
    wr  { CREATE TABLE t1(a INT PRIMARY KEY, b, c) WITHOUT ROWID }
    pk  { CREATE TABLE t1(a INT PRIMARY KEY, b, c) }
  } {

    foreach {tn2 rbusql r1 r3} {
      1 {
        CREATE TABLE data0_t1(a, b, c, rbu_control);
        INSERT INTO data0_t1 VALUES(15, 15, 15, 0);
        INSERT INTO data0_t1 VALUES(20, 20, 20, 0);
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 2); 
      } 
      {2500 5000 7500 10000}
      {1666 3333 5000 6666 8333 10000}

      2 {
        CREATE TABLE data0_t1(a, b, c, rbu_control);
        INSERT INTO data0_t1 VALUES(10, 10, 10, 2);
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 1); 
      } 
      {3333 6666 10000}
      {2000 4000 6000 8000 10000}

      3 {
        CREATE TABLE data0_t1(a, b, c, rbu_control);
        INSERT INTO data0_t1 VALUES(7, 7, 7, 2);
        INSERT INTO data0_t1 VALUES(10, 10, 10, 2);
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 2); 
      } 
      {2500 4000 6000 8000 10000}
      {1666 2500 3750 5000 6250 7500 8750 10000}

    } {

      reset_db ; execsql $tbl
      do_test 4.$tn.$bReopen.$tn2.0 {
        execsql {
          CREATE INDEX t1c ON t1(c);
          INSERT INTO t1 VALUES(1, 1, 1);
          INSERT INTO t1 VALUES(5, 5, 5);
          INSERT INTO t1 VALUES(10, 10, 10);
        }
        create_db_file rbu.db $rbusql
      } {}

      set R(ipk) $r1
      set R(wr) $r1
      set R(pk) $r3
      do_sp_test 4.$tn.$bReopen.$tn2.1 $bReopen test.db rbu.db $R($tn)
    }
  }
}

foreach {bReopen} { 0 1 } {
  foreach {tn tbl} {
    nopk { 
      CREATE TABLE t1(a, b, c);
      CREATE INDEX t1c ON t1(c);
    }
    vtab { 
      CREATE VIRTUAL TABLE t1 USING fts5(a, b, c);
    }
  } {

    foreach {tn2 rbusql r1 r2} {
      1 {
        CREATE TABLE data0_t1(a, b, c, rbu_rowid, rbu_control);
        INSERT INTO data0_t1 VALUES(15, 15, 15, 4, 0);
        INSERT INTO data0_t1 VALUES(20, 20, 20, 5, 0);
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 2); 
      } 
      {2500 5000 7500 10000}
      {5000 10000}

      2 {
        CREATE TABLE data0_t1(rbu_rowid, a, b, c, rbu_control);
        INSERT INTO data0_t1 VALUES(0, 7, 7, 7, 2);
        INSERT INTO data0_t1 VALUES(2, 10, 10, 10, 2);
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 2); 
      } 
      {2500 4000 6000 8000 10000}
      {5000 10000}

      3 {
        CREATE TABLE data0_t1(rbu_rowid, a, b, c, rbu_control);
        INSERT INTO data0_t1 VALUES(1, NULL, NULL, NULL, 1);
        INSERT INTO data0_t1 VALUES(2, NULL, NULL, 7, '..x');
        CREATE TABLE rbu_count(tbl, cnt);
        INSERT INTO rbu_count VALUES('data0_t1', 2); 
      } 
      {2500 4000 6000 8000 10000}
      {5000 10000}
    } {

      reset_db ; execsql $tbl
      do_test 5.$tn.$bReopen.$tn2.0 {
        execsql {
          INSERT INTO t1 VALUES(1, 1, 1);
          INSERT INTO t1 VALUES(5, 5, 5);
          INSERT INTO t1 VALUES(10, 10, 10);
        }
        create_db_file rbu.db $rbusql
      } {}

      set R(nopk) $r1
      set R(vtab) $r2
      do_sp_test 5.$tn.$bReopen.$tn2.1 $bReopen test.db rbu.db $R($tn)
    }
  }
}


finish_test

Changes to ext/rbu/sqlite3rbu.c.

** RBU_STATE_COOKIE:
**   Valid if STAGE==1. The current change-counter cookie value in the 
**   target db file.
**
** RBU_STATE_OALSZ:
**   Valid if STAGE==1. The size in bytes of the *-oal file.
*/
#define RBU_STATE_STAGE       1
#define RBU_STATE_TBL         2
#define RBU_STATE_IDX         3
#define RBU_STATE_ROW         4
#define RBU_STATE_PROGRESS    5
#define RBU_STATE_CKPT        6
#define RBU_STATE_COOKIE      7
#define RBU_STATE_OALSZ       8


#define RBU_STAGE_OAL         1
#define RBU_STAGE_MOVE        2
#define RBU_STAGE_CAPTURE     3
#define RBU_STAGE_CKPT        4
#define RBU_STAGE_DONE        5

................................................................................
  char *zTbl;
  char *zIdx;
  i64 iWalCksum;
  int nRow;
  i64 nProgress;
  u32 iCookie;
  i64 iOalSz;

};

struct RbuUpdateStmt {
  char *zMask;                    /* Copy of update mask used with pUpdate */
  sqlite3_stmt *pUpdate;          /* Last update statement (or NULL) */
  RbuUpdateStmt *pNext;
};
................................................................................
  int bCleanup;                   /* True in "cleanup" state */
  const char *zTbl;               /* Name of target db table */
  const char *zDataTbl;           /* Name of rbu db table (or null) */
  const char *zIdx;               /* Name of target db index (or null) */
  int iTnum;                      /* Root page of current object */
  int iPkTnum;                    /* If eType==EXTERNAL, root of PK index */
  int bUnique;                    /* Current index is unique */


  /* Statements created by rbuObjIterPrepareAll() */
  int nCol;                       /* Number of columns in current object */
  sqlite3_stmt *pSelect;          /* Source data */
  sqlite3_stmt *pInsert;          /* Statement for INSERT operations */
  sqlite3_stmt *pDelete;          /* Statement for DELETE ops */
  sqlite3_stmt *pTmpInsert;       /* Insert into rbu_tmp_$zDataTbl */
................................................................................
struct RbuFrame {
  u32 iDbPage;
  u32 iWalFrame;
};

/*
** RBU handle.





































*/
struct sqlite3rbu {
  int eStage;                     /* Value of RBU_STATE_STAGE field */
  sqlite3 *dbMain;                /* target database handle */
  sqlite3 *dbRbu;                 /* rbu database handle */
  char *zTarget;                  /* Path to target db */
  char *zRbu;                     /* Path to rbu db */
................................................................................
  char *zErrmsg;                  /* Error message if rc!=SQLITE_OK */
  int nStep;                      /* Rows processed for current object */
  int nProgress;                  /* Rows processed for all objects */
  RbuObjIter objiter;             /* Iterator for skipping through tbl/idx */
  const char *zVfsName;           /* Name of automatically created rbu vfs */
  rbu_file *pTargetFd;            /* File handle open on target db */
  i64 iOalSz;


  /* The following state variables are used as part of the incremental
  ** checkpoint stage (eStage==RBU_STAGE_CKPT). See comments surrounding
  ** function rbuSetupCheckpoint() for details.  */
  u32 iMaxFrame;                  /* Largest iWalFrame value in aFrame[] */
  u32 mLock;
  int nFrame;                     /* Entries in aFrame[] array */
................................................................................
  if( p->rc==SQLITE_OK ){
    memcpy(pIter->abIndexed, pIter->abTblPk, sizeof(u8)*pIter->nTblCol);
    p->rc = prepareFreeAndCollectError(p->dbMain, &pList, &p->zErrmsg,
        sqlite3_mprintf("PRAGMA main.index_list = %Q", pIter->zTbl)
    );
  }


  while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pList) ){
    const char *zIdx = (const char*)sqlite3_column_text(pList, 1);
    sqlite3_stmt *pXInfo = 0;
    if( zIdx==0 ) break;
    p->rc = prepareFreeAndCollectError(p->dbMain, &pXInfo, &p->zErrmsg,
        sqlite3_mprintf("PRAGMA main.index_xinfo = %Q", zIdx)
    );
    while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pXInfo) ){
      int iCid = sqlite3_column_int(pXInfo, 1);
      if( iCid>=0 ) pIter->abIndexed[iCid] = 1;
    }
    rbuFinalize(p, pXInfo);
    bIndex = 1;






  }

  rbuFinalize(p, pList);
  if( bIndex==0 ) pIter->abIndexed = 0;
}


................................................................................
        iOrder++;
      }
    }

    rbuFinalize(p, pStmt);
    rbuObjIterCacheIndexedCols(p, pIter);
    assert( pIter->eType!=RBU_PK_VTAB || pIter->abIndexed==0 );

  }

  return p->rc;
}

/*
** This function constructs and returns a pointer to a nul-terminated 
................................................................................
  sqlite3_context *pCtx, 
  int nVal,
  sqlite3_value **apVal
){
  sqlite3rbu *p = sqlite3_user_data(pCtx);
  int rc = SQLITE_OK;
  int i;









  for(i=0; rc==SQLITE_OK && i<nVal; i++){
    rc = sqlite3_bind_value(p->objiter.pTmpInsert, i+1, apVal[i]);
  }
  if( rc==SQLITE_OK ){
    sqlite3_step(p->objiter.pTmpInsert);
    rc = sqlite3_reset(p->objiter.pTmpInsert);
................................................................................
  RbuObjIter *pIter = &p->objiter;
  sqlite3_value *pVal;
  sqlite3_stmt *pWriter;
  int i;

  assert( p->rc==SQLITE_OK );
  assert( eType!=RBU_DELETE || pIter->zIdx==0 );












  if( eType==RBU_IDX_DELETE || eType==RBU_DELETE ){
    pWriter = pIter->pDelete;
  }else{
    pWriter = pIter->pInsert;
  }

................................................................................
  if( eType ){
    assert( eType==RBU_INSERT     || eType==RBU_DELETE
         || eType==RBU_REPLACE    || eType==RBU_IDX_DELETE
         || eType==RBU_IDX_INSERT || eType==RBU_UPDATE
    );
    assert( eType!=RBU_UPDATE || pIter->zIdx==0 );

    if( pIter->zIdx==0 && eType==RBU_IDX_DELETE ){
      rbuBadControlError(p);
    }
    else if( eType==RBU_REPLACE ){


      if( pIter->zIdx==0 ) rbuStepOneOp(p, RBU_DELETE);

      if( p->rc==SQLITE_OK ) rbuStepOneOp(p, RBU_INSERT);
    }
    else if( eType!=RBU_UPDATE ){
      rbuStepOneOp(p, eType);
    }
    else{
      sqlite3_value *pVal;
      sqlite3_stmt *pUpdate = 0;
      assert( eType==RBU_UPDATE );

      rbuGetUpdateStmt(p, pIter, zMask, &pUpdate);
      if( pUpdate ){
        int i;
        for(i=0; p->rc==SQLITE_OK && i<pIter->nCol; i++){
          char c = zMask[pIter->aiSrcOrder[i]];
          pVal = sqlite3_column_value(pIter->pSelect, i);
          if( pIter->abTblPk[i] || c!='.' ){
................................................................................
          "(%d, %d), "
          "(%d, %Q), "
          "(%d, %Q), "
          "(%d, %d), "
          "(%d, %d), "
          "(%d, %lld), "
          "(%d, %lld), "

          "(%d, %lld) ",
          p->zStateDb,
          RBU_STATE_STAGE, eStage,
          RBU_STATE_TBL, p->objiter.zTbl, 
          RBU_STATE_IDX, p->objiter.zIdx, 
          RBU_STATE_ROW, p->nStep, 
          RBU_STATE_PROGRESS, p->nProgress,
          RBU_STATE_CKPT, p->iWalCksum,
          RBU_STATE_COOKIE, (i64)p->pTargetFd->iCookie,
          RBU_STATE_OALSZ, p->iOalSz

      )
    );
    assert( pInsert==0 || rc==SQLITE_OK );

    if( rc==SQLITE_OK ){
      sqlite3_step(pInsert);
      rc = sqlite3_finalize(pInsert);
................................................................................
        pRet->iCookie = (u32)sqlite3_column_int64(pStmt, 1);
        break;

      case RBU_STATE_OALSZ:
        pRet->iOalSz = (u32)sqlite3_column_int64(pStmt, 1);
        break;





      default:
        rc = SQLITE_CORRUPT;
        break;
    }
  }
  rc2 = sqlite3_finalize(pStmt);
  if( rc==SQLITE_OK ) rc = rc2;
................................................................................
*/
static void rbuDeleteVfs(sqlite3rbu *p){
  if( p->zVfsName ){
    sqlite3rbu_destroy_vfs(p->zVfsName);
    p->zVfsName = 0;
  }
}































































































/*
** Open and return a new RBU handle. 
*/
sqlite3rbu *sqlite3rbu_open(
  const char *zTarget, 
  const char *zRbu,
................................................................................
    if( p->rc==SQLITE_OK ){
      pState = rbuLoadState(p);
      assert( pState || p->rc!=SQLITE_OK );
      if( p->rc==SQLITE_OK ){

        if( pState->eStage==0 ){ 
          rbuDeleteOalFile(p);

          p->eStage = RBU_STAGE_OAL;
        }else{
          p->eStage = pState->eStage;

        }
        p->nProgress = pState->nProgress;
        p->iOalSz = pState->iOalSz;
      }
    }
    assert( p->rc!=SQLITE_OK || p->eStage!=0 );

................................................................................
** Return the total number of key-value operations (inserts, deletes or 
** updates) that have been performed on the target database since the
** current RBU update was started.
*/
sqlite3_int64 sqlite3rbu_progress(sqlite3rbu *pRbu){
  return pRbu->nProgress;
}





































int sqlite3rbu_savestate(sqlite3rbu *p){
  int rc = p->rc;
  
  if( rc==SQLITE_DONE ) return SQLITE_OK;

  assert( p->eStage>=RBU_STAGE_OAL && p->eStage<=RBU_STAGE_DONE );








|
|
|
|
|
|
|
|
>







 







>







 







>







 







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>







 







>







 







>













>
>
>
>
>
>







 







>







 







>
>
>
>
>
>
>
>







 







>
>
>
>
>
>
>
>
>
>
>







 







|



>
>
|
>









>







 







>









|
>







 







>
>
>
>







 







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>







 







>



>







 







>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

** RBU_STATE_COOKIE:
**   Valid if STAGE==1. The current change-counter cookie value in the 
**   target db file.
**
** RBU_STATE_OALSZ:
**   Valid if STAGE==1. The size in bytes of the *-oal file.
*/
#define RBU_STATE_STAGE        1
#define RBU_STATE_TBL          2
#define RBU_STATE_IDX          3
#define RBU_STATE_ROW          4
#define RBU_STATE_PROGRESS     5
#define RBU_STATE_CKPT         6
#define RBU_STATE_COOKIE       7
#define RBU_STATE_OALSZ        8
#define RBU_STATE_PHASEONESTEP 9

#define RBU_STAGE_OAL         1
#define RBU_STAGE_MOVE        2
#define RBU_STAGE_CAPTURE     3
#define RBU_STAGE_CKPT        4
#define RBU_STAGE_DONE        5

................................................................................
  char *zTbl;
  char *zIdx;
  i64 iWalCksum;
  int nRow;
  i64 nProgress;
  u32 iCookie;
  i64 iOalSz;
  i64 nPhaseOneStep;
};

struct RbuUpdateStmt {
  char *zMask;                    /* Copy of update mask used with pUpdate */
  sqlite3_stmt *pUpdate;          /* Last update statement (or NULL) */
  RbuUpdateStmt *pNext;
};
................................................................................
  int bCleanup;                   /* True in "cleanup" state */
  const char *zTbl;               /* Name of target db table */
  const char *zDataTbl;           /* Name of rbu db table (or null) */
  const char *zIdx;               /* Name of target db index (or null) */
  int iTnum;                      /* Root page of current object */
  int iPkTnum;                    /* If eType==EXTERNAL, root of PK index */
  int bUnique;                    /* Current index is unique */
  int nIndex;                     /* Number of aux. indexes on table zTbl */

  /* Statements created by rbuObjIterPrepareAll() */
  int nCol;                       /* Number of columns in current object */
  sqlite3_stmt *pSelect;          /* Source data */
  sqlite3_stmt *pInsert;          /* Statement for INSERT operations */
  sqlite3_stmt *pDelete;          /* Statement for DELETE ops */
  sqlite3_stmt *pTmpInsert;       /* Insert into rbu_tmp_$zDataTbl */
................................................................................
struct RbuFrame {
  u32 iDbPage;
  u32 iWalFrame;
};

/*
** RBU handle.
**
** nPhaseOneStep:
**   If the RBU database contains an rbu_count table, this value is set to
**   a running estimate of the number of b-tree operations required to 
**   finish populating the *-oal file. This allows the sqlite3_bp_progress()
**   API to calculate the permyriadage progress of populating the *-oal file
**   using the formula:
**
**     permyriadage = (10000 * nProgress) / nPhaseOneStep
**
**   nPhaseOneStep is initialized to the sum of:
**
**     nRow * (nIndex + 1)
**
**   for all source tables in the RBU database, where nRow is the number
**   of rows in the source table and nIndex the number of indexes on the
**   corresponding target database table.
**
**   This estimate is accurate if the RBU update consists entirely of
**   INSERT operations. However, it is inaccurate if:
**
**     * the RBU update contains any UPDATE operations. If the PK specified
**       for an UPDATE operation does not exist in the target table, then
**       no b-tree operations are required on index b-trees. Or if the 
**       specified PK does exist, then (nIndex*2) such operations are
**       required (one delete and one insert on each index b-tree).
**
**     * the RBU update contains any DELETE operations for which the specified
**       PK does not exist. In this case no operations are required on index
**       b-trees.
**
**     * the RBU update contains REPLACE operations. These are similar to
**       UPDATE operations.
**
**   nPhaseOneStep is updated to account for the conditions above during the
**   first pass of each source table. The updated nPhaseOneStep value is
**   stored in the rbu_state table if the RBU update is suspended.
*/
struct sqlite3rbu {
  int eStage;                     /* Value of RBU_STATE_STAGE field */
  sqlite3 *dbMain;                /* target database handle */
  sqlite3 *dbRbu;                 /* rbu database handle */
  char *zTarget;                  /* Path to target db */
  char *zRbu;                     /* Path to rbu db */
................................................................................
  char *zErrmsg;                  /* Error message if rc!=SQLITE_OK */
  int nStep;                      /* Rows processed for current object */
  int nProgress;                  /* Rows processed for all objects */
  RbuObjIter objiter;             /* Iterator for skipping through tbl/idx */
  const char *zVfsName;           /* Name of automatically created rbu vfs */
  rbu_file *pTargetFd;            /* File handle open on target db */
  i64 iOalSz;
  i64 nPhaseOneStep;

  /* The following state variables are used as part of the incremental
  ** checkpoint stage (eStage==RBU_STAGE_CKPT). See comments surrounding
  ** function rbuSetupCheckpoint() for details.  */
  u32 iMaxFrame;                  /* Largest iWalFrame value in aFrame[] */
  u32 mLock;
  int nFrame;                     /* Entries in aFrame[] array */
................................................................................
  if( p->rc==SQLITE_OK ){
    memcpy(pIter->abIndexed, pIter->abTblPk, sizeof(u8)*pIter->nTblCol);
    p->rc = prepareFreeAndCollectError(p->dbMain, &pList, &p->zErrmsg,
        sqlite3_mprintf("PRAGMA main.index_list = %Q", pIter->zTbl)
    );
  }

  pIter->nIndex = 0;
  while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pList) ){
    const char *zIdx = (const char*)sqlite3_column_text(pList, 1);
    sqlite3_stmt *pXInfo = 0;
    if( zIdx==0 ) break;
    p->rc = prepareFreeAndCollectError(p->dbMain, &pXInfo, &p->zErrmsg,
        sqlite3_mprintf("PRAGMA main.index_xinfo = %Q", zIdx)
    );
    while( p->rc==SQLITE_OK && SQLITE_ROW==sqlite3_step(pXInfo) ){
      int iCid = sqlite3_column_int(pXInfo, 1);
      if( iCid>=0 ) pIter->abIndexed[iCid] = 1;
    }
    rbuFinalize(p, pXInfo);
    bIndex = 1;
    pIter->nIndex++;
  }

  if( pIter->eType==RBU_PK_WITHOUT_ROWID ){
    /* "PRAGMA index_list" includes the main PK b-tree */
    pIter->nIndex--;
  }

  rbuFinalize(p, pList);
  if( bIndex==0 ) pIter->abIndexed = 0;
}


................................................................................
        iOrder++;
      }
    }

    rbuFinalize(p, pStmt);
    rbuObjIterCacheIndexedCols(p, pIter);
    assert( pIter->eType!=RBU_PK_VTAB || pIter->abIndexed==0 );
    assert( pIter->eType!=RBU_PK_VTAB || pIter->nIndex==0 );
  }

  return p->rc;
}

/*
** This function constructs and returns a pointer to a nul-terminated 
................................................................................
  sqlite3_context *pCtx, 
  int nVal,
  sqlite3_value **apVal
){
  sqlite3rbu *p = sqlite3_user_data(pCtx);
  int rc = SQLITE_OK;
  int i;

  assert( sqlite3_value_int(apVal[0])!=0
      || p->objiter.eType==RBU_PK_EXTERNAL 
      || p->objiter.eType==RBU_PK_NONE 
  );
  if( sqlite3_value_int(apVal[0])!=0 ){
    p->nPhaseOneStep += p->objiter.nIndex;
  }

  for(i=0; rc==SQLITE_OK && i<nVal; i++){
    rc = sqlite3_bind_value(p->objiter.pTmpInsert, i+1, apVal[i]);
  }
  if( rc==SQLITE_OK ){
    sqlite3_step(p->objiter.pTmpInsert);
    rc = sqlite3_reset(p->objiter.pTmpInsert);
................................................................................
  RbuObjIter *pIter = &p->objiter;
  sqlite3_value *pVal;
  sqlite3_stmt *pWriter;
  int i;

  assert( p->rc==SQLITE_OK );
  assert( eType!=RBU_DELETE || pIter->zIdx==0 );
  assert( eType==RBU_DELETE || eType==RBU_IDX_DELETE
       || eType==RBU_INSERT || eType==RBU_IDX_INSERT
  );

  /* If this is a delete, decrement nPhaseOneStep by nIndex. If the DELETE
  ** statement below does actually delete a row, nPhaseOneStep will be
  ** incremented by the same amount when SQL function rbu_tmp_insert()
  ** is invoked by the trigger.  */
  if( eType==RBU_DELETE ){
    p->nPhaseOneStep -= p->objiter.nIndex;
  }

  if( eType==RBU_IDX_DELETE || eType==RBU_DELETE ){
    pWriter = pIter->pDelete;
  }else{
    pWriter = pIter->pInsert;
  }

................................................................................
  if( eType ){
    assert( eType==RBU_INSERT     || eType==RBU_DELETE
         || eType==RBU_REPLACE    || eType==RBU_IDX_DELETE
         || eType==RBU_IDX_INSERT || eType==RBU_UPDATE
    );
    assert( eType!=RBU_UPDATE || pIter->zIdx==0 );

    if( pIter->zIdx==0 && (eType==RBU_IDX_DELETE || eType==RBU_IDX_INSERT) ){
      rbuBadControlError(p);
    }
    else if( eType==RBU_REPLACE ){
      if( pIter->zIdx==0 ){
        p->nPhaseOneStep += p->objiter.nIndex;
        rbuStepOneOp(p, RBU_DELETE);
      }
      if( p->rc==SQLITE_OK ) rbuStepOneOp(p, RBU_INSERT);
    }
    else if( eType!=RBU_UPDATE ){
      rbuStepOneOp(p, eType);
    }
    else{
      sqlite3_value *pVal;
      sqlite3_stmt *pUpdate = 0;
      assert( eType==RBU_UPDATE );
      p->nPhaseOneStep -= p->objiter.nIndex;
      rbuGetUpdateStmt(p, pIter, zMask, &pUpdate);
      if( pUpdate ){
        int i;
        for(i=0; p->rc==SQLITE_OK && i<pIter->nCol; i++){
          char c = zMask[pIter->aiSrcOrder[i]];
          pVal = sqlite3_column_value(pIter->pSelect, i);
          if( pIter->abTblPk[i] || c!='.' ){
................................................................................
          "(%d, %d), "
          "(%d, %Q), "
          "(%d, %Q), "
          "(%d, %d), "
          "(%d, %d), "
          "(%d, %lld), "
          "(%d, %lld), "
          "(%d, %lld), "
          "(%d, %lld) ",
          p->zStateDb,
          RBU_STATE_STAGE, eStage,
          RBU_STATE_TBL, p->objiter.zTbl, 
          RBU_STATE_IDX, p->objiter.zIdx, 
          RBU_STATE_ROW, p->nStep, 
          RBU_STATE_PROGRESS, p->nProgress,
          RBU_STATE_CKPT, p->iWalCksum,
          RBU_STATE_COOKIE, (i64)p->pTargetFd->iCookie,
          RBU_STATE_OALSZ, p->iOalSz,
          RBU_STATE_PHASEONESTEP, p->nPhaseOneStep
      )
    );
    assert( pInsert==0 || rc==SQLITE_OK );

    if( rc==SQLITE_OK ){
      sqlite3_step(pInsert);
      rc = sqlite3_finalize(pInsert);
................................................................................
        pRet->iCookie = (u32)sqlite3_column_int64(pStmt, 1);
        break;

      case RBU_STATE_OALSZ:
        pRet->iOalSz = (u32)sqlite3_column_int64(pStmt, 1);
        break;

      case RBU_STATE_PHASEONESTEP:
        pRet->nPhaseOneStep = sqlite3_column_int64(pStmt, 1);
        break;

      default:
        rc = SQLITE_CORRUPT;
        break;
    }
  }
  rc2 = sqlite3_finalize(pStmt);
  if( rc==SQLITE_OK ) rc = rc2;
................................................................................
*/
static void rbuDeleteVfs(sqlite3rbu *p){
  if( p->zVfsName ){
    sqlite3rbu_destroy_vfs(p->zVfsName);
    p->zVfsName = 0;
  }
}

/*
** This user-defined SQL function is invoked with a single argument - the
** name of a table expected to appear in the target database. It returns
** the number of auxilliary indexes on the table.
*/
static void rbuIndexCntFunc(
  sqlite3_context *pCtx, 
  int nVal,
  sqlite3_value **apVal
){
  sqlite3rbu *p = (sqlite3rbu*)sqlite3_user_data(pCtx);
  sqlite3_stmt *pStmt = 0;
  char *zErrmsg = 0;
  int rc;

  assert( nVal==1 );
  
  rc = prepareFreeAndCollectError(p->dbMain, &pStmt, &zErrmsg, 
      sqlite3_mprintf("SELECT count(*) FROM sqlite_master "
        "WHERE type='index' AND tbl_name = %Q", sqlite3_value_text(apVal[0]))
  );
  if( rc!=SQLITE_OK ){
    sqlite3_result_error(pCtx, zErrmsg, -1);
  }else{
    int nIndex = 0;
    if( SQLITE_ROW==sqlite3_step(pStmt) ){
      nIndex = sqlite3_column_int(pStmt, 0);
    }
    rc = sqlite3_finalize(pStmt);
    if( rc==SQLITE_OK ){
      sqlite3_result_int(pCtx, nIndex);
    }else{
      sqlite3_result_error(pCtx, sqlite3_errmsg(p->dbMain), -1);
    }
  }

  sqlite3_free(zErrmsg);
}

/*
** If the RBU database contains the rbu_count table, use it to initialize
** the sqlite3rbu.nPhaseOneStep variable. The schema of the rbu_count table
** is assumed to contain the same columns as:
**
**   CREATE TABLE rbu_count(tbl TEXT PRIMARY KEY, cnt INTEGER) WITHOUT ROWID;
**
** There should be one row in the table for each data_xxx table in the
** database. The 'tbl' column should contain the name of a data_xxx table,
** and the cnt column the number of rows it contains.
**
** sqlite3rbu.nPhaseOneStep is initialized to the sum of (1 + nIndex) * cnt
** for all rows in the rbu_count table, where nIndex is the number of 
** indexes on the corresponding target database table.
*/
static void rbuInitPhaseOneSteps(sqlite3rbu *p){
  if( p->rc==SQLITE_OK ){
    sqlite3_stmt *pStmt = 0;
    int bExists = 0;                /* True if rbu_count exists */

    p->nPhaseOneStep = -1;

    p->rc = sqlite3_create_function(p->dbRbu, 
        "rbu_index_cnt", 1, SQLITE_UTF8, (void*)p, rbuIndexCntFunc, 0, 0
    );
  
    /* Check for the rbu_count table. If it does not exist, or if an error
    ** occurs, nPhaseOneStep will be left set to -1. */
    if( p->rc==SQLITE_OK ){
      p->rc = prepareAndCollectError(p->dbRbu, &pStmt, &p->zErrmsg,
          "SELECT 1 FROM sqlite_master WHERE tbl_name = 'rbu_count'"
      );
    }
    if( p->rc==SQLITE_OK ){
      if( SQLITE_ROW==sqlite3_step(pStmt) ){
        bExists = 1;
      }
      p->rc = sqlite3_finalize(pStmt);
    }
  
    if( p->rc==SQLITE_OK && bExists ){
      p->rc = prepareAndCollectError(p->dbRbu, &pStmt, &p->zErrmsg,
          "SELECT sum(cnt * (1 + rbu_index_cnt(rbu_target_name(tbl))))"
          "FROM rbu_count"
      );
      if( p->rc==SQLITE_OK ){
        if( SQLITE_ROW==sqlite3_step(pStmt) ){
          p->nPhaseOneStep = sqlite3_column_int64(pStmt, 0);
        }
        p->rc = sqlite3_finalize(pStmt);
      }
    }
  }
}

/*
** Open and return a new RBU handle. 
*/
sqlite3rbu *sqlite3rbu_open(
  const char *zTarget, 
  const char *zRbu,
................................................................................
    if( p->rc==SQLITE_OK ){
      pState = rbuLoadState(p);
      assert( pState || p->rc!=SQLITE_OK );
      if( p->rc==SQLITE_OK ){

        if( pState->eStage==0 ){ 
          rbuDeleteOalFile(p);
          rbuInitPhaseOneSteps(p);
          p->eStage = RBU_STAGE_OAL;
        }else{
          p->eStage = pState->eStage;
          p->nPhaseOneStep = pState->nPhaseOneStep;
        }
        p->nProgress = pState->nProgress;
        p->iOalSz = pState->iOalSz;
      }
    }
    assert( p->rc!=SQLITE_OK || p->eStage!=0 );

................................................................................
** Return the total number of key-value operations (inserts, deletes or 
** updates) that have been performed on the target database since the
** current RBU update was started.
*/
sqlite3_int64 sqlite3rbu_progress(sqlite3rbu *pRbu){
  return pRbu->nProgress;
}

/*
** Return permyriadage progress indications for the two main stages of
** an RBU update.
*/
void sqlite3rbu_bp_progress(sqlite3rbu *p, int *pnOne, int *pnTwo){
  const int MAX_PROGRESS = 10000;
  switch( p->eStage ){
    case RBU_STAGE_OAL:
      if( p->nPhaseOneStep>0 ){
        *pnOne = (int)(MAX_PROGRESS * (i64)p->nProgress/(i64)p->nPhaseOneStep);
      }else{
        *pnOne = -1;
      }
      *pnTwo = 0;
      break;

    case RBU_STAGE_MOVE:
      *pnOne = MAX_PROGRESS;
      *pnTwo = 0;
      break;

    case RBU_STAGE_CKPT:
      *pnOne = MAX_PROGRESS;
      *pnTwo = (int)(MAX_PROGRESS * (i64)p->nStep / (i64)p->nFrame);
      break;

    case RBU_STAGE_DONE:
      *pnOne = MAX_PROGRESS;
      *pnTwo = MAX_PROGRESS;
      break;

    default:
      assert( 0 );
  }
}

int sqlite3rbu_savestate(sqlite3rbu *p){
  int rc = p->rc;
  
  if( rc==SQLITE_DONE ) return SQLITE_OK;

  assert( p->eStage>=RBU_STAGE_OAL && p->eStage<=RBU_STAGE_DONE );

Changes to ext/rbu/sqlite3rbu.h.

/*
** Return the total number of key-value operations (inserts, deletes or 
** updates) that have been performed on the target database since the
** current RBU update was started.
*/
sqlite3_int64 sqlite3rbu_progress(sqlite3rbu *pRbu);











































/*
** Create an RBU VFS named zName that accesses the underlying file-system
** via existing VFS zParent. Or, if the zParent parameter is passed NULL, 
** then the new RBU VFS uses the default system VFS to access the file-system.
** The new object is registered as a non-default VFS with SQLite before 
** returning.
**








>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>

/*
** Return the total number of key-value operations (inserts, deletes or 
** updates) that have been performed on the target database since the
** current RBU update was started.
*/
sqlite3_int64 sqlite3rbu_progress(sqlite3rbu *pRbu);

/*
** Obtain permyriadage (permyriadage is to 10000 as percentage is to 100) 
** progress indications for the two stages of an RBU update. This API may
** be useful for driving GUI progress indicators and similar.
**
** An RBU update is divided into two stages:
**
**   * Stage 1, in which changes are accumulated in an oal/wal file, and
**   * Stage 2, in which the contents of the wal file are copied into the
**     main database.
**
** The update is visible to non-RBU clients during stage 2. During stage 1
** non-RBU reader clients may see the original database.
**
** If this API is called during stage 2 of the update, output variable 
** (*pnOne) is set to 10000 to indicate that stage 1 has finished and (*pnTwo)
** to a value between 0 and 10000 to indicate the permyriadage progress of
** stage 2. A value of 5000 indicates that stage 2 is half finished, 
** 9000 indicates that it is 90% finished, and so on.
**
** If this API is called during stage 1 of the update, output variable 
** (*pnTwo) is set to 0 to indicate that stage 2 has not yet started. The
** value to which (*pnOne) is set depends on whether or not the RBU 
** database contains an "rbu_count" table. The rbu_count table, if it 
** exists, must contain the same columns as the following:
**
**   CREATE TABLE rbu_count(tbl TEXT PRIMARY KEY, cnt INTEGER) WITHOUT ROWID;
**
** There must be one row in the table for each source (data_xxx) table within
** the RBU database. The 'tbl' column should contain the name of the source
** table. The 'cnt' column should contain the number of rows within the
** source table.
**
** If the rbu_count table is present and populated correctly and this
** API is called during stage 1, the *pnOne output variable is set to the
** permyriadage progress of the same stage. If the rbu_count table does
** not exist, then (*pnOne) is set to -1 during stage 1. If the rbu_count
** table exists but is not correctly populated, the value of the *pnOne
** output variable during stage 1 is undefined.
*/
void sqlite3rbu_bp_progress(sqlite3rbu *pRbu, int *pnOne, int *pnTwo);

/*
** Create an RBU VFS named zName that accesses the underlying file-system
** via existing VFS zParent. Or, if the zParent parameter is passed NULL, 
** then the new RBU VFS uses the default system VFS to access the file-system.
** The new object is registered as a non-default VFS with SQLite before 
** returning.
**

Changes to ext/rbu/test_rbu.c.

    const char *zUsage;
  } aCmd[] = {
    {"step", 2, ""},              /* 0 */
    {"close", 2, ""},             /* 1 */
    {"create_rbu_delta", 2, ""},  /* 2 */
    {"savestate", 2, ""},         /* 3 */
    {"dbMain_eval", 3, "SQL"},    /* 4 */

    {0,0,0}
  };
  int iCmd;

  if( objc<2 ){
    Tcl_WrongNumArgs(interp, 1, objv, "METHOD");
    return TCL_ERROR;
................................................................................
      int rc = sqlite3_exec(db, Tcl_GetString(objv[2]), 0, 0, 0);
      if( rc!=SQLITE_OK ){
        Tcl_SetObjResult(interp, Tcl_NewStringObj(sqlite3_errmsg(db), -1));
        ret = TCL_ERROR;
      }
      break;
    }













    default: /* seems unlikely */
      assert( !"cannot happen" );
      break;
  }

  return ret;








>







 







>
>
>
>
>
>
>
>
>
>
>
>

    const char *zUsage;
  } aCmd[] = {
    {"step", 2, ""},              /* 0 */
    {"close", 2, ""},             /* 1 */
    {"create_rbu_delta", 2, ""},  /* 2 */
    {"savestate", 2, ""},         /* 3 */
    {"dbMain_eval", 3, "SQL"},    /* 4 */
    {"bp_progress", 2, ""},    /* 5 */
    {0,0,0}
  };
  int iCmd;

  if( objc<2 ){
    Tcl_WrongNumArgs(interp, 1, objv, "METHOD");
    return TCL_ERROR;
................................................................................
      int rc = sqlite3_exec(db, Tcl_GetString(objv[2]), 0, 0, 0);
      if( rc!=SQLITE_OK ){
        Tcl_SetObjResult(interp, Tcl_NewStringObj(sqlite3_errmsg(db), -1));
        ret = TCL_ERROR;
      }
      break;
    }

    case 5: /* bp_progress */ {
      int one, two;
      Tcl_Obj *pObj;
      sqlite3rbu_bp_progress(pRbu, &one, &two);

      pObj = Tcl_NewObj();
      Tcl_ListObjAppendElement(interp, pObj, Tcl_NewIntObj(one));
      Tcl_ListObjAppendElement(interp, pObj, Tcl_NewIntObj(two));
      Tcl_SetObjResult(interp, pObj);
      break;
    }

    default: /* seems unlikely */
      assert( !"cannot happen" );
      break;
  }

  return ret;

Changes to src/loadext.c.

  sqlite3_bind_zeroblob64,
  /* Version 3.9.0 and later */
  sqlite3_value_subtype,
  sqlite3_result_subtype,
  /* Version 3.10.0 and later */
  sqlite3_status64,
  sqlite3_strlike,
  sqlite3_db_cacheflush


};

/*
** Attempt to load an SQLite extension library contained in the file
** zFile.  The entry point is zProc.  zProc may be 0 in which case a
** default entry point name (sqlite3_extension_init) is used.  Use
** of the default name is recommended.








|
>
>

  sqlite3_bind_zeroblob64,
  /* Version 3.9.0 and later */
  sqlite3_value_subtype,
  sqlite3_result_subtype,
  /* Version 3.10.0 and later */
  sqlite3_status64,
  sqlite3_strlike,
  sqlite3_db_cacheflush,
  /* Version 3.12.0 and later */
  sqlite3_system_errno
};

/*
** Attempt to load an SQLite extension library contained in the file
** zFile.  The entry point is zProc.  zProc may be 0 in which case a
** default entry point name (sqlite3_extension_init) is used.  Use
** of the default name is recommended.

Changes to src/main.c.

    return SQLITE_MISUSE_BKPT;
  }
  if( !db || db->mallocFailed ){
    return SQLITE_NOMEM_BKPT;
  }
  return db->errCode;
}




/*
** Return a string that describes the kind of error specified in the
** argument.  For now, this simply calls the internal sqlite3ErrStr()
** function.
*/
const char *sqlite3_errstr(int rc){








>
>
>

    return SQLITE_MISUSE_BKPT;
  }
  if( !db || db->mallocFailed ){
    return SQLITE_NOMEM_BKPT;
  }
  return db->errCode;
}
int sqlite3_system_errno(sqlite3 *db){
  return db ? db->iSysErrno : 0;
}  

/*
** Return a string that describes the kind of error specified in the
** argument.  For now, this simply calls the internal sqlite3ErrStr()
** function.
*/
const char *sqlite3_errstr(int rc){

Changes to src/os.c.

#endif /* SQLITE_OMIT_LOAD_EXTENSION */
int sqlite3OsRandomness(sqlite3_vfs *pVfs, int nByte, char *zBufOut){
  return pVfs->xRandomness(pVfs, nByte, zBufOut);
}
int sqlite3OsSleep(sqlite3_vfs *pVfs, int nMicro){
  return pVfs->xSleep(pVfs, nMicro);
}



int sqlite3OsCurrentTimeInt64(sqlite3_vfs *pVfs, sqlite3_int64 *pTimeOut){
  int rc;
  /* IMPLEMENTATION-OF: R-49045-42493 SQLite will use the xCurrentTimeInt64()
  ** method to get the current date and time if that method is available
  ** (if iVersion is 2 or greater and the function pointer is not NULL) and
  ** will fall back to xCurrentTime() if xCurrentTimeInt64() is
  ** unavailable.








>
>
>

#endif /* SQLITE_OMIT_LOAD_EXTENSION */
int sqlite3OsRandomness(sqlite3_vfs *pVfs, int nByte, char *zBufOut){
  return pVfs->xRandomness(pVfs, nByte, zBufOut);
}
int sqlite3OsSleep(sqlite3_vfs *pVfs, int nMicro){
  return pVfs->xSleep(pVfs, nMicro);
}
int sqlite3OsGetLastError(sqlite3_vfs *pVfs){
  return pVfs->xGetLastError ? pVfs->xGetLastError(pVfs, 0, 0) : 0;
}
int sqlite3OsCurrentTimeInt64(sqlite3_vfs *pVfs, sqlite3_int64 *pTimeOut){
  int rc;
  /* IMPLEMENTATION-OF: R-49045-42493 SQLite will use the xCurrentTimeInt64()
  ** method to get the current date and time if that method is available
  ** (if iVersion is 2 or greater and the function pointer is not NULL) and
  ** will fall back to xCurrentTime() if xCurrentTimeInt64() is
  ** unavailable.

Changes to src/os.h.

void *sqlite3OsDlOpen(sqlite3_vfs *, const char *);
void sqlite3OsDlError(sqlite3_vfs *, int, char *);
void (*sqlite3OsDlSym(sqlite3_vfs *, void *, const char *))(void);
void sqlite3OsDlClose(sqlite3_vfs *, void *);
#endif /* SQLITE_OMIT_LOAD_EXTENSION */
int sqlite3OsRandomness(sqlite3_vfs *, int, char *);
int sqlite3OsSleep(sqlite3_vfs *, int);

int sqlite3OsCurrentTimeInt64(sqlite3_vfs *, sqlite3_int64*);

/*
** Convenience functions for opening and closing files using 
** sqlite3_malloc() to obtain space for the file-handle structure.
*/
int sqlite3OsOpenMalloc(sqlite3_vfs *, const char *, sqlite3_file **, int,int*);
int sqlite3OsCloseFree(sqlite3_file *);

#endif /* _SQLITE_OS_H_ */

void *sqlite3OsDlOpen(sqlite3_vfs *, const char *);
void sqlite3OsDlError(sqlite3_vfs *, int, char *);
void (*sqlite3OsDlSym(sqlite3_vfs *, void *, const char *))(void);
void sqlite3OsDlClose(sqlite3_vfs *, void *);
#endif /* SQLITE_OMIT_LOAD_EXTENSION */
int sqlite3OsRandomness(sqlite3_vfs *, int, char *);
int sqlite3OsSleep(sqlite3_vfs *, int);
int sqlite3OsGetLastError(sqlite3_vfs*);
int sqlite3OsCurrentTimeInt64(sqlite3_vfs *, sqlite3_int64*);

/*
** Convenience functions for opening and closing files using 
** sqlite3_malloc() to obtain space for the file-handle structure.
*/
int sqlite3OsOpenMalloc(sqlite3_vfs *, const char *, sqlite3_file **, int,int*);
int sqlite3OsCloseFree(sqlite3_file *);

#endif /* _SQLITE_OS_H_ */

Changes to src/os_unix.c.

  *prNow = i/86400000.0;
  return rc;
}
#else
# define unixCurrentTime 0
#endif

#ifndef SQLITE_OMIT_DEPRECATED
/*
** We added the xGetLastError() method with the intention of providing
** better low-level error messages when operating-system problems come up
** during SQLite operation.  But so far, none of that has been implemented
** in the core.  So this routine is never called.  For now, it is merely
** a place-holder.

*/
static int unixGetLastError(sqlite3_vfs *NotUsed, int NotUsed2, char *NotUsed3){
  UNUSED_PARAMETER(NotUsed);
  UNUSED_PARAMETER(NotUsed2);
  UNUSED_PARAMETER(NotUsed3);
  return 0;
}
#else
# define unixGetLastError 0
#endif


/*
************************ End of sqlite3_vfs methods ***************************
******************************************************************************/

/******************************************************************************








<

|
|
|
<
<
>





|

<
<
<

  *prNow = i/86400000.0;
  return rc;
}
#else
# define unixCurrentTime 0
#endif


/*
** The xGetLastError() method is designed to return a better
** low-level error message when operating-system problems come up
** during SQLite operation.  Only the integer return code is currently


** used.
*/
static int unixGetLastError(sqlite3_vfs *NotUsed, int NotUsed2, char *NotUsed3){
  UNUSED_PARAMETER(NotUsed);
  UNUSED_PARAMETER(NotUsed2);
  UNUSED_PARAMETER(NotUsed3);
  return errno;
}





/*
************************ End of sqlite3_vfs methods ***************************
******************************************************************************/

/******************************************************************************

Changes to src/os_win.c.

**   }
**
** However if an error message is supplied, it will be incorporated
** by sqlite into the error message available to the user using
** sqlite3_errmsg(), possibly making IO errors easier to debug.
*/
static int winGetLastError(sqlite3_vfs *pVfs, int nBuf, char *zBuf){

  UNUSED_PARAMETER(pVfs);
  return winGetLastErrorMsg(osGetLastError(), nBuf, zBuf);

}

/*
** Initialize and deinitialize the operating system interface.
*/
int sqlite3_os_init(void){
  static sqlite3_vfs winVfs = {








>

|
>

**   }
**
** However if an error message is supplied, it will be incorporated
** by sqlite into the error message available to the user using
** sqlite3_errmsg(), possibly making IO errors easier to debug.
*/
static int winGetLastError(sqlite3_vfs *pVfs, int nBuf, char *zBuf){
  DWORD e = osGetLastError();
  UNUSED_PARAMETER(pVfs);
  if( nBuf>0 ) winGetLastErrorMsg(e, nBuf, zBuf);
  return e;
}

/*
** Initialize and deinitialize the operating system interface.
*/
int sqlite3_os_init(void){
  static sqlite3_vfs winVfs = {

Changes to src/sqlite.h.in.

** ^Otherwise, if no error occurs, [sqlite3_db_cacheflush()] returns SQLITE_OK.
**
** ^This function does not set the database handle error code or message
** returned by the [sqlite3_errcode()] and [sqlite3_errmsg()] functions.
*/
int sqlite3_db_cacheflush(sqlite3*);













/*
** CAPI3REF: Database Snapshot
** KEYWORDS: {snapshot}
** EXPERIMENTAL
**
** An instance of the snapshot object records the state of a [WAL mode]
** database for some specific point in history.








>
>
>
>
>
>
>
>
>
>
>
>

** ^Otherwise, if no error occurs, [sqlite3_db_cacheflush()] returns SQLITE_OK.
**
** ^This function does not set the database handle error code or message
** returned by the [sqlite3_errcode()] and [sqlite3_errmsg()] functions.
*/
int sqlite3_db_cacheflush(sqlite3*);

/*
** CAPI3REF: Low-level system error code
**
** ^Attempt to return the underlying operating system error code or error
** number that caused the most reason I/O error or failure to open a file.
** The return value is OS-dependent.  For example, on unix systems, after
** [sqlite3_open_v2()] returns [SQLITE_CANTOPEN], this interface could be
** called to get back the underlying "errno" that caused the problem, such
** as ENOSPC, EAUTH, EISDIR, and so forth.  
*/
int sqlite3_system_errno(sqlite3*);

/*
** CAPI3REF: Database Snapshot
** KEYWORDS: {snapshot}
** EXPERIMENTAL
**
** An instance of the snapshot object records the state of a [WAL mode]
** database for some specific point in history.

Changes to src/sqlite3ext.h.

  /* Version 3.9.0 and later */
  unsigned int (*value_subtype)(sqlite3_value*);
  void (*result_subtype)(sqlite3_context*,unsigned int);
  /* Version 3.10.0 and later */
  int (*status64)(int,sqlite3_int64*,sqlite3_int64*,int);
  int (*strlike)(const char*,const char*,unsigned int);
  int (*db_cacheflush)(sqlite3*);


};

/*
** The following macros redefine the API routines so that they are
** redirected through the global sqlite3_api structure.
**
** This header file is also used by the loadext.c source file
................................................................................
/* Version 3.9.0 and later */
#define sqlite3_value_subtype          sqlite3_api->value_subtype
#define sqlite3_result_subtype         sqlite3_api->result_subtype
/* Version 3.10.0 and later */
#define sqlite3_status64               sqlite3_api->status64
#define sqlite3_strlike                sqlite3_api->strlike
#define sqlite3_db_cacheflush          sqlite3_api->db_cacheflush


#endif /* !defined(SQLITE_CORE) && !defined(SQLITE_OMIT_LOAD_EXTENSION) */

#if !defined(SQLITE_CORE) && !defined(SQLITE_OMIT_LOAD_EXTENSION)
  /* This case when the file really is being compiled as a loadable 
  ** extension */
# define SQLITE_EXTENSION_INIT1     const sqlite3_api_routines *sqlite3_api=0;
# define SQLITE_EXTENSION_INIT2(v)  sqlite3_api=v;

  /* Version 3.9.0 and later */
  unsigned int (*value_subtype)(sqlite3_value*);
  void (*result_subtype)(sqlite3_context*,unsigned int);
  /* Version 3.10.0 and later */
  int (*status64)(int,sqlite3_int64*,sqlite3_int64*,int);
  int (*strlike)(const char*,const char*,unsigned int);
  int (*db_cacheflush)(sqlite3*);
  /* Version 3.12.0 and later */
  int (*system_errno)(sqlite3*);
};

/*
** The following macros redefine the API routines so that they are
** redirected through the global sqlite3_api structure.
**
** This header file is also used by the loadext.c source file
................................................................................
/* Version 3.9.0 and later */
#define sqlite3_value_subtype          sqlite3_api->value_subtype
#define sqlite3_result_subtype         sqlite3_api->result_subtype
/* Version 3.10.0 and later */
#define sqlite3_status64               sqlite3_api->status64
#define sqlite3_strlike                sqlite3_api->strlike
#define sqlite3_db_cacheflush          sqlite3_api->db_cacheflush
/* Version 3.12.0 and later */
#define sqlite3_system_errno           sqlite3_api->system_errno
#endif /* !defined(SQLITE_CORE) && !defined(SQLITE_OMIT_LOAD_EXTENSION) */

#if !defined(SQLITE_CORE) && !defined(SQLITE_OMIT_LOAD_EXTENSION)
  /* This case when the file really is being compiled as a loadable 
  ** extension */
# define SQLITE_EXTENSION_INIT1     const sqlite3_api_routines *sqlite3_api=0;
# define SQLITE_EXTENSION_INIT2(v)  sqlite3_api=v;

Changes to src/sqliteInt.h.

  int nDb;                      /* Number of backends currently in use */
  int flags;                    /* Miscellaneous flags. See below */
  i64 lastRowid;                /* ROWID of most recent insert (see above) */
  i64 szMmap;                   /* Default mmap_size setting */
  unsigned int openFlags;       /* Flags passed to sqlite3_vfs.xOpen() */
  int errCode;                  /* Most recent error code (SQLITE_*) */
  int errMask;                  /* & result codes with this before returning */

  u16 dbOptFlags;               /* Flags to enable/disable optimizations */
  u8 enc;                       /* Text encoding */
  u8 autoCommit;                /* The auto-commit flag. */
  u8 temp_store;                /* 1: file 2: memory 0: default */
  u8 mallocFailed;              /* True if we have seen a malloc failure */
  u8 bBenignMalloc;             /* Do not require OOMs if true */
  u8 dfltLockMode;              /* Default locking-mode for attached dbs */
................................................................................
char sqlite3CompareAffinity(Expr *pExpr, char aff2);
int sqlite3IndexAffinityOk(Expr *pExpr, char idx_affinity);
char sqlite3ExprAffinity(Expr *pExpr);
int sqlite3Atoi64(const char*, i64*, int, u8);
int sqlite3DecOrHexToI64(const char*, i64*);
void sqlite3ErrorWithMsg(sqlite3*, int, const char*,...);
void sqlite3Error(sqlite3*,int);

void *sqlite3HexToBlob(sqlite3*, const char *z, int n);
u8 sqlite3HexToInt(int h);
int sqlite3TwoPartName(Parse *, Token *, Token *, Token **);

#if defined(SQLITE_NEED_ERR_NAME)
const char *sqlite3ErrName(int);
#endif

  int nDb;                      /* Number of backends currently in use */
  int flags;                    /* Miscellaneous flags. See below */
  i64 lastRowid;                /* ROWID of most recent insert (see above) */
  i64 szMmap;                   /* Default mmap_size setting */
  unsigned int openFlags;       /* Flags passed to sqlite3_vfs.xOpen() */
  int errCode;                  /* Most recent error code (SQLITE_*) */
  int errMask;                  /* & result codes with this before returning */
  int iSysErrno;                /* Errno value from last system error */
  u16 dbOptFlags;               /* Flags to enable/disable optimizations */
  u8 enc;                       /* Text encoding */
  u8 autoCommit;                /* The auto-commit flag. */
  u8 temp_store;                /* 1: file 2: memory 0: default */
  u8 mallocFailed;              /* True if we have seen a malloc failure */
  u8 bBenignMalloc;             /* Do not require OOMs if true */
  u8 dfltLockMode;              /* Default locking-mode for attached dbs */
................................................................................
char sqlite3CompareAffinity(Expr *pExpr, char aff2);
int sqlite3IndexAffinityOk(Expr *pExpr, char idx_affinity);
char sqlite3ExprAffinity(Expr *pExpr);
int sqlite3Atoi64(const char*, i64*, int, u8);
int sqlite3DecOrHexToI64(const char*, i64*);
void sqlite3ErrorWithMsg(sqlite3*, int, const char*,...);
void sqlite3Error(sqlite3*,int);
void sqlite3SystemError(sqlite3*,int);
void *sqlite3HexToBlob(sqlite3*, const char *z, int n);
u8 sqlite3HexToInt(int h);
int sqlite3TwoPartName(Parse *, Token *, Token *, Token **);

#if defined(SQLITE_NEED_ERR_NAME)
const char *sqlite3ErrName(int);
#endif

Changes to src/test1.c.

    Tcl_SetResult(interp, (char *)sqlite3ErrStr(rc), TCL_STATIC);
    return TCL_ERROR;
  }

  Tcl_ResetResult(interp);
  return TCL_OK;
}
























/*
** Usage:  sqlite3_db_filename DB DBNAME
**
** Return the name of a file associated with a database.
*/
static int test_db_filename(
................................................................................
     { "sqlite3_stmt_readonly",         test_stmt_readonly ,0 },
     { "sqlite3_stmt_busy",             test_stmt_busy     ,0 },
     { "uses_stmt_journal",             uses_stmt_journal ,0 },

     { "sqlite3_release_memory",        test_release_memory,     0},
     { "sqlite3_db_release_memory",     test_db_release_memory,  0},
     { "sqlite3_db_cacheflush",         test_db_cacheflush,      0},

     { "sqlite3_db_filename",           test_db_filename,        0},
     { "sqlite3_db_readonly",           test_db_readonly,        0},
     { "sqlite3_soft_heap_limit",       test_soft_heap_limit,    0},
     { "sqlite3_thread_cleanup",        test_thread_cleanup,     0},
     { "sqlite3_pager_refcounts",       test_pager_refcounts,    0},

     { "sqlite3_load_extension",        test_load_extension,     0},








>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>
>







 







>

    Tcl_SetResult(interp, (char *)sqlite3ErrStr(rc), TCL_STATIC);
    return TCL_ERROR;
  }

  Tcl_ResetResult(interp);
  return TCL_OK;
}

/*
** Usage:  sqlite3_system_errno DB
**
** Return the low-level system errno value.
*/
static int test_system_errno(
  void * clientData,
  Tcl_Interp *interp,
  int objc,
  Tcl_Obj *CONST objv[]
){
  sqlite3 *db;
  int iErrno;
  if( objc!=2 ){
    Tcl_WrongNumArgs(interp, 1, objv, "DB");
    return TCL_ERROR;
  }
  if( getDbPointer(interp, Tcl_GetString(objv[1]), &db) ) return TCL_ERROR;
  iErrno = sqlite3_system_errno(db);
  Tcl_SetObjResult(interp, Tcl_NewIntObj(iErrno));
  return TCL_OK;
}

/*
** Usage:  sqlite3_db_filename DB DBNAME
**
** Return the name of a file associated with a database.
*/
static int test_db_filename(
................................................................................
     { "sqlite3_stmt_readonly",         test_stmt_readonly ,0 },
     { "sqlite3_stmt_busy",             test_stmt_busy     ,0 },
     { "uses_stmt_journal",             uses_stmt_journal ,0 },

     { "sqlite3_release_memory",        test_release_memory,     0},
     { "sqlite3_db_release_memory",     test_db_release_memory,  0},
     { "sqlite3_db_cacheflush",         test_db_cacheflush,      0},
     { "sqlite3_system_errno",          test_system_errno,       0},
     { "sqlite3_db_filename",           test_db_filename,        0},
     { "sqlite3_db_readonly",           test_db_readonly,        0},
     { "sqlite3_soft_heap_limit",       test_soft_heap_limit,    0},
     { "sqlite3_thread_cleanup",        test_thread_cleanup,     0},
     { "sqlite3_pager_refcounts",       test_pager_refcounts,    0},

     { "sqlite3_load_extension",        test_load_extension,     0},

Changes to src/util.c.

/*
** The string z[] is followed immediately by another string.  Return
** a poiner to that other string.
*/
const char *sqlite3StrNext(const char *z){
  return z + strlen(z) + 1;
}











/*
** Set the current error code to err_code and clear any prior error message.


*/
void sqlite3Error(sqlite3 *db, int err_code){
  assert( db!=0 );
  db->errCode = err_code;
  if( db->pErr ) sqlite3ValueSetNull(db->pErr);












}

/*
** Set the most recent error code and error string for the sqlite
** handle "db". The error code is set to "err_code".
**
** If it is not NULL, string zFormat specifies the format of the
................................................................................
** To clear the most recent error for sqlite handle "db", sqlite3Error
** should be called with err_code set to SQLITE_OK and zFormat set
** to NULL.
*/
void sqlite3ErrorWithMsg(sqlite3 *db, int err_code, const char *zFormat, ...){
  assert( db!=0 );
  db->errCode = err_code;

  if( zFormat==0 ){
    sqlite3Error(db, err_code);
  }else if( db->pErr || (db->pErr = sqlite3ValueNew(db))!=0 ){
    char *z;
    va_list ap;
    va_start(ap, zFormat);
    z = sqlite3VMPrintf(db, zFormat, ap);








>
>
>
>
>
>
>
>
>
>



>
>




|
>
>
>
>
>
>
>
>
>
>
>
>







 







>

/*
** The string z[] is followed immediately by another string.  Return
** a poiner to that other string.
*/
const char *sqlite3StrNext(const char *z){
  return z + strlen(z) + 1;
}

/*
** Helper function for sqlite3Error() - called rarely.  Broken out into
** a separate routine to avoid unnecessary register saves on entry to
** sqlite3Error().
*/
static SQLITE_NOINLINE void  sqlite3ErrorFinish(sqlite3 *db, int err_code){
  if( db->pErr ) sqlite3ValueSetNull(db->pErr);
  sqlite3SystemError(db, err_code);
}

/*
** Set the current error code to err_code and clear any prior error message.
** Also set iSysErrno (by calling sqlite3System) if the err_code indicates
** that would be appropriate.
*/
void sqlite3Error(sqlite3 *db, int err_code){
  assert( db!=0 );
  db->errCode = err_code;
  if( err_code || db->pErr ) sqlite3ErrorFinish(db, err_code);
}

/*
** Load the sqlite3.iSysErrno field if that is an appropriate thing
** to do based on the SQLite error code in rc.
*/
void sqlite3SystemError(sqlite3 *db, int rc){
  if( rc==SQLITE_IOERR_NOMEM ) return;
  rc &= 0xff;
  if( rc==SQLITE_CANTOPEN || rc==SQLITE_IOERR ){
    db->iSysErrno = sqlite3OsGetLastError(db->pVfs);
  }
}

/*
** Set the most recent error code and error string for the sqlite
** handle "db". The error code is set to "err_code".
**
** If it is not NULL, string zFormat specifies the format of the
................................................................................
** To clear the most recent error for sqlite handle "db", sqlite3Error
** should be called with err_code set to SQLITE_OK and zFormat set
** to NULL.
*/
void sqlite3ErrorWithMsg(sqlite3 *db, int err_code, const char *zFormat, ...){
  assert( db!=0 );
  db->errCode = err_code;
  sqlite3SystemError(db, err_code);
  if( zFormat==0 ){
    sqlite3Error(db, err_code);
  }else if( db->pErr || (db->pErr = sqlite3ValueNew(db))!=0 ){
    char *z;
    va_list ap;
    va_start(ap, zFormat);
    z = sqlite3VMPrintf(db, zFormat, ap);

Changes to src/vdbe.c.

  **     different sized allocations. Memory cells provide growable
  **     allocations.
  **
  **   * When using ENABLE_MEMORY_MANAGEMENT, memory cell buffers can
  **     be freed lazily via the sqlite3_release_memory() API. This
  **     minimizes the number of malloc calls made by the system.
  **
  ** Memory cells for cursors are allocated at the top of the address
  ** space. Memory cell (p->nMem) corresponds to cursor 0. Space for
  ** cursor 1 is managed by memory cell (p->nMem-1), etc.
  */
  Mem *pMem = &p->aMem[p->nMem-iCur];

  int nByte;
  VdbeCursor *pCx = 0;
  nByte = 
      ROUND8(sizeof(VdbeCursor)) + 2*sizeof(u32)*nField + 
      (eCurType==CURTYPE_BTREE?sqlite3BtreeCursorSize():0);

  assert( iCur<p->nCursor );
  if( p->apCsr[iCur] ){
    sqlite3VdbeFreeCursor(p, p->apCsr[iCur]);
    p->apCsr[iCur] = 0;
  }
  if( SQLITE_OK==sqlite3VdbeMemClearAndResize(pMem, nByte) ){
    p->apCsr[iCur] = pCx = (VdbeCursor*)pMem->z;
    memset(pCx, 0, sizeof(VdbeCursor));
................................................................................
  sqlite3VdbeMemSetNull(pOut);
  pOut->flags = MEM_Int;
  return pOut;
}
static Mem *out2Prerelease(Vdbe *p, VdbeOp *pOp){
  Mem *pOut;
  assert( pOp->p2>0 );
  assert( pOp->p2<=(p->nMem-p->nCursor) );
  pOut = &p->aMem[pOp->p2];
  memAboutToChange(p, pOut);
  if( VdbeMemDynamic(pOut) ){
    return out2PrereleaseWithClear(pOut);
  }else{
    pOut->flags = MEM_Int;
    return pOut;
................................................................................
#endif

    /* Sanity checking on other operands */
#ifdef SQLITE_DEBUG
    assert( pOp->opflags==sqlite3OpcodeProperty[pOp->opcode] );
    if( (pOp->opflags & OPFLG_IN1)!=0 ){
      assert( pOp->p1>0 );
      assert( pOp->p1<=(p->nMem-p->nCursor) );
      assert( memIsValid(&aMem[pOp->p1]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p1]) );
      REGISTER_TRACE(pOp->p1, &aMem[pOp->p1]);
    }
    if( (pOp->opflags & OPFLG_IN2)!=0 ){
      assert( pOp->p2>0 );
      assert( pOp->p2<=(p->nMem-p->nCursor) );
      assert( memIsValid(&aMem[pOp->p2]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p2]) );
      REGISTER_TRACE(pOp->p2, &aMem[pOp->p2]);
    }
    if( (pOp->opflags & OPFLG_IN3)!=0 ){
      assert( pOp->p3>0 );
      assert( pOp->p3<=(p->nMem-p->nCursor) );
      assert( memIsValid(&aMem[pOp->p3]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p3]) );
      REGISTER_TRACE(pOp->p3, &aMem[pOp->p3]);
    }
    if( (pOp->opflags & OPFLG_OUT2)!=0 ){
      assert( pOp->p2>0 );
      assert( pOp->p2<=(p->nMem-p->nCursor) );
      memAboutToChange(p, &aMem[pOp->p2]);
    }
    if( (pOp->opflags & OPFLG_OUT3)!=0 ){
      assert( pOp->p3>0 );
      assert( pOp->p3<=(p->nMem-p->nCursor) );
      memAboutToChange(p, &aMem[pOp->p3]);
    }
#endif
#if defined(SQLITE_DEBUG) || defined(VDBE_PROFILE)
    pOrigOp = pOp;
#endif
  
................................................................................

/* Opcode:  Gosub P1 P2 * * *
**
** Write the current address onto register P1
** and then jump to address P2.
*/
case OP_Gosub: {            /* jump */
  assert( pOp->p1>0 && pOp->p1<=(p->nMem-p->nCursor) );
  pIn1 = &aMem[pOp->p1];
  assert( VdbeMemDynamic(pIn1)==0 );
  memAboutToChange(p, pIn1);
  pIn1->flags = MEM_Int;
  pIn1->u.i = (int)(pOp-aOp);
  REGISTER_TRACE(pOp->p1, pIn1);

................................................................................
** If P2!=0 then the coroutine implementation immediately follows
** this opcode.  So jump over the coroutine implementation to
** address P2.
**
** See also: EndCoroutine
*/
case OP_InitCoroutine: {     /* jump */
  assert( pOp->p1>0 &&  pOp->p1<=(p->nMem-p->nCursor) );
  assert( pOp->p2>=0 && pOp->p2<p->nOp );
  assert( pOp->p3>=0 && pOp->p3<p->nOp );
  pOut = &aMem[pOp->p1];
  assert( !VdbeMemDynamic(pOut) );
  pOut->u.i = pOp->p3 - 1;
  pOut->flags = MEM_Int;
  if( pOp->p2 ) goto jump_to_p2;
................................................................................
  pOut->z = pOp->p4.z;
  pOut->n = pOp->p1;
  pOut->enc = encoding;
  UPDATE_MAX_BLOBSIZE(pOut);
#ifndef SQLITE_LIKE_DOESNT_MATCH_BLOBS
  if( pOp->p5 ){
    assert( pOp->p3>0 );
    assert( pOp->p3<=(p->nMem-p->nCursor) );
    pIn3 = &aMem[pOp->p3];
    assert( pIn3->flags & MEM_Int );
    if( pIn3->u.i ) pOut->flags = MEM_Blob|MEM_Static|MEM_Term;
  }
#endif
  break;
}
................................................................................
** OP_Ne or OP_Eq.
*/
case OP_Null: {           /* out2 */
  int cnt;
  u16 nullFlag;
  pOut = out2Prerelease(p, pOp);
  cnt = pOp->p3-pOp->p2;
  assert( pOp->p3<=(p->nMem-p->nCursor) );
  pOut->flags = nullFlag = pOp->p1 ? (MEM_Null|MEM_Cleared) : MEM_Null;
  while( cnt>0 ){
    pOut++;
    memAboutToChange(p, pOut);
    sqlite3VdbeMemSetNull(pOut);
    pOut->flags = nullFlag;
    cnt--;
................................................................................
**
** Set register P1 to have the value NULL as seen by the OP_MakeRecord
** instruction, but do not free any string or blob memory associated with
** the register, so that if the value was a string or blob that was
** previously copied using OP_SCopy, the copies will continue to be valid.
*/
case OP_SoftNull: {
  assert( pOp->p1>0 && pOp->p1<=(p->nMem-p->nCursor) );
  pOut = &aMem[pOp->p1];
  pOut->flags = (pOut->flags|MEM_Null)&~MEM_Undefined;
  break;
}

/* Opcode: Blob P1 P2 * P4 *
** Synopsis: r[P2]=P4 (len=P1)
................................................................................
  p2 = pOp->p2;
  assert( n>0 && p1>0 && p2>0 );
  assert( p1+n<=p2 || p2+n<=p1 );

  pIn1 = &aMem[p1];
  pOut = &aMem[p2];
  do{
    assert( pOut<=&aMem[(p->nMem-p->nCursor)] );
    assert( pIn1<=&aMem[(p->nMem-p->nCursor)] );
    assert( memIsValid(pIn1) );
    memAboutToChange(p, pOut);
    sqlite3VdbeMemMove(pOut, pIn1);
#ifdef SQLITE_DEBUG
    if( pOut->pScopyFrom>=&aMem[p1] && pOut->pScopyFrom<pOut ){
      pOut->pScopyFrom += pOp->p2 - p1;
    }
................................................................................
** the result row.
*/
case OP_ResultRow: {
  Mem *pMem;
  int i;
  assert( p->nResColumn==pOp->p2 );
  assert( pOp->p1>0 );
  assert( pOp->p1+pOp->p2<=(p->nMem-p->nCursor)+1 );

#ifndef SQLITE_OMIT_PROGRESS_CALLBACK
  /* Run the progress counter just before returning.
  */
  if( db->xProgress!=0
   && nVmStep>=nProgressLimit
   && db->xProgress(db->pProgressArg)!=0
................................................................................
*/
case OP_Function0: {
  int n;
  sqlite3_context *pCtx;

  assert( pOp->p4type==P4_FUNCDEF );
  n = pOp->p5;
  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  assert( n==0 || (pOp->p2>0 && pOp->p2+n<=(p->nMem-p->nCursor)+1) );
  assert( pOp->p3<pOp->p2 || pOp->p3>=pOp->p2+n );
  pCtx = sqlite3DbMallocRawNN(db, sizeof(*pCtx) + (n-1)*sizeof(sqlite3_value*));
  if( pCtx==0 ) goto no_mem;
  pCtx->pOut = 0;
  pCtx->pFunc = pOp->p4.pFunc;
  pCtx->iOp = (int)(pOp - aOp);
  pCtx->pVdbe = p;
................................................................................
  assert( pKeyInfo!=0 );
  p1 = pOp->p1;
  p2 = pOp->p2;
#if SQLITE_DEBUG
  if( aPermute ){
    int k, mx = 0;
    for(k=0; k<n; k++) if( aPermute[k]>mx ) mx = aPermute[k];
    assert( p1>0 && p1+mx<=(p->nMem-p->nCursor)+1 );
    assert( p2>0 && p2+mx<=(p->nMem-p->nCursor)+1 );
  }else{
    assert( p1>0 && p1+n<=(p->nMem-p->nCursor)+1 );
    assert( p2>0 && p2+n<=(p->nMem-p->nCursor)+1 );
  }
#endif /* SQLITE_DEBUG */
  for(i=0; i<n; i++){
    idx = aPermute ? aPermute[i] : i;
    assert( memIsValid(&aMem[p1+idx]) );
    assert( memIsValid(&aMem[p2+idx]) );
    REGISTER_TRACE(p1+idx, &aMem[p1+idx]);
................................................................................

  pC = p->apCsr[pOp->p1];
  p2 = pOp->p2;

  /* If the cursor cache is stale, bring it up-to-date */
  rc = sqlite3VdbeCursorMoveto(&pC, &p2);

  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  pDest = &aMem[pOp->p3];
  memAboutToChange(p, pDest);
  assert( pOp->p1>=0 && pOp->p1<p->nCursor );
  assert( pC!=0 );
  assert( p2<pC->nField );
  aOffset = pC->aOffset;
  assert( pC->eCurType!=CURTYPE_VTAB );
................................................................................
  char cAff;               /* A single character of affinity */

  zAffinity = pOp->p4.z;
  assert( zAffinity!=0 );
  assert( zAffinity[pOp->p2]==0 );
  pIn1 = &aMem[pOp->p1];
  while( (cAff = *(zAffinity++))!=0 ){
    assert( pIn1 <= &p->aMem[(p->nMem-p->nCursor)] );
    assert( memIsValid(pIn1) );
    applyAffinity(pIn1, cAff, encoding);
    pIn1++;
  }
  break;
}

................................................................................
  ** of the record to data0.
  */
  nData = 0;         /* Number of bytes of data space */
  nHdr = 0;          /* Number of bytes of header space */
  nZero = 0;         /* Number of zero bytes at the end of the record */
  nField = pOp->p1;
  zAffinity = pOp->p4.z;
  assert( nField>0 && pOp->p2>0 && pOp->p2+nField<=(p->nMem-p->nCursor)+1 );
  pData0 = &aMem[nField];
  nField = pOp->p2;
  pLast = &pData0[nField-1];
  file_format = p->minWriteFileFormat;

  /* Identify the output register */
  assert( pOp->p3<pOp->p1 || pOp->p3>=pOp->p1+pOp->p2 );
................................................................................
    /* EVIDENCE-OF: R-64536-51728 The values for each column in the record
    ** immediately follow the header. */
    j += sqlite3VdbeSerialPut(&zNewRecord[j], pRec, serial_type); /* content */
  }while( (++pRec)<=pLast );
  assert( i==nHdr );
  assert( j==nByte );

  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  pOut->n = (int)nByte;
  pOut->flags = MEM_Blob;
  if( nZero ){
    pOut->u.nZero = nZero;
    pOut->flags |= MEM_Zero;
  }
  pOut->enc = SQLITE_UTF8;  /* In case the blob is ever converted to text */
................................................................................
      p->minWriteFileFormat = pDb->pSchema->file_format;
    }
  }else{
    wrFlag = 0;
  }
  if( pOp->p5 & OPFLAG_P2ISREG ){
    assert( p2>0 );
    assert( p2<=(p->nMem-p->nCursor) );
    pIn2 = &aMem[p2];
    assert( memIsValid(pIn2) );
    assert( (pIn2->flags & MEM_Int)!=0 );
    sqlite3VdbeMemIntegerify(pIn2);
    p2 = (int)pIn2->u.i;
    /* The p2 value always comes from a prior OP_CreateTable opcode and
    ** that opcode will always set the p2 value to 2 or more or else fail.
................................................................................
      if( p->pFrame ){
        for(pFrame=p->pFrame; pFrame->pParent; pFrame=pFrame->pParent);
        /* Assert that P3 is a valid memory cell. */
        assert( pOp->p3<=pFrame->nMem );
        pMem = &pFrame->aMem[pOp->p3];
      }else{
        /* Assert that P3 is a valid memory cell. */
        assert( pOp->p3<=(p->nMem-p->nCursor) );
        pMem = &aMem[pOp->p3];
        memAboutToChange(p, pMem);
      }
      assert( memIsValid(pMem) );

      REGISTER_TRACE(pOp->p3, pMem);
      sqlite3VdbeMemIntegerify(pMem);
................................................................................
case OP_IdxDelete: {
  VdbeCursor *pC;
  BtCursor *pCrsr;
  int res;
  UnpackedRecord r;

  assert( pOp->p3>0 );
  assert( pOp->p2>0 && pOp->p2+pOp->p3<=(p->nMem-p->nCursor)+1 );
  assert( pOp->p1>=0 && pOp->p1<p->nCursor );
  pC = p->apCsr[pOp->p1];
  assert( pC!=0 );
  assert( pC->eCurType==CURTYPE_BTREE );
  pCrsr = pC->uc.pCursor;
  assert( pCrsr!=0 );
  assert( pOp->p5==0 );
................................................................................
  Mem *pnErr;     /* Register keeping track of errors remaining */

  assert( p->bIsReader );
  nRoot = pOp->p2;
  aRoot = pOp->p4.ai;
  assert( nRoot>0 );
  assert( aRoot[nRoot]==0 );
  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  pnErr = &aMem[pOp->p3];
  assert( (pnErr->flags & MEM_Int)!=0 );
  assert( (pnErr->flags & (MEM_Str|MEM_Blob))==0 );
  pIn1 = &aMem[pOp->p1];
  assert( pOp->p5<db->nDb );
  assert( DbMaskTest(p->btreeMask, pOp->p5) );
  z = sqlite3BtreeIntegrityCheck(db->aDb[pOp->p5].pBt, aRoot, nRoot,
................................................................................
  if( (pRt->flags&MEM_Frame)==0 ){
    /* SubProgram.nMem is set to the number of memory cells used by the 
    ** program stored in SubProgram.aOp. As well as these, one memory
    ** cell is required for each cursor used by the program. Set local
    ** variable nMem (and later, VdbeFrame.nChildMem) to this value.
    */
    nMem = pProgram->nMem + pProgram->nCsr;


    nByte = ROUND8(sizeof(VdbeFrame))
              + nMem * sizeof(Mem)
              + pProgram->nCsr * sizeof(VdbeCursor *)
              + pProgram->nOnce * sizeof(u8);
    pFrame = sqlite3DbMallocZero(db, nByte);
    if( !pFrame ){
      goto no_mem;
................................................................................
    pEnd = &VdbeFrameMem(pFrame)[pFrame->nChildMem];
    for(pMem=VdbeFrameMem(pFrame); pMem!=pEnd; pMem++){
      pMem->flags = MEM_Undefined;
      pMem->db = db;
    }
  }else{
    pFrame = pRt->u.pFrame;
    assert( pProgram->nMem+pProgram->nCsr==pFrame->nChildMem );

    assert( pProgram->nCsr==pFrame->nChildCsr );
    assert( (int)(pOp - aOp)==pFrame->pc );
  }

  p->nFrame++;
  pFrame->pParent = p->pFrame;
  pFrame->lastRowid = lastRowid;
................................................................................
  pFrame->nChange = p->nChange;
  pFrame->nDbChange = p->db->nChange;
  assert( pFrame->pAuxData==0 );
  pFrame->pAuxData = p->pAuxData;
  p->pAuxData = 0;
  p->nChange = 0;
  p->pFrame = pFrame;
  p->aMem = aMem = &VdbeFrameMem(pFrame)[-1];
  p->nMem = pFrame->nChildMem;
  p->nCursor = (u16)pFrame->nChildCsr;
  p->apCsr = (VdbeCursor **)&aMem[p->nMem+1];
  p->aOp = aOp = pProgram->aOp;
  p->nOp = pProgram->nOp;
  p->aOnceFlag = (u8 *)&p->apCsr[p->nCursor];
  p->nOnceFlag = pProgram->nOnce;
#ifdef SQLITE_ENABLE_STMT_SCANSTATUS
  p->anExec = 0;
#endif
................................................................................
*/
case OP_AggStep0: {
  int n;
  sqlite3_context *pCtx;

  assert( pOp->p4type==P4_FUNCDEF );
  n = pOp->p5;
  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  assert( n==0 || (pOp->p2>0 && pOp->p2+n<=(p->nMem-p->nCursor)+1) );
  assert( pOp->p3<pOp->p2 || pOp->p3>=pOp->p2+n );
  pCtx = sqlite3DbMallocRawNN(db, sizeof(*pCtx) + (n-1)*sizeof(sqlite3_value*));
  if( pCtx==0 ) goto no_mem;
  pCtx->pMem = 0;
  pCtx->pFunc = pOp->p4.pFunc;
  pCtx->iOp = (int)(pOp - aOp);
  pCtx->pVdbe = p;
................................................................................
** argument is not used by this opcode.  It is only there to disambiguate
** functions that can take varying numbers of arguments.  The
** P4 argument is only needed for the degenerate case where
** the step function was not previously called.
*/
case OP_AggFinal: {
  Mem *pMem;
  assert( pOp->p1>0 && pOp->p1<=(p->nMem-p->nCursor) );
  pMem = &aMem[pOp->p1];
  assert( (pMem->flags & ~(MEM_Null|MEM_Agg))==0 );
  rc = sqlite3VdbeMemFinalize(pMem, pOp->p4.pFunc);
  if( rc ){
    sqlite3VdbeError(p, "%s", sqlite3_value_text(pMem));
    goto abort_due_to_error;
  }
................................................................................
  sqlite3_vtab *pVtab;
  const sqlite3_module *pModule;
  Mem *pDest;
  sqlite3_context sContext;

  VdbeCursor *pCur = p->apCsr[pOp->p1];
  assert( pCur->eCurType==CURTYPE_VTAB );
  assert( pOp->p3>0 && pOp->p3<=(p->nMem-p->nCursor) );
  pDest = &aMem[pOp->p3];
  memAboutToChange(p, pDest);
  if( pCur->nullRow ){
    sqlite3VdbeMemSetNull(pDest);
    break;
  }
  pVtab = pCur->uc.pVCur->pVtab;
................................................................................
abort_due_to_error:
  if( db->mallocFailed ) rc = SQLITE_NOMEM_BKPT;
  assert( rc );
  if( p->zErrMsg==0 && rc!=SQLITE_IOERR_NOMEM ){
    sqlite3VdbeError(p, "%s", sqlite3ErrStr(rc));
  }
  p->rc = rc;

  testcase( sqlite3GlobalConfig.xLog!=0 );
  sqlite3_log(rc, "statement aborts at %d: [%s] %s", 
                   (int)(pOp - aOp), p->zSql, p->zErrMsg);
  sqlite3VdbeHalt(p);
  if( rc==SQLITE_IOERR_NOMEM ) sqlite3OomFault(db);
  rc = SQLITE_ERROR;
  if( resetSchemaOnFault>0 ){

  **     different sized allocations. Memory cells provide growable
  **     allocations.
  **
  **   * When using ENABLE_MEMORY_MANAGEMENT, memory cell buffers can
  **     be freed lazily via the sqlite3_release_memory() API. This
  **     minimizes the number of malloc calls made by the system.
  **
  ** The memory cell for cursor 0 is aMem[0]. The rest are allocated from
  ** the top of the register space.  Cursor 1 is at Mem[p->nMem-1].
  ** Cursor 2 is at Mem[p->nMem-2]. And so forth.
  */
  Mem *pMem = iCur>0 ? &p->aMem[p->nMem-iCur] : p->aMem;

  int nByte;
  VdbeCursor *pCx = 0;
  nByte = 
      ROUND8(sizeof(VdbeCursor)) + 2*sizeof(u32)*nField + 
      (eCurType==CURTYPE_BTREE?sqlite3BtreeCursorSize():0);

  assert( iCur>=0 && iCur<p->nCursor );
  if( p->apCsr[iCur] ){
    sqlite3VdbeFreeCursor(p, p->apCsr[iCur]);
    p->apCsr[iCur] = 0;
  }
  if( SQLITE_OK==sqlite3VdbeMemClearAndResize(pMem, nByte) ){
    p->apCsr[iCur] = pCx = (VdbeCursor*)pMem->z;
    memset(pCx, 0, sizeof(VdbeCursor));
................................................................................
  sqlite3VdbeMemSetNull(pOut);
  pOut->flags = MEM_Int;
  return pOut;
}
static Mem *out2Prerelease(Vdbe *p, VdbeOp *pOp){
  Mem *pOut;
  assert( pOp->p2>0 );
  assert( pOp->p2<=(p->nMem+1 - p->nCursor) );
  pOut = &p->aMem[pOp->p2];
  memAboutToChange(p, pOut);
  if( VdbeMemDynamic(pOut) ){
    return out2PrereleaseWithClear(pOut);
  }else{
    pOut->flags = MEM_Int;
    return pOut;
................................................................................
#endif

    /* Sanity checking on other operands */
#ifdef SQLITE_DEBUG
    assert( pOp->opflags==sqlite3OpcodeProperty[pOp->opcode] );
    if( (pOp->opflags & OPFLG_IN1)!=0 ){
      assert( pOp->p1>0 );
      assert( pOp->p1<=(p->nMem+1 - p->nCursor) );
      assert( memIsValid(&aMem[pOp->p1]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p1]) );
      REGISTER_TRACE(pOp->p1, &aMem[pOp->p1]);
    }
    if( (pOp->opflags & OPFLG_IN2)!=0 ){
      assert( pOp->p2>0 );
      assert( pOp->p2<=(p->nMem+1 - p->nCursor) );
      assert( memIsValid(&aMem[pOp->p2]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p2]) );
      REGISTER_TRACE(pOp->p2, &aMem[pOp->p2]);
    }
    if( (pOp->opflags & OPFLG_IN3)!=0 ){
      assert( pOp->p3>0 );
      assert( pOp->p3<=(p->nMem+1 - p->nCursor) );
      assert( memIsValid(&aMem[pOp->p3]) );
      assert( sqlite3VdbeCheckMemInvariants(&aMem[pOp->p3]) );
      REGISTER_TRACE(pOp->p3, &aMem[pOp->p3]);
    }
    if( (pOp->opflags & OPFLG_OUT2)!=0 ){
      assert( pOp->p2>0 );
      assert( pOp->p2<=(p->nMem+1 - p->nCursor) );
      memAboutToChange(p, &aMem[pOp->p2]);
    }
    if( (pOp->opflags & OPFLG_OUT3)!=0 ){
      assert( pOp->p3>0 );
      assert( pOp->p3<=(p->nMem+1 - p->nCursor) );
      memAboutToChange(p, &aMem[pOp->p3]);
    }
#endif
#if defined(SQLITE_DEBUG) || defined(VDBE_PROFILE)
    pOrigOp = pOp;
#endif
  
................................................................................

/* Opcode:  Gosub P1 P2 * * *
**
** Write the current address onto register P1
** and then jump to address P2.
*/
case OP_Gosub: {            /* jump */
  assert( pOp->p1>0 && pOp->p1<=(p->nMem+1 - p->nCursor) );
  pIn1 = &aMem[pOp->p1];
  assert( VdbeMemDynamic(pIn1)==0 );
  memAboutToChange(p, pIn1);
  pIn1->flags = MEM_Int;
  pIn1->u.i = (int)(pOp-aOp);
  REGISTER_TRACE(pOp->p1, pIn1);

................................................................................
** If P2!=0 then the coroutine implementation immediately follows
** this opcode.  So jump over the coroutine implementation to
** address P2.
**
** See also: EndCoroutine
*/
case OP_InitCoroutine: {     /* jump */
  assert( pOp->p1>0 &&  pOp->p1<=(p->nMem+1 - p->nCursor) );
  assert( pOp->p2>=0 && pOp->p2<p->nOp );
  assert( pOp->p3>=0 && pOp->p3<p->nOp );
  pOut = &aMem[pOp->p1];
  assert( !VdbeMemDynamic(pOut) );
  pOut->u.i = pOp->p3 - 1;
  pOut->flags = MEM_Int;
  if( pOp->p2 ) goto jump_to_p2;
................................................................................
  pOut->z = pOp->p4.z;
  pOut->n = pOp->p1;
  pOut->enc = encoding;
  UPDATE_MAX_BLOBSIZE(pOut);
#ifndef SQLITE_LIKE_DOESNT_MATCH_BLOBS
  if( pOp->p5 ){
    assert( pOp->p3>0 );
    assert( pOp->p3<=(p->nMem+1 - p->nCursor) );
    pIn3 = &aMem[pOp->p3];
    assert( pIn3->flags & MEM_Int );
    if( pIn3->u.i ) pOut->flags = MEM_Blob|MEM_Static|MEM_Term;
  }
#endif
  break;
}
................................................................................
** OP_Ne or OP_Eq.
*/
case OP_Null: {           /* out2 */
  int cnt;
  u16 nullFlag;
  pOut = out2Prerelease(p, pOp);
  cnt = pOp->p3-pOp->p2;
  assert( pOp->p3<=(p->nMem+1 - p->nCursor) );
  pOut->flags = nullFlag = pOp->p1 ? (MEM_Null|MEM_Cleared) : MEM_Null;
  while( cnt>0 ){
    pOut++;
    memAboutToChange(p, pOut);
    sqlite3VdbeMemSetNull(pOut);
    pOut->flags = nullFlag;
    cnt--;
................................................................................
**
** Set register P1 to have the value NULL as seen by the OP_MakeRecord
** instruction, but do not free any string or blob memory associated with
** the register, so that if the value was a string or blob that was
** previously copied using OP_SCopy, the copies will continue to be valid.
*/
case OP_SoftNull: {
  assert( pOp->p1>0 && pOp->p1<=(p->nMem+1 - p->nCursor) );
  pOut = &aMem[pOp->p1];
  pOut->flags = (pOut->flags|MEM_Null)&~MEM_Undefined;
  break;
}

/* Opcode: Blob P1 P2 * P4 *
** Synopsis: r[P2]=P4 (len=P1)
................................................................................
  p2 = pOp->p2;
  assert( n>0 && p1>0 && p2>0 );
  assert( p1+n<=p2 || p2+n<=p1 );

  pIn1 = &aMem[p1];
  pOut = &aMem[p2];
  do{
    assert( pOut<=&aMem[(p->nMem+1 - p->nCursor)] );
    assert( pIn1<=&aMem[(p->nMem+1 - p->nCursor)] );
    assert( memIsValid(pIn1) );
    memAboutToChange(p, pOut);
    sqlite3VdbeMemMove(pOut, pIn1);
#ifdef SQLITE_DEBUG
    if( pOut->pScopyFrom>=&aMem[p1] && pOut->pScopyFrom<pOut ){
      pOut->pScopyFrom += pOp->p2 - p1;
    }
................................................................................
** the result row.
*/
case OP_ResultRow: {
  Mem *pMem;
  int i;
  assert( p->nResColumn==pOp->p2 );
  assert( pOp->p1>0 );
  assert( pOp->p1+pOp->p2<=(p->nMem+1 - p->nCursor)+1 );

#ifndef SQLITE_OMIT_PROGRESS_CALLBACK
  /* Run the progress counter just before returning.
  */
  if( db->xProgress!=0
   && nVmStep>=nProgressLimit
   && db->xProgress(db->pProgressArg)!=0
................................................................................
*/
case OP_Function0: {
  int n;
  sqlite3_context *pCtx;

  assert( pOp->p4type==P4_FUNCDEF );
  n = pOp->p5;
  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  assert( n==0 || (pOp->p2>0 && pOp->p2+n<=(p->nMem+1 - p->nCursor)+1) );
  assert( pOp->p3<pOp->p2 || pOp->p3>=pOp->p2+n );
  pCtx = sqlite3DbMallocRawNN(db, sizeof(*pCtx) + (n-1)*sizeof(sqlite3_value*));
  if( pCtx==0 ) goto no_mem;
  pCtx->pOut = 0;
  pCtx->pFunc = pOp->p4.pFunc;
  pCtx->iOp = (int)(pOp - aOp);
  pCtx->pVdbe = p;
................................................................................
  assert( pKeyInfo!=0 );
  p1 = pOp->p1;
  p2 = pOp->p2;
#if SQLITE_DEBUG
  if( aPermute ){
    int k, mx = 0;
    for(k=0; k<n; k++) if( aPermute[k]>mx ) mx = aPermute[k];
    assert( p1>0 && p1+mx<=(p->nMem+1 - p->nCursor)+1 );
    assert( p2>0 && p2+mx<=(p->nMem+1 - p->nCursor)+1 );
  }else{
    assert( p1>0 && p1+n<=(p->nMem+1 - p->nCursor)+1 );
    assert( p2>0 && p2+n<=(p->nMem+1 - p->nCursor)+1 );
  }
#endif /* SQLITE_DEBUG */
  for(i=0; i<n; i++){
    idx = aPermute ? aPermute[i] : i;
    assert( memIsValid(&aMem[p1+idx]) );
    assert( memIsValid(&aMem[p2+idx]) );
    REGISTER_TRACE(p1+idx, &aMem[p1+idx]);
................................................................................

  pC = p->apCsr[pOp->p1];
  p2 = pOp->p2;

  /* If the cursor cache is stale, bring it up-to-date */
  rc = sqlite3VdbeCursorMoveto(&pC, &p2);

  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  pDest = &aMem[pOp->p3];
  memAboutToChange(p, pDest);
  assert( pOp->p1>=0 && pOp->p1<p->nCursor );
  assert( pC!=0 );
  assert( p2<pC->nField );
  aOffset = pC->aOffset;
  assert( pC->eCurType!=CURTYPE_VTAB );
................................................................................
  char cAff;               /* A single character of affinity */

  zAffinity = pOp->p4.z;
  assert( zAffinity!=0 );
  assert( zAffinity[pOp->p2]==0 );
  pIn1 = &aMem[pOp->p1];
  while( (cAff = *(zAffinity++))!=0 ){
    assert( pIn1 <= &p->aMem[(p->nMem+1 - p->nCursor)] );
    assert( memIsValid(pIn1) );
    applyAffinity(pIn1, cAff, encoding);
    pIn1++;
  }
  break;
}

................................................................................
  ** of the record to data0.
  */
  nData = 0;         /* Number of bytes of data space */
  nHdr = 0;          /* Number of bytes of header space */
  nZero = 0;         /* Number of zero bytes at the end of the record */
  nField = pOp->p1;
  zAffinity = pOp->p4.z;
  assert( nField>0 && pOp->p2>0 && pOp->p2+nField<=(p->nMem+1 - p->nCursor)+1 );
  pData0 = &aMem[nField];
  nField = pOp->p2;
  pLast = &pData0[nField-1];
  file_format = p->minWriteFileFormat;

  /* Identify the output register */
  assert( pOp->p3<pOp->p1 || pOp->p3>=pOp->p1+pOp->p2 );
................................................................................
    /* EVIDENCE-OF: R-64536-51728 The values for each column in the record
    ** immediately follow the header. */
    j += sqlite3VdbeSerialPut(&zNewRecord[j], pRec, serial_type); /* content */
  }while( (++pRec)<=pLast );
  assert( i==nHdr );
  assert( j==nByte );

  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  pOut->n = (int)nByte;
  pOut->flags = MEM_Blob;
  if( nZero ){
    pOut->u.nZero = nZero;
    pOut->flags |= MEM_Zero;
  }
  pOut->enc = SQLITE_UTF8;  /* In case the blob is ever converted to text */
................................................................................
      p->minWriteFileFormat = pDb->pSchema->file_format;
    }
  }else{
    wrFlag = 0;
  }
  if( pOp->p5 & OPFLAG_P2ISREG ){
    assert( p2>0 );
    assert( p2<=(p->nMem+1 - p->nCursor) );
    pIn2 = &aMem[p2];
    assert( memIsValid(pIn2) );
    assert( (pIn2->flags & MEM_Int)!=0 );
    sqlite3VdbeMemIntegerify(pIn2);
    p2 = (int)pIn2->u.i;
    /* The p2 value always comes from a prior OP_CreateTable opcode and
    ** that opcode will always set the p2 value to 2 or more or else fail.
................................................................................
      if( p->pFrame ){
        for(pFrame=p->pFrame; pFrame->pParent; pFrame=pFrame->pParent);
        /* Assert that P3 is a valid memory cell. */
        assert( pOp->p3<=pFrame->nMem );
        pMem = &pFrame->aMem[pOp->p3];
      }else{
        /* Assert that P3 is a valid memory cell. */
        assert( pOp->p3<=(p->nMem+1 - p->nCursor) );
        pMem = &aMem[pOp->p3];
        memAboutToChange(p, pMem);
      }
      assert( memIsValid(pMem) );

      REGISTER_TRACE(pOp->p3, pMem);
      sqlite3VdbeMemIntegerify(pMem);
................................................................................
case OP_IdxDelete: {
  VdbeCursor *pC;
  BtCursor *pCrsr;
  int res;
  UnpackedRecord r;

  assert( pOp->p3>0 );
  assert( pOp->p2>0 && pOp->p2+pOp->p3<=(p->nMem+1 - p->nCursor)+1 );
  assert( pOp->p1>=0 && pOp->p1<p->nCursor );
  pC = p->apCsr[pOp->p1];
  assert( pC!=0 );
  assert( pC->eCurType==CURTYPE_BTREE );
  pCrsr = pC->uc.pCursor;
  assert( pCrsr!=0 );
  assert( pOp->p5==0 );
................................................................................
  Mem *pnErr;     /* Register keeping track of errors remaining */

  assert( p->bIsReader );
  nRoot = pOp->p2;
  aRoot = pOp->p4.ai;
  assert( nRoot>0 );
  assert( aRoot[nRoot]==0 );
  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  pnErr = &aMem[pOp->p3];
  assert( (pnErr->flags & MEM_Int)!=0 );
  assert( (pnErr->flags & (MEM_Str|MEM_Blob))==0 );
  pIn1 = &aMem[pOp->p1];
  assert( pOp->p5<db->nDb );
  assert( DbMaskTest(p->btreeMask, pOp->p5) );
  z = sqlite3BtreeIntegrityCheck(db->aDb[pOp->p5].pBt, aRoot, nRoot,
................................................................................
  if( (pRt->flags&MEM_Frame)==0 ){
    /* SubProgram.nMem is set to the number of memory cells used by the 
    ** program stored in SubProgram.aOp. As well as these, one memory
    ** cell is required for each cursor used by the program. Set local
    ** variable nMem (and later, VdbeFrame.nChildMem) to this value.
    */
    nMem = pProgram->nMem + pProgram->nCsr;
    assert( nMem>0 );
    if( pProgram->nCsr==0 ) nMem++;
    nByte = ROUND8(sizeof(VdbeFrame))
              + nMem * sizeof(Mem)
              + pProgram->nCsr * sizeof(VdbeCursor *)
              + pProgram->nOnce * sizeof(u8);
    pFrame = sqlite3DbMallocZero(db, nByte);
    if( !pFrame ){
      goto no_mem;
................................................................................
    pEnd = &VdbeFrameMem(pFrame)[pFrame->nChildMem];
    for(pMem=VdbeFrameMem(pFrame); pMem!=pEnd; pMem++){
      pMem->flags = MEM_Undefined;
      pMem->db = db;
    }
  }else{
    pFrame = pRt->u.pFrame;
    assert( pProgram->nMem+pProgram->nCsr==pFrame->nChildMem 
        || (pProgram->nCsr==0 && pProgram->nMem+1==pFrame->nChildMem) );
    assert( pProgram->nCsr==pFrame->nChildCsr );
    assert( (int)(pOp - aOp)==pFrame->pc );
  }

  p->nFrame++;
  pFrame->pParent = p->pFrame;
  pFrame->lastRowid = lastRowid;
................................................................................
  pFrame->nChange = p->nChange;
  pFrame->nDbChange = p->db->nChange;
  assert( pFrame->pAuxData==0 );
  pFrame->pAuxData = p->pAuxData;
  p->pAuxData = 0;
  p->nChange = 0;
  p->pFrame = pFrame;
  p->aMem = aMem = VdbeFrameMem(pFrame);
  p->nMem = pFrame->nChildMem;
  p->nCursor = (u16)pFrame->nChildCsr;
  p->apCsr = (VdbeCursor **)&aMem[p->nMem];
  p->aOp = aOp = pProgram->aOp;
  p->nOp = pProgram->nOp;
  p->aOnceFlag = (u8 *)&p->apCsr[p->nCursor];
  p->nOnceFlag = pProgram->nOnce;
#ifdef SQLITE_ENABLE_STMT_SCANSTATUS
  p->anExec = 0;
#endif
................................................................................
*/
case OP_AggStep0: {
  int n;
  sqlite3_context *pCtx;

  assert( pOp->p4type==P4_FUNCDEF );
  n = pOp->p5;
  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  assert( n==0 || (pOp->p2>0 && pOp->p2+n<=(p->nMem+1 - p->nCursor)+1) );
  assert( pOp->p3<pOp->p2 || pOp->p3>=pOp->p2+n );
  pCtx = sqlite3DbMallocRawNN(db, sizeof(*pCtx) + (n-1)*sizeof(sqlite3_value*));
  if( pCtx==0 ) goto no_mem;
  pCtx->pMem = 0;
  pCtx->pFunc = pOp->p4.pFunc;
  pCtx->iOp = (int)(pOp - aOp);
  pCtx->pVdbe = p;
................................................................................
** argument is not used by this opcode.  It is only there to disambiguate
** functions that can take varying numbers of arguments.  The
** P4 argument is only needed for the degenerate case where
** the step function was not previously called.
*/
case OP_AggFinal: {
  Mem *pMem;
  assert( pOp->p1>0 && pOp->p1<=(p->nMem+1 - p->nCursor) );
  pMem = &aMem[pOp->p1];
  assert( (pMem->flags & ~(MEM_Null|MEM_Agg))==0 );
  rc = sqlite3VdbeMemFinalize(pMem, pOp->p4.pFunc);
  if( rc ){
    sqlite3VdbeError(p, "%s", sqlite3_value_text(pMem));
    goto abort_due_to_error;
  }
................................................................................
  sqlite3_vtab *pVtab;
  const sqlite3_module *pModule;
  Mem *pDest;
  sqlite3_context sContext;

  VdbeCursor *pCur = p->apCsr[pOp->p1];
  assert( pCur->eCurType==CURTYPE_VTAB );
  assert( pOp->p3>0 && pOp->p3<=(p->nMem+1 - p->nCursor) );
  pDest = &aMem[pOp->p3];
  memAboutToChange(p, pDest);
  if( pCur->nullRow ){
    sqlite3VdbeMemSetNull(pDest);
    break;
  }
  pVtab = pCur->uc.pVCur->pVtab;
................................................................................
abort_due_to_error:
  if( db->mallocFailed ) rc = SQLITE_NOMEM_BKPT;
  assert( rc );
  if( p->zErrMsg==0 && rc!=SQLITE_IOERR_NOMEM ){
    sqlite3VdbeError(p, "%s", sqlite3ErrStr(rc));
  }
  p->rc = rc;
  sqlite3SystemError(db, rc);
  testcase( sqlite3GlobalConfig.xLog!=0 );
  sqlite3_log(rc, "statement aborts at %d: [%s] %s", 
                   (int)(pOp - aOp), p->zSql, p->zErrMsg);
  sqlite3VdbeHalt(p);
  if( rc==SQLITE_IOERR_NOMEM ) sqlite3OomFault(db);
  rc = SQLITE_ERROR;
  if( resetSchemaOnFault>0 ){

Changes to src/vdbeaux.c.

  */
  assert( p->nOp>0 );

  /* Set the magic to VDBE_MAGIC_RUN sooner rather than later. */
  p->magic = VDBE_MAGIC_RUN;

#ifdef SQLITE_DEBUG
  for(i=1; i<p->nMem; i++){
    assert( p->aMem[i].db==p->db );
  }
#endif
  p->pc = -1;
  p->rc = SQLITE_OK;
  p->errorAction = OE_Abort;
  p->nChange = 0;
................................................................................
  nVar = pParse->nVar;
  nMem = pParse->nMem;
  nCursor = pParse->nTab;
  nArg = pParse->nMaxArg;
  nOnce = pParse->nOnce;
  if( nOnce==0 ) nOnce = 1; /* Ensure at least one byte in p->aOnceFlag[] */
  
  /* For each cursor required, also allocate a memory cell. Memory
  ** cells (nMem+1-nCursor)..nMem, inclusive, will never be used by
  ** the vdbe program. Instead they are used to allocate memory for
  ** VdbeCursor/BtCursor structures. The blob of memory associated with 
  ** cursor 0 is stored in memory cell nMem. Memory cell (nMem-1)
  ** stores the blob of memory associated with cursor 1, etc.
  **
  ** See also: allocateCursor().
  */
  nMem += nCursor;


  /* Figure out how much reusable memory is available at the end of the
  ** opcode array.  This extra memory will be reallocated for other elements
  ** of the prepared statement.
  */
  n = ROUND8(sizeof(Op)*p->nOp);              /* Bytes of opcode memory used */
  x.pSpace = &((u8*)p->aOp)[n];               /* Unused opcode memory */
................................................................................
    }
  }
  p->nzVar = pParse->nzVar;
  p->azVar = pParse->azVar;
  pParse->nzVar =  0;
  pParse->azVar = 0;
  if( p->aMem ){
    p->aMem--;                      /* aMem[] goes from 1..nMem */
    p->nMem = nMem;                 /*       not from 0..nMem-1 */
    for(n=1; n<=nMem; n++){
      p->aMem[n].flags = MEM_Undefined;
      p->aMem[n].db = db;
    }
  }
  p->explain = pParse->explain;
  sqlite3VdbeRewind(p);
}
................................................................................
    sqlite3VdbeFrameRestore(pFrame);
    p->pFrame = 0;
    p->nFrame = 0;
  }
  assert( p->nFrame==0 );
  closeCursorsInFrame(p);
  if( p->aMem ){
    releaseMemArray(&p->aMem[1], p->nMem);
  }
  while( p->pDelFrame ){
    VdbeFrame *pDel = p->pDelFrame;
    p->pDelFrame = pDel->pParent;
    sqlite3VdbeFrameDelete(pDel);
  }

................................................................................

#ifdef SQLITE_DEBUG
  /* Execute assert() statements to ensure that the Vdbe.apCsr[] and 
  ** Vdbe.aMem[] arrays have already been cleaned up.  */
  int i;
  if( p->apCsr ) for(i=0; i<p->nCursor; i++) assert( p->apCsr[i]==0 );
  if( p->aMem ){
    for(i=1; i<=p->nMem; i++) assert( p->aMem[i].flags==MEM_Undefined );
  }
#endif

  sqlite3DbFree(db, p->zErrMsg);
  p->zErrMsg = 0;
  p->pResultSet = 0;
}

  */
  assert( p->nOp>0 );

  /* Set the magic to VDBE_MAGIC_RUN sooner rather than later. */
  p->magic = VDBE_MAGIC_RUN;

#ifdef SQLITE_DEBUG
  for(i=0; i<p->nMem; i++){
    assert( p->aMem[i].db==p->db );
  }
#endif
  p->pc = -1;
  p->rc = SQLITE_OK;
  p->errorAction = OE_Abort;
  p->nChange = 0;
................................................................................
  nVar = pParse->nVar;
  nMem = pParse->nMem;
  nCursor = pParse->nTab;
  nArg = pParse->nMaxArg;
  nOnce = pParse->nOnce;
  if( nOnce==0 ) nOnce = 1; /* Ensure at least one byte in p->aOnceFlag[] */
  
  /* Each cursor uses a memory cell.  The first cursor (cursor 0) can
  ** use aMem[0] which is not otherwise used by the VDBE program.  Allocate
  ** space at the end of aMem[] for cursors 1 and greater.




  ** See also: allocateCursor().
  */
  nMem += nCursor;
  if( nCursor==0 && nMem>0 ) nMem++;  /* Space for aMem[0] even if not used */

  /* Figure out how much reusable memory is available at the end of the
  ** opcode array.  This extra memory will be reallocated for other elements
  ** of the prepared statement.
  */
  n = ROUND8(sizeof(Op)*p->nOp);              /* Bytes of opcode memory used */
  x.pSpace = &((u8*)p->aOp)[n];               /* Unused opcode memory */
................................................................................
    }
  }
  p->nzVar = pParse->nzVar;
  p->azVar = pParse->azVar;
  pParse->nzVar =  0;
  pParse->azVar = 0;
  if( p->aMem ){

    p->nMem = nMem;
    for(n=0; n<nMem; n++){
      p->aMem[n].flags = MEM_Undefined;
      p->aMem[n].db = db;
    }
  }
  p->explain = pParse->explain;
  sqlite3VdbeRewind(p);
}
................................................................................
    sqlite3VdbeFrameRestore(pFrame);
    p->pFrame = 0;
    p->nFrame = 0;
  }
  assert( p->nFrame==0 );
  closeCursorsInFrame(p);
  if( p->aMem ){
    releaseMemArray(p->aMem, p->nMem);
  }
  while( p->pDelFrame ){
    VdbeFrame *pDel = p->pDelFrame;
    p->pDelFrame = pDel->pParent;
    sqlite3VdbeFrameDelete(pDel);
  }

................................................................................

#ifdef SQLITE_DEBUG
  /* Execute assert() statements to ensure that the Vdbe.apCsr[] and 
  ** Vdbe.aMem[] arrays have already been cleaned up.  */
  int i;
  if( p->apCsr ) for(i=0; i<p->nCursor; i++) assert( p->apCsr[i]==0 );
  if( p->aMem ){
    for(i=0; i<p->nMem; i++) assert( p->aMem[i].flags==MEM_Undefined );
  }
#endif

  sqlite3DbFree(db, p->zErrMsg);
  p->zErrMsg = 0;
  p->pResultSet = 0;
}

Changes to src/vdbemem.c.

**
** This is used for testing and debugging only - to make sure shallow
** copies are not misused.
*/
void sqlite3VdbeMemAboutToChange(Vdbe *pVdbe, Mem *pMem){
  int i;
  Mem *pX;
  for(i=1, pX=&pVdbe->aMem[1]; i<=pVdbe->nMem; i++, pX++){
    if( pX->pScopyFrom==pMem ){
      pX->flags |= MEM_Undefined;
      pX->pScopyFrom = 0;
    }
  }
  pMem->pScopyFrom = 0;
}

**
** This is used for testing and debugging only - to make sure shallow
** copies are not misused.
*/
void sqlite3VdbeMemAboutToChange(Vdbe *pVdbe, Mem *pMem){
  int i;
  Mem *pX;
  for(i=0, pX=pVdbe->aMem; i<pVdbe->nMem; i++, pX++){
    if( pX->pScopyFrom==pMem ){
      pX->flags |= MEM_Undefined;
      pX->pScopyFrom = 0;
    }
  }
  pMem->pScopyFrom = 0;
}

Changes to test/capi3.test.

do_test capi3-3.2 {
  sqlite3_close $db2
} {SQLITE_OK}
do_test capi3-3.3 {
  catch {
    set db2 [sqlite3_open /bogus/path/test.db {}]
  }

  sqlite3_extended_errcode $db2
} {SQLITE_CANTOPEN}
do_test capi3-3.4 {
  sqlite3_errmsg $db2
} {unable to open database file}
do_test capi3-3.5 {
  sqlite3_close $db2
} {SQLITE_OK}
if {[clang_sanitize_address]==0} {
  do_test capi3-3.6.1-misuse {
    sqlite3_close $db2
  } {SQLITE_MISUSE}
  do_test capi3-3.6.2-misuse {
    sqlite3_errmsg $db2
  } {library routine called out of sequence}








>
|
|




|
|

do_test capi3-3.2 {
  sqlite3_close $db2
} {SQLITE_OK}
do_test capi3-3.3 {
  catch {
    set db2 [sqlite3_open /bogus/path/test.db {}]
  }
  set ::capi3_errno [sqlite3_system_errno $db2]
  list [sqlite3_extended_errcode $db2] [expr {$::capi3_errno!=0}]
} {SQLITE_CANTOPEN 1}
do_test capi3-3.4 {
  sqlite3_errmsg $db2
} {unable to open database file}
do_test capi3-3.5 {
  list [sqlite3_system_errno $db2] [sqlite3_close $db2]
} [list $::capi3_errno SQLITE_OK]
if {[clang_sanitize_address]==0} {
  do_test capi3-3.6.1-misuse {
    sqlite3_close $db2
  } {SQLITE_MISUSE}
  do_test capi3-3.6.2-misuse {
    sqlite3_errmsg $db2
  } {library routine called out of sequence}

Changes to test/exclusive.test.

  }
} {}
do_test exclusive-5.1 {
  # Three files are open: The db, journal and statement-journal.
  # (2016-03-04) The statement-journal is now opened lazily
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} [expr 2 - ($TEMP_STORE>=2)]
do_test exclusive-5.2 {
  execsql {
    COMMIT;
  }
  # One file open: the db.
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
................................................................................
  execsql {
    INSERT INTO abc SELECT a+10, b+10, c+10 FROM abc;
  }
  # Three files are open: The db, journal and statement-journal.
  # 2016-03-04: The statement-journal open is deferred
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} [expr 2 - ($TEMP_STORE>=2)]
do_test exclusive-5.5 {
  execsql {
    COMMIT;
  }
  # Three files are still open: The db, journal and statement-journal.
  # 2016-03-04: The statement-journal open is deferred
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} [expr 2 - ($TEMP_STORE>=2)]
do_test exclusive-5.6 {
  execsql {
    PRAGMA locking_mode = normal;
    SELECT * FROM abc;
  }
} {normal 1 2 3 2 3 4 5 6 7 11 12 13 12 13 14 15 16 17}
do_test exclusive-5.7 {

  }
} {}
do_test exclusive-5.1 {
  # Three files are open: The db, journal and statement-journal.
  # (2016-03-04) The statement-journal is now opened lazily
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} {2}
do_test exclusive-5.2 {
  execsql {
    COMMIT;
  }
  # One file open: the db.
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
................................................................................
  execsql {
    INSERT INTO abc SELECT a+10, b+10, c+10 FROM abc;
  }
  # Three files are open: The db, journal and statement-journal.
  # 2016-03-04: The statement-journal open is deferred
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} {2}
do_test exclusive-5.5 {
  execsql {
    COMMIT;
  }
  # Three files are still open: The db, journal and statement-journal.
  # 2016-03-04: The statement-journal open is deferred
  set sqlite_open_file_count
  expr $sqlite_open_file_count-$extrafds
} {2}
do_test exclusive-5.6 {
  execsql {
    PRAGMA locking_mode = normal;
    SELECT * FROM abc;
  }
} {normal 1 2 3 2 3 4 5 6 7 11 12 13 12 13 14 15 16 17}
do_test exclusive-5.7 {

Changes to test/sqldiff1.test.

#***********************************************************************
#
# Quick tests for the sqldiff tool
#
set testdir [file dirname $argv0]
source $testdir/tester.tcl

if {$tcl_platform(platform)=="windows"} {
  set PROG "sqldiff.exe"
} else {
  set PROG "./sqldiff"
}
if {![file exe $PROG]} {
  puts "sqldiff cannot run because $PROG is not available"
  finish_test
  return
}
db close
forcedelete test.db test2.db
sqlite3 db test.db

do_test sqldiff-1.0 {
  db eval {
    CREATE TABLE t1(a INTEGER PRIMARY KEY, b);








<
<
<
|
|
<
<
<
<
<

#***********************************************************************
#
# Quick tests for the sqldiff tool
#
set testdir [file dirname $argv0]
source $testdir/tester.tcl




set PROG [test_find_sqldiff]






db close
forcedelete test.db test2.db
sqlite3 db test.db

do_test sqldiff-1.0 {
  db eval {
    CREATE TABLE t1(a INTEGER PRIMARY KEY, b);

Changes to test/tester.tcl.

  sqlite3_shutdown
  eval sqlite3_config_pagecache $::old_pagecache_config
  unset ::old_pagecache_config 
  sqlite3_initialize
  autoinstall_test_functions
  sqlite3 db test.db
}















# Find the name of the 'shell' executable (e.g. "sqlite3.exe") to use for
# the tests in shell[1-5].test. If no such executable can be found, invoke
# [finish_test ; return] in the callers context.
#
proc test_find_cli {} {
  if {$::tcl_platform(platform)=="windows"} {
    set ret "sqlite3.exe"
  } else {
    set ret "sqlite3"
  }

  set ret [file normalize [file join $::cmdlinearg(TESTFIXTURE_HOME) $ret]]
  if {![file executable $ret]} {
    finish_test



    return -code return

  }
  return $ret
}

# If the library is compiled with the SQLITE_DEFAULT_AUTOVACUUM macro set
# to non-zero, then set the global variable $AUTOVACUUM to 1.
set AUTOVACUUM $sqlite_options(default_autovacuum)

# Make sure the FTS enhanced query syntax is disabled.
set sqlite_fts3_enable_parentheses 0








>
>
>
>
>
>
>
>
>
>
>
>
>
>






|
|
|
<
|
>
|
|
|
>
>
>
|
>
|
<
|

  sqlite3_shutdown
  eval sqlite3_config_pagecache $::old_pagecache_config
  unset ::old_pagecache_config 
  sqlite3_initialize
  autoinstall_test_functions
  sqlite3 db test.db
}

proc test_find_binary {nm} {
  if {$::tcl_platform(platform)=="windows"} {
    set ret "$nm.exe"
  } else {
    set ret $nm
  }
  set ret [file normalize [file join $::cmdlinearg(TESTFIXTURE_HOME) $ret]]
  if {![file executable $ret]} {
    finish_test
    return ""
  }
  return $ret
}

# Find the name of the 'shell' executable (e.g. "sqlite3.exe") to use for
# the tests in shell[1-5].test. If no such executable can be found, invoke
# [finish_test ; return] in the callers context.
#
proc test_find_cli {} {
  set prog [test_find_binary sqlite3]
  if {$prog==""} { return -code return }
  return $prog

}

# Find the name of the 'sqldiff' executable (e.g. "sqlite3.exe") to use for
# the tests in sqldiff tests. If no such executable can be found, invoke
# [finish_test ; return] in the callers context.
#
proc test_find_sqldiff {} {
  set prog [test_find_binary sqldiff]
  if {$prog==""} { return -code return }
  return $prog
}



# If the library is compiled with the SQLITE_DEFAULT_AUTOVACUUM macro set
# to non-zero, then set the global variable $AUTOVACUUM to 1.
set AUTOVACUUM $sqlite_options(default_autovacuum)

# Make sure the FTS enhanced query syntax is disabled.
set sqlite_fts3_enable_parentheses 0

Changes to tool/sqldiff.c.

  char **azCol;                   /* NULL terminated array of col names */
  int i;
  int nCol;
  Str ct = {0, 0, 0};             /* The "CREATE TABLE data_xxx" statement */
  Str sql = {0, 0, 0};            /* Query to find differences */
  Str insert = {0, 0, 0};         /* First part of output INSERT statement */
  sqlite3_stmt *pStmt = 0;


  /* --rbu mode must use real primary keys. */
  g.bSchemaPK = 1;

  /* Check that the schemas of the two tables match. Exit early otherwise. */
  checkSchemasMatch(zTab);

................................................................................
    if( ct.z ){
      fprintf(out, "%s\n", ct.z);
      strFree(&ct);
    }

    /* Output the first part of the INSERT statement */
    fprintf(out, "%s", insert.z);


    if( sqlite3_column_type(pStmt, nCol)==SQLITE_INTEGER ){
      for(i=0; i<=nCol; i++){
        if( i>0 ) fprintf(out, ", ");
        printQuoted(out, sqlite3_column_value(pStmt, i));
      }
    }else{
................................................................................
    }

    /* And the closing bracket of the insert statement */
    fprintf(out, ");\n");
  }

  sqlite3_finalize(pStmt);







  strFree(&ct);
  strFree(&sql);
  strFree(&insert);
}

/*
................................................................................
  }
  rc = sqlite3_exec(g.db, "SELECT * FROM aux.sqlite_master", 0, 0, &zErrMsg);
  if( rc || zErrMsg ){
    cmdlineError("\"%s\" does not appear to be a valid SQLite database", zDb2);
  }

  if( neverUseTransaction ) useTransaction = 0;
  if( useTransaction ) printf("BEGIN TRANSACTION;\n");






  if( zTab ){
    xDiff(zTab, out);
  }else{
    /* Handle tables one by one */
    pStmt = db_prepare(
      "SELECT name FROM main.sqlite_master\n"
      " WHERE type='table' AND sql NOT LIKE 'CREATE VIRTUAL%%'\n"

  char **azCol;                   /* NULL terminated array of col names */
  int i;
  int nCol;
  Str ct = {0, 0, 0};             /* The "CREATE TABLE data_xxx" statement */
  Str sql = {0, 0, 0};            /* Query to find differences */
  Str insert = {0, 0, 0};         /* First part of output INSERT statement */
  sqlite3_stmt *pStmt = 0;
  int nRow = 0;                   /* Total rows in data_xxx table */

  /* --rbu mode must use real primary keys. */
  g.bSchemaPK = 1;

  /* Check that the schemas of the two tables match. Exit early otherwise. */
  checkSchemasMatch(zTab);

................................................................................
    if( ct.z ){
      fprintf(out, "%s\n", ct.z);
      strFree(&ct);
    }

    /* Output the first part of the INSERT statement */
    fprintf(out, "%s", insert.z);
    nRow++;

    if( sqlite3_column_type(pStmt, nCol)==SQLITE_INTEGER ){
      for(i=0; i<=nCol; i++){
        if( i>0 ) fprintf(out, ", ");
        printQuoted(out, sqlite3_column_value(pStmt, i));
      }
    }else{
................................................................................
    }

    /* And the closing bracket of the insert statement */
    fprintf(out, ");\n");
  }

  sqlite3_finalize(pStmt);
  if( nRow>0 ){
    Str cnt = {0, 0, 0};
    strPrintf(&cnt, "INSERT INTO rbu_count VALUES('data_%q', %d);", zTab, nRow);
    fprintf(out, "%s\n", cnt.z);
    strFree(&cnt);
  }

  strFree(&ct);
  strFree(&sql);
  strFree(&insert);
}

/*
................................................................................
  }
  rc = sqlite3_exec(g.db, "SELECT * FROM aux.sqlite_master", 0, 0, &zErrMsg);
  if( rc || zErrMsg ){
    cmdlineError("\"%s\" does not appear to be a valid SQLite database", zDb2);
  }

  if( neverUseTransaction ) useTransaction = 0;
  if( useTransaction ) fprintf(out, "BEGIN TRANSACTION;\n");
  if( xDiff==rbudiff_one_table ){
    fprintf(out, "CREATE TABLE IF NOT EXISTS rbu_count"
           "(tbl TEXT PRIMARY KEY COLLATE NOCASE, cnt INTEGER) "
           "WITHOUT ROWID;\n"
    );
  }
  if( zTab ){
    xDiff(zTab, out);
  }else{
    /* Handle tables one by one */
    pStmt = db_prepare(
      "SELECT name FROM main.sqlite_master\n"
      " WHERE type='table' AND sql NOT LIKE 'CREATE VIRTUAL%%'\n"