Documentation Source Text

<title>SQLite Database File Format</title>
<tcl>

hd_keywords {first edition file format document}
source [file join $::DOC pages fancyformat.tcl]
fancyformat_document "SQLite Database File Format" hlr30000.txt {

[h1 "Document Overview"]

  [h2 "Scope and Purpose"]

  <p>
    
    This document provides an engineering guide to the file formats used by 
    SQLite to store databases on disk. It also contains a description of the
    file locking protocol used by SQLite to control read and write access to 
    the files and other protocols for safely modifying the database in a live
    system (one that may contain other database clients). It is intended that 
    this document shall provide all the information required to create an 
    system that reads and writes SQLite databases in a way that is completely 
    compatible with SQLite itself. There are two broad purposes for providing 
    this information:

  <ul>
    <li><p> To make it easier to maintain, test and improve the SQLite 
            software library.

    <li><p> To facilitate the development of external (non-SQLite) software that may 
            operate directly on SQLite databases stored within a file-system. For
	    example a database space analysis utility or a competing database
	    client implementation.
  </ul>

  <p>
    A shorter and more recent \[second edition file format document\]
    also available.  The two file format descriptions are independently
    written and hence serve as cross-checks of one another.  Any
    incompatibilities between the two documents should be considered a bug.

  <p>
    The availability of this information makes an SQLite database an even safer
    choice for long-term data storage. If at some point in the future the
    SQLite software library cannot be used to access an SQLite database that
    contains useful data, a procedure or software module may be developed based
    on the content of this document to extract the required data.

  <p>
    None of the information contained in this document is required by programmers
    wishing to use the SQLite library in applications. The intended audience is
    engineers working on SQLite itself or those interested in creating alternative
    methods of accessing SQLite databases (without using SQLite).


  [h2 "Document and Requirements Organization"]

    <p>
      The content of this document is divided into three sections.

    <p>
      <b>Section <cite>database_file_format</cite></b> describes the format 
      of a database image. A database image is the serialized form of an 
      SQLite database that is stored on disk.

    <p>
      Usually, an SQLite database image is stored in a single file on disk, 
      an SQLite database file. However, while the database image as stored 
      on disk is being modified, it may be temporarily stored in a more
      convoluted format, distributed between two files, the database file
      and a journal file. If a failure occurs while modifying a database image
      in this fashion, then the database image must be extracted from the
      database and journal files found in the file-system following recovery
      (other documentation refers to this as "hot journal rollback"). <b>Section 
      <cite>file_system_usage</cite></b> describes the format used by the
      journal file and the rules for correctly reading a database image from
      the combination of a database file and journal file. 
      

    <p><b>Section <cite>interoperability_requirements</cite></b> contains 
      descriptions of and software requirements related to other protocols that
      must be observed by software that reads and writes SQLite databases
      within a live system, including:


    <ul>
      <li>requirements governing the integrity of database file-system representations,
      <li>the locking protocol used by SQLite to manage read and write access
          to the database and journal files within the file-system, and
      <li>the change-counter and schema-cookie protocols that must be followed
          by all database writers to facilitate the implementation of
          efficient in-memory caches of the database schema and content by
          readers and writers.
    </ul>

  [h2 "Glossary"]
    <table id=glossary>
      <tr><td>Auto-vacuum last root-page<td>
        A page number stored as 32-bit integer at byte offset 52 of the
        database header (see section <cite>database_header</cite>). In
        an auto-vacuum database, this is the numerically largest 
        <i>root-page</i> number in the database. Additionally, all pages that
        occur before this page in the database are either B-Tree <i>root
        pages</i>, <i>pointer-map pages</i> or the <i>locking page</i>.

      <tr><td>Auto-vacuum database      <td>
        Each database is either an auto-vacuum database or a non auto-vacuum
        database. Auto-vacuum databases feature pointer-map pages (section
        <cite>pointer_map_pages</cite>) and have a non-zero value stored
        as a 4-byte big-endian integer at offset 52 of the database header (section
        <cite>database_header</cite>).
      <tr><td>B-Tree                    <td>
        A B-Tree is a tree structure optimized for offline storage. The table
        and index data in an SQLite database file is stored in B-Tree
        structures.

      <tr><td>B-Tree cell               <td>
        Each database page that is part of a B-Tree structure contains zero
        or more B-Tree cells. A B-Tree cell contains a single B-Tree key value
        (either an integer or database record) and optionally an associated
        database record value.

      <tr><td>B-Tree page               <td>
        A database page that is part of a B-Tree tree structure (not an
        overflow page).

      <tr><td>(B-Tree) page header      <td>
        The 8-byte (leaf pages) or 12-byte (internal node pages) header that
        occurs at the start of each B-Tree page.

      <tr><td>Cell content area         <td>
        The area within a B-Tree page in which the B-Tree cells are stored.

      <tr><td>(Database) text encoding  <td>
        The text encoding used for all text values in the database file. One
        of UTF-8, big-endian UTF-16 and little-endian UTF-16. The database
        text encoding is defined by a 4 byte field stored at byte offset
        56 of the database header (see section <cite>database_header</cite>).

      [Glossary "Database header" {
        The first 100 bytes of an SQLite database image constitute the
	database header. See section <cite>database_header</cite> for details.
      }]

      <tr><td>(Database) page size      <td>
        An SQLite database file is divided into one or more pages of
        page-size bytes each.

      <tr><td>Database record           <td>
        A database record is a blob of data containing the serialized
        representation of an ordered list of one or more SQL values.

      <tr><td>Database record header    <td>
        The first part of each database record contains the database
        record header. It encodes the types and lengths of values stored
        in the record (see section <cite>record_format</cite>).

      <tr><td>Database record data area <td>
        Following the database record header in each database record is
        the database record data area. It contains the actual data (string
        content, numeric value etc.) of all values in the record 
        (see section <cite>record_format</cite>).

      <tr><td>Default pager cache size  <td>
        A 32-bit integer field stored at byte offset 48 of the database file
        header (see section <cite>database_header</cite>).

      <tr><td style="white-space:nowrap">(Database) usable page size <td>
        The number of bytes of each database page that is usable. This
        is the page-size less the number of bytes left unused at the end
        of each page. The number of bytes left unused is governed by the
        value stored at offset 20 of the database header (see section
        <cite>database_header</cite>).

      <tr><td>File format read version  <td>
        Single byte field stored at byte offset 20 of the database header
        (see section <cite>database_header</cite>).

      <tr><td>File format write version  <td>
        Single byte field stored at byte offset 19 of the database header
        (see section <cite>database_header</cite>).

      <tr><td>File change counter       <td>
        A 32-bit integer field stored at byte offset 24 of the database file
        header (see section <cite>database_header</cite>). Normally, SQLite
        increments this value each time it commits a transaction.

      <tr><td>Fragment                  <td>
        A block of 3 or less bytes of unused space within the cell content
        area of a B-Tree page.

      <tr><td>Free block                <td>
        A block of 4 or more bytes of unused space within the cell content
        area of a B-Tree page.

      <tr><td>Free block list           <td>
        The linked list of all free blocks on a single B-Tree page (see 
        section <cite>index_btree_page_format</cite>).

      <tr><td>Free page                 <td>
        A page that is not currently being used to store any database data
        or meta data. Part of the free-page list.

      <tr><td>Free page list            <td>
        A data structure within an SQLite database file that links all the
        free-pages together.

      <tr><td>Index B-Tree              <td>
        One of two variants on the B-Tree data structure used within SQLite
        database files. An index B-Tree (section <cite>index_btrees</cite>)
        uses database records as keys.

      <tr><td>Incremental Vacuum flag   <td>
        A 32-bit integer field stored at byte offset 64 of the database file
        header (see section <cite>database_header</cite>). In auto-vacuum 
        databases, if this field is non-zero then the database is not
        automatically compacted at the end of each transaction.

      <tr><td>Locking page              <td>
        The database page that begins at the 1GB (2<sup>30</sup> byte)
        boundary. This page is always left unused.

      <tr><td>Logical database          <td>
        An SQLite database file is a serialized representation of a logical
        database. A logical database consists of the SQL database schema,
        the content of the various tables in the database, and assorted
        database properties that may be set by the user (auto-vacuum,
        page-size, user-cookie value etc.),

      <tr><td>Non-auto-vacuum database  <td>
        Any database that is not an auto-vacuum database. A non-auto-vacuum
        database contains no pointer-map pages and has a zero value stored
        in the 4-byte big-endian integer field at offset 52 of the database
        database header (section <cite>database_header</cite>).

      <tr><td>Overflow chain             <td>
        A linked list of overflow pages across which a single (large)
        database record is stored (see section 
        <cite>overflow_page_chains</cite>).

      <tr><td>Overflow page             <td>
        If a B-Tree cell is too large to store within a B-Tree page, a
        portion of it is stored using a chain of one or more overflow pages
        (see section <cite>overflow_page_chains</cite>).

      <tr><td>Pointer-map page          <td>
        A database page used to store meta data only present in auto-vacuum
        databases (see section <cite>pointer_map_pages</cite>).

      <tr><td>Right child page          <td>
        Each internal B-Tree node page has one or more child pages. The
        rightmost of these (the one containing the largest key values) is
        known as the right child page.

      <tr><td>Root page                 <td>
        A root page is a database page used to store the root node of a
        B-Tree data structure.

      <tr><td>Schema layer file format  <td>
        An integer between 1 and 4 stored as a 4 byte big-endian integer at
        offset 44 of the database header (section <cite>database_header</cite>).
        Certain file format constructions may only be present in databases
        with a certain minimum schema layer file format value.

      <tr><td>Schema table              <td>
        The table B-Tree with root-page 1 used to store database records
        describing the database schema. Accessible as the "sqlite_master" 
        table from within SQLite.

      <tr><td>Schema version            <td>
        A 32-bit integer field stored at byte offset 40 of the database file
        header (see section <cite>database_header</cite>). Normally, SQLite
        increments this value each time it modifies the database schema.

      <tr><td>Table B-Tree              <td>
        One of two variants on the B-Tree data structure used within SQLite
        database files. A table B-Tree (section <cite>table_btrees</cite>)
        uses 64 bit integers as key values and stores an associated database
        record along with each key value.

      <tr><td>User cookie               <td>
        A 32-bit integer field stored at byte offset 60 of the database file
        header (see section <cite>database_header</cite>). This value can be
        set and queried using the user_version PRAGMA but is not otherwise
        used by SQLite.

      <tr><td>Variable Length Integer   <td>
        A format used for storing 64-bit signed integer values in SQLite 
        database files. Consumes between 1 and 9 bytes of space, depending
        on the precise value being stored.

      <tr><td>Well formed database file <td>
        An SQLite database file that meets all the criteria laid out in
        section <cite>database_file_format</cite> of this document.

      [Glossary "Database image" {
        A serialized blob of data representing an SQLite database. The
        contents of a database file are usually a valid database image.
      }]
      [Glossary "Database file" {
        A database file is a file on disk that usually, but not always,
        contains a well-formed database image.
      }]
      [Glossary "Journal file" {
        For each database file, there may exist an associated journal file
	stored in the same file-system directory. Under some circumstances,
	the database image may be distributed between the database and journal
	files (instead of being stored wholly within the database file).
      }]
      [Glossary "Page size" {
        An SQLite database image is divided into fixed size pages, each 
        "page size" bytes in size.
      }]
      [Glossary "Sector size" {
        In this document, the term "sector size" refers to a field in a
	journal header which determines some aspects of the layout of the
	journal file. It is set by SQLite (or a compatible) application
	based on the properties of the underlying file-system that the journal
	file is being written to.
      }]
      [Glossary "Journal Section" {
	A journal file may contain multiple journal sections. A journal section
	consists of a journal header followed by zero or more journal records.
      }]
      [Glossary "Journal Header" {
	A journal header is a control block sector-size bytes in size that
	appears at the start of each journal section within a journal file.
      }]
      [Glossary "Journal Record" {
	A journal record is a structure used to store data for a single
	database page within a journal file. A single journal file may contain
	many journal records.
      }]
      [Glossary "Master Journal Pointer" {
        A master journal pointer is a structure that may appear at the end of
	a journal file. It contains a full file-system path identifying 
	a master-journal file.
      }]
      [Glossary "Database File-System Representation" {
        A file or files within the file-system used to store an SQLite 
        database image.
      }]

      [Glossary "Database user-cookie" {
        An SQLite database contains a single 32-bit signed integer field known
	as the database user-cookie. Applications may read and write this field
	for any purpose.
      }]

    </table>

<!--
h1 "SQLite Database Files" sqlite_database_files
 
  <p>
    The bulk of this document, section <cite>database_file_format</cite>,
    contains the definition of a <i>well-formed SQLite database file</i>.
    SQLite is required to create database files that meet this definition.

          [fancyformat_import_requirement H30010]

  <p>
    Additionally, the database file should contain a serialized version
    of the logical database produced by the transaction. For all but the
    most trivial logical databases, there are many possible serial 
    representations.

          [fancyformat_import_requirement H30020]
-->

<!--
  <p>
    Section <cite>database_file_manipulation</cite> contains requirements
    describing in more detail the way in which SQLite manipulates the
    fields and data structures described in section
    <cite>database_file_format</cite> under various circumstances. These
    requirements are to a certain extent derived from the requirements 
    in this section.
-->
  

[h1 "Database Image Format Details" database_file_format]

  <p>
    This section describes the various fields and sub-structures that make up
    the format used by SQLite to serialize a logical SQL database. A serialized
    logical database is referred to as a database image. Section
    <cite>file_system_usage</cite> describes the way a database image is stored
    in the file-system. Most of the time a database image is stored in a single
    file, the database file. So while reading this section, the term database 
    image may be understood to mean "contents of the database file". However,
    it is important to remember that there are exceptions to this.

  <p>
    This section does not contain requirements governing the behaviour of any
    software system. Instead, along with the plain language description of the
    file format are a series of succinct, testable statements describing the
    properties of "well-formed SQLite database files".  Some of these
    statements describe the contents of the database file in terms of the
    contents of the logical SQL database that it is a serialization of. e.g.
    "For each SQL table in the database, the database file shall...". The 
    contents of a logical database consist of:

  <ul>
    <li>The database schema: The set of database tables, virtual tables, 
        indexes, triggers and views stored in the database.

    <li>The database contents: The set of tuples (rows) stored in
	each database table.

    <li>Other database properties, as follows:
      <ol>
	<li>The page-size of the database.
	<li>The text-encoding of the database.
	<li>A flag indicating whether or not the database is an auto-vacuum 
	    database.
	<li>The value of the database user-cookie.
	<li>If the database is an auto-vacuum database, a flag indicating 
	    whether or not the database is in incremental vacuum mode or not.
	<li>The default page cache size in pages to use with the database (an 
	    integer field).
      </ol>
  </ul>

  <p>
    Of the six database properties enumerated above, the values taken by the
    initial three dramatically affect the structure of the database image. Any
    software system that handles SQLite database images will need to understand
    and interpret them. Properties 4 to 6 may be considered advisory. Although
    properties 5 and 6 modify the operation of the SQLite library in 
    well-defined manners, an alternative SQLite database client is free to
    interpret them differently, or not interpret them at all. 

  <p class=todo>
    The concept of a logical database and its contents should be defined
    properly in some requirements document so that it can be referenced from
    here and other places. The definition will be something like the list of
    bullet points above.

  <p>
    Many of the numbered requirements in the following sub-sections describe 
    the relationship between the contents of the logical database, as itemized
    above, and the contents of the serialized database image. Others describe
    the relationships between various database image substructures, invariants
    that are true for all well-formed database images.

  <p>
    A well-formed SQLite database image is defined as an image for which
    all of the statements itemized as requirements within this section
    are true. <span class=todo>mention the requirements numbering scheme
    here.</span> A software system that wishes to inter-operate with other
    systems using the SQLite database image format should only ever
    output well-formed SQLite databases. In the case of SQLite itself,
    the system should ensure that the database file contains a well-formed 
    database image the conclusion of each transaction.

  [h2 "Image Format Overview" "fileformat_overview"]
    <p>
      A B-Tree is a data structure designed for offline storage of a set of
      unique key values. It is structured so as to support fast querying 
      for a single key or range of keys. As implemented in SQLite, each
      entry may be associated with a blob of data that is not part of the
      key. For the canonical introduction to the B-Tree and its variants, 
      refer to reference <cite>ref_comer_btree</cite>. The B-Tree 
      implementation in SQLite also adopts some of the enhancements 
      suggested in <cite>ref_knuth_btree</cite>.
    <p>
      An SQLite database image contains one or more B-Tree structures. Each
      B-Tree structure stores the data for a single database table or 
      index. Hence each database file contains a single B-Tree to store
      the contents of the <i>sqlite_master</i> table, and one B-Tree
      for each database table or index created by the user. If the database
      uses auto-increment integer primary keys, then the database file
      also contains a B-Tree to store the contents of the automatically 
      created <i>sqlite_sequence</i> table.
    <p>
      SQLite uses two distinct variants of the B-Tree structure. One variant,
      hereafter referred to as a "table B-Tree" uses signed 64-bit integer
      values as keys. Each entry has an associated variable length blob of 
      data used to store a database record (see section
      <cite>record_format</cite>). Each SQLite database file contains one 
      table B-Tree for the schema table and one table B-Tree for each
      additional database table created by the user. If it is present, the
      sqlite_sequence table is also stored as a table B-Tree.
    <p>
      A database record is a blob of data containing an ordered list of
      SQL values (integers, real numbers, NULL values, blobs or strings).
      For each row in each table in the logical database, there is an 
      entry in the corresponding table B-Tree structure in the database 
      image. The entry's integer key value is same as the SQL "rowid" or 
      "integer primary key" field of the table row. The associated database 
      record is made up of the row's column values, in declaration (CREATE 
      TABLE) order.
    <p>
      The other B-Tree variant used by SQLite, hereafter an "index B-Tree"
      uses database records (section <cite>record_format</cite>) as keys.
      For this kind of B-Tree, there is no additional data associated with
      each entry. SQLite databases contain an index B-Tree for each database
      index created by the user. Database indexes may be created by CREATE
      INDEX statements, or by UNIQUE or PRIMARY KEY (but not INTEGER PRIMARY
      KEY) clauses added to CREATE TABLE statements. 
    <p>
      Index B-Tree structures contain one entry for each row in the 
      associated table in the logical SQL database. The database record used 
      as the key consists of the row's value for each of the indexed columns in
      declaration (CREATE INDEX) order, followed by the row's "rowid" or
      "integer primary key" column value.
    <p>
      For example, the following SQL script:
    <pre>
      CREATE TABLE t1(a INTEGER PRIMARY KEY, b, c, d);
      CREATE INDEX i1 ON t1(d, c);

      INSERT INTO t1 VALUES(1, 'triangle', 3, 180, 'green');
      INSERT INTO t1 VALUES(2, 'square',   4, 360, 'gold');
      INSERT INTO t1 VALUES(3, 'pentagon', 5, 540, 'grey');
      ...</pre>
    <p>
      Creates a database image containing three B-Tree structures: one table
      B-Tree to store the <i>sqlite_master</i> table, one table B-Tree to 
      store table "t1", and one index B-Tree to store index "i1". The
      B-Tree structures created for the user table and index are populated
      as shown in figure <cite>figure_examplepop</cite>.

      [Figure examplepop.gif figure_examplepop "Example B-Tree Data"]

    <p>
      The following sections and sub-sections describe precisely the format
      used to serialize the B-Tree structures within an SQLite database image.

  [h2 "Global Structure"]

    [h3 "Database Header" "database_header"]
      <p>
        An SQLite database image begins with a 100-byte database header. The database 
        header consists of a well known 16-byte sequence followed by a series of
        1, 2 and 4 byte unsigned integers. All integers in the database header (as
        well as the rest of the database file) are stored in big-endian format.
        
      <p>
        The well known 16-byte sequence that begins every SQLite database file
        is:
      <pre>
          0x53 0x51 0x4c 0x69 0x74 0x65 0x20 0x66 0x6f 0x72 0x6d 0x61 0x74 0x20 0x33 0x00</pre>

      <p>
        Interpreted as UTF-8 encoded text, this byte sequence corresponds 
        to the string "SQLite format 3" followed by a nul-terminator byte.

          [fancyformat_import_requirement H30030]

      <p>
        The 1, 2 and 4 byte unsigned integers that make up the rest of the
        database header are described in the following table.

      [Table]
        [Tr]<th>Byte Range <th>Byte Size <th width=100%>Description <th>Reqs
	[Tr]<td>16..17 <td>2<td>
            Database page size in bytes. See section 
            <cite>pages_and_page_types</cite> for details.
            <td>H30190

        [Tr]<td>18     <td>1<td>
            <p style="margin-top:0">
            File-format "write version". Currently, this field
            is always set to 1. If a value greater than 1 is read by SQLite,
            then the library will only open the file for read-only access.

            <p style="margin-bottom:0">
            This field and the next one are intended to be used for 
            forwards compatibility, should the need ever arise. If in the
            future a version of SQLite is created that uses a file format
            that may be safely read but not written by older versions of
            SQLite, then this field will be set to a value greater than 1
            to prevent older SQLite versions from writing to a file that
            uses the new format. 
            <td>H30040

        [Tr]<td>19     <td>1<td>
            <p style="margin-top:0">
             File-format "read version". Currently, this 
            field is always set to 1. If a value greater than 1 is read 
            by SQLite, then the library will refuse to open the database 

            <p style="margin-bottom:0">
            Like the "write version" described above, this field exists
            to facilitate some degree of forwards compatibility, in case
            it is ever required. If a version of SQLite created in the 
            future uses a file format that may not be safely read by older
            SQLite versions, then this field will be set to a value greater
            than 1.
            <td>H30040

        [Tr]<td>20     <td>1<td>
            Number of bytes of unused space at the end of each database
            page. Usually this field is set to 0. If it is non-zero, then 
            it contains the number of bytes that are left unused at the
            end of every database page (see section
            <cite>pages_and_page_types</cite> for a description of a
            database page).
            <td>H30040

        [Tr]<td>21     <td>1<td>
             Maximum fraction of an index tree page to use for 
            embedded content. This value is used to determine the maximum
            size of a B-Tree cell to store as embedded content on a
            page that is part of an index B-Tree. Refer to section 
            <cite>index_btree_cell_format</cite> for details.
            <td>H30040

        [Tr]<td>22     <td>1<td>
            Minimum fraction of an index B-Tree page to use for
            embedded content when an entry uses one or more overflow pages.
            This value is used to determine the portion of a B-Tree cell 
            that requires one or more overflow pages to store as embedded
            content on a page that is part of an index B-Tree. Refer to
            section <cite>index_btree_cell_format</cite> for details.
            <td>H30040

        [Tr]<td>23     <td>1<td>
            Minimum fraction of an table B-Tree leaf page to use for
            embedded content when an entry uses one or more overflow pages.
            This value is used to determine the portion of a B-Tree cell 
            that requires one or more overflow pages to store as embedded
            content on a page that is a leaf of a table B-Tree. Refer to
            section <cite>table_btree_cell_format</cite> for details.
            <td>H30040

        [Tr]<td>24..27 <td>4<td>
            <p style="margin-top:0">
            The file change counter. Each time a database transaction is
            committed, the value of the 32-bit unsigned integer stored in
            this field is incremented.
            <p style="margin-bottom:0">
            SQLite uses this field to test the validity of its internal
            cache. After unlocking the database file, SQLite may retain
            a portion of the file cached in memory. However, since the file
            is unlocked, another process may use SQLite to modify the 
            contents of the file, invalidating the internal cache of the
            first process. When the file is relocked, the first process can
            check if the value of the file change counter has been modified
            since the file was unlocked. If it has not, then the internal
            cache may be assumed to be valid and may be reused.
            <td>H33040

        [Tr]<td>28..31 <td>4<td>
            <p style="margin-top:0">
            The in-header database size.  This field holds the logical size
            of the database file in pages.  This field is only valid if it
            is nonzero and if the file change counter at offset 24 exactly
            matches the version-valid-for at offset 92.  The in-header database
            size will only be valid when the database was last written by
            SQLite version 3.7.0 or later.  If the in-header database size
            is valid, then it is used as the logical size of the database.
            If the in-header database size is not valid, then the actual
            database file size is examined to determine the logical database
            size.
            <td>

        [Tr]<td>32..35 <td>4<td>
            Page number of first freelist trunk page. 
            For more details, refer to section <cite>free_page_list</cite>.
            <td>H31320

        [Tr]<td>36..39 <td>4<td>
            Number of free pages in the database file.
            For more details, refer to section <cite>free_page_list</cite>.
            <td>H31310

        [Tr]<td>40..43 <td>4<td>
            The schema version. Each time the database schema is modified (by
            creating or deleting a database table, index, trigger or view)
            the value of the 32-bit unsigned integer stored in this field
            is incremented.
            <td>H33050

        [Tr]<td>44..47 <td>4<td>
            <p style="margin-top:0">
            Schema layer file-format. This value is similar to the
            "read-version" and "write-version" fields at offsets 18 and 19
            of the database header. If SQLite encounters a database
            with a schema layer file-format value greater than the file-format
            that it understands (currently 4), then SQLite will refuse to
            access the database.
            <p>
            Usually, this value is set to 1. However, if any of the following
            file-format features are used, then the schema layer file-format
            must be set to the corresponding value or greater:
            <ol start=2 style="margin-bottom:0">
              <li> Implicit NULL values at the end of table records 
                   (see section <cite>table_btree_content</cite>).
              <li> Implicit default (non-NULL) values at the end of table
                   records (see section <cite>table_btree_content</cite>).
              <li> Descending indexes (see section
                   <cite>index_btree_compare_func</cite>) and Boolean values
                   in database records (see section <cite>record_format</cite>,
                   serial types 8 and 9).
            </ol>

            <p class=todo>
              Turns out SQLite can be tricked into violating this. If you delete
	      all tables from a database and then VACUUM the database, the
              schema layer file-format field somehow gets set to 0.
            <td>H30120

        [Tr]<td>48..51 <td>4<td>
            Default pager cache size. This field is used by SQLite to store
            the recommended pager cache size to use for the database.
            <td>H30130

        [Tr]<td>52..55 <td>4<td>
            For auto-vacuum capable databases, the numerically largest 
            root-page number in the database. Since page 1 is always the
            root-page of the schema table (section <cite>schema_table</cite>),
            this value is always non-zero for auto-vacuum databases. For
            non-auto-vacuum databases, this value is always zero.
            <td>H30140, H30141

        [Tr]<td>56..59 <td>4<td>
            (constant) Database text encoding. A value of 1 means all 
            text values are stored using UTF-8 encoding. 2 indicates
            little-endian UTF-16 text. A value of 3 means that the database
            contains big-endian UTF-16 text.  
            <td>H30150

        [Tr]<td>60..63 <td>4<td>
            The user-cookie value. A 32-bit integer value available to the
            user for read/write access.
            <td>H30160

        [Tr]<td>64..67 <td>4<td>
            The incremental-vacuum flag. In non-auto-vacuum databases this
            value is always zero. In auto-vacuum databases, this field is
            set to 1 if "incremental vacuum" mode is enabled. If incremental
            vacuum mode is not enabled, then the database file is reorganized
            so that it contains no free pages (section
            <cite>free_page_list</cite>) at the end of each database
            transaction. If incremental vacuum mode is enabled, then the
            reorganization is not performed until explicitly requested
            by the user.
            <td>H30171

        [Tr]<td>92..95 <td>4<td>
            The version-valid-for integer.  This is a copy of the
            file change counter (offset 24) for when the SQLite version number
            in offset 96 was written.  This integer is also used to determine
            if the in-header database size (offset 28) is valid.
            <td>

        [Tr]<td>99..99 <td>4<td>
            The SQLite version number (obtained from \[SQLITE_VERSION_NUMBER\])
            for the instance of SQLite that last wrote to the database file.
            Only SQLite versions 3.7.0 and later will update this number.
            <td>

      </table>

      <p>
        The four byte block beginning at offset 28 stores a big-endian integer
        which is the number of pages in the database.  Older versions of
        SQLite set this integer to zero.  For compatibility, SQLite database
        readers should be able to deal with either value.
      </p>

      <p>
        The 32 byte block beginning at offset 68 is unused in SQLite versions
        up to and including 3.6.23.1.  The 8 bytes at offset 92 came into use
        beginning with SQLite version 3.7.0.  The 24 bytes between offset 68
        and offset 91 might come into use in a future release of SQLite.
      </p>

      <p>
        The following requirements state that certain database header
        fields must contain defined constant values, even though the sqlite 
        database file format is designed to allow various values. These fields
        were intended to be flexible when the SQLite database image format
        was designed, but it has since been determined that it is faster and
        safer to require these parameters to be populated with well-known 
        values. Specifically, in a well-formed database, the following header
        fields are always set to well-known values:

      <ul>
        <li> The file-format write version (single byte field, byte offset 18), 
             is always set to 0x01.
        <li> The file-format read version (single byte field, byte offset 19), 
             is always set to 0x01.
        <li> The number of unused bytes on each page (single byte field, byte 
             offset 20), is always set to 0x00.
        <li> The maximum fraction of an index B-Tree page to use for embedded content 
	     (single byte field, byte offset 21), is always set to 0x40.  <li>
	     The minimum fraction of an index B-Tree page to use for embedded
	     content when using overflow pages (single byte field, byte 
             offset 22), is always set to 0x20.
	<li> The minimum fraction of a table B-Tree page to use for embedded
	     content when using overflow pages (single byte field, byte offset 23),
	     is always set to 0x20.
      </ul>

      <p>
        The following requirement encompasses all of the above.

          [fancyformat_import_requirement H30040]

      <p>
        Section <cite>database_file_format</cite> identifies six persistent
        user-visible properties of an SQLite database. The following 
        requirements describe the way in which these properties are stored.

          [fancyformat_import_requirement H30190]
          [fancyformat_import_requirement H30191]
          [fancyformat_import_requirement H30150]
          [fancyformat_import_requirement H30140]
          [fancyformat_import_requirement H30141]
          [fancyformat_import_requirement H30160]
          [fancyformat_import_requirement H30170]
          [fancyformat_import_requirement H30171]
          [fancyformat_import_requirement H30130]

      <p>
        The following requirement describes the valid range of values for the
        schema layer file format field.

          [fancyformat_import_requirement H30120]

      <p class=todo>
        See the note to do with the schema file format version above. Turns
        out this field may also be set to 0 by SQLite.

    [h3 "Pages and Page Types" "pages_and_page_types"]
      <p>
        The entire database file is divided into pages, each page consisting
        of <i>page-size</i> bytes, where <i>page-size</i> is the 2-byte 
        integer value stored at offset 16 of the database header (see above).
        The <i>page-size</i> is always a power of two between 512 
        (2<sup>9</sup>) and 32768 (2<sup>15</sup>) or the value 1 used to 
        represent a 65536-byte page. This field can equivalently be viewed
        as a little-endian number which is page size divided by 256.
        SQLite database files
        always consist of an exact number of pages.
      <p>
        Pages are numbered beginning from 1, not 0. Page 1 consists of
        the first <i>page-size</i> bytes of the database file. 
        The database header described in the previous section consumes
        the first 100 bytes of page 1.
      <p>
        Each page of the database file is one of the following:
      <ul>
        <li><b>A B-Tree page</b>. B-Tree pages are part of the tree 
            structures used to store database tables and indexes.
        <li><b>An overflow page</b>. Overflow pages are used by particularly
            large database records that do not fit on a single B-Tree page.
        <li><b>A free page</b>. Free pages are pages within the database file
            that are not being used to store meaningful data.
        <li><b>A "pointer-map" page</b>. In auto-vacuum capable databases
            (databases for which the 4 byte big-endian integer stored at
            byte offset 52 of the database header is non-zero) some pages are
            permanently designated "pointer-map" pages. See section 
            <cite>pointer_map_pages</cite> for details.
        <li><b>The locking page</b>. The database page that starts at
            byte offset 2<sup>30</sup>, if it is large enough to contain
            such a page, is always left unused.
      </ul>

          [fancyformat_import_requirement H30200]
          [fancyformat_import_requirement H30210]
          [fancyformat_import_requirement H30220]
        

    [h3 "The Schema Table" schema_table]
      <p>
        Apart from being the page that contains the file-header, page 1 of
        a database image is special because it is the root page of the
        B-Tree structure that contains the schema table data. From the SQL
        level, the schema table is accessible via the name "sqlite_master".
      <p>
        The exact format of the B-Tree structure and the meaning of the term
        "root page" is discussed in section <cite>btree_structures</cite>.
        For now, it is sufficient to know that the B-Tree structure is a
        data structure that stores a set of records. Each record is an
        ordered set of SQL values (the format of which is described in
        section <cite>record_format</cite>). Given the root page number of
        the B-Tree structure (which is well known for the schema table), it
        is possible to iterate through the set of records.
      <p>
        The schema table contains a record for each SQL table (including
        virtual tables) except for sqlite_master, and for each index, trigger
        and view in the logical database.  There is also an entry for each
        UNIQUE or PRIMARY KEY clause present in the definition of a database
        table. Each record in the schema table contains exactly 5 values, in
        the following order:

      [Table]
        [Tr]<th>Field<th>Description
        [Tr]<td>Schema item type.
            <td>A string value. One of "table", "index", "trigger" or "view",
                according to the schema item type. Entries associated with
                UNIQUE or PRIMARY KEY clauses have this field set to "index".
        [Tr]<td>Schema item name.
            <td>A string value. The name of the database schema item (table,
                index, trigger or view) associated with this record, if any.
                Entries associated with UNIQUE or PRIMARY KEY clauses have
                this field set to a string of the form
                "sqlite_autoindex_&lt;name&gt;_&lt;idx&gt;" where &lt;name&gt;
                is the name of the SQL table and &lt;idx&gt; is an integer
                value.

        [Tr]<td style="white-space:nowrap">Associated table name.
            <td>A string value. For "table" 
            or "view" records this is a copy of the second (previous) value. 
            For "index" and "trigger" records, this field is set to the name 
            of the associated database table.
        [Tr]<td style="white-space:nowrap">The "root page" number. 
            <td>For "trigger" and "view" records, as well as "table" records
	        associated with virtual tables, this is set to integer value 0. 
		For other "table" and "index" records (including those associated 
		with UNIQUE or PRIMARY KEY clauses), this field contains the root
                page number (an integer) of the B-Tree structure that contains
                the table or index data.
        [Tr]<td>The SQL statement.
            <td>A string value. The SQL statement used to create the schema
                item (i.e.  the complete text of an SQL "CREATE TABLE"
                statement). This field contains an empty string for table
                entries associated with PRIMARY KEY or UNIQUE clauses.
                <span class=todo>Refer to some document that describes these
                SQL statements more precisely.</span>
      </table>
      <p>
        Logical database schema items other than non-virtual tables and indexes
        (including indexes created by UNIQUE or PRIMARY key constraints) do not
        require any other data structures to be created within the database
        file.

      <p>
        Tables and indexes on the other hand, are represented within the
        database file by both an entry in the schema table and a B-Tree
        structure stored elsewhere in the file. The specific B-Tree associated
        with each database table or index is identified by its root page
        number, which is stored in the 4th field of the schema table record.
        In a non-auto-vacuum database, the B-Tree root pages may be stored
        anywhere within the database file. For an auto-vacuum database, all
        B-Tree root pages must at all times form a contiguous set starting
        at page 3 of the database file, skipping any pages that are required to
        be used as pointer-map pages (see section
        <cite>pointer_map_pages</cite>).
      <p>
        As noted in section <cite>database_header</cite>, in an auto-vacuum
        database the page number of the page immediately following the
        final root page in the contiguous set of root pages is stored
        as a 4 byte big-endian integer at byte offset 52 of the database
        header. Unless that page is itself a pointer-map page, in which
        case the page number of the page following it is stored instead.

      <p>
        For example, if the schema of a logical database is created using
        the following SQL statements:
      <pre>
          CREATE TABLE abc(a, b, c);
          CREATE INDEX i1 ON abc(b, c);
          CREATE TABLE main.def(a PRIMARY KEY, b, c, UNIQUE(b, c));
          CREATE VIEW v1 AS SELECT * FROM abc;
      </pre>
      <p>
        Then the schema table would contain a total of 7 records, as follows:

      [Table]
        [Tr]<th>Field 1<th>Field 2<th>Field 3<th>Field 4<th>Field 5
        [Tr]<td>table <td>abc <td>abc <td>2 <td>CREATE TABLE abc(a, b, c)
        [Tr]<td>index <td>i1 <td>abc <td>3 <td>CREATE INDEX i1 ON abc(b, c)
        [Tr]<td>table <td>def <td>def <td>4 <td>CREATE TABLE def(a PRIMARY KEY, b, c, UNIQUE(b, c))
        [Tr]<td>index <td>sqlite_autoindex_def_1 <td>def <td>5 <td>
        [Tr]<td>index <td>sqlite_autoindex_def_2 <td>def <td>6 <td>
        [Tr]<td>view <td>v1 <td>v1 <td>0 <td>CREATE VIEW v1 AS SELECT * FROM abc
      </table>

          [fancyformat_import_requirement H30230]
          [fancyformat_import_requirement H30240]

      <p>The following requirements describe "table" records.

          [fancyformat_import_requirement H30250]
          [fancyformat_import_requirement H30260]
          [fancyformat_import_requirement H30270]
          [fancyformat_import_requirement H30280]
          [fancyformat_import_requirement H30290]
          [fancyformat_import_requirement H30300]
          [fancyformat_import_requirement H30310]

      <p>The following requirements describe "implicit index" records.

          [fancyformat_import_requirement H30320]
          [fancyformat_import_requirement H30330]
          [fancyformat_import_requirement H30340]
          [fancyformat_import_requirement H30350]

      <p>The following requirements describe "explicit index" records.

          [fancyformat_import_requirement H30360]
          [fancyformat_import_requirement H30370]
          [fancyformat_import_requirement H30380]
          [fancyformat_import_requirement H30390]

      <p>The following requirements describe "view" records.

          [fancyformat_import_requirement H30400]
          [fancyformat_import_requirement H30410]
          [fancyformat_import_requirement H30420]
          [fancyformat_import_requirement H30430]

      <p>The following requirements describe "trigger" records.

          [fancyformat_import_requirement H30440]
          [fancyformat_import_requirement H30450]
          [fancyformat_import_requirement H30460]
          [fancyformat_import_requirement H30470]

      <p>The following requirements describe the placement of B-Tree root 
         pages in auto-vacuum databases.

          [fancyformat_import_requirement H30480]
          [fancyformat_import_requirement H30490]

 
  [h2 "B-Tree Structures" "btree_structures"]
    <p>
      A large part of any SQLite database file is given over to one or more
      B-Tree structures. A single B-Tree structure is stored using one or more
      database pages. Each page contains a single B-Tree node.
      The pages used to store a single B-Tree structure need not form a
      contiguous block. The page that contains the root node of a B-Tree
      structure is known as the "root page".

    <p>
      SQLite uses two distinct variants of the B-Tree structure:
    <ul>
      <li>The <b>table B-Tree</b>, which uses 64-bit integer values for keys. 
          In a table B-Tree, an associated database record (section
          <cite>record_format</cite>) is stored along with each entry. Table
          B-Tree structures are described in detail in section 
          <cite>table_btrees</cite>.
      <li>The <b>index B-Tree</b>, which uses database records as keys. Index
          B-Tree structures are described in detail in section 
          <cite>index_btrees</cite>.
    </ul>

          [fancyformat_import_requirement H30500]
          [fancyformat_import_requirement H30510]

    [h3 "Variable Length Integer Format" "varint_format"]
      <p>
        In several parts of the B-Tree structure, 64-bit twos-complement signed
        integer values are stored in the "variable length integer format"
        described here.
      <p>
        A variable length integer consumes from one to nine bytes of space,
        depending on the value stored. Seven bits are used from each of
        the first eight bytes present, and, if present, all eight from
        the final ninth byte. Unless the full nine byte format is used, the
        serialized form consists of all bytes up to and including the first
        byte with the 0x80 bit cleared.
      <p>
        The number of bytes present depends on the position of the most
        significant set bit in the 64-bit word. Negative numbers always have
        the most significant bit of the word (the sign bit) set and so are
        always encoded using the full nine bytes. Positive integers may be
        encoded using less space. The following table shows the 9 different
        length formats available for storing a variable length integer
        value.

      [Table]
        [Tr]<th>Bytes<th>Value Range<th>Bit Pattern
        [Tr]<td>1<td>7 bit<td>0xxxxxxx
        [Tr]<td>2<td>14 bit<td>1xxxxxxx 0xxxxxxx
        [Tr]<td>3<td>21 bit<td>1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>4<td>28 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>5<td>35 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>6<td>42 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>7<td>49 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>8<td>56 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 0xxxxxxx
        [Tr]<td>9<td>64 bit<td>1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx 1xxxxxxx xxxxxxxx
      </table>
      <p>
        When using the full 9 byte representation, the first byte contains
        the 7 most significant bits of the 64-bit value. The final byte of
        the 9 byte representation contains the 8 least significant bits of
        the 64-bit value. When using one of the other representations, the
        final byte contains the 7 least significant bits of the 64-bit value.
        The second last byte, if present, contains the 7 next least significant
        bits of the value, and so on. The significant bits of the 64-bit
        value for which no storage is provided are assumed to be zero.
      <p>
        When encoding a variable length integer, SQLite usually selects the
        most compact representation that provides enough storage to accommodate
        the most significant set bit of the value. This is not required
        however, using more bytes than is strictly necessary when encoding
        an integer is valid.

      [Table]
        [Tr]<th>Decimal<th>Hexadecimal        <th>Variable Length Integer
        [Tr]<td>43     <td>0x000000000000002B <td>0x2B
        [Tr]<td>200815 <td>0x000000000003106F <td>0x8C 0xA0 0x6F
        [Tr]<td>-1     <td>0xFFFFFFFFFFFFFFFF 
            <td>0xFF 0xFF 0xFF 0xFF 0xFF 0xFF 0xFF 0xFF 0xFF
        [Tr]<td>-78506 <td>0xFFFFFFFFFFFECD56
            <td>0xFF 0xFF 0xFF 0xFF 0xFF 0xFF 0xFD 0xCD 0x56
      </table>

          [fancyformat_import_requirement H30520]
          [fancyformat_import_requirement H30530]
          [fancyformat_import_requirement H30540]
          [fancyformat_import_requirement H30550]
      

    [h3 "Database Record Format" "record_format"]
      <p>
        A database record is a blob of data that represents an ordered
        list of one or more SQL values. Database records are used in two
        places in SQLite database files - as the associated data for entries
        in table B-Tree structures, and as the key values in index B-Tree
        structures. The size (number of bytes consumed by) a database record
        depends on the values it contains.
      <p>
        Each database record consists of a short record header followed by 
        a data area. The record header consists of <i>N+1</i> variable
        length integers (see section <cite>varint_format</cite>), where
        <i>N</i> is the number of values stored in the record.
      <p>
        The first variable length integer in a record header contains the
        size of the record header in bytes. The following <i>N</i> variable
        length integer values each describe the type and size of the 
	corresponding SQL value within the record (the second integer in the 
	record header describes the first value in the record, etc.). The
	second and subsequent integer values in a record header are interpreted
	according to the following table:
      [Table]
        [Tr]<th>Header Value <th>Data type and size
        [Tr]<td>0 
            <td>An SQL NULL value (type SQLITE_NULL). This value
                consumes zero bytes of space in the record's data area.
        [Tr]<td>1
            <td>An SQL integer value (type SQLITE_INTEGER), stored as a
                big-endian 1-byte signed integer.
        [Tr]<td>2
            <td>An SQL integer value (type SQLITE_INTEGER), stored as a
                big-endian 2-byte signed integer.
        [Tr]<td>3
            <td>An SQL integer value (type SQLITE_INTEGER), stored as a
                big-endian 3-byte signed integer.
        [Tr]<td>4
            <td>An SQL integer value (type SQLITE_INTEGER), stored as a
                big-endian 4-byte signed integer.
        [Tr]<td>5
            <td>An SQL integer value (type SQLITE_INTEGER), stored as a
                big-endian 6-byte signed integer.
        [Tr]<td>6
            <td>An SQL integer value (type SQLITE_INTEGER), stored as an
                big-endian 8-byte signed integer.
        [Tr]<td>7
            <td>An SQL real value (type SQLITE_FLOAT), stored as an
                8-byte IEEE floating point value.
        [Tr]<td>8
            <td>The literal SQL integer 0 (type SQLITE_INTEGER). The value 
                consumes zero bytes of space in the record's data area.
                Values of this type are only present in databases with
                a schema file format (the 32-bit integer at byte offset 44
                of the database header) value of 4 or greater.

        [Tr]<td>9
            <td>The literal SQL integer 1 (type SQLITE_INTEGER). The value
                consumes zero bytes of space in the record's data area.
                Values of this type are only present in databases with
                a schema file format (the 32-bit integer at byte offset 44
                of the database header) value of 4 or greater.

        [Tr]<td style="white-space:nowrap"><i>bytes</i> * 2 + 12
	    <td>Even values greater than or equal to 12 are used to signify a 
	        blob of data (type SQLITE_BLOB) (<i>n</i>-12)/2 bytes in length, 
	        where <i>n</i> is the integer value stored in the record header.
                
        [Tr]<td style="white-space:nowrap"><i>bytes</i> * 2 + 13
            <td>Odd values greater than 12 are used to signify a string
                (type SQLITE_TEXT) (<i>n</i>-13)/2 bytes in length, where
                <i>n</i> is the integer value stored in the record header.
      </table>
      <p>
        Immediately following the record header is the data for each
        of the record's values. A record containing <i>N</i> values is
        depicted in figure <cite>figure_recordformat</cite>.

        [Figure recordformat.gif figure_recordformat "Database Record Format"]
      
      <p>
        For each SQL value in the record, there is a blob of data stored
        in the records data area. If the corresponding integer type value
        in the record header is 0 (NULL), 8 (integer value 0) or 9 (integer
        value 1), then the blob of data is zero bytes in length. Otherwise,
        the length of the data field is as described in the table above.
      <p>
        The data field associated with a string value contains the string
        encoded using the database encoding, as defined in the database
        header (see section <cite>database_header</cite>). No 
        nul-terminator character is stored in the database.

          [fancyformat_import_requirement H30560]
          [fancyformat_import_requirement H30570]
          [fancyformat_import_requirement H30580]
          [fancyformat_import_requirement H30590]
          [fancyformat_import_requirement H30600]
          [fancyformat_import_requirement H30610]
          [fancyformat_import_requirement H30620]
          [fancyformat_import_requirement H30630]
          [fancyformat_import_requirement H30640]
          [fancyformat_import_requirement H30650]
          [fancyformat_import_requirement H30660]
          [fancyformat_import_requirement H30670]
          [fancyformat_import_requirement H30680]
          [fancyformat_import_requirement H30690]
          [fancyformat_import_requirement H30700]

      <p>
        The following database file properties define restrictions on the 
        integer values that may be stored within a 
        <i>database record header</i>.

          [fancyformat_import_requirement H30710]
          [fancyformat_import_requirement H30720]

    [h3 "Index B-Trees" index_btrees]
      <p>
        As specified in section <cite>fileformat_overview</cite>, index 
        B-Tree structures store a unique set of the database records described
        in the previous section. While in some cases, when there are very
        few entries in the B-Tree, the entire structure may fit on a single
        database page, usually the database records must be spread across
        two or more pages. In this case, the pages are organized into a
        tree structure with a single "root" page at the head of the tree.
      <p>
        Within the tree structure, each page is either an internal tree 
        node containing an ordered list of N references to child nodes 
        (page numbers) and N-1 database records, or a leaf node containing 
        M database records. The value of N may be different for each page, but
        is always two or greater. Similarly, each leaf page may have a
        different non-zero positive value for M. The tree is always of
        uniform height, meaning the number of intermediate levels between 
        each leaf node page and the root page is the same.
      <p>
        Within both internal and leaf node pages, the records are stored in
        sorted order. The comparison function used to determine the sort order
        is described in section <cite>index_btree_compare_func</cite>.
      <p>
        Records are distributed throughout the tree such that for each 
        internal node, all records stored in the sub-tree headed by 
        the first child node ( C(0) ) are considered less than 
        the first record stored on the internal node ( R(0) ) by the 
        comparison function described in section
        <cite>index_btree_compare_func</cite>. Similarly all records stored 
        in the sub-tree headed by C(n) are considered greater than R(n-1) but
        less than R(n) for values of n between 1 and N-2, inclusive. All
        records in the sub-tree headed by C(N-1) are greater than the 
        largest record stored on the internal node.

        [Figure indextree.gif figure_indextree "Index B-Tree Tree Structure"]

      <p>
        Figure <cite>figure_indextree</cite> depicts one possible record
        distribution for an index B-Tree containing records R1 to R26, assuming
        that for all values of N, <i>R(N+1)&gt;R(N)</i>. In total the B-Tree
        structure uses 11 database file pages. Internal tree nodes contain
        database records and references to child node pages. Leaf nodes contain
        database records only.

          [fancyformat_import_requirement H30730]
          [fancyformat_import_requirement H30740]
          [fancyformat_import_requirement H30750]
          [fancyformat_import_requirement H30760]

      <p>
        The precise way in which index B-Tree pages and cells are formatted is
        described in subsequent sections.


        [h4 "Index B-Tree Content"]
          <p>
            The database file contains one index B-Tree for each database index
            in the logical database, including those created by UNIQUE or
            PRIMARY KEY clauses in table declarations. Each record stored in
            an index B-Tree contains the same number of fields, the number of
            indexed columns in the database index declaration plus one. 
          <p>
            An index B-Tree contains an entry for each row in its associated
            database table. The fields of the record used as the index B-Tree
            key are copies of each of the indexed columns of the associated 
            database row, in order, followed by the rowid value of the same 
            row. See figure <cite>figure_examplepop</cite> for an example.

          [fancyformat_import_requirement H30770]
          [fancyformat_import_requirement H30780]
          [fancyformat_import_requirement H30790]
          [fancyformat_import_requirement H30800]
 
      [h4 "Record Sort Order" "index_btree_compare_func"]
        <p>
          This section defines the comparison function used when database
          records are used as B-Tree keys for index B-Trees. The comparison
          function is only defined when both database records contain the same
          number of fields.
        <p>
          When comparing two database records, the first field of one
          record is compared to the first field of the other. If they
          are not equal, the next pair of fields are compared, and so
          on. If all the fields in the database records are equal, then
          the two records are considered equal. Otherwise, the result
          of the comparison is determined by the first pair of unequal 
          fields.
        <p>
          Two database record fields (SQL values) are compared using the
          following rules:
        <ol>
          <li>If both values are NULL, then they are considered equal.
          <li>If one value is a NULL and the other is not, it is considered
              the lesser of the two.
          <li>If both values are either real or integer values, then the
              comparison is done numerically.
          <li>If one value is a real or integer value, and the other is
              a text or blob value, then the numeric value is considered 
              lesser.
          <li>If both values are text, then the collation function is used
              to compare them. The collation function is a property of the
              index column in which the values are found. <span class=todo>
              Link to document with CREATE INDEX syntax.</span>
          <li>If one value is text and the other a blob, the text value
              is considered lesser.
          <li>If both values are blobs, memcmp() is used to determine the 
              results of the comparison function. If one blob is a prefix
              of the other, the shorter blob is considered lesser.
        </ol>
        <p>
          Each column of a database index may be declared as "descending".
          <span class=todo>Link to document with CREATE INDEX syntax.</span>
          In SQLite database files with a schema layer file-format equal
          to 4, this modifies the order in which the records are stored in
          the corresponding index B-Tree structure. For each index column
          declared as descending, the results of the above comparison 
          procedure are inverted.
        <p>
          The columns of database indexes created by UNIQUE or PRIMARY
          KEY clauses are never treated as descending.

        <p class=todo>
          Need requirements style statements for this information. Easier
          to do once collation sequences have been defined somewhere.


      [h4 "Index B-Tree Page Format" index_btree_page_format]
        <p>
          Each index B-Tree page is divided into four sections that occur
          in order on the page:
        <ul>
          <li> The 8-byte (leaf node pages) or 12-byte (internal tree
               node pages) page-header.
          <li> The cell offset array. This is a series of N big-endian 2-byte
               integer values, where N is the number of records stored on 
               the page.
          <li> A block of unused space. This may be 0 bytes in size.
          <li> The cell content area consumes the remaining space on the page.
        </ul>
        [Figure indexpage.gif figure_indexpage "Index B-Tree Page Data"]
        <p>
          The 8-byte (leaf node pages) or 12-byte (internal tree node pages)
          page header that begins each index B-Tree page is made up
          of a series of 
          1, 2 and 4 byte unsigned integer values as shown in the following
          table. All values are stored in big-endian byte order.

      [Table]
        [Tr]<th>Byte Range <th>Byte Size <th width=100%>Description
        [Tr]<td>0     <td>1<td>B-Tree page flags. For an index B-Tree internal 
                               tree node page, this is set to 0x02. For a
                               leaf node page, 0x0A.
        [Tr]<td>1..2  <td>2<td>Byte offset of first block of free space on 
                               this page. If there are no free blocks on this
                               page, this field is set to 0.
        [Tr]<td>3..4  <td>2<td>Number of cells (entries) on this page.
        [Tr]<td>5..6  <td>2<td>Byte offset of the first byte of the cell
                               content area (see figure 
                               <cite>figure_indexpage</cite>), relative to the 
                               start of the page.  If this value is zero, then
                               it should be interpreted as 65536.
        [Tr]<td>7     <td>1<td>Number of fragmented free bytes on page.
        [Tr]<td>8..11 <td>4<td>Page number of rightmost child-page (the
                               child-page that heads the sub-tree in which all
                               records are larger than all records stored on
                               this page). This field is not present for leaf
                               node pages.
      </table>
      <p>
        The cell content area, which occurs last on the page, contains one
        B-Tree cell for each record stored on the B-Tree page. On a leaf node
        page, each cell is responsible for storing a database record only. On
        an internal tree node page, each cell contains a database record and
        the corresponding child page number ((R(0) and C(0)) are stored 
        together, for example - the cell record is considered greater than
        all records stored in the sub-tree headed by the child page). The
        final child page number is stored as part of the page header.
      <p>
        The B-Tree cells may be distributed throughout the cell content area
        and may be interspersed with blocks of unused space. They are not
        sorted within the cell content area in any particular order. The
        serialized format of a B-Tree cell is described in detail in 
        section <cite>index_btree_cell_format</cite>.
      <p>
        The byte offset of each cell in the cell content area, relative
        to the start of the page, is stored in the cell offset array. The
        offsets are in sorted order according to the database records stored
        in the corresponding cells. The first offset in the array is the 
        offset of the cell containing the smallest record on the page,
        according to the comparison function defined in section 
        <cite>index_btree_compare_func</cite>.
      <p>
        As well as the block of unused space between the cell offset array and
        the cell content area, which may be any size, there may be small blocks
        of free space interspersed with the B-Tree cells within the cell
        content area. These are classified into two classes, depending on their
        size:
      <ul>
        <li>Blocks of free-space consisting of 3 bytes or less are called
            <b>fragments</b>. The total number of bytes consumed by all
            fragments on a page is stored in the 1 byte unsigned integer at
            byte offset 7 of the page header. The total number of fragmented
            bytes on a single page is never greater than 255.
        <li>Blocks of free-space consisting of more than 3 bytes of contiguous
            space are called <b>free blocks</b>. All free blocks on a single
            page are linked together into a singly linked list. The byte
            offset (relative to the start of the page) of the first block in 
            the list is stored in the 2 byte unsigned integer stored at byte
            offset 1 of the page header. The first two bytes of each free
            block contain the byte offset (again relative to the start of
            the page) of the next block in the list stored as a big-endian
            unsigned integer. The first two bytes of the final block in the 
            list are set to zero. The third and fourth bytes of each free
            block contain the total size of the free block in bytes, stored
            as a 2 byte big-endian unsigned integer.
      </ul>

      <p class=todo>
	The list of free blocks is kept in order, sorted by offset. Right?
	Later: True statement. SQLite function sqlite3BtreeInitPage() returns
        SQLITE_CORRUPT if they are not.

          [fancyformat_import_requirement H30810]
          [fancyformat_import_requirement H30820]

      <p>
        The following requirements describe the <i>B-Tree page header</i>
        present at the start of both index and table B-Tree pages.

          [fancyformat_import_requirement H30830]
          [fancyformat_import_requirement H30840]
          [fancyformat_import_requirement H30850]
          [fancyformat_import_requirement H30860]

      <p>
        This requirement describes the cell content offset array. It applies
        to both B-Tree variants.

          [fancyformat_import_requirement H30870]
          [fancyformat_import_requirement H30880]
          [fancyformat_import_requirement H30890]
          [fancyformat_import_requirement H30900]
          [fancyformat_import_requirement H30910]

      <p>
        The following requirements govern management of free-space within the
        page content area (both table and index B-Tree pages).

          [fancyformat_import_requirement H30920]
          [fancyformat_import_requirement H30930]
          [fancyformat_import_requirement H30940]
          [fancyformat_import_requirement H30950]
          [fancyformat_import_requirement H30960]

      [h4 "Index B-Tree Cell Format" index_btree_cell_format]
        <p> 
          For index B-Tree internal tree node pages, each B-Tree cell begins
          with a child page-number, stored as a 4-byte big-endian unsigned
          integer. This field is omitted for leaf pages, which have no 
          children.
        <p> 
          Following the child page number is the total number of bytes 
          consumed by the cell's record, stored as a variable length integer
          (see section <cite>varint_format</cite>). 
        <p> 
          If the record is small enough, it is stored verbatim in the cell.
          A record is deemed to be small enough to be completely stored in
          the cell if it consists of less than or equal to:
        <pre>
            <i>max-local</i> := (<i>usable-size</i> - 12) * <i>max-embedded-fraction</i> / 255 - 23
</pre>
        <p>
          bytes. In the formula above, <i>usable-size</i> is the page-size
          in bytes less the number of unused bytes left at the end of every
          page (as read from byte offset 20 of the database header), and
          <i>max-embedded-fraction</i> is the value read from byte offset 
          21 of the database header.
        [Figure indexshortrecord.gif figure_indexshortrecord "Small Record Index B-Tree Cell"]
        <p>
          If the cell record is larger than the maximum size identified by
          the formula above, then only the first part of the record is stored
          within the cell. The remainder is stored in an overflow-chain (see
          section <cite>overflow_page_chains</cite> for details). Following 
          the part of the record stored within the cell is the page number 
          of the first page in the overflow chain, stored as a 4 byte 
          big-endian unsigned integer. The size of the part of the record 
          stored within the B-Tree cell (<i>local-size</i> in figure 
          <cite>figure_indexlongrecord</cite>) is calculated according to the 
          following algorithm:
        <pre>
            <i>min-local</i> := (<i>usable-size</i> - 12) * <i>min-embedded-fraction</i> / 255 - 23
            <i>max-local</i> := (<i>usable-size</i> - 12) * <i>max-embedded-fraction</i> / 255 - 23
            <i>local-size</i> := <i>min-local</i> + (<i>record-size</i> - <i>min-local</i>) % (<i>usable-size</i> - 4)
            if( <i>local-size</i> &gt; <i>max-local</i> )
                <i>local-size</i> := <i>min-local</i>
</pre>
        <p>
          In the formula above, <i>usable-size</i> is the page-size
          in bytes less the number of unused bytes left at the end of every
          page (as read from byte offset 20 of the database header), and
          <i>max-embedded-fraction</i> and <i>min-embedded-fraction</i> are
          the values read from byte offsets 21 and 22 of the database header,
          respectively.
        [Figure indexlongrecord.gif figure_indexlongrecord "Large Record Index B-Tree Cell"]

          [fancyformat_import_requirement H30970]
          [fancyformat_import_requirement H30980]
          [fancyformat_import_requirement H30990]
          [fancyformat_import_requirement H31000]
          [fancyformat_import_requirement H31010]

      <p>
        Requirements H31010 and H30990 are similar to the algorithms 
        presented in the text above. However instead of 
        <i>min-embedded-fraction</i> and <i>max-embedded-fraction</i> the
        requirements use the constant values 32 and 64, as well-formed 
        database files are required by H30080 and H30070 to store these 
        values in the relevant database database header fields.

    [h3 "Table B-Trees" table_btrees]
      <p>
        As noted in section <cite>fileformat_overview</cite>, table B-Trees
        store a set of unique 64-bit signed integer keys. Associated with
        each key is a database record. As with index B-Trees, the database
        file pages that make up a table B-Tree are organized into a tree
        structure with a single "root" page at the head of the tree.
      <p>
        Unlike index B-Tree structures, where entries are stored on both
        internal and leaf nodes, all entries in a table B-Tree are stored
        in the leaf nodes. Within each leaf node, keys are stored in sorted
        order.
      <p>
        Each internal tree node contains an ordered list of N references 
        to child pages, where N is some number greater than one. In a 
        similar manner to the way in which an index B-Tree page would 
        contain N-1 records, each internal table B-Tree node page also 
        contains a list of N-1 64-bit signed integer values in sorted order. 
        The keys are distributed throughout the tree such that for all internal
        tree nodes, integer I(n) is equal to the largest key value stored in
        the sub-tree headed by child page C(n) for values of n between 0 and
        N-2, inclusive. Additionally, all keys stored in the sub-tree headed
        by child page C(n+1) have values larger than that of I(n), for values
        of n in the same range.

        [Figure tabletree.gif figure_tabletree "Table B-Tree Tree Structure"]

      <p>
        Figure <cite>figure_tabletree</cite> depicts a table B-Tree containing
        a contiguous set of 14 integer keys starting with 1. Each key <i>n</i>
        has an associated database record R<i>n</i>. All the keys and their
        associated records are stored in the leaf pages. The internal node
        pages contain no database data, their only purpose is to provide
        a way to navigate the tree structure.

          [fancyformat_import_requirement H31020]
          [fancyformat_import_requirement H31030]
          [fancyformat_import_requirement H31040]
          [fancyformat_import_requirement H31050]

      <p class=todo>
        The special case for root page 1. Root page 1 may contain zero cells,
        just a right-child pointer to the only other b-tree page in the tree.

      <p>
        The precise way in which table B-Tree pages and cells are formatted is
        described in subsequent sections.

      [h4 "Table B-Tree Content" table_btree_content]
        <p>
          The database file contains one table B-Tree for each database table
          in the logical database. Although some data may be duplicated in
          index B-Tree structures, the table B-Tree is the primary location
          of table data.
        <p>
          The table B-Tree contains exactly one entry for each row in the
          database table. The integer key value used for the B-Tree entry is
          the value of the "rowid" field of the corresponding logical row 
          in the database table. The database row fields are stored in the
          record associated with the table B-Tree entry, in the same order
          as they appear in the logical database table. The first field in
          the record (see section <cite>record_format</cite>) contains the
          value of the leftmost field in the database row, and so on.
        <p>
          If a database table column is declared as an INTEGER PRIMARY KEY,
          then it is an alias for the rowid field, which is stored as the 
          table B-Tree key value. Instead of duplicating the integer value
          in the associated record, the record field associated with the
          INTEGER PRIMARY KEY column is always set to an SQL NULL.
        <p>
          Finally, if the schema layer file-format is greater than or equal 
          to 2, some of the records stored in table B-Trees may contain
          less fields than the associated logical database table does columns.
          If the schema layer file-format is exactly 2, then the logical
          database table column values associated with the "missing" fields 
          are SQL NULL. If the schema layer file-format is greater than
          2, then the values associated with the "missing" fields are 
          determined by the default value of the associated database table 
          columns.
          <span class=todo>Reference to CREATE TABLE syntax. How are default
          values determined?</span>

          [fancyformat_import_requirement H31060]
          [fancyformat_import_requirement H31070]
          [fancyformat_import_requirement H31080]
          [fancyformat_import_requirement H31090]

        <p>The following database properties discuss table B-Tree records 
           with implicit (default) values.

          [fancyformat_import_requirement H31100]
          [fancyformat_import_requirement H31110]
          [fancyformat_import_requirement H31120]

      [h4 "Table B-Tree Page Format"]
        <p>
          Table B-Tree structures use the same page format as index B-Tree 
          structures, described in section <cite>index_btree_page_format</cite>,
          with the following differences:
        <ul>
          <li>The first byte of the page-header, the "flags" field, is set to 
              0x05 for internal tree node pages, and 0x0D for leaf pages.
          <li>The content and format of the B-Tree cells is different. See
              section <cite>table_btree_cell_format</cite> for details.
          <li>The format of page 1 is the same as any other table B-Tree,
              except that 100 bytes less than usual is available for content.
              The first 100 bytes of page 1 is consumed by the database
              header.
        </ul>

          [fancyformat_import_requirement H31130]
          [fancyformat_import_requirement H31140]
        
      <p>
        Most of the requirements specified in section 
        <cite>index_btree_page_format</cite> also apply to table B-Tree 
        pages. The wording of the requirements make it clear when this is
        the case, either by referring to generic "B-Tree pages" or by
        explicitly stating that the statement applies to both "table and
        index B-Tree pages".

      [h4 "Table B-Tree Cell Format" table_btree_cell_format]
        <p>
          Cells stored on internal table B-Tree nodes consist of exactly two 
          fields. The associated child page number, stored as a 4-byte
          big-endian unsigned integer, followed by the 64-bit signed integer
          value, stored as a variable length integer (section 
          <cite>varint_format</cite>). This is depicted graphically in figure
          <cite>figure_tablenodecell</cite>.
        [Figure tablenodecell.gif figure_tablenodecell "Table B-Tree Internal Node Cell"]
        <p>
          Cells of table B-Tree leaf pages are required to store a 64-bit
          signed integer key and its associated database record. The first
          two fields of all table B-Tree leaf page cells are the size of
          the database record, stored as a <i>variable length integer</i>
          (see section <cite>varint_format</cite>), followed by the key
          value, also stored as a <i>variable length integer</i>. For 
          sufficiently small records, the entire record is stored in the 
          B-Tree cell following the record-size field. In this case, 
          sufficiently small is defined as less than or equal to:
        <pre>
          max-local := <i>usable-size</i> - 35
</pre>
        <p>
          bytes. Where <i>usable-size</i> is defined as the page-size
          in bytes less the number of unused bytes left at the end of every
          page (as read from byte offset 20 of the database header). 
          This scenario, where the entire record is
          stored within the B-Tree cell, is depicted in figure
          <cite>figure_tableshortrecord</cite>.
        [Figure tableshortrecord.gif figure_tableshortrecord "Table B-Tree Small Record Leaf Node Cell"]

        <p>
          If the record is too large to be stored entirely within the B-Tree
          cell, then the first part of it is stored within the cell and the
          remainder in an overflow chain (see section
          <cite>overflow_page_chains</cite>). The size of the part of the 
          record stored within the B-Tree cell (<i>local-size</i> in figure
          <cite>figure_tablelongrecord</cite>) is calculated according to
          the following algorithm (a similar procedure to that used to
          calculate the portion of an index B-Tree key to store within the cell
          when an overflow chain is required):
        <pre>
            <i>min-local</i> := (<i>usable-size</i> - 12) * <i>min-embedded-fraction</i> / 255 - 23
            <i>max-local</i> := <i>usable-size</i> - 35
            <i>local-size</i> := <i>min-local</i> + (<i>record-size</i> - <i>min-local</i>) % (<i>usable-size</i> - 4)
            if( <i>local-size</i> &gt; <i>max-local</i> )
                <i>local-size</i> := <i>min-local</i>
</pre>
        <p>
          In this case, <i>min-embedded-fraction</i> is the value read from
          byte offset 22 of the database header. The layout of the cell in this
          case, when an overflow-chain is required, is shown in figure
          <cite>figure_tablelongrecord</cite>.

        [Figure tablelongrecord.gif figure_tablelongrecord "Table B-Tree Large Record Leaf Node Cell"]

        <p>
          If the leaf page is page 1, then the value of <i>usable-size</i> is
          as it would be for any other B-Tree page, even though the actual
          usable size is 100 bytes less than this for page 1 (because the
          first 100 bytes of the page is consumed by the database file
          header).

        <p>
          The following requirements describe the format of table B-Tree 
          cells, and the distribution thereof between B-Tree and overflow
          pages.

          [fancyformat_import_requirement H31150]
          [fancyformat_import_requirement H31160]
          [fancyformat_import_requirement H31170]
          [fancyformat_import_requirement H31180]
          [fancyformat_import_requirement H31190]
        
        <p>
          Requirement H31190 is very similar to the algorithm presented in
          the text above. Instead of <i>min-embedded-fraction</i>, it uses
          the constant value 32, as well-formed database files are required
          by H30090 to store this value in the relevant database file 
          header field.

    [h3 "Overflow Page Chains" "overflow_page_chains"]
      <p>
        Sometimes, a database record stored in either an index or table 
        B-Trees is too large to fit entirely within a B-Tree cell. In this
        case part of the record is stored within the B-Tree cell and the
        remainder stored on one or more overflow pages. The overflow pages
        are chained together using a singly linked list. The first 4 bytes
        of each overflow page is a big-endian unsigned integer value 
        containing the page number of the next page in the list. The 
        remaining usable database page space is available for record data.

        [Figure overflowpage.gif figure_overflowpage "Overflow Page Format"]

      <p>
        The scenarios in which overflow pages are required and the number
        of bytes stored within the B-Tree cell in each are described for
        index and table B-Trees in sections 
        <cite>index_btree_cell_format</cite> and
        <cite>table_btree_cell_format</cite> respectively. In each case 
        the B-Tree cell also stores the page number of the first page in
        a linked list of overflow pages.
      <p>
        The amount of space available for record data on each overflow 
        page is:
      <pre>
        <i>available-space</i> := <i>usable-size</i> - 4
</pre>
      <p>
        Where <i>usable-size</i> is defined as the page-size in bytes less the
        number of unused bytes left at the end of every page (as read from 
        byte offset 20 of the database header).
      <p>
        Each overflow page except for the last one in the linked list 
        contains <i>available-space</i> bytes of record data. The last
        page in the list contains the remaining data, starting at byte
        offset 4. The value of the "next page" field on the last page
        in an overflow chain is undefined.

          [fancyformat_import_requirement H31200]
          [fancyformat_import_requirement H31210]
          [fancyformat_import_requirement H31220]
          [fancyformat_import_requirement H31230]

  [h2 "The Free Page List" free_page_list]
    <p>
      Sometimes, after deleting data from the database, SQLite removes pages
      from B-Tree structures. If these pages are not immediately required
      for some other purpose, they are placed on the free page list. The
      free page list contains those pages that are not currently being
      used to store any valid data.
    <p>
      Each page in the free-list is classified as a free-list trunk page 
      or a free-list leaf page. All trunk pages are linked together into
      a singly linked list (in the same way as pages in an overflow chain
      are - see section <cite>overflow_page_chains</cite>). The first four
      bytes of each trunk page contain the page number of the next trunk
      page in the list, formatted as an unsigned big-endian integer. If
      the trunk page is the last page in the linked list, the first four
      bytes are set to zero.
    <p>
      Bytes 4 to 7 of each free-list trunk page contain the number of
      references to free-list leaf pages (page numbers) stored on the
      free-list trunk page. Each leaf page on the free-list is referenced
      by exactly one trunk page.
    <p>
      The remaining space on a free-list trunk page is used to store the
      page numbers of free-list leaf pages as 4 byte big-endian integers. 
      Each free-list trunk page contains up to:
    <pre>
        <i>max-leaf-pointers</i> := (<i>usable-size</i> - 8) / 4
</pre>
    <p>
      pointers, where <i>usable-size</i> is defined as the page-size in bytes
      less the number of unused bytes left at the end of every page (as read
      from byte offset 20 of the database header).

      [Figure freelistpage.gif figure_freelistpage "Free List Trunk Page Format"]
    <p>
      All trunk pages in the free-list except for the first contain the 
      maximum possible number of references to leaf pages. <span class=todo>Is this actually true in an auto-vacuum capable database? Later: No, not even nearly true. It is a false statement.</span> The page number
      of the first page in the linked list of free-list trunk pages is 
      stored as a 4-byte big-endian unsigned integer at offset 32 of the
      database header (section <cite>database_header</cite>).

          [fancyformat_import_requirement H31240]
          [fancyformat_import_requirement H31250]
          [fancyformat_import_requirement H31260]
          [fancyformat_import_requirement H31270]
          [fancyformat_import_requirement H31280]
          [fancyformat_import_requirement H31290]
          [fancyformat_import_requirement H31300]

    <p>The following statements govern the two 4-byte big-endian integers
       associated with the <i>free page list</i> structure in the database
       header.

          [fancyformat_import_requirement H31310]
          [fancyformat_import_requirement H31320]
  

  [h2 "Pointer Map Pages" pointer_map_pages]
    <p>
      Pointer map pages are only present in auto-vacuum capable databases.
      A database is an auto-vacuum capable database if the value stored 
      at byte offset 52 of the file-header is non-zero.
    <p>
      If they are present, the pointer-map pages together form a lookup 
      table that can be used to determine the type and "parent page" of
      any page in the database, given its page number. The lookup table
      classifies pages into the following categories:
    [Table]
      [Tr]<th>Page Type <th>Byte Value <th>Description
      [Tr]<td style="white-space:nowrap">B-Tree Root Page<td>0x01
          <td>The page is the root page of a table or index B-Tree structure.
              There is no parent page number in this case, the value stored
              in the pointer map lookup table is always zero.
      [Tr]<td>Free Page<td>0x02
          <td>The page is part of the free page list (section
              <cite>free_page_list</cite>). There is no parent page in this
              case, zero is stored in the lookup table instead of a parent
              page number.
      [Tr]<td>Overflow type 1<td>0x03
          <td>The page is the first page in an overflow chain. The parent
              page is the B-Tree page containing the B-Tree cell to which
              the overflow chain belongs.
      [Tr]<td style="white-space:nowrap">Overflow type 2<td>0x04
          <td>The page is part of an overflow chain, but is not the first
              page in that chain. The parent page is the previous page in
              the overflow chain linked-list.
      [Tr]<td>B-Tree Page<td>0x05
          <td>The page is part of a table or index B-Tree structure, and is 
              not an overflow page or root page. The parent page is the page
              containing the parent tree node in the B-Tree structure.
    </table>
    <p>
      Pointer map pages themselves do not appear in the pointer-map lookup
      table. Page 1 does not appear in the pointer-map lookup table either.

    [Figure pointermapentry.gif figure_pointermapentry "Pointer Map Entry Format"]
    <p>
      Each pointer-map lookup table entry consumes 5 bytes of space. 
      The first byte of each entry indicates the page type, according to the 
      key described in the table above. The following 4 bytes store the 
      parent page number as a big-endian unsigned integer. This format is
      depicted in figure <cite>figure_pointermapentry</cite>. Each 
      pointer-map page may therefore contain:
    <pre>
        <i>num-entries</i> := <i>usable-size</i> / 5
</pre>
    <p>
      entries, where <i>usable-size</i> is defined as the page-size in bytes
      less the number of unused bytes left at the end of every page (as read
      from byte offset 20 of the database header).
    <p>
      Assuming the database is auto-vacuum capable, page 2 is always a 
      pointer map page. It contains the pointer map lookup table entries for
      pages 3 through (2 + <i>num-entries</i>), inclusive. The first 5 bytes
      of page 2 contain the pointer map lookup table entry for page 3. Bytes
      5 through 9, inclusive, contain the pointer map lookup table entry
      for page 4, and so on.
    <p>
      The next pointer map page in the database is page number (3 +
      <i>num-entries</i>), which contains the pointer map entries for pages
      (4 + <i>num-entries</i>) through (3 + 2 * <i>num-entries</i>) 
      inclusive. In general, for any value of <i>n</i> greater than zero, 
      the following page is a pointer-map page that contains lookup 
      table entries for the <i>num-entries</i> pages that follow it in the
      database file:
    <pre>
        <i>pointer-map-page-number</i> := 2 + <i>n</i> * <i>num-entries</i>
</pre>


          [fancyformat_import_requirement H31330]
          [fancyformat_import_requirement H31340]
          [fancyformat_import_requirement H31350]
          [fancyformat_import_requirement H31360]
          [fancyformat_import_requirement H31370]

    <p>
      The following requirements govern the content of pointer-map entries.

          [fancyformat_import_requirement H31380]
          [fancyformat_import_requirement H31390]
          [fancyformat_import_requirement H31400]
          [fancyformat_import_requirement H31410]
          [fancyformat_import_requirement H31420]

[h1 "Database File-System Representation" file_system_usage]

    <p>
      The previous section, section <cite>database_file_format</cite> 
      describes the format of an SQLite database image. A database
      image is the serialized form of a logical SQLite database. Normally,
      a database image is stored within the file-system in a single
      file, a database file. In this case no other data is stored
      within the database file. The first byte of the <i>database
      file</i> is the first byte of the database image, and the last 
      byte of the database file is the last byte of the <i>database 
      image</i>. For this reason, SQLite is often described as a "single-file 
      database system". However, an SQLite database image is not always 
      stored in a single file within the file-system. It is also possible 
      for it to be distributed between the database file and a journal file. A
      third file, a <i>master-journal file</i> may also be part of the
      file-system representation. Although a <i>master-journal file</i> never
      contains any part of the <i>database image</i>, it can contain meta-data
      that helps determine which parts of the database image are stored within
      the database file, and which parts are stored within the journal file.

    <p>
      In other words, the file-system representation of an SQLite database
      consists of the following:

    <ul>
      <li> <p>A main <b>database file</b>. The database file is
           always present. It may be zero bytes in size, but it is always
           present.
           
      <li> <p>Optionally, a <b>journal file</b>. If present, the <i>journal
           file</i> is stored in the same file-system directory as the
           database file. The name of the journal file is the
           same as that of the database file with the string "-journal"
           appended to it.

      <li> <p>Optionally, a <b>master-journal file</b> may also be part of the
           file-system representation of a database image. A master-journal
           file may only be part of the representation if the journal file 
           is present. A <i>master-journal file</i> may be located anywhere
           within the file-system and may take any name. If present, the
           <i>master-journal</i> is identified by the <i>master-journal
           pointer</i> stored in the journal file (see section
           <cite>master_journal_ptr</cite> for details).
    </ul>

    <p>
      Usually, a database image is stored entirely within the database
      file. Other configurations, where the database image data
      is distributed between the database file and its journal
      file, are used as interim states when modifying the contents of
      the database image to commit a database transaction. In practice,
      a database reader only encounters such a configuration if a previous
      attempt to modify the database image on disk was interrupted by an
      application, OS or power failure. The most practical approach (and
      that taken by SQLite) is to extract the subset of the database image
      currently stored within the journal file and write it into the database
      file, thus restoring the system to a state where the database file
      contains the entire database image. Other SQLite documentation, and 
      the comments in the SQLite source code, identify this process as <i>hot 
      journal rollback</i>. Instead of focusing on the <i>hot journal
      rollback</i> process, this document describes how journal and
      master-journal files must be interpreted in order to extract the 
      current database image from the file-system representation in the
      general case.

    <p>
      Sub-section <cite>journal_file_formats</cite> describes the formats 
      used by <i>journal</i> and <i>master-journal</i> files.

    <p>
      Sub-section <cite>reading_from_files</cite> contains a precise 
      description of the various ways a database image may be
      distributed between the database file and journal file, 
      and the rules that must be followed to extract it. In other words, a 
      description of how SQLite or compatible software reads the database 
      image from the file-system.

[h2 "Journal File Formats" journal_file_formats]

  <p>
    The following sub-sections describe the formats used by SQLite journal
    files (section <cite>journal_file_format</cite>) and master journal files
    (section <cite>masterjournal_file_format</cite>).


[h3 "Journal File Details" journal_file_format]

    <p>
      This section describes the format used by an SQLite journal file.

    <p>
      A journal file consists of one or more journal sections, optionally
      followed by a master journal pointer field. The first journal section
      starts at the beginning of the journal file. There is no limit to the
      number of journal sections that may be present in a single journal file.

    <p>
      Each journal section consists of a journal header immediately followed
      by zero or more journal records. The format of journal header and journal
      records are described in sections <cite>journal_header_format</cite> and
      <cite>journal_record_format</cite> respectively. One of the numeric fields 
      stored in a journal header is the sector size field. Each journal section 
      in a journal file must be an integer multiple of the sector size stored
      in the first journal header of the journal file (the value of the sector
      size field in the second and subsequent journal headers is not used). If
      the sum of the sizes of the journal header and journal records in a journal
      section is not an integer multiple of the sector size, then up to 
      (sector-size - 1) bytes of unused space (padding) follow the end of the
      last journal record to make up the required length.

    <p>
      Figure <cite>figure_journal_format</cite> illustrates a journal file that 
      contains <i>N</i> journal sections and a master journal pointer. The first
      journal section in the file is depicted as containing <i>M</i> journal
      records.

    [Figure journal_format.gif figure_journal_format "Journal File Format"]

    <p>
      The following requirements define a well-formed journal section. This concept
      is used in section <cite>reading_from_files</cite>. 

        [fancyformat_import_requirement H32210]
        [fancyformat_import_requirement H32220]
        [fancyformat_import_requirement H32230]
        [fancyformat_import_requirement H32240]

    <p>
      Note that a journal section that is not strictly speaking a well-formed
      journal section often contains important data. For example, many journal 
      files created by SQLite that consist of a single journal section and no
      master journal pointer contain a journal section that is not well-formed
      according to requirement H32240. See section <cite>reading_from_files</cite> 
      for details on when well-formedness is an important property of journal
      sections and when it is not.

    [h4 "Journal Header Format" journal_header_format]

    <p>
      A journal header is sector-size bytes in size, where sector-size is the 
      value stored as a 32-bit big-endian unsigned integer at byte offset 20 of
      the first journal header that occurs in the journal file. The sector-size
      must be an integer power of two greater than or equal to 512. The
      sector-size is chosen by the process that creates the journal file based
      on the considerations described in section <cite>writing_to_files</cite>.
      Only the first 28 bytes of the journal header are used, the remainder may
      contain garbage data. The first 28 bytes of each <i>journal header</i>
      consists of an eight byte block set to a well-known value, followed by
      five big-endian 32-bit unsigned integer fields.
     
    [Figure journal_header.gif figure_journal_header "Journal Header Format"]

    <p>
      Figure <cite>figure_journal_header</cite> graphically depicts the layout
      of a <i>journal header</i>. The individual fields are described in
      the following table. The offsets in the 'byte offset' column of the
      table are relative to the start of the <i>journal header</i>.

    [Table]
      [Tr]<th>Byte offset<th>Size in bytes<th width=100%>Description
      [Tr]<td>0<td>8<td>The <b>journal magic</b> field always contains a
                        well-known 8-byte string value used to identify SQLite
                        journal files. The well-known sequence of byte values
                        is:
                        <pre>0xd9 0xd5 0x05 0xf9 0x20 0xa1 0x63 0xd7</pre>
      [Tr]<td>8<td>4<td>This field, the <b>record count</b>, is set to the
                        number of <i>journal records</i> that follow this
                        <i>journal header</i> in the journal file.
      [Tr]<td>12<td>4<td>The <b>checksum initializer</b> field is set to a 
                         pseudo-random value. It is used as part of the
                         algorithm to calculate the checksum for all <i>journal
                         records</i> that follow this <i>journal header</i>.
      [Tr]<td>16<td>4<td>This field, the <b>database page count</b>, is set
                         to the number of pages that the database file
                         contained before any modifications associated with
                         <i>write transaction</i> are applied.
      [Tr]<td>20<td>4<td>This field, the <b>sector size</b>, is set to the
                         <i>sector size</i> of the device on which the 
                         journal file was created, in bytes. This value
                         is required when reading the journal file to determine
                         the size of each <i>journal header</i>.
      [Tr]<td>24<td>4<td>The <b>page size</b> field contains the database page
                         size used by the corresponding database file
                         when the journal file was created, in bytes.
    </table>

    <p>
      Because a journal header always occurs at the start of a journal 
      section, and because the size of each journal section is always a
      multiple of sector-size bytes, journal headers are always positioned
      in the file such that they start at a sector-size aligned offset.

    <p>
      The following requirements define a "well-formed journal header". This
      concept is used in the following sections. A well-formed journal header
      is defined as a blob of 28 bytes for which the journal magic field is set
      correctly and for which both the page size and sector size fields are set
      to power of two values greater than 512. Because there are no
      restrictions on the values that may be stored in the record count,
      checksum initializer or database page count fields, they do not enter
      into the definition of a well-formed journal header.

      [fancyformat_import_requirement H32090]
      [fancyformat_import_requirement H32180]
      [fancyformat_import_requirement H32190]
      [fancyformat_import_requirement H32200]

  [h4 "Journal Record Format" journal_record_format]

    <p>
    Each <i>journal record</i> contains the data for a single database page,
    a page number identifying the page, and a checksum value used to help
    detect journal file corruption.

    [Figure journal_record.gif figure_journal_record "Journal Record Format"]

    <p>
      A <i>journal record</i>, depicted graphically by figure
      <cite>figure_journal_record</cite>, contains three fields, as described
      in the following table. Byte offsets are relative to the start of the
      <i>journal record</i>.

    [Table]
      [Tr]<th>Byte offset<th>Size in bytes<th width=100%>Description
      [Tr]<td>0<td>4<td>The page number of the database page associated with
                        this <i>journal record</i>, stored as a 4 byte
                        big-endian unsigned integer.
      [Tr]<td>4<td><i>page-size<td>
                        This field contains the original data for the page,
                        exactly as it appeared in the database file before the
                        <i>write transaction</i> began.
      [Tr]<td style="white-space: nowrap">4 + <i>page-size</i><td>4<td>
                        This field contains a checksum value, calculated based
                        on the contents of the journaled database page data
                        (the previous field) and the values stored in the
                        <i>checksum initializer</i> field of the preceding
                        <i>journal header</i>.
    </table>

    <p>
      The checksum value stored in each journal record is calculated based
      on the contents of the page data field of the record and the value
      stored in the checksum initializer field of the journal header that
      occurs immediately before the journal record. The checksum initializer
      field is interpreted as a 32-bit unsigned integer. To this value is
      added the value stored in every 200th byte of the page data field,
      interpreted as an 8-bit unsigned integer, beginning with the byte
      at offset (page-size % 200). The sum is accumulated in a 32-bit 
      unsigned integer. Overflow is handled by wrapping around to zero.

      <div style="padding: 0 1ex; float:right">
      <div style="padding: 0 1ex; border:1px solid black">
      Example Checksum Calculation:
      <pre>
  Sum of values:
       0xFFFFFFE1 + 
       0x00000023 +
       0x00000032 +
       0x0000009E +
       0x00000062 +
       0x0000001F
      -----------
      0x100000155

  Truncated to 32-bits: 
       0x00000155</pre>
      </div></div>

    <p>
      For example, if the page-size is 1024 bytes, then the offsets within
      the page of the bytes added to the checksum initializer value are
      24, 224, 424, 624 and 824 (the first byte of the page is offset 0, the
      last byte is offset 1023). If the values of the bytes at these offsets
      are 0x23, 0x32, 0x9E, 0x62 and 0x1F, and the value of the checksum
      initializer field is 0xFFFFFFE1, then the value stored in the checksum
      field of the journal record is 0x00000155.


    <p>
      The set of <i>journal records</i> that follow a <i>journal header</i>
      in a journal file are packed tightly together. There are no alignment 
      requirements for <i>journal records</i>.

      [fancyformat_import_requirement H32100]
      [fancyformat_import_requirement H32110]
      [fancyformat_import_requirement H32120]


  [h4 "Master Journal Pointer" master_journal_ptr]

    <p>
      If present, a master journal pointer occurs at the end of a journal file.
      There may or may not be unused space between the end of the final journal 
      section and the start of the master journal pointer.

    <p>
      A <i>master journal pointer</i> contains the full path of a 
      <i>master journal-file</i> along with a check-sum and some well-known values 
      that allow the <i>master journal pointer</i> to be unambiguously distinguished 
      from a journal record or journal header.

    [Figure master_journal_ptr.gif figure_master_journal_ptr "Master Journal Pointer Format"]

    <p>
      A <i>master journal pointer</i>, depicted graphically by figure
      <cite>figure_master_journal_ptr</cite>, contains five fields, as 
      described in the following table. Byte offsets are relative to the 
      start of the <i>master journal pointer</i>.

    [Table]
      [Tr]<th>Byte offset<th>Size in bytes<th width=100%>Description
      [Tr]<td>0<td>4<td>This field, the <b>locking page number</b>, is always
               set to the page number of the database <i>locking page</i>
               stored as a 4-byte big-endian integer. The <i>locking page</i>
               is the page that begins at byte offset 2<sup>30</sup> of the
               database file. Even if the database file is large enough to
               contain the <i>locking page</i>, the <i>locking page</i> is
               never used to store any data and so the first four bytes of of a
               valid <i>journal record</i> will never contain this value.

      [Tr]<td>4<td><i>name-length</i><td>
               The <b>master journal name</b> field contains the name of the
               master journal file, encoded as a utf-8 string. There is no
               nul-terminator appended to the string.
      [Tr]<td>4 + <i>name-length</i><td><i>4<td>
               The <b>name-length</b> field contains the length of the 
               previous field in bytes, formatted as a 4-byte big-endian 
               unsigned integer.
      [Tr]<td>8 + <i>name-length</i><td><i>4<td>
               The <b>checksum</b> field contains a checksum value stored as
               a 4-byte big-endian signed integer. The checksum value is
               calculated as the sum of the bytes that make up the <i>
               master journal name</i> field, interpreting each byte as
               an 8-bit signed integer.
      [Tr]<td style="white-space: nowrap">12 + <i>name-length</i><td><i>8<td>
               Finally, the <b>journal magic</b> field always contains a
               well-known 8-byte string value; the same value stored in the
               first 8 bytes of a <i>journal header</i>. The well-known
               sequence of bytes is:
                 <pre>0xd9 0xd5 0x05 0xf9 0x20 0xa1 0x63 0xd7</pre>
    </table>

      [fancyformat_import_requirement H32140]
      [fancyformat_import_requirement H32150]
      [fancyformat_import_requirement H32160]
      [fancyformat_import_requirement H32170]

[h3 "Master-Journal File Details" masterjournal_file_format]

  <p>
    A <i>master-journal file</i> contains the full paths to two or more
    <i>journal files</i>, each encoded using UTF-8 encoding and terminated
    by a single nul character (byte value 0x00). There is no padding 
    between the journal paths, each UTF-8 encoded path begins immediately
    after the nul character that terminates the previous one.

  <p class=todo>
    Note that the contents of a master-journal is not really all that
    important, and is not required at all to read the database image. 
    Used for cleanup only.

[h2 "Reading an SQLite Database" reading_from_files]

  <p>
    As described in section <cite>pages_and_page_types</cite> of this document,
    an SQLite database image is a set of contiguously numbered fixed size 
    pages. The numbering starts at 1, not 0. Page 1 contains the 
    database header and the root page of the <i>schema table</i>, 
    and all other pages within the database image are somehow referenced
    by number, either directly or indirectly, from page 1, either directly 
    or indirectly. In order to be able to read the database image from within
    the file-system, a database reader needs to be able to ascertain:

  <ol>
    <li> The <i>page-size</i> used by the database image,
    <li> The number of pages in the database image, and
    <li> The content of each database page.
  </ol>

  <p>
    Usually, the database image is simply the contents of the database file. 
    In this case, reading the database image is straightforward. The
    page-size used by the database image can be read from the 2-byte
    big-endian integer field stored at byte offset 16 of
    the database file (see section <cite>database_header</cite>). The number of
    pages in the database image can be determined by querying the size of
    the database file in bytes and then dividing by the <i>page-size</i>.
    Reading the contents of a <i>database page</i> is a simple matter of 
    reading a block of <i>page-size</i> bytes from an offset calculated from
    the page-number of the required page:
    <pre>
        <i>offset</i> := (<i>page-number</i> - 1) * page-size
</pre>
    
  <p>
    However, if there is a valid journal file corresponding to the 
    database file present within the file-system then the situation 
    is more complicated. The file-system is considered to contain a valid 
    journal file if each of the following conditions are met:

  <ul>
    <li> A journal file is present in the file system, and
    <li> the journal file either does not end with a well-formed 
         <i>master-journal pointer</i> (see section 
         <cite>master_journal_ptr</cite>) or the <i>master-journal file</i> 
         referred to by the <i>master-journal pointer</i> is present in
         the file-system, and
    <li> the first 28 bytes of the journal file contain a 
         well-formed <i>journal header</i> (see section
         <cite>journal_header_format</cite>).  
  </ul>

  <p>
    If the file system contains a valid journal file, then the
    <i>page-size</i> used by and the number of pages in the <i>database
    image</i> are stored in the first <i>journal header</i> of the 
    journal file. Specifically, the page-size is stored as a 4-byte
    big-endian unsigned integer at byte offset 24 of the journal file, and the
    number of pages in the database image is stored as a 4-byte big-endian
    unsigned integer at byte offset of 16 of the same file.
    
  <p>
    The current data for each page of the database image may be stored 
    within the database file at a file offset based on its page number as 
    it normally is, or the current version of the data may be stored 
    somewhere within the journal file. For each page within the database 
    image, if the journal file contains a valid journal record for the
    corresponding page-number, then the current content of the database 
    image page is the blob of data stored in the page data field of the
    journal record. If the journal file does not contain a valid journal
    record for a page, then the current content of the database image page
    is the blob of data currently stored in the corresponding region of
    the database file.

  <p>
    Not all journal records within a journal file are valid. A journal 
    record is said to be valid if:

  <ul>
    <li> The journal file that contains the journal record is a valid 
         journal file, and
    <li> all journal sections that occur before the journal section 
         containing the journal record are well-formed, and
    <li> the journal section that contains the journal record begins with 
         a well-formed journal header, and
    <li> the journal record itself and all journal records that occur before
         it in the same journal section are well-formed.
  </ul>

  <p>
    Note that it is not necessary for a journal record to be part of a
    well-formed journal section to be considered valid.

  <p>
    Figure <cite>figure_filesystem1</cite> illustrates two distinct ways
    to store a database image within the file system. In this example, the
    database image consists of 4 pages of <i>page-size</i> bytes each. The
    content of each of the 4 pages is designated A, B, C and D, respectively.
    Representation 1 uses only the database file. In this case the entire
    database image is stored in the database file.

  <p>
    In representation 2 of figure <cite>figure_filesystem1</cite>, the current
    database images is stored using both the journal file and the database 
    file. The size and page-size of the database image are both stored in
    the first (in this case only) journal header in the journal file. 
    Following the journal header are two valid journal records. These contain
    the data for pages 3 and 4 of the database image. Because there are no
    valid journal records for pages 1 and 2 of the database image, the content
    for each of these is stored in the database file. Even though the contents
    of the file-system is quite different in representation 2 as in
    representation 1, the stored database image is the same in each case: 4
    pages of page-size bytes each, content A, B, C and D respectively.

    [Figure filesystem1.gif figure_filesystem1 "Two ways to store the same database image"]

  <p class=todo>
    The requirements that follow talk about "well-formed" journal sections,
    records and master-journal-pointers. There should be some kind of reference
    back to the definitions of these things. Either in the requirements
    themselves (refer to other requirements by number) or in the surrounding
    text (point to document sections). Or, better, both.

  <p>
    These requirements describe the way a database reader must determine
    whether or not there is a valid journal file within the 
    file-system.

    [fancyformat_import_requirement H32000]
    [fancyformat_import_requirement H32010]
    [fancyformat_import_requirement H32020]

  <p>
    If there is a valid journal file within the file-system, the 
    following requirements govern how a reader should determine the set
    of valid <i>journal records</i> that it contains.

    [fancyformat_import_requirement H32250]
    [fancyformat_import_requirement H32260]
    [fancyformat_import_requirement H32270]
    [fancyformat_import_requirement H32280]

  <p>
    The following requirements dictate the way in which database
    <i>page-size</i> and the number of pages in the database image
    should be determined by the reader.

    [fancyformat_import_requirement H32030]
    [fancyformat_import_requirement H32040]
    [fancyformat_import_requirement H32050]
    [fancyformat_import_requirement H32060]

  <p>
    The following requirements dictate the way in which the data for each
    page of the database image can be located within the file-system
    by a database reader.

    [fancyformat_import_requirement H32070]
    [fancyformat_import_requirement H32080]

[h1 "SQLite Interoperability Requirements" interoperability_requirements]

  <p>
    This section contains requirements that further constrains the behaviour
    of software that accesses (reads and/or writes) SQLite databases stored
    within the file-system. These requirements need only be implemented by
    systems that access databases while other clients may also be doing so.
    More specifically, they need only be implemented by software operating
    within a system where one or more of the database clients writes to the
    database. If the database file-system representation remains constant
    at all times, or if there is only ever a single database client for each
    database within the system, the requirements in this section can be 
    ignored.

  <p>
    The requirements in this section fall into three categories:

  <ul>
    <li> <p><b>Database Writer Requirements</b>. Section <cite>writing_database</cite>
         contains notes on and requirements that must be observed by software 
	 systems that update an existing SQLite database image within the file-system.

    <li> <p><b>Locking Requirements</b>. Section <cite>locking_protocol</cite>
         contains a description of the file-system locks that must be obtained
         on the database file, and how locks placed by other database clients 
         should be interpreted.

    <li> <p><b>Header Cookie Requirements</b>. An SQLite database image header 
         (see section <cite>database_header</cite>) contains two "cookie" values
         that must sometimes be incremented when the database image stored in
         the file-system is updated. Section 
         <cite>database_header_cookies_protocol</cite> contains requirements
         identifying exactly when the cookie values must be incremented, and
         how they can be used by a database client to determine if cached
         data is valid or not.
  </ul> 

  [h2 "Writing to an SQLite Database File" writing_database]

  <p>
    When writing to an SQLite database, the database representation on disk
    must be modified to reflect the new, modified, database image. Exactly
    how this is done in terms of raw IO operations depends on the 
    characteristics of the file-system in which the database is stored and
    the degree to which the application is required to handle failures within
    the system. A failure may be an application crash, an operating system
    crash, a power failure or other unexpected event that terminates 
    processing. For example, SQLite itself runs in several different modes
    with various levels of guarantees on how failures are handled as follows:

  <ul>
    <li> <b>In-memory journal mode</b> (PRAGMA journal_mode=memory). In this
      mode any failure may cause database file-system corruption, including an
      application crash or unexpected exit.
    <li> <b>Non-synchronous mode</b> (PRAGMA synchronous=off). In this mode 
      an application crash or unexpected exit may not cause database 
      corruption, however an operating system crash or power failure may.
    <li> <b>Synchronous mode</b> (PRAGMA synchronous=full). In this mode 
      neither an application crash, operating system crash or power failure 
      may cause database file-system corruption.
  </ul>

  <p>
    If a process attempts to modify a database so as to replace database
    image A with database image B and a failure occurs while doing so, 
    then following recovery the file-system must contain a database image 
    equivalent to A or B. Otherwise, the database file-system is considered 
    corrupt.
    
  <p>
    Two database images are considered to be equivalent if each of the 
    following are true:

    <ul>
      <li> <p> The two database images have the same page-size.
      <li> <p> The two database images have the same number of pages.
      <li> <p> The content of each page in the first database image that is not
               a free-list leaf page is identical to the corresponding page in
               the second database image.
    </ul>

  <p> 
    The exception for free-list leaf pages (see section
    <cite>free_page_list</cite>) in the third bullet point above is made
    because free-list leaf pages contain no valid data and are never read
    by SQLite database readers. Since the blob of data stored on such a
    page is never read for any purpose, two database images may have a
    different blob stored on a free-list leaf page and still be considered
    equivalent. This concept can sometimes be exploited to more efficiently
    update an SQLite database file-system representation.

    [fancyformat_import_requirement H32290]

  <p>
    The following requirement constrains the way in which a database 
    file-system representation may be updated. In many ways, it is
    equivalent to "do not corrupt the database file-system representation
    under those conditions where the file-system should not be corrupted".
    The definition of "handled failure" depends on the mode that SQLite
    is running in (or on the requirements of the external system accessing
    the database file-system representation).

    [fancyformat_import_requirement H32300]

  <p>
    The following two sections, <cite>rollback_journal_method</cite>
    and <cite>atomic_write_method</cite>, are somewhat advisory in nature.
    They contain descriptions of two different methods used by SQLite to
    modify a database image within a database file-system representation in
    accordance with the above requirements. They are not the only methods
    that can be used. So long as the above requirements (and
    those in sections <cite>locking_protocol</cite> and 
    <cite>database_header_cookies_protocol</cite>) are honoured, any method may
    be used by an SQLite database writer to update the database file-system
    representation. Sections <cite>rollback_journal_method</cite> and 
    <cite>atomic_write_method</cite> do not contain formal requirements. Formal
    requirements governing the way in which SQLite safely updates database
    file-system representations may be found in <span class=todo>Not available yet!</span>.
    An informal description is available in <cite>atomic_commit_page</cite>.

  [h3 "The Rollback-Journal Method" rollback_journal_method]

    <p>
      This section describes the method usually used by SQLite to update a database 
      image within a database file-system representation. This is one way
      to modify a database image in accordance with the requirements in the
      parent and other sections. When overwriting database image A with database 
      image B using this method, assuming that to begin with database image A is 
      entirely contained within the database file and that the page-size of
      database image B is the same as that of database image A, the following 
      steps are taken:

    <ol>
      <li> <p>The start of the journal file is populated with data that is not
           a valid journal header.
      <li> <p>For each page in database image A that is not a free-list leaf 
           page and either does not exist in database image B or exists but
           is populated with different content, a record is written to the
           journal file. The record contains a copy of the original database
           image A page.
      <li> <p>The start of the journal file is populated with a valid journal
           header. The page-count field of the journal header is set to the
           number of pages in database image A. The record-count is set to the 
           number of records written to the journal file in step 2.
      <li> <p>The content of each page of database image B that is either not
           present or populated differently in database image A is copied
           into the database file. If database image B is smaller than database
           image A, the database file is truncated to the size required by
           database image B.
      <li> <p>One of several file-system operations that cause the journal file
           to become invalid is performed. For example:
           <ul>
             <li>The journal file is deleted from the file-system, or
             <li>The journal file is truncated to zero bytes in size, or
             <li>Some or all of the first 8 bytes of the journal file are
                 overwritten so that the journal file no longer begins with 
                 a well-formed journal header (and is therefore not valid).
           </ul>
    </ol>

    <p>
      During steps 1 and 2 of the above procedure, the database file-system 
      representation clearly contains database image A. The database file
      itself has not been modified, and the journal file is not valid (since
      it does not begin with a valid journal file header). Following step 3,
      the database file-system representation still contains database image
      A. The number of pages in the database image and the content of some
      pages now resides in the journal file, but the database image remains
      unchanged. During and following step 4, the current database image is
      still database image A. Although some or all of the pages in the 
      database file may have been overwritten or truncated away, a valid 
      journal records containing the original database image A data exists
      for all such pages that were not free-list leaf pages in database 
      image A. And although the size of the database file may have been
      modified, the size of the current database image, database image A, 
      is stored in the journal header.

    <p>
      Once step 5 of the above procedure is performed, the database file-system
      representation contains database image B. The journal file is no longer
      valid, so the database image consists of the contents of the database
      file, database image B.
  <p>
    Figure <cite>figure_filesystem2</cite> depicts a possible interim state 
    of the database file-system representation used while committing a transaction
    that replaces a four page database image with a three page database image.
    The contents of the initial database image pages are A, B, C and D respectively. 
    The final database image content is A, E and C. The interim state depicted
    is that reached at the end of step 4 in the above procedure. In this state, 
    the file-system contains the initial database image, ABCD. However, if the 
    journal file were to be somehow invalidated, then the file-system would 
    contain the final database image, AEC.

    [Figure filesystem2.gif figure_filesystem2 "Interim file-system state used to atomically overwrite database image ABCD with AEC"]

  <p>
    The procedure described above can be onerous to implement, as it requires
    that the data for all modified pages of database image B be available
    (presumably in main memory) at the same time, when step 4 is performed.
    For transactions that write to a large number of database pages, this 
    may be undesirable. A solution is to create a journal-file containing
    two or more journal headers. If, while modifying a database image within
    main-memory, a client wishes to reduce the amount of data held in memory,
    it may perform steps 3 and 4 of the above procedure in order to write
    modified content out to the file-system. Once the modified pages have been
    written into the database file, the in-memory copies may be discarded.
    The writer process may then continue accumulating changes in memory. When
    it is ready to write these changes out to the file-system, either to free
    up main-memory or because all changes associated with the transaction have
    been prepared, it adds a second (or subsequent) journal header to the 
    journal file, followed by journal records containing the original data
    for pages about to be modified. It may then write the changes accumulated
    in-memory to the database file, as described in step 4 above.

  <p>
    This technique can also be modified to support atomic modification of 
    multiple databases. In this case the first 4 steps of the procedure outlined
    above are followed for each individual database. Following this a
    master-journal file is created somewhere within the file-system and a
    master-journal pointer added to each individual journal file. Since
    a journal-file that contains a master-journal pointer to a master-journal
    file that does not exist is considered invalid (requirement H32000),
    all journal-files may be simultaneously invalidated by deleting the
    master-journal file from the file-system. This delete operation takes the
    place of step 5 of the procedure as outlined above.

  [h3 "The Atomic-Write Method" atomic_write_method]

    <p>
      On some systems, SQLite is able to overwrite a single page of the
      database file as an atomic operation. If, while updating the page,
      a failure occurs, the system guarantees that following recovery, the 
      page will be found to have been correctly and completely updated or 
      not modified at all. When running in such an environment, if SQLite
      is required to update a database image so that only a single page
      is modified, it can do so simply by overwriting the page.

    <p>
      Assuming the database page being updated is not page 1, if requirement 
      H33040 requires that the database header change counter be updated, then 
      the database image modification is no longer confined to a single page.
      In this case it can be split in two: SQLite first atomically updates
      page 1 of the database file to increment the database header change 
      counter, then updates the page that it is actually required to update
      using a second atomic write operation. If a failure occurs some time
      between the two write operations, following recovery the database 
      image may be found to be in its original state except for the value
      of the database header change counter <span class=todo>It would be good
      to have some requirement to say that that is Ok. Some modification to
      the definition of equivalent databases perhaps.</span>

<!--

  <p>
    The following requirements require that the journal header at the start of
    a journal file is set to contain the original database page-size and 
    page-count and written to non-volatile storage before the size of the
    database file is modified. And that once the size of the database file has
    been modified, the journal header does not become unstable and the page-size 
    and page-count values stored therein are not modified until the end of
    the transaction.

       [fancyformat_import_requirement H32320]
       [fancyformat_import_requirement H32330]

  <p>
    Journal before overwrite:

       [fancyformat_import_requirement H32340]
       [fancyformat_import_requirement H32350]

  <p>
    Journal before truncate:

       [fancyformat_import_requirement H32360]
       [fancyformat_import_requirement H32370]
      

  h4 "Multiple Database Transactions" multi_db_transactions

  <p>
    SQLite is required 
    to do make all modifications associated with the transaction such that 
    the database image is modified atomically. If an application, OS or 
    power failure occurs while SQLite is updating the database, upon recovery 
    the contents of the database must reflect either that all modifications 
    associated with the database transaction were successfully applied, or 
    that none of the modifications were applied and the contents of the 
    database are as they were before the failed attempt to modify the database.

  <p>
    Therefore, when modifying the file-system representation of a database
    image so as to commit a transaction that modifies the database image
    from state A to state B, it must be ensured that the file-system at
    all times contains a database image in either state A or state B. And
    that if an OS or power failure occurs before, during or after any IO
    operation, following recovery the file-system must contain a database
    image in either state A or state B.

  <p class=todo>
    Should introduce requirements here - that the file-system be modified
    such that the file is always in state A or state B.

  <p>
    Some operations on a file-system may be considered atomic. For example
    deleting a file, or on some systems writing to a single disk sector.
    However, in general there exists no atomic file-system operation
    that may be used to update an SQLite database file with the effects
    of an arbitrary database transaction, which may remove, modify or
    add multiple database rows, tables or indexes. Therefore, a two stage
    approach to writing an SQLite database (or indeed, modifying the logical
    contents of any on-disk database) is required:

  <ol>
    <li> The file-system representation of the database is manipulated to
         a state where a single atomic operation may be used to transform
         the logical contents of the database from its initial state to
         the required final state.
    <li> The required atomic operation is applied.
  </ol>

  <p class=todo>
    The paragraph below is not quite accurate. Each interim state must
    correspond to either state A or state B. Not necessarily state A.
    And the above is not completely general either, for the same reason.

  <p>
    Step 1 of the above must be accomplished such that all interim states
    of the file-system correspond to the logical contents of the database
    as they were before the procedure began. This way, if an application,
    OS or power failure occurs during step 1, upon recovery the database
    contents remain unchanged. It is not possible for such a failure to
    occur "during" step 2, as step 2 consists of a single atomic operation.

  <p>
    SQLite is also required to support atomic database transactions that 
    involve updates to multiple database files. If an application, OS or
    power failure occurs while committing the transaction, it is required
    that following recovery either the logical contents of all affected
    databases have been completely updated, or that the contents of each
    of them remains unchanged. Whether or not a transaction involves 
    multiple database files, the principle remains the same: the file-system
    must be manipulated into a state whereby a single atomic file-system
    operation can be used to effect all required changes to the logical
    database contents.

  <p>
    The following two sub-sections describe the specific ways in which 
    SQLite achieves this for single and multiple database transactions.

  h4 "Single Database Transactions" single_db_transactions

  <p>
    In order to atomically modify the database image stored in the 
    file-system from database image A to database image B, the file-system must
    first be manipulated to a state where it contains the database image A,
    but can by a single atomic operation be modified to contain database 
    image B. A file-system state that has the following properties satisfies
    this requirement:

  <ol>
    <li> The database file contains database image B.
    <li> There exists a valid journal file.
    <li> The first header of the journal file contains the page-size and 
         number of pages in database image A.
    <li> The journal file contains a valid journal record for each page of
         of database image A that either does not exist in database image B
         (because image B is smaller than image A), or does exist but has
         different content in database image B than it does in database 
         image A.
  </ol>

  <p>
    In this state, the file-system contains database image A. However, if the
    journal file is somehow made invalid, then the file-system will then
    contain database image B. There are several possibilities for IO 
    operations that will cause the journal file to become invalid, for 
    example:

  <ul>
    <li> Deleting the journal file from the file-system, or
    <li> Truncating the journal file to zero bytes in size, or
    <li> Overwriting some or all of the first 8 bytes in the journal file
         so that the journal file no longer contains a well-formed journal
         header.
  </ul>


  h4 "Multiple Database Transactions" multi_db_transactions

  <p class=todo>
    Deleting the master-journal is used as the atomic operation.
-->

  [h2 "SQLite Locking Protocol" locking_protocol]

    <p>
      An SQLite database client may hold at any time one of four different types 
      of locks on a database file-system representation. This document does not
      describe how these locks are to be implemented. Possible implementation
      techniques include mapping the four SQLite locks to operating system file
      locks, using an external software module to manage locks, or by creating
      special "lock files" within the file-system. Regardless of how the locks
      are implemented, it is important that all database clients in a system 
      use the same implementation. The following table summarizes the four 
      types of locks used by SQLite:

    [Table]
      [Tr] <th> Lock type <th> Description <th> Blocks <th> Blocked By
      [Tr] <td> SHARED    <td> 
          It is only possible to obtain a SHARED lock if no other client is
          holding a PENDING or EXCLUSIVE lock. Holding a SHARED lock prevents
          any other client from obtaining an EXCLUSIVE lock.
          <td> EXCLUSIVE <td> PENDING, EXCLUSIVE

      [Tr] <td> RESERVED  <td> 
          A RESERVED lock may only be obtained if no other client holds a 
          RESERVED, PENDING or EXCLUSIVE lock on the database. While a
          client holds a RESERVED lock, other clients may obtain new SHARED
          locks, but may not obtain new RESERVED, PENDING or EXCLUSIVE locks.
          <td> RESERVED, PENDING, EXCLUSIVE <td> RESERVED, PENDING, EXCLUSIVE

      [Tr] <td> PENDING <td>
          It is only possible to obtain a PENDING lock if no other client holds
          a RESERVED, PENDING or EXCLUSIVE lock. While a database client is 
          holding a PENDING lock, no other client may obtain any new lock. 
          <td> All <td> RESERVED, PENDING, EXCLUSIVE

      [Tr] <td> EXCLUSIVE <td>
          An EXCLUSIVE lock may only be obtained if no other client holds any
          lock on the database. While an EXCLUSIVE lock is held, no other 
          client may obtain any kind of lock on the database.
          <td> All <td> All

    </table>
      
    <p>
      The most important types of locks are SHARED and EXCLUSIVE. Before any 
      part of the database file is read, a database client must obtain a SHARED 
      lock or greater.

    [fancyformat_import_requirement H33000]

    <p>
      Before the database file may be written to, a database client must
      be holding an EXCLUSIVE lock. Because holding an EXCLUSIVE lock 
      guarantees that no other client is holding a SHARED lock, it also
      guarantees that no other client may be reading from the database file
      as it is being written.

    [fancyformat_import_requirement H33010]

    <p>
      The two requirements above govern reading from and writing to the
      database file. In order to write to a journal file, a database client
      must obtain at least a RESERVED lock.

    [fancyformat_import_requirement H33020]

    <p>
      The requirement above implies that a database writer may write to the
      journal file at the same time as a reader is reading from the database
      file. This improves concurrency in environments that feature multiple
      clients, as a database writer may perform part of its IO before locking
      the database file-system representation with an EXCLUSIVE lock. In order
      for this to work though, the following must be true:

    <ul>
      <li> <p>Database readers must recognize that when a writer holds a RESERVED 
           or PENDING lock on the database file-system representation the
           writer may be manipulating the journal file and as a result it is
           not safe to read.

      <li> <p>Database writers may only obtain a RESERVED or PENDING lock on the
           database file-system representation when it would be safe for a
           database reader to assume that the contents of the database file
           represents the current database image.
    </ul>

    <p>
      The following requirements formally restate the above bullet points.

    [fancyformat_import_requirement H33030]
    [fancyformat_import_requirement H33060]
    [fancyformat_import_requirement H33080]

  [h2 "SQLite Database Header Cookie Protocol" database_header_cookies_protocol]

    <p>
      While a database reader is holding a SHARED lock on the database
      file-system representation, it may freely cache data in main memory
      since there is no way that another client can modify the database
      image. However, if a client relinquishes all locks on a database
      file-system representation and then re-establishes a SHARED lock
      at some point in the future, any cached data may or may not be
      valid (as the database image may have been modified while the client
      was not holding a lock). The requirements in this section dictate
      the way in which database writers must update two fields of the database
      image header (the "cookies") in order to enable readers to determine
      when cached data can be safely reused and when it must be discarded.

    <p>
      SQLite clients may cache two types of data from a database image in
      main-memory:

    <ul>
      <li> <p>The <b>database schema</b>. In order to access database content,
           the contents of the schema table must be parsed (see section 
           <cite>schema_table</cite>). Since this is a relatively expensive
           process, it is advantageous for clients to cache the parsed 
           representation in memory.

      <li> <p>Database image <b>page content</b>. Clients may also cache raw
           page content in order to reduce the number of file-system read 
           operations required when reading the database image.
    </ul>

    <p>
      Similar mechanisms are used to support cache validation for each class
      of data. If a database writer changes the database schema in any way, it
      is also required to increment the value stored in the database schema
      version field of the database image header (see section 
      <cite>database_header</cite>). This way, when a database reader establishes
      a SHARED lock on a database file-system representation, it may validate
      any cached schema data by checking if the value of the database schema 
      version field has changed since the data was cached. If the value has not
      changed, then the cached schema data may be retained and reused. 
      Otherwise, if the value of the database schema version field is not the
      same as it was when the schema data was last cached, then the reader
      can deduce that some other database client has modified the database
      schema in some way and it must be reparsed.

    <p>
      Each time a database image stored within a database file-system 
      representation is modified, the database writer is required to increment
      the value stored in the change counter field of the database image header
      (see section <cite>database_header</cite>). This allows database readers to
      validate any cache of raw database image page content that may be present
      when a database reader establishes a SHARED (or other) lock on the 
      database file-system representation. If the value stored in the change
      counter field of the database image has not changed since the cached
      data was read, then it may be safely reused. Otherwise, if the change
      counter value has changed, then any cached page content data must be
      deemed untrustworthy and discarded.

    <p>
      If a database image is modified more than once while a writer is holding
      an EXCLUSIVE lock, then each header value need only be updated once, as
      part of the first image modification that modifies the associated class
      of data. Specifically, the change counter field need only be incremented
      as part of the first image modification that takes place, and the 
      database schema version need only be incremented as part of the first
      modification that includes a schema change. 

    [fancyformat_import_requirement H33040]
    [fancyformat_import_requirement H33050]
    [fancyformat_import_requirement H33070]


[h1 References]

  <table id="refs" style="width:auto; margin: 1em 5ex">
  [Ref 1 ref_comer_btree {
     Douglas Comer, <u>Ubiquitous B-Tree</u>, ACM Computing Surveys (CSUR),
     v.11 n.2, pages 121-137, June 1979.
  }]
  [Ref 2 ref_knuth_btree {
     Donald E. Knuth, <u>The Art Of Computer Programming, Volume 3:
     "Sorting And Searching"</u>, pages 473-480. Addison-Wesley
     Publishing Company, Reading, Massachusetts.
  }]
  [Ref 3 atomic_commit_page {
    SQLite Online Documentation,<u>How SQLite Implements Atomic Commit</u>,
    <a href="atomiccommit.html">http://www.sqlite.org/atomiccommit.html</a>.
  }]
  </table>
}
</tcl>