AryaWu/sqlite
0
1/*2** 2004 April 63**4** The author disclaims copyright to this source code. In place of5** a legal notice, here is a blessing:6**7** May you do good and not evil.8** May you find forgiveness for yourself and forgive others.9** May you share freely, never taking more than you give.10**11*************************************************************************12** This file implements an external (disk-based) database using BTrees.13** For a detailed discussion of BTrees, refer to14**15** Donald E. Knuth, THE ART OF COMPUTER PROGRAMMING, Volume 3:16** "Sorting And Searching", pages 473-480. Addison-Wesley17** Publishing Company, Reading, Massachusetts.18**19** The basic idea is that each page of the file contains N database20** entries and N+1 pointers to subpages.21**22** ----------------------------------------------------------------23** | Ptr(0) | Key(0) | Ptr(1) | Key(1) | ... | Key(N-1) | Ptr(N) |24** ----------------------------------------------------------------25**26** All of the keys on the page that Ptr(0) points to have values less27** than Key(0). All of the keys on page Ptr(1) and its subpages have28** values greater than Key(0) and less than Key(1). All of the keys29** on Ptr(N) and its subpages have values greater than Key(N-1). And30** so forth.31**32** Finding a particular key requires reading O(log(M)) pages from the 33** disk where M is the number of entries in the tree.34**35** In this implementation, a single file can hold one or more separate 36** BTrees. Each BTree is identified by the index of its root page. The37** key and data for any entry are combined to form the "payload". A38** fixed amount of payload can be carried directly on the database39** page. If the payload is larger than the preset amount then surplus40** bytes are stored on overflow pages. The payload for an entry41** and the preceding pointer are combined to form a "Cell". Each 42** page has a small header which contains the Ptr(N) pointer and other43** information such as the size of key and data.44**45** FORMAT DETAILS46**47** The file is divided into pages. The first page is called page 1,48** the second is page 2, and so forth. A page number of zero indicates49** "no such page". The page size can be any power of 2 between 512 and 65536.50** Each page can be either a btree page, a freelist page, an overflow51** page, or a pointer-map page.52**53** The first page is always a btree page. The first 100 bytes of the first54** page contain a special header (the "file header") that describes the file.55** The format of the file header is as follows:56**57** OFFSET SIZE DESCRIPTION58** 0 16 Header string: "SQLite format 3\000"59** 16 2 Page size in bytes. (1 means 65536)60** 18 1 File format write version61** 19 1 File format read version62** 20 1 Bytes of unused space at the end of each page63** 21 1 Max embedded payload fraction (must be 64)64** 22 1 Min embedded payload fraction (must be 32)65** 23 1 Min leaf payload fraction (must be 32)66** 24 4 File change counter67** 28 4 The size of the database in pages68** 32 4 First freelist page69** 36 4 Number of freelist pages in the file70** 40 60 15 4-byte meta values passed to higher layers71**72** 40 4 Schema cookie73** 44 4 File format of schema layer74** 48 4 Size of page cache75** 52 4 Largest root-page (auto/incr_vacuum)76** 56 4 1=UTF-8 2=UTF16le 3=UTF16be77** 60 4 User version78** 64 4 Incremental vacuum mode79** 68 4 Application-ID80** 72 20 unused81** 92 4 The version-valid-for number82** 96 4 SQLITE_VERSION_NUMBER83**84** All of the integer values are big-endian (most significant byte first).85**86** The file change counter is incremented when the database is changed87** This counter allows other processes to know when the file has changed88** and thus when they need to flush their cache.89**90** The max embedded payload fraction is the amount of the total usable91** space in a page that can be consumed by a single cell for standard92** B-tree (non-LEAFDATA) tables. A value of 255 means 100%. The default93** is to limit the maximum cell size so that at least 4 cells will fit94** on one page. Thus the default max embedded payload fraction is 64.95**96** If the payload for a cell is larger than the max payload, then extra97** payload is spilled to overflow pages. Once an overflow page is allocated,98** as many bytes as possible are moved into the overflow pages without letting99** the cell size drop below the min embedded payload fraction.100**101** The min leaf payload fraction is like the min embedded payload fraction102** except that it applies to leaf nodes in a LEAFDATA tree. The maximum103** payload fraction for a LEAFDATA tree is always 100% (or 255) and it104** not specified in the header.105**106** Each btree pages is divided into three sections: The header, the107** cell pointer array, and the cell content area. Page 1 also has a 100-byte108** file header that occurs before the page header.109**110** |----------------|111** | file header | 100 bytes. Page 1 only.112** |----------------|113** | page header | 8 bytes for leaves. 12 bytes for interior nodes114** |----------------|115** | cell pointer | | 2 bytes per cell. Sorted order.116** | array | | Grows downward117** | | v118** |----------------|119** | unallocated |120** | space |121** |----------------| ^ Grows upwards122** | cell content | | Arbitrary order interspersed with freeblocks.123** | area | | and free space fragments.124** |----------------|125**126** The page headers looks like this:127**128** OFFSET SIZE DESCRIPTION129** 0 1 Flags. 1: intkey, 2: zerodata, 4: leafdata, 8: leaf130** 1 2 byte offset to the first freeblock131** 3 2 number of cells on this page132** 5 2 first byte of the cell content area133** 7 1 number of fragmented free bytes134** 8 4 Right child (the Ptr(N) value). Omitted on leaves.135**136** The flags define the format of this btree page. The leaf flag means that137** this page has no children. The zerodata flag means that this page carries138** only keys and no data. The intkey flag means that the key is an integer139** which is stored in the key size entry of the cell header rather than in140** the payload area.141**142** The cell pointer array begins on the first byte after the page header.143** The cell pointer array contains zero or more 2-byte numbers which are144** offsets from the beginning of the page to the cell content in the cell145** content area. The cell pointers occur in sorted order. The system strives146** to keep free space after the last cell pointer so that new cells can147** be easily added without having to defragment the page.148**149** Cell content is stored at the very end of the page and grows toward the150** beginning of the page.151**152** Unused space within the cell content area is collected into a linked list of153** freeblocks. Each freeblock is at least 4 bytes in size. The byte offset154** to the first freeblock is given in the header. Freeblocks occur in155** increasing order. Because a freeblock must be at least 4 bytes in size,156** any group of 3 or fewer unused bytes in the cell content area cannot157** exist on the freeblock chain. A group of 3 or fewer free bytes is called158** a fragment. The total number of bytes in all fragments is recorded.159** in the page header at offset 7.160**161** SIZE DESCRIPTION162** 2 Byte offset of the next freeblock163** 2 Bytes in this freeblock164**165** Cells are of variable length. Cells are stored in the cell content area at166** the end of the page. Pointers to the cells are in the cell pointer array167** that immediately follows the page header. Cells is not necessarily168** contiguous or in order, but cell pointers are contiguous and in order.169**170** Cell content makes use of variable length integers. A variable171** length integer is 1 to 9 bytes where the lower 7 bits of each 172** byte are used. The integer consists of all bytes that have bit 8 set and173** the first byte with bit 8 clear. The most significant byte of the integer174** appears first. A variable-length integer may not be more than 9 bytes long.175** As a special case, all 8 bits of the 9th byte are used as data. This176** allows a 64-bit integer to be encoded in 9 bytes.177**178** 0x00 becomes 0x00000000179** 0x7f becomes 0x0000007f180** 0x81 0x00 becomes 0x00000080181** 0x82 0x00 becomes 0x00000100182** 0x80 0x7f becomes 0x0000007f183** 0x81 0x91 0xd1 0xac 0x78 becomes 0x12345678184** 0x81 0x81 0x81 0x81 0x01 becomes 0x10204081185**186** Variable length integers are used for rowids and to hold the number of187** bytes of key and data in a btree cell.188**189** The content of a cell looks like this:190**191** SIZE DESCRIPTION192** 4 Page number of the left child. Omitted if leaf flag is set.193** var Number of bytes of data. Omitted if the zerodata flag is set.194** var Number of bytes of key. Or the key itself if intkey flag is set.195** * Payload196** 4 First page of the overflow chain. Omitted if no overflow197**198** Overflow pages form a linked list. Each page except the last is completely199** filled with data (pagesize - 4 bytes). The last page can have as little200** as 1 byte of data.201**202** SIZE DESCRIPTION203** 4 Page number of next overflow page204** * Data205**206** Freelist pages come in two subtypes: trunk pages and leaf pages. The207** file header points to the first in a linked list of trunk page. Each trunk208** page points to multiple leaf pages. The content of a leaf page is209** unspecified. A trunk page looks like this:210**211** SIZE DESCRIPTION212** 4 Page number of next trunk page213** 4 Number of leaf pointers on this page214** * zero or more pages numbers of leaves215*/216#include "sqliteInt.h"217 218 219/* The following value is the maximum cell size assuming a maximum page220** size give above.221*/222#define MX_CELL_SIZE(pBt) ((int)(pBt->pageSize-8))223 224/* The maximum number of cells on a single page of the database. This225** assumes a minimum cell size of 6 bytes (4 bytes for the cell itself226** plus 2 bytes for the index to the cell in the page header). Such227** small cells will be rare, but they are possible.228*/229#define MX_CELL(pBt) ((pBt->pageSize-8)/6)230 231/* Forward declarations */232typedef struct MemPage MemPage;233typedef struct BtLock BtLock;234typedef struct CellInfo CellInfo;235 236/*237** This is a magic string that appears at the beginning of every238** SQLite database in order to identify the file as a real database.239**240** You can change this value at compile-time by specifying a241** -DSQLITE_FILE_HEADER="..." on the compiler command-line. The242** header must be exactly 16 bytes including the zero-terminator so243** the string itself should be 15 characters long. If you change244** the header, then your custom library will not be able to read 245** databases generated by the standard tools and the standard tools246** will not be able to read databases created by your custom library.247*/248#ifndef SQLITE_FILE_HEADER /* 123456789 123456 */249# define SQLITE_FILE_HEADER "SQLite format 3"250#endif251 252/*253** Page type flags. An ORed combination of these flags appear as the254** first byte of on-disk image of every BTree page.255*/256#define PTF_INTKEY 0x01257#define PTF_ZERODATA 0x02258#define PTF_LEAFDATA 0x04259#define PTF_LEAF 0x08260 261/*262** An instance of this object stores information about each a single database263** page that has been loaded into memory. The information in this object264** is derived from the raw on-disk page content.265**266** As each database page is loaded into memory, the pager allocates an267** instance of this object and zeros the first 8 bytes. (This is the268** "extra" information associated with each page of the pager.)269**270** Access to all fields of this structure is controlled by the mutex271** stored in MemPage.pBt->mutex.272*/273struct MemPage {274 u8 isInit; /* True if previously initialized. MUST BE FIRST! */275 u8 intKey; /* True if table b-trees. False for index b-trees */276 u8 intKeyLeaf; /* True if the leaf of an intKey table */277 Pgno pgno; /* Page number for this page */278 /* Only the first 8 bytes (above) are zeroed by pager.c when a new page279 ** is allocated. All fields that follow must be initialized before use */280 u8 leaf; /* True if a leaf page */281 u8 hdrOffset; /* 100 for page 1. 0 otherwise */282 u8 childPtrSize; /* 0 if leaf==1. 4 if leaf==0 */283 u8 max1bytePayload; /* min(maxLocal,127) */284 u8 nOverflow; /* Number of overflow cell bodies in aCell[] */285 u16 maxLocal; /* Copy of BtShared.maxLocal or BtShared.maxLeaf */286 u16 minLocal; /* Copy of BtShared.minLocal or BtShared.minLeaf */287 u16 cellOffset; /* Index in aData of first cell pointer */288 int nFree; /* Number of free bytes on the page. -1 for unknown */289 u16 nCell; /* Number of cells on this page, local and ovfl */290 u16 maskPage; /* Mask for page offset */291 u16 aiOvfl[4]; /* Insert the i-th overflow cell before the aiOvfl-th292 ** non-overflow cell */293 u8 *apOvfl[4]; /* Pointers to the body of overflow cells */294 BtShared *pBt; /* Pointer to BtShared that this page is part of */295 u8 *aData; /* Pointer to disk image of the page data */296 u8 *aDataEnd; /* One byte past the end of the entire page - not just297 ** the usable space, the entire page. Used to prevent298 ** corruption-induced buffer overflow. */299 u8 *aCellIdx; /* The cell index area */300 u8 *aDataOfst; /* Same as aData for leaves. aData+4 for interior */301 DbPage *pDbPage; /* Pager page handle */302 u16 (*xCellSize)(MemPage*,u8*); /* cellSizePtr method */303 void (*xParseCell)(MemPage*,u8*,CellInfo*); /* btreeParseCell method */304};305 306/*307** A linked list of the following structures is stored at BtShared.pLock.308** Locks are added (or upgraded from READ_LOCK to WRITE_LOCK) when a cursor 309** is opened on the table with root page BtShared.iTable. Locks are removed310** from this list when a transaction is committed or rolled back, or when311** a btree handle is closed.312*/313struct BtLock {314 Btree *pBtree; /* Btree handle holding this lock */315 Pgno iTable; /* Root page of table */316 u8 eLock; /* READ_LOCK or WRITE_LOCK */317 BtLock *pNext; /* Next in BtShared.pLock list */318};319 320/* Candidate values for BtLock.eLock */321#define READ_LOCK 1322#define WRITE_LOCK 2323 324/* A Btree handle325**326** A database connection contains a pointer to an instance of327** this object for every database file that it has open. This structure328** is opaque to the database connection. The database connection cannot329** see the internals of this structure and only deals with pointers to330** this structure.331**332** For some database files, the same underlying database cache might be 333** shared between multiple connections. In that case, each connection334** has it own instance of this object. But each instance of this object335** points to the same BtShared object. The database cache and the336** schema associated with the database file are all contained within337** the BtShared object.338**339** All fields in this structure are accessed under sqlite3.mutex.340** The pBt pointer itself may not be changed while there exists cursors 341** in the referenced BtShared that point back to this Btree since those342** cursors have to go through this Btree to find their BtShared and343** they often do so without holding sqlite3.mutex.344*/345struct Btree {346 sqlite3 *db; /* The database connection holding this btree */347 BtShared *pBt; /* Sharable content of this btree */348 u8 inTrans; /* TRANS_NONE, TRANS_READ or TRANS_WRITE */349 u8 sharable; /* True if we can share pBt with another db */350 u8 locked; /* True if db currently has pBt locked */351 u8 hasIncrblobCur; /* True if there are one or more Incrblob cursors */352 int wantToLock; /* Number of nested calls to sqlite3BtreeEnter() */353 int nBackup; /* Number of backup operations reading this btree */354 u32 iBDataVersion; /* Combines with pBt->pPager->iDataVersion */355 Btree *pNext; /* List of other sharable Btrees from the same db */356 Btree *pPrev; /* Back pointer of the same list */357#ifdef SQLITE_DEBUG358 u64 nSeek; /* Calls to sqlite3BtreeMovetoUnpacked() */359#endif360#ifndef SQLITE_OMIT_SHARED_CACHE361 BtLock lock; /* Object used to lock page 1 */362#endif363};364 365/*366** Btree.inTrans may take one of the following values.367**368** If the shared-data extension is enabled, there may be multiple users369** of the Btree structure. At most one of these may open a write transaction,370** but any number may have active read transactions.371**372** These values must match SQLITE_TXN_NONE, SQLITE_TXN_READ, and373** SQLITE_TXN_WRITE374*/375#define TRANS_NONE 0376#define TRANS_READ 1377#define TRANS_WRITE 2378 379#if TRANS_NONE!=SQLITE_TXN_NONE380# error wrong numeric code for no-transaction381#endif382#if TRANS_READ!=SQLITE_TXN_READ383# error wrong numeric code for read-transaction384#endif385#if TRANS_WRITE!=SQLITE_TXN_WRITE386# error wrong numeric code for write-transaction387#endif388 389 390/*391** An instance of this object represents a single database file.392** 393** A single database file can be in use at the same time by two394** or more database connections. When two or more connections are395** sharing the same database file, each connection has it own396** private Btree object for the file and each of those Btrees points397** to this one BtShared object. BtShared.nRef is the number of398** connections currently sharing this database file.399**400** Fields in this structure are accessed under the BtShared.mutex401** mutex, except for nRef and pNext which are accessed under the402** global SQLITE_MUTEX_STATIC_MAIN mutex. The pPager field403** may not be modified once it is initially set as long as nRef>0.404** The pSchema field may be set once under BtShared.mutex and405** thereafter is unchanged as long as nRef>0.406**407** isPending:408**409** If a BtShared client fails to obtain a write-lock on a database410** table (because there exists one or more read-locks on the table),411** the shared-cache enters 'pending-lock' state and isPending is412** set to true.413**414** The shared-cache leaves the 'pending lock' state when either of415** the following occur:416**417** 1) The current writer (BtShared.pWriter) concludes its transaction, OR418** 2) The number of locks held by other connections drops to zero.419**420** while in the 'pending-lock' state, no connection may start a new421** transaction.422**423** This feature is included to help prevent writer-starvation.424*/425struct BtShared {426 Pager *pPager; /* The page cache */427 sqlite3 *db; /* Database connection currently using this Btree */428 BtCursor *pCursor; /* A list of all open cursors */429 MemPage *pPage1; /* First page of the database */430 u8 openFlags; /* Flags to sqlite3BtreeOpen() */431#ifndef SQLITE_OMIT_AUTOVACUUM432 u8 autoVacuum; /* True if auto-vacuum is enabled */433 u8 incrVacuum; /* True if incr-vacuum is enabled */434 u8 bDoTruncate; /* True to truncate db on commit */435#endif436 u8 inTransaction; /* Transaction state */437 u8 max1bytePayload; /* Maximum first byte of cell for a 1-byte payload */438 u8 nReserveWanted; /* Desired number of extra bytes per page */439 u16 btsFlags; /* Boolean parameters. See BTS_* macros below */440 u16 maxLocal; /* Maximum local payload in non-LEAFDATA tables */441 u16 minLocal; /* Minimum local payload in non-LEAFDATA tables */442 u16 maxLeaf; /* Maximum local payload in a LEAFDATA table */443 u16 minLeaf; /* Minimum local payload in a LEAFDATA table */444 u32 pageSize; /* Total number of bytes on a page */445 u32 usableSize; /* Number of usable bytes on each page */446 int nTransaction; /* Number of open transactions (read + write) */447 u32 nPage; /* Number of pages in the database */448 void *pSchema; /* Pointer to space allocated by sqlite3BtreeSchema() */449 void (*xFreeSchema)(void*); /* Destructor for BtShared.pSchema */450 sqlite3_mutex *mutex; /* Non-recursive mutex required to access this object */451 Bitvec *pHasContent; /* Set of pages moved to free-list this transaction */452#ifndef SQLITE_OMIT_SHARED_CACHE453 int nRef; /* Number of references to this structure */454 BtShared *pNext; /* Next on a list of sharable BtShared structs */455 BtLock *pLock; /* List of locks held on this shared-btree struct */456 Btree *pWriter; /* Btree with currently open write transaction */457#endif458 u8 *pTmpSpace; /* Temp space sufficient to hold a single cell */459 int nPreformatSize; /* Size of last cell written by TransferRow() */460};461 462/*463** Allowed values for BtShared.btsFlags464*/465#define BTS_READ_ONLY 0x0001 /* Underlying file is readonly */466#define BTS_PAGESIZE_FIXED 0x0002 /* Page size can no longer be changed */467#define BTS_SECURE_DELETE 0x0004 /* PRAGMA secure_delete is enabled */468#define BTS_OVERWRITE 0x0008 /* Overwrite deleted content with zeros */469#define BTS_FAST_SECURE 0x000c /* Combination of the previous two */470#define BTS_INITIALLY_EMPTY 0x0010 /* Database was empty at trans start */471#define BTS_NO_WAL 0x0020 /* Do not open write-ahead-log files */472#define BTS_EXCLUSIVE 0x0040 /* pWriter has an exclusive lock */473#define BTS_PENDING 0x0080 /* Waiting for read-locks to clear */474 475/*476** An instance of the following structure is used to hold information477** about a cell. The parseCellPtr() function fills in this structure478** based on information extract from the raw disk page.479*/480struct CellInfo {481 i64 nKey; /* The key for INTKEY tables, or nPayload otherwise */482 u8 *pPayload; /* Pointer to the start of payload */483 u32 nPayload; /* Bytes of payload */484 u16 nLocal; /* Amount of payload held locally, not on overflow */485 u16 nSize; /* Size of the cell content on the main b-tree page */486};487 488/*489** Maximum depth of an SQLite B-Tree structure. Any B-Tree deeper than490** this will be declared corrupt. This value is calculated based on a491** maximum database size of 2^31 pages a minimum fanout of 2 for a492** root-node and 3 for all other internal nodes.493**494** If a tree that appears to be taller than this is encountered, it is495** assumed that the database is corrupt.496*/497#define BTCURSOR_MAX_DEPTH 20498 499/*500** Maximum amount of storage local to a database page, regardless of501** page size.502*/503#define BT_MAX_LOCAL 65501 /* 65536 - 35 */504 505/*506** A cursor is a pointer to a particular entry within a particular507** b-tree within a database file.508**509** The entry is identified by its MemPage and the index in510** MemPage.aCell[] of the entry.511**512** A single database file can be shared by two more database connections,513** but cursors cannot be shared. Each cursor is associated with a514** particular database connection identified BtCursor.pBtree.db.515**516** Fields in this structure are accessed under the BtShared.mutex517** found at self->pBt->mutex. 518**519** skipNext meaning:520** The meaning of skipNext depends on the value of eState:521**522** eState Meaning of skipNext523** VALID skipNext is meaningless and is ignored524** INVALID skipNext is meaningless and is ignored525** SKIPNEXT sqlite3BtreeNext() is a no-op if skipNext>0 and526** sqlite3BtreePrevious() is no-op if skipNext<0.527** REQUIRESEEK restoreCursorPosition() restores the cursor to528** eState=SKIPNEXT if skipNext!=0529** FAULT skipNext holds the cursor fault error code.530*/531struct BtCursor {532 u8 eState; /* One of the CURSOR_XXX constants (see below) */533 u8 curFlags; /* zero or more BTCF_* flags defined below */534 u8 curPagerFlags; /* Flags to send to sqlite3PagerGet() */535 u8 hints; /* As configured by CursorSetHints() */536 int skipNext; /* Prev() is noop if negative. Next() is noop if positive.537 ** Error code if eState==CURSOR_FAULT */538 Btree *pBtree; /* The Btree to which this cursor belongs */539 Pgno *aOverflow; /* Cache of overflow page locations */540 void *pKey; /* Saved key that was cursor last known position */541 /* All fields above are zeroed when the cursor is allocated. See542 ** sqlite3BtreeCursorZero(). Fields that follow must be manually543 ** initialized. */544#define BTCURSOR_FIRST_UNINIT pBt /* Name of first uninitialized field */545 BtShared *pBt; /* The BtShared this cursor points to */546 BtCursor *pNext; /* Forms a linked list of all cursors */547 CellInfo info; /* A parse of the cell we are pointing at */548 i64 nKey; /* Size of pKey, or last integer key */549 Pgno pgnoRoot; /* The root page of this tree */550 i8 iPage; /* Index of current page in apPage */551 u8 curIntKey; /* Value of apPage[0]->intKey */552 u16 ix; /* Current index for apPage[iPage] */553 u16 aiIdx[BTCURSOR_MAX_DEPTH-1]; /* Current index in apPage[i] */554 struct KeyInfo *pKeyInfo; /* Arg passed to comparison function */555 MemPage *pPage; /* Current page */556 MemPage *apPage[BTCURSOR_MAX_DEPTH-1]; /* Stack of parents of current page */557};558 559/*560** Legal values for BtCursor.curFlags561*/562#define BTCF_WriteFlag 0x01 /* True if a write cursor */563#define BTCF_ValidNKey 0x02 /* True if info.nKey is valid */564#define BTCF_ValidOvfl 0x04 /* True if aOverflow is valid */565#define BTCF_AtLast 0x08 /* Cursor is pointing to the last entry */566#define BTCF_Incrblob 0x10 /* True if an incremental I/O handle */567#define BTCF_Multiple 0x20 /* Maybe another cursor on the same btree */568#define BTCF_Pinned 0x40 /* Cursor is busy and cannot be moved */569 570/*571** Potential values for BtCursor.eState.572**573** CURSOR_INVALID:574** Cursor does not point to a valid entry. This can happen (for example) 575** because the table is empty or because BtreeCursorFirst() has not been576** called.577**578** CURSOR_VALID:579** Cursor points to a valid entry. getPayload() etc. may be called.580**581** CURSOR_SKIPNEXT:582** Cursor is valid except that the Cursor.skipNext field is non-zero583** indicating that the next sqlite3BtreeNext() or sqlite3BtreePrevious()584** operation should be a no-op.585**586** CURSOR_REQUIRESEEK:587** The table that this cursor was opened on still exists, but has been 588** modified since the cursor was last used. The cursor position is saved589** in variables BtCursor.pKey and BtCursor.nKey. When a cursor is in 590** this state, restoreCursorPosition() can be called to attempt to591** seek the cursor to the saved position.592**593** CURSOR_FAULT:594** An unrecoverable error (an I/O error or a malloc failure) has occurred595** on a different connection that shares the BtShared cache with this596** cursor. The error has left the cache in an inconsistent state.597** Do nothing else with this cursor. Any attempt to use the cursor598** should return the error code stored in BtCursor.skipNext599*/600#define CURSOR_VALID 0601#define CURSOR_INVALID 1602#define CURSOR_SKIPNEXT 2603#define CURSOR_REQUIRESEEK 3604#define CURSOR_FAULT 4605 606/* 607** The database page the PENDING_BYTE occupies. This page is never used.608*/609#define PENDING_BYTE_PAGE(pBt) ((Pgno)((PENDING_BYTE/((pBt)->pageSize))+1))610 611/*612** These macros define the location of the pointer-map entry for a 613** database page. The first argument to each is the number of usable614** bytes on each page of the database (often 1024). The second is the615** page number to look up in the pointer map.616**617** PTRMAP_PAGENO returns the database page number of the pointer-map618** page that stores the required pointer. PTRMAP_PTROFFSET returns619** the offset of the requested map entry.620**621** If the pgno argument passed to PTRMAP_PAGENO is a pointer-map page,622** then pgno is returned. So (pgno==PTRMAP_PAGENO(pgsz, pgno)) can be623** used to test if pgno is a pointer-map page. PTRMAP_ISPAGE implements624** this test.625*/626#define PTRMAP_PAGENO(pBt, pgno) ptrmapPageno(pBt, pgno)627#define PTRMAP_PTROFFSET(pgptrmap, pgno) (5*(pgno-pgptrmap-1))628#define PTRMAP_ISPAGE(pBt, pgno) (PTRMAP_PAGENO((pBt),(pgno))==(pgno))629 630/*631** The pointer map is a lookup table that identifies the parent page for632** each child page in the database file. The parent page is the page that633** contains a pointer to the child. Every page in the database contains634** 0 or 1 parent pages. (In this context 'database page' refers635** to any page that is not part of the pointer map itself.) Each pointer map636** entry consists of a single byte 'type' and a 4 byte parent page number.637** The PTRMAP_XXX identifiers below are the valid types.638**639** The purpose of the pointer map is to facility moving pages from one640** position in the file to another as part of autovacuum. When a page641** is moved, the pointer in its parent must be updated to point to the642** new location. The pointer map is used to locate the parent page quickly.643**644** PTRMAP_ROOTPAGE: The database page is a root-page. The page-number is not645** used in this case.646**647** PTRMAP_FREEPAGE: The database page is an unused (free) page. The page-number 648** is not used in this case.649**650** PTRMAP_OVERFLOW1: The database page is the first page in a list of 651** overflow pages. The page number identifies the page that652** contains the cell with a pointer to this overflow page.653**654** PTRMAP_OVERFLOW2: The database page is the second or later page in a list of655** overflow pages. The page-number identifies the previous656** page in the overflow page list.657**658** PTRMAP_BTREE: The database page is a non-root btree page. The page number659** identifies the parent page in the btree.660*/661#define PTRMAP_ROOTPAGE 1662#define PTRMAP_FREEPAGE 2663#define PTRMAP_OVERFLOW1 3664#define PTRMAP_OVERFLOW2 4665#define PTRMAP_BTREE 5666 667/* A bunch of assert() statements to check the transaction state variables668** of handle p (type Btree*) are internally consistent.669*/670#define btreeIntegrity(p) \671 assert( p->pBt->inTransaction!=TRANS_NONE || p->pBt->nTransaction==0 ); \672 assert( p->pBt->inTransaction>=p->inTrans ); 673 674 675/*676** The ISAUTOVACUUM macro is used within balance_nonroot() to determine677** if the database supports auto-vacuum or not. Because it is used678** within an expression that is an argument to another macro 679** (sqliteMallocRaw), it is not possible to use conditional compilation.680** So, this macro is defined instead.681*/682#ifndef SQLITE_OMIT_AUTOVACUUM683#define ISAUTOVACUUM(pBt) (pBt->autoVacuum)684#else685#define ISAUTOVACUUM(pBt) 0686#endif687 688 689/*690** This structure is passed around through all the PRAGMA integrity_check691** checking routines in order to keep track of some global state information.692**693** The aRef[] array is allocated so that there is 1 bit for each page in694** the database. As the integrity-check proceeds, for each page used in695** the database the corresponding bit is set. This allows integrity-check to 696** detect pages that are used twice and orphaned pages (both of which 697** indicate corruption).698*/699typedef struct IntegrityCk IntegrityCk;700struct IntegrityCk {701 BtShared *pBt; /* The tree being checked out */702 Pager *pPager; /* The associated pager. Also accessible by pBt->pPager */703 u8 *aPgRef; /* 1 bit per page in the db (see above) */704 Pgno nCkPage; /* Pages in the database. 0 for partial check */705 int mxErr; /* Stop accumulating errors when this reaches zero */706 int nErr; /* Number of messages written to zErrMsg so far */707 int rc; /* SQLITE_OK, SQLITE_NOMEM, or SQLITE_INTERRUPT */708 u32 nStep; /* Number of steps into the integrity_check process */709 const char *zPfx; /* Error message prefix */710 Pgno v0; /* Value for first %u substitution in zPfx (root page) */711 Pgno v1; /* Value for second %u substitution in zPfx (current pg) */712 int v2; /* Value for third %d substitution in zPfx */713 StrAccum errMsg; /* Accumulate the error message text here */714 u32 *heap; /* Min-heap used for analyzing cell coverage */715 sqlite3 *db; /* Database connection running the check */716 i64 nRow; /* Number of rows visited in current tree */717};718 719/*720** Routines to read or write a two- and four-byte big-endian integer values.721*/722#define get2byte(x) ((x)[0]<<8 | (x)[1])723#define put2byte(p,v) ((p)[0] = (u8)((v)>>8), (p)[1] = (u8)(v))724#define get4byte sqlite3Get4byte725#define put4byte sqlite3Put4byte726 727/*728** get2byteAligned(), unlike get2byte(), requires that its argument point to a729** two-byte aligned address. get2byteAligned() is only used for accessing the730** cell addresses in a btree header.731*/732#if SQLITE_BYTEORDER==4321733# define get2byteAligned(x) (*(u16*)(x))734#elif SQLITE_BYTEORDER==1234 && GCC_VERSION>=4008000735# define get2byteAligned(x) __builtin_bswap16(*(u16*)(x))736#elif SQLITE_BYTEORDER==1234 && MSVC_VERSION>=1300737# define get2byteAligned(x) _byteswap_ushort(*(u16*)(x))738#else739# define get2byteAligned(x) ((x)[0]<<8 | (x)[1])740#endif741