Many hyperlinks are disabled.
Use anonymous login
to enable hyperlinks.
Overview
Comment: | :-) (CVS 218) |
---|---|
Downloads: | Tarball | ZIP archive |
Timelines: | family | ancestors | descendants | both | trunk |
Files: | files | file ages | folders |
SHA1: |
523d52dfa6ae3028cbcc88d406501f3e |
User & Date: | drh 2001-05-21 13:45:10.000 |
Context
2001-05-24
| ||
21:06 | Continued work on btree (CVS 219) (check-in: 18500cdcc1 user: drh tags: trunk) | |
2001-05-21
| ||
13:45 | :-) (CVS 218) (check-in: 523d52dfa6 user: drh tags: trunk) | |
2001-05-15
| ||
00:39 | :-) (CVS 217) (check-in: ee6760fb62 user: drh tags: trunk) | |
Changes
Changes to src/btree.c.
︙ | ︙ | |||
17 18 19 20 21 22 23 | ** Boston, MA 02111-1307, USA. ** ** Author contact information: ** drh@hwaci.com ** http://www.hwaci.com/drh/ ** ************************************************************************* | | | | < > > > > < < | > > | < < > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > | > > > > > > > | | | | | | | | | | | | 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 | ** Boston, MA 02111-1307, USA. ** ** Author contact information: ** drh@hwaci.com ** http://www.hwaci.com/drh/ ** ************************************************************************* ** $Id: btree.c,v 1.6 2001/05/21 13:45:10 drh Exp $ */ #include "sqliteInt.h" #include "pager.h" #include "btree.h" #include <assert.h> /* ** The maximum number of database entries that can be held in a single ** page of the database. */ #define MX_CELL ((SQLITE_PAGE_SIZE-sizeof(PageHdr))/sizeof(Cell)) /* ** The maximum amount of data (in bytes) that can be stored locally for a ** database entry. If the entry contains more data than this, the ** extra goes onto overflow pages. */ #define MX_LOCAL_PAYLOAD \ ((SQLITE_PAGE_SIZE-sizeof(PageHdr)-4*(sizeof(Cell)+sizeof(Pgno)))/4) /* ** The in-memory image of a disk page has the auxiliary information appended ** to the end. EXTRA_SIZE is the number of bytes of space needed to hold ** that extra information. */ #define EXTRA_SIZE (sizeof(MemPage)-SQLITE_PAGE_SIZE) /* ** Number of bytes on a single overflow page. */ #define OVERFLOW_SIZE (SQLITE_PAGE_SIZE-sizeof(Pgno)) /* ** Primitive data types. u32 must be 4 bytes and u16 must be 2 bytes. ** Change these typedefs when porting to new architectures. */ typedef unsigned int u32; typedef unsigned short int u16; /* ** Forward declarations of structures used only in this file. */ typedef struct Page1Header Page1Header; typedef struct MemPage MemPage; typedef struct PageHdr PageHdr; typedef struct Cell Cell; typedef struct FreeBlk FreeBlk; typedef struct OverflowPage OverflowPage; /* ** All structures on a database page are aligned to 4-byte boundries. ** This routine rounds up a number of bytes to the next multiple of 4. ** ** This might need to change for computer architectures that require ** and 8-byte alignment boundry for structures. */ #define ROUNDUP(X) ((X+3) & ~3) /* ** The first pages of the database file contains some additional ** information used for housekeeping and sanity checking. Otherwise, ** the first page is just like any other. The additional information ** found on the first page is described by the following structure. */ struct Page1Header { u32 magic1; /* A magic number for sanity checking */ u32 magic2; /* A second magic number for sanity checking */ Pgno firstList; /* First free page in a list of all free pages */ }; #define MAGIC_1 0x7264dc61 #define MAGIC_2 0x54e55d9e /* ** Each database page has a header as follows: ** ** page1_header Extra numbers found on page 1 only. ** rightmost_pgno Page number of the right-most child page ** first_cell Index into MemPage.aPage of first cell ** first_free Index of first free block ** ** MemPage.pStart always points to the rightmost_pgno. First_free is ** 0 if there is no free space on this page. Otherwise, first_free is ** the index in MemPage.aPage[] of a FreeBlk structure that describes ** the first block of free space. All free space is defined by a linked ** list of FreeBlk structures. ** ** Data is stored in a linked list of Cell structures. First_cell is ** the index into MemPage.aPage[] of the first cell on the page. The ** Cells are kept in sorted order. */ struct PageHdr { Pgno pgno; /* Child page that comes after all cells on this page */ u16 firstCell; /* Index in MemPage.aPage[] of the first cell */ u16 firstFree; /* Index in MemPage.aPage[] of the first free block */ }; /* ** Data on a database page is stored as a linked list of Cell structures. ** Both the key and the data are stored in aData[]. The key always comes ** first. The aData[] field grows as necessary to hold the key and data, ** up to a maximum of MX_LOCAL_PAYLOAD bytes. If the size of the key and ** data combined exceeds MX_LOCAL_PAYLOAD bytes, then the 4 bytes beginning ** at Cell.aData[MX_LOCAL_PAYLOAD] are the page number of the first overflow ** page. */ struct Cell { Pgno pgno; /* Child page that comes before this cell */ u16 nKey; /* Number of bytes in the key */ u16 iNext; /* Index in MemPage.aPage[] of next cell in sorted order */ u32 nData; /* Number of bytes of data */ char aData[4]; /* Key and data */ }; /* ** Free space on a page is remembered using a linked list of the FreeBlk ** structures. Space on a database page is allocated in increments of ** at least 4 bytes and is always aligned to a 4-byte boundry. */ struct FreeBlk { u16 iSize; /* Number of u32-sized slots in the block of free space */ u16 iNext; /* Index in MemPage.aPage[] of the next free block */ }; /* ** When the key and data for a single entry in the BTree will not fit in ** the MX_LOACAL_PAYLOAD bytes of space available on the database page, ** then all extra data is written to a linked list of overflow pages. ** Each overflow page is an instance of the following structure. ** ** Unused pages in the database are also represented by instances of ** the OverflowPage structure. The Page1Header.freeList field is the ** page number of the first page in a linked list of unused database ** pages. */ struct OverflowPage { Pgno next; char aData[SQLITE_PAGE_SIZE-sizeof(Pgno)]; }; /* ** For every page in the database file, an instance of the following structure ** is stored in memory. The aPage[] array contains the data obtained from ** the disk. The rest is auxiliary data that held in memory only. The ** auxiliary data is only valid for regular database pages - the auxiliary ** data is meaningless for overflow pages and pages on the freelist. ** ** Of particular interest in the auxiliary data is the aCell[] entry. Each ** aCell[] entry is a pointer to a Cell structure in aPage[]. The cells ** put in this array so that they can be accessed in constant time, rather ** than in linear time which would be needed if we walked the linked list. */ struct MemPage { char aPage[SQLITE_PAGE_SIZE]; /* Page data stored on disk */ unsigned char isInit; /* True if auxiliary data is initialized */ unsigned char validUp; /* True if MemPage.up is valid */ unsigned char validLeft; /* True if MemPage.left is valid */ unsigned char validRight; /* True if MemPage.right is valid */ Pgno up; /* The parent page. 0 means this is the root */ Pgno left; /* Left sibling page. 0==none */ Pgno right; /* Right sibling page. 0==none */ int idxStart; /* Index in aPage[] of real data */ PageHdr *pStart; /* Points to aPage[idxStart] */ int nFree; /* Number of free bytes in aPage[] */ int nCell; /* Number of entries on this page */ Cell *aCell[MX_CELL]; /* All data entires in sorted order */ } /* ** Everything we need to know about an open database */ struct Btree { Pager *pPager; /* The page cache */ BtCursor *pCursor; /* A list of all open cursors */ MemPage *page1; /* First page of the database */ int inTrans; /* True if a transaction is in progress */ }; typedef Btree Bt; /* ** A cursor is a pointer to a particular entry in the BTree. ** The entry is identified by its MemPage and the index in ** MemPage.aCell[] of the entry. |
︙ | ︙ | |||
209 210 211 212 213 214 215 | } /* ** Allocate space on a page. The space needs to be at least ** nByte bytes in size. (Actually, all allocations are rounded ** up to the next even multiple of 4.) Return the index into ** pPage->aPage[] of the first byte of the new allocation. | | > | 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 | } /* ** Allocate space on a page. The space needs to be at least ** nByte bytes in size. (Actually, all allocations are rounded ** up to the next even multiple of 4.) Return the index into ** pPage->aPage[] of the first byte of the new allocation. ** Or return 0 if there is not enough free space on the page to ** satisfy the allocation request. ** ** This routine will call defragmentPage if necessary to consolidate ** free space. */ static int allocSpace(MemPage *pPage, int nByte){ FreeBlk *p; u16 *pIdx; |
︙ | ︙ | |||
245 246 247 248 249 250 251 | pPage->nFree -= nByte; return start; } /* ** Return a section of the MemPage.aPage[] to the freelist. ** The first byte of the new free block is pPage->aPage[start] | | > > > | 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 | pPage->nFree -= nByte; return start; } /* ** Return a section of the MemPage.aPage[] to the freelist. ** The first byte of the new free block is pPage->aPage[start] ** and the size of the block is "size". ** ** Most of the effort here is involved in coalesing adjacent ** free blocks into a single big free block. */ static void freeSpace(MemPage *pPage, int start, int size){ int end = start + size; u16 *pIdx, idx; FreeBlk *pFBlk; FreeBlk *pNew; FreeBlk *pNext; |
︙ | ︙ | |||
303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 | pPage->isInit = 1; pPage->validUp = 1; pPage->up = pgnoParent; pPage->nCell = 0; idx = pPage->pStart->firstCell; while( idx!=0 ){ if( idx>SQLITE_PAGE_SIZE-sizeof(Cell) ) goto page_format_error; pCell = (Cell*)&pPage->aPage[idx]; pPage->aCell[pPage->nCell++] = pCell; idx = pCell->iNext; } pPage->nFree = 0; idx = pPage->pStart->firstFree; while( idx!=0 ){ if( idx>SQLITE_PAGE_SIZE-sizeof(FreeBlk) ) goto page_format_error; pFBlk = (FreeBlk*)&pPage->aPage[idx]; pPage->nFree += pFBlk->iSize; if( pFBlk->iNext <= idx ) goto page_format_error; idx = pFBlk->iNext; } return SQLITE_OK; page_format_error: return SQLITE_CORRUPT; } /* | > > | > > > > | 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 | pPage->isInit = 1; pPage->validUp = 1; pPage->up = pgnoParent; pPage->nCell = 0; idx = pPage->pStart->firstCell; while( idx!=0 ){ if( idx>SQLITE_PAGE_SIZE-sizeof(Cell) ) goto page_format_error; if( idx<pPage->idxStart + sizeof(PageHeader) ) goto page_format_error; pCell = (Cell*)&pPage->aPage[idx]; pPage->aCell[pPage->nCell++] = pCell; idx = pCell->iNext; } pPage->nFree = 0; idx = pPage->pStart->firstFree; while( idx!=0 ){ if( idx>SQLITE_PAGE_SIZE-sizeof(FreeBlk) ) goto page_format_error; if( idx<pPage->idxStart + sizeof(PageHeader) ) goto page_format_error; pFBlk = (FreeBlk*)&pPage->aPage[idx]; pPage->nFree += pFBlk->iSize; if( pFBlk->iNext <= idx ) goto page_format_error; idx = pFBlk->iNext; } return SQLITE_OK; page_format_error: return SQLITE_CORRUPT; } /* ** Open a new database. ** ** Actually, this routine just sets up the internal data structures ** for accessing the database. We do not actually open the database ** file until the first page is loaded. */ int sqliteBtreeOpen(const char *zFilename, int mode, Btree **ppBtree){ Btree *pBt; pBt = sqliteMalloc( sizeof(*pBt) ); if( pBt==0 ){ **ppBtree = 0; |
︙ | ︙ | |||
357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 | while( pBt->pCursor ){ sqliteBtreeCloseCursor(pBt->pCursor); } sqlitepager_close(pBt->pPager); sqliteFree(pBt); return SQLITE_OK; } /* ** Start a new transaction */ int sqliteBtreeBeginTrans(Btree *pBt){ int rc; if( pBt->inTrans ) return SQLITE_ERROR; if( pBt->page1==0 ){ rc = lockBtree(pBt); if( rc!=SQLITE_OK ) return rc; } rc = sqlitepager_write(pBt->page1); if( rc==SQLITE_OK ){ pBt->inTrans = 1; } return rc; } | > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > < < < < < < < < < < < < < < < < < < < | 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 | while( pBt->pCursor ){ sqliteBtreeCloseCursor(pBt->pCursor); } sqlitepager_close(pBt->pPager); sqliteFree(pBt); return SQLITE_OK; } /* ** Get a reference to page1 of the database file. This will ** also acquire a readlock on that file. ** ** SQLITE_OK is returned on success. If the file is not a ** well-formed database file, then SQLITE_CORRUPT is returned. ** SQLITE_BUSY is returned if the database is locked. SQLITE_NOMEM ** is returned if we run out of memory. SQLITE_PROTOCOL is returned ** if there is a locking protocol violation. */ static int lockBtree(Btree *pBt){ int rc; if( pBt->page1 ) return SQLITE_OK; rc = sqlitepager_get(pBt->pPager, 1, &pBt->page1); if( rc!=SQLITE_OK ) return rc; rc = initPage(pBt->page1, 1, 0); if( rc!=SQLITE_OK ) goto lock_failed; /* Do some checking to help insure the file we opened really is ** a valid database file. */ if( sqlitepager_pagecount(pBt->pPager)>0 ){ Page1Header *pP1 = (Page1Header*)pBt->page1; if( pP1->magic1!=MAGIC_1 || pP1->magic2!=MAGIC_2 ){ rc = SQLITE_CORRUPT; goto lock_failed; } } return rc; lock_failed: sqlitepager_unref(pBt->page1); pBt->page1 = 0; } /* ** Start a new transaction */ int sqliteBtreeBeginTrans(Btree *pBt){ int rc; if( pBt->inTrans ) return SQLITE_ERROR; if( pBt->page1==0 ){ rc = lockBtree(pBt); if( rc!=SQLITE_OK ) return rc; } rc = sqlitepager_write(pBt->page1); if( rc==SQLITE_OK ){ pBt->inTrans = 1; } return rc; } /* ** Remove the last reference to the database file. This will ** remove the read lock. */ static void unlockBtree(Btree *pBt){ if( pBt->pCursor==0 && pBt->page1!=0 ){ sqlitepager_unref(pBt->page1); |
︙ | ︙ |
Changes to src/pager.c.
︙ | ︙ | |||
21 22 23 24 25 26 27 | ** http://www.hwaci.com/drh/ ** ************************************************************************* ** This is the implementation of the page cache subsystem. ** ** The page cache is used to access a database file. The pager journals ** all writes in order to support rollback. Locking is used to limit | | | | 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 | ** http://www.hwaci.com/drh/ ** ************************************************************************* ** This is the implementation of the page cache subsystem. ** ** The page cache is used to access a database file. The pager journals ** all writes in order to support rollback. Locking is used to limit ** access to one or more reader or one writer. ** ** @(#) $Id: pager.c,v 1.6 2001/05/21 13:45:10 drh Exp $ */ #include "sqliteInt.h" #include "pager.h" #include <fcntl.h> #include <sys/stat.h> #include <unistd.h> #include <assert.h> |
︙ | ︙ | |||
52 53 54 55 56 57 58 | ** file at the same time. ** ** SQLITE_WRITELOCK The page cache is writing the database. ** Access is exclusive. No other processes or ** threads can be reading or writing while one ** process is writing. ** | | | | > > > | | | 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 | ** file at the same time. ** ** SQLITE_WRITELOCK The page cache is writing the database. ** Access is exclusive. No other processes or ** threads can be reading or writing while one ** process is writing. ** ** The page cache comes up in SQLITE_UNLOCK. The first time a ** sqlite_page_get() occurs, the state transitions to SQLITE_READLOCK. ** After all pages have been released using sqlite_page_unref(), ** the state transitions back to SQLITE_UNLOCK. The first time ** that sqlite_page_write() is called, the state transitions to ** SQLITE_WRITELOCK. (Note that sqlite_page_write() can only be ** called on an outstanding page which means that the pager must ** be in SQLITE_READLOCK before it transitions to SQLITE_WRITELOCK.) ** The sqlite_page_rollback() and sqlite_page_commit() functions ** transition the state from SQLITE_WRITELOCK back to SQLITE_READLOCK. */ #define SQLITE_UNLOCK 0 #define SQLITE_READLOCK 1 #define SQLITE_WRITELOCK 2 /* |
︙ | ︙ | |||
92 93 94 95 96 97 98 | */ #define PGHDR_TO_DATA(P) ((void*)(&(P)[1])) #define DATA_TO_PGHDR(D) (&((PgHdr*)(D))[-1]) #define PGHDR_TO_EXTRA(P) ((void*)&((char*)(&(P)[1]))[SQLITE_PAGE_SIZE]) /* ** How big to make the hash table used for locating in-memory pages | | | 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 | */ #define PGHDR_TO_DATA(P) ((void*)(&(P)[1])) #define DATA_TO_PGHDR(D) (&((PgHdr*)(D))[-1]) #define PGHDR_TO_EXTRA(P) ((void*)&((char*)(&(P)[1]))[SQLITE_PAGE_SIZE]) /* ** How big to make the hash table used for locating in-memory pages ** by page number. Knuth says this should be a prime number. */ #define N_PG_HASH 101 /* ** A open page cache is an instance of the following structure. */ struct Pager { |
︙ | ︙ | |||
332 333 334 335 336 337 338 | ** Playback the journal and thus restore the database file to ** the state it was in before we started making changes. ** ** The journal file format is as follows: There is an initial ** file-type string for sanity checking. Then there is a single ** Pgno number which is the number of pages in the database before ** changes were made. The database is truncated to this size. | | | > | 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 | ** Playback the journal and thus restore the database file to ** the state it was in before we started making changes. ** ** The journal file format is as follows: There is an initial ** file-type string for sanity checking. Then there is a single ** Pgno number which is the number of pages in the database before ** changes were made. The database is truncated to this size. ** Next come zero or more page records where each page record ** consists of a Pgno and SQLITE_PAGE_SIZE bytes of data. See ** the PageRecord structure for details. ** ** For playback, the pages have to be read from the journal in ** reverse order and put back into the original database file. ** ** If the file opened as the journal file is not a well-formed ** journal file (as determined by looking at the magic number ** at the beginning) then this routine returns SQLITE_PROTOCOL. |
︙ | ︙ | |||
563 564 565 566 567 568 569 570 571 572 573 574 575 576 | } /* ** Acquire a page. ** ** A read lock is obtained for the first page acquired. The lock ** is dropped when the last page is released. ** ** The acquisition might fail for several reasons. In all cases, ** an appropriate error code is returned and *ppPage is set to NULL. ** ** See also sqlitepager_lookup(). Both this routine and _lookup() attempt ** to find a page in the in-memory cache first. If the page is not already ** in cache, this routine goes to disk to read it in whereas _lookup() | > > > > > > | 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 | } /* ** Acquire a page. ** ** A read lock is obtained for the first page acquired. The lock ** is dropped when the last page is released. ** ** A _get works for any page number greater than 0. If the database ** file is smaller than the requested page, then no actual disk ** read occurs and the memory image of the page is initialized to ** all zeros. The extra data appended to a page is always initialized ** to zeros the first time a page is loaded into memory. ** ** The acquisition might fail for several reasons. In all cases, ** an appropriate error code is returned and *ppPage is set to NULL. ** ** See also sqlitepager_lookup(). Both this routine and _lookup() attempt ** to find a page in the in-memory cache first. If the page is not already ** in cache, this routine goes to disk to read it in whereas _lookup() |
︙ | ︙ | |||
723 724 725 726 727 728 729 | h = pager_hash(pgno); pPg->pNextHash = pPager->aHash[h]; pPager->aHash[h] = pPg; if( pPg->pNextHash ){ assert( pPg->pNextHash->pPrevHash==0 ); pPg->pNextHash->pPrevHash = pPg; } | > > > > | | > | | | 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 | h = pager_hash(pgno); pPg->pNextHash = pPager->aHash[h]; pPager->aHash[h] = pPg; if( pPg->pNextHash ){ assert( pPg->pNextHash->pPrevHash==0 ); pPg->pNextHash->pPrevHash = pPg; } if( pPager->dbSize<0 ) sqlitepager_pagecount(pPager); if( pPager->dbSize<pgno ){ memset(PGHDR_TO_DATA(pPg), 0, SQLITE_PAGE_SIZE); }else{ pager_seek(pPager->fd, (pgno-1)*SQLITE_PAGE_SIZE); pager_read(pPager->fd, PGHDR_TO_DATA(pPg), SQLITE_PAGE_SIZE); } if( pPager->nExtra>0 ){ memset(PGHDR_TO_EXTRA(pPg), 0, pPager->nExtra); } }else{ /* The requested page is in the page cache. */ pPager->nHit++; sqlitepager_ref(pPg); } *ppPage = PGHDR_TO_DATA(pPg); return SQLITE_OK; } /* ** Acquire a page if it is already in the in-memory cache. Do ** not read the page from disk. Return a pointer to the page, ** or 0 if the page is not in cache. ** ** See also sqlitepager_get(). The difference between this routine ** and sqlitepager_get() is that _get() will go to the disk and read ** in the page if the page is not already in cache. This routine ** returns NULL if the page is not in cache of if a disk I/O has ever ** happened. */ void *sqlitepager_lookup(Pager *pPager, Pgno pgno){ PgHdr *pPg; /* Make sure we have not hit any critical errors. */ if( pPager==0 || pgno==0 ){ |
︙ | ︙ | |||
820 821 822 823 824 825 826 | ** Mark a data page as writeable. The page is written into the journal ** if it is not there already. This routine must be called before making ** changes to a page. ** ** The first time this routine is called, the pager creates a new ** journal and acquires a write lock on the database. If the write ** lock could not be acquired, this routine returns SQLITE_BUSY. The | | | 835 836 837 838 839 840 841 842 843 844 845 846 847 848 849 | ** Mark a data page as writeable. The page is written into the journal ** if it is not there already. This routine must be called before making ** changes to a page. ** ** The first time this routine is called, the pager creates a new ** journal and acquires a write lock on the database. If the write ** lock could not be acquired, this routine returns SQLITE_BUSY. The ** calling routine must check for that return value and be careful not to ** change any page data until this routine returns SQLITE_OK. ** ** If the journal file could not be written because the disk is full, ** then this routine returns SQLITE_FULL and does an immediate rollback. ** All subsequent write attempts also return SQLITE_FULL until there ** is a call to sqlitepager_commit() or sqlitepager_rollback() to ** reset. |
︙ | ︙ | |||
885 886 887 888 889 890 891 892 893 894 895 896 897 898 | if( rc!=SQLITE_OK ){ sqlitepager_rollback(pPager); pPager->errMask |= PAGER_ERR_FULL; return rc; } } pPg->inJournal = 1; return rc; } /* ** Commit all changes to the database and release the write lock. ** ** If the commit fails for any reason, a rollback attempt is made | > > > | 900 901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 | if( rc!=SQLITE_OK ){ sqlitepager_rollback(pPager); pPager->errMask |= PAGER_ERR_FULL; return rc; } } pPg->inJournal = 1; if( pPager->dbSize<pPg->pgno ){ pPager->dbSize = pPg->pgno; } return rc; } /* ** Commit all changes to the database and release the write lock. ** ** If the commit fails for any reason, a rollback attempt is made |
︙ | ︙ |
Changes to test/pager.test.
︙ | ︙ | |||
19 20 21 22 23 24 25 | # drh@hwaci.com # http://www.hwaci.com/drh/ # #*********************************************************************** # This file implements regression tests for SQLite library. The # focus of this script is page cache subsystem. # | | | 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 | # drh@hwaci.com # http://www.hwaci.com/drh/ # #*********************************************************************** # This file implements regression tests for SQLite library. The # focus of this script is page cache subsystem. # # $Id: pager.test,v 1.4 2001/05/21 13:45:10 drh Exp $ set testdir [file dirname $argv0] source $testdir/tester.tcl if {$dbprefix!="mem:" && [info commands pager_open]!=""} { |
︙ | ︙ | |||
71 72 73 74 75 76 77 | set ::g1 [page_get $::p1 1] } msg] if {$v} {lappend v $msg} set v } {0} do_test pager-2.4 { pager_stats $::p1 | | | 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 | set ::g1 [page_get $::p1 1] } msg] if {$v} {lappend v $msg} set v } {0} do_test pager-2.4 { pager_stats $::p1 } {ref 1 page 1 max 10 size 0 state 1 err 0 hit 0 miss 1 ovfl 0} do_test pager-2.5 { pager_pagecount $::p1 } {0} do_test pager-2.6 { pager_stats $::p1 } {ref 1 page 1 max 10 size 0 state 1 err 0 hit 0 miss 1 ovfl 0} do_test pager-2.7 { |
︙ | ︙ | |||
99 100 101 102 103 104 105 | expr {$::g1!=0} } {1} do_test pager-2.12 { page_number $::g1 } {1} do_test pager-2.13 { pager_stats $::p1 | | | | 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 | expr {$::g1!=0} } {1} do_test pager-2.12 { page_number $::g1 } {1} do_test pager-2.13 { pager_stats $::p1 } {ref 1 page 1 max 10 size 0 state 1 err 0 hit 0 miss 2 ovfl 0} do_test pager-2.14 { set v [catch { page_write $::g1 "Page-One" } msg] lappend v $msg } {0 {}} do_test pager-2.15 { pager_stats $::p1 } {ref 1 page 1 max 10 size 1 state 2 err 0 hit 0 miss 2 ovfl 0} do_test pager-2.16 { page_read $::g1 } {Page-One} do_test pager-2.17 { set v [catch { pager_commit $::p1 } msg] |
︙ | ︙ |
Changes to www/index.tcl.
1 2 3 | # # Run this TCL script to generate HTML for the index.html file. # | | | 1 2 3 4 5 6 7 8 9 10 11 | # # Run this TCL script to generate HTML for the index.html file. # set rcsid {$Id: index.tcl,v 1.37 2001/05/21 13:45:10 drh Exp $} puts {<html> <head><title>SQLite: An SQL Database Library Built Atop GDBM</title></head> <body bgcolor=white> <h1 align=center>SQLite: An SQL Database Library Built Atop <a href="http://www.gnu.org/software/gdbm/gdbm.html">GDBM</a></h1> <p align=center>} |
︙ | ︙ | |||
65 66 67 68 69 70 71 | <ul> <li><p> The LIKE operator is suppose to ignore case. But it only ignores case for 7-bit Latin characters. The case of 8-bit iso8859 characters or UTF-8 characters is signification. Hence, <b>'a' LIKE 'A'</b> returns | | | 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 | <ul> <li><p> The LIKE operator is suppose to ignore case. But it only ignores case for 7-bit Latin characters. The case of 8-bit iso8859 characters or UTF-8 characters is signification. Hence, <b>'a' LIKE 'A'</b> returns TRUE but <b>'æ' LIKE 'Æ'</b> returns FALSE. </p></li> </ul> <p><b>Important Note:</b> Serious bugs have been found in versions 1.0.22 on Unix and 1.0.26 on Windows. Users of these or earlier versions of SQLite should upgrade.</p> |
︙ | ︙ | |||
143 144 145 146 147 148 149 | <p>Instructions for building SQLite for WindowsNT are found <a href="crosscompile.html">here</a>. } puts {<h2>Command-line Usage Example</h2> <p>Download the source archive and compile the <b>sqlite</b> | | | 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 | <p>Instructions for building SQLite for WindowsNT are found <a href="crosscompile.html">here</a>. } puts {<h2>Command-line Usage Example</h2> <p>Download the source archive and compile the <b>sqlite</b> program as described above. Then type:</p> <blockquote><pre> bash$ sqlite ~/newdb <i>Directory ~/newdb created automatically</i> sqlite> create table t1( ...> a int, ...> b varchar(20) ...> c text |
︙ | ︙ |
Changes to www/lang.tcl.
1 2 3 | # # Run this Tcl script to generate the sqlite.html file. # | | | 1 2 3 4 5 6 7 8 9 10 11 | # # Run this Tcl script to generate the sqlite.html file. # set rcsid {$Id: lang.tcl,v 1.8 2001/05/21 13:45:10 drh Exp $} puts {<html> <head> <title>Query Language Understood By SQLite</title> </head> <body bgcolor=white> <h1 align=center> |
︙ | ︙ | |||
380 381 382 383 384 385 386 | puts { <p>The LIKE operator does a wildcard comparision. The operand to the right contains the wildcards.} puts "A percent symbol [Operator %] in the right operand matches any sequence of zero or more characters on the left. An underscore [Operator _] on the right | | > | > > > > > | | | 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 | puts { <p>The LIKE operator does a wildcard comparision. The operand to the right contains the wildcards.} puts "A percent symbol [Operator %] in the right operand matches any sequence of zero or more characters on the left. An underscore [Operator _] on the right matches any single character on the left." puts {The LIKE operator is not case sensitive and will match upper case characters on one side against lower case characters on the other. (A bug: SQLite only understands upper/lower case for 7-bit Latin characters. Hence the LIKE operator is case sensitive for 8-bit iso8859 characters or UTF-8 characters. For example, the expression <b>'a' LIKE 'A'</b> is TRUE but <b>'æ' LIKE 'Æ'</b> is FALSE.) </p> <p>The GLOB operator is similar to LIKE but uses the Unix file globbing syntax for its wildcards. Also, GLOB is case sensitive, unlike LIKE. Both GLOB and LIKE may be preceded by the NOT keyword to invert the sense of the test.</p> <p>A column name can be any of the names defined in the CREATE TABLE statement or one of the following special identifiers: "<b>ROWID</b>", "<b>OID</b>", or "<b>_ROWID_</b>". These special identifiers all describe the unique random integer key (the "row key") associated with every row of every table. The special identifiers only refer to the row key if the CREATE TABLE statement does not define a real column with the same name. Row keys act like read-only columns. A row key can be used anywhere a regular column can be used, except that you cannot change the value of a row key in an UPDATE or INSERT statement. "SELECT * ..." does not return the row key.</p> |
︙ | ︙ |