Add //third_party/sqlite to dirs_to_snapshot, remove net_sql.patch

third_party/sqlite/sqlite-src-3080704/src/test_onefile.c

Index: third_party/sqlite/sqlite-src-3080704/src/test_onefile.c
diff --git a/third_party/sqlite/sqlite-src-3080704/src/test_onefile.c b/third_party/sqlite/sqlite-src-3080704/src/test_onefile.c
new file mode 100644
index 0000000000000000000000000000000000000000..69867441b8cc9a621af32413fc63da80993764f4
--- /dev/null
+++ b/third_party/sqlite/sqlite-src-3080704/src/test_onefile.c
@@ -0,0 +1,830 @@
+/*
+** 2007 September 14
+**
+** The author disclaims copyright to this source code.  In place of
+** a legal notice, here is a blessing:
+**
+**    May you do good and not evil.
+**    May you find forgiveness for yourself and forgive others.
+**    May you share freely, never taking more than you give.
+**
+*************************************************************************
+**
+** OVERVIEW:
+**
+**   This file contains some example code demonstrating how the SQLite 
+**   vfs feature can be used to have SQLite operate directly on an 
+**   embedded media, without using an intermediate file system.
+**
+**   Because this is only a demo designed to run on a workstation, the
+**   underlying media is simulated using a regular file-system file. The
+**   size of the file is fixed when it is first created (default size 10 MB).
+**   From SQLite's point of view, this space is used to store a single
+**   database file and the journal file. 
+**
+**   Any statement journal created is stored in volatile memory obtained 
+**   from sqlite3_malloc(). Any attempt to create a temporary database file 
+**   will fail (SQLITE_IOERR). To prevent SQLite from attempting this,
+**   it should be configured to store all temporary database files in 
+**   main memory (see pragma "temp_store" or the SQLITE_TEMP_STORE compile 
+**   time option).
+**
+** ASSUMPTIONS:
+**
+**   After it has been created, the blob file is accessed using the
+**   following three functions only:
+**
+**       mediaRead();            - Read a 512 byte block from the file.
+**       mediaWrite();           - Write a 512 byte block to the file.
+**       mediaSync();            - Tell the media hardware to sync.
+**
+**   It is assumed that these can be easily implemented by any "real"
+**   media vfs driver adapting this code.
+**
+** FILE FORMAT:
+**
+**   The basic principle is that the "database file" is stored at the
+**   beginning of the 10 MB blob and grows in a forward direction. The 
+**   "journal file" is stored at the end of the 10MB blob and grows
+**   in the reverse direction. If, during a transaction, insufficient
+**   space is available to expand either the journal or database file,
+**   an SQLITE_FULL error is returned. The database file is never allowed
+**   to consume more than 90% of the blob space. If SQLite tries to
+**   create a file larger than this, SQLITE_FULL is returned.
+**
+**   No allowance is made for "wear-leveling", as is required by.
+**   embedded devices in the absence of equivalent hardware features.
+**
+**   The first 512 block byte of the file is reserved for storing the
+**   size of the "database file". It is updated as part of the sync()
+**   operation. On startup, it can only be trusted if no journal file
+**   exists. If a journal-file does exist, then it stores the real size
+**   of the database region. The second and subsequent blocks store the 
+**   actual database content.
+**
+**   The size of the "journal file" is not stored persistently in the 
+**   file. When the system is running, the size of the journal file is
+**   stored in volatile memory. When recovering from a crash, this vfs
+**   reports a very large size for the journal file. The normal journal
+**   header and checksum mechanisms serve to prevent SQLite from 
+**   processing any data that lies past the logical end of the journal.
+**
+**   When SQLite calls OsDelete() to delete the journal file, the final
+**   512 bytes of the blob (the area containing the first journal header)
+**   are zeroed.
+**
+** LOCKING:
+**
+**   File locking is a no-op. Only one connection may be open at any one
+**   time using this demo vfs.
+*/
+
+#include "sqlite3.h"
+#include <assert.h>
+#include <string.h>
+
+/*
+** Maximum pathname length supported by the fs backend.
+*/
+#define BLOCKSIZE 512
+#define BLOBSIZE 10485760
+
+/*
+** Name used to identify this VFS.
+*/
+#define FS_VFS_NAME "fs"
+
+typedef struct fs_real_file fs_real_file;
+struct fs_real_file {
+  sqlite3_file *pFile;
+  const char *zName;
+  int nDatabase;              /* Current size of database region */
+  int nJournal;               /* Current size of journal region */
+  int nBlob;                  /* Total size of allocated blob */
+  int nRef;                   /* Number of pointers to this structure */
+  fs_real_file *pNext;
+  fs_real_file **ppThis;
+};
+
+typedef struct fs_file fs_file;
+struct fs_file {
+  sqlite3_file base;
+  int eType;
+  fs_real_file *pReal;
+};
+
+typedef struct tmp_file tmp_file;
+struct tmp_file {
+  sqlite3_file base;
+  int nSize;
+  int nAlloc;
+  char *zAlloc;
+};
+
+/* Values for fs_file.eType. */
+#define DATABASE_FILE   1
+#define JOURNAL_FILE    2
+
+/*
+** Method declarations for fs_file.
+*/
+static int fsClose(sqlite3_file*);
+static int fsRead(sqlite3_file*, void*, int iAmt, sqlite3_int64 iOfst);
+static int fsWrite(sqlite3_file*, const void*, int iAmt, sqlite3_int64 iOfst);
+static int fsTruncate(sqlite3_file*, sqlite3_int64 size);
+static int fsSync(sqlite3_file*, int flags);
+static int fsFileSize(sqlite3_file*, sqlite3_int64 *pSize);
+static int fsLock(sqlite3_file*, int);
+static int fsUnlock(sqlite3_file*, int);
+static int fsCheckReservedLock(sqlite3_file*, int *pResOut);
+static int fsFileControl(sqlite3_file*, int op, void *pArg);
+static int fsSectorSize(sqlite3_file*);
+static int fsDeviceCharacteristics(sqlite3_file*);
+
+/*
+** Method declarations for tmp_file.
+*/
+static int tmpClose(sqlite3_file*);
+static int tmpRead(sqlite3_file*, void*, int iAmt, sqlite3_int64 iOfst);
+static int tmpWrite(sqlite3_file*, const void*, int iAmt, sqlite3_int64 iOfst);
+static int tmpTruncate(sqlite3_file*, sqlite3_int64 size);
+static int tmpSync(sqlite3_file*, int flags);
+static int tmpFileSize(sqlite3_file*, sqlite3_int64 *pSize);
+static int tmpLock(sqlite3_file*, int);
+static int tmpUnlock(sqlite3_file*, int);
+static int tmpCheckReservedLock(sqlite3_file*, int *pResOut);
+static int tmpFileControl(sqlite3_file*, int op, void *pArg);
+static int tmpSectorSize(sqlite3_file*);
+static int tmpDeviceCharacteristics(sqlite3_file*);
+
+/*
+** Method declarations for fs_vfs.
+*/
+static int fsOpen(sqlite3_vfs*, const char *, sqlite3_file*, int , int *);
+static int fsDelete(sqlite3_vfs*, const char *zName, int syncDir);
+static int fsAccess(sqlite3_vfs*, const char *zName, int flags, int *);
+static int fsFullPathname(sqlite3_vfs*, const char *zName, int nOut,char *zOut);
+static void *fsDlOpen(sqlite3_vfs*, const char *zFilename);
+static void fsDlError(sqlite3_vfs*, int nByte, char *zErrMsg);
+static void (*fsDlSym(sqlite3_vfs*,void*, const char *zSymbol))(void);
+static void fsDlClose(sqlite3_vfs*, void*);
+static int fsRandomness(sqlite3_vfs*, int nByte, char *zOut);
+static int fsSleep(sqlite3_vfs*, int microseconds);
+static int fsCurrentTime(sqlite3_vfs*, double*);
+
+
+typedef struct fs_vfs_t fs_vfs_t;
+struct fs_vfs_t {
+  sqlite3_vfs base;
+  fs_real_file *pFileList;
+  sqlite3_vfs *pParent;
+};
+
+static fs_vfs_t fs_vfs = {
+  {
+    1,                                          /* iVersion */
+    0,                                          /* szOsFile */
+    0,                                          /* mxPathname */
+    0,                                          /* pNext */
+    FS_VFS_NAME,                                /* zName */
+    0,                                          /* pAppData */
+    fsOpen,                                     /* xOpen */
+    fsDelete,                                   /* xDelete */
+    fsAccess,                                   /* xAccess */
+    fsFullPathname,                             /* xFullPathname */
+    fsDlOpen,                                   /* xDlOpen */
+    fsDlError,                                  /* xDlError */
+    fsDlSym,                                    /* xDlSym */
+    fsDlClose,                                  /* xDlClose */
+    fsRandomness,                               /* xRandomness */
+    fsSleep,                                    /* xSleep */
+    fsCurrentTime,                              /* xCurrentTime */
+    0                                           /* xCurrentTimeInt64 */
+  }, 
+  0,                                            /* pFileList */
+  0                                             /* pParent */
+};
+
+static sqlite3_io_methods fs_io_methods = {
+  1,                            /* iVersion */
+  fsClose,                      /* xClose */
+  fsRead,                       /* xRead */
+  fsWrite,                      /* xWrite */
+  fsTruncate,                   /* xTruncate */
+  fsSync,                       /* xSync */
+  fsFileSize,                   /* xFileSize */
+  fsLock,                       /* xLock */
+  fsUnlock,                     /* xUnlock */
+  fsCheckReservedLock,          /* xCheckReservedLock */
+  fsFileControl,                /* xFileControl */
+  fsSectorSize,                 /* xSectorSize */
+  fsDeviceCharacteristics,      /* xDeviceCharacteristics */
+  0,                            /* xShmMap */
+  0,                            /* xShmLock */
+  0,                            /* xShmBarrier */
+  0                             /* xShmUnmap */
+};
+
+
+static sqlite3_io_methods tmp_io_methods = {
+  1,                            /* iVersion */
+  tmpClose,                     /* xClose */
+  tmpRead,                      /* xRead */
+  tmpWrite,                     /* xWrite */
+  tmpTruncate,                  /* xTruncate */
+  tmpSync,                      /* xSync */
+  tmpFileSize,                  /* xFileSize */
+  tmpLock,                      /* xLock */
+  tmpUnlock,                    /* xUnlock */
+  tmpCheckReservedLock,         /* xCheckReservedLock */
+  tmpFileControl,               /* xFileControl */
+  tmpSectorSize,                /* xSectorSize */
+  tmpDeviceCharacteristics,     /* xDeviceCharacteristics */
+  0,                            /* xShmMap */
+  0,                            /* xShmLock */
+  0,                            /* xShmBarrier */
+  0                             /* xShmUnmap */
+};
+
+/* Useful macros used in several places */
+#define MIN(x,y) ((x)<(y)?(x):(y))
+#define MAX(x,y) ((x)>(y)?(x):(y))
+
+
+/*
+** Close a tmp-file.
+*/
+static int tmpClose(sqlite3_file *pFile){
+  tmp_file *pTmp = (tmp_file *)pFile;
+  sqlite3_free(pTmp->zAlloc);
+  return SQLITE_OK;
+}
+
+/*
+** Read data from a tmp-file.
+*/
+static int tmpRead(
+  sqlite3_file *pFile, 
+  void *zBuf, 
+  int iAmt, 
+  sqlite_int64 iOfst
+){
+  tmp_file *pTmp = (tmp_file *)pFile;
+  if( (iAmt+iOfst)>pTmp->nSize ){
+    return SQLITE_IOERR_SHORT_READ;
+  }
+  memcpy(zBuf, &pTmp->zAlloc[iOfst], iAmt);
+  return SQLITE_OK;
+}
+
+/*
+** Write data to a tmp-file.
+*/
+static int tmpWrite(
+  sqlite3_file *pFile, 
+  const void *zBuf, 
+  int iAmt, 
+  sqlite_int64 iOfst
+){
+  tmp_file *pTmp = (tmp_file *)pFile;
+  if( (iAmt+iOfst)>pTmp->nAlloc ){
+    int nNew = (int)(2*(iAmt+iOfst+pTmp->nAlloc));
+    char *zNew = sqlite3_realloc(pTmp->zAlloc, nNew);
+    if( !zNew ){
+      return SQLITE_NOMEM;
+    }
+    pTmp->zAlloc = zNew;
+    pTmp->nAlloc = nNew;
+  }
+  memcpy(&pTmp->zAlloc[iOfst], zBuf, iAmt);
+  pTmp->nSize = (int)MAX(pTmp->nSize, iOfst+iAmt);
+  return SQLITE_OK;
+}
+
+/*
+** Truncate a tmp-file.
+*/
+static int tmpTruncate(sqlite3_file *pFile, sqlite_int64 size){
+  tmp_file *pTmp = (tmp_file *)pFile;
+  pTmp->nSize = (int)MIN(pTmp->nSize, size);
+  return SQLITE_OK;
+}
+
+/*
+** Sync a tmp-file.
+*/
+static int tmpSync(sqlite3_file *pFile, int flags){
+  return SQLITE_OK;
+}
+
+/*
+** Return the current file-size of a tmp-file.
+*/
+static int tmpFileSize(sqlite3_file *pFile, sqlite_int64 *pSize){
+  tmp_file *pTmp = (tmp_file *)pFile;
+  *pSize = pTmp->nSize;
+  return SQLITE_OK;
+}
+
+/*
+** Lock a tmp-file.
+*/
+static int tmpLock(sqlite3_file *pFile, int eLock){
+  return SQLITE_OK;
+}
+
+/*
+** Unlock a tmp-file.
+*/
+static int tmpUnlock(sqlite3_file *pFile, int eLock){
+  return SQLITE_OK;
+}
+
+/*
+** Check if another file-handle holds a RESERVED lock on a tmp-file.
+*/
+static int tmpCheckReservedLock(sqlite3_file *pFile, int *pResOut){
+  *pResOut = 0;
+  return SQLITE_OK;
+}
+
+/*
+** File control method. For custom operations on a tmp-file.
+*/
+static int tmpFileControl(sqlite3_file *pFile, int op, void *pArg){
+  return SQLITE_OK;
+}
+
+/*
+** Return the sector-size in bytes for a tmp-file.
+*/
+static int tmpSectorSize(sqlite3_file *pFile){
+  return 0;
+}
+
+/*
+** Return the device characteristic flags supported by a tmp-file.
+*/
+static int tmpDeviceCharacteristics(sqlite3_file *pFile){
+  return 0;
+}
+
+/*
+** Close an fs-file.
+*/
+static int fsClose(sqlite3_file *pFile){
+  int rc = SQLITE_OK;
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+
+  /* Decrement the real_file ref-count. */
+  pReal->nRef--;
+  assert(pReal->nRef>=0);
+
+  /* When the ref-count reaches 0, destroy the structure */
+  if( pReal->nRef==0 ){
+    *pReal->ppThis = pReal->pNext;
+    if( pReal->pNext ){
+      pReal->pNext->ppThis = pReal->ppThis;
+    }
+    rc = pReal->pFile->pMethods->xClose(pReal->pFile);
+    sqlite3_free(pReal);
+  }
+
+  return rc;
+}
+
+/*
+** Read data from an fs-file.
+*/
+static int fsRead(
+  sqlite3_file *pFile, 
+  void *zBuf, 
+  int iAmt, 
+  sqlite_int64 iOfst
+){
+  int rc = SQLITE_OK;
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+  sqlite3_file *pF = pReal->pFile;
+
+  if( (p->eType==DATABASE_FILE && (iAmt+iOfst)>pReal->nDatabase)
+   || (p->eType==JOURNAL_FILE && (iAmt+iOfst)>pReal->nJournal)
+  ){
+    rc = SQLITE_IOERR_SHORT_READ;
+  }else if( p->eType==DATABASE_FILE ){
+    rc = pF->pMethods->xRead(pF, zBuf, iAmt, iOfst+BLOCKSIZE);
+  }else{
+    /* Journal file. */
+    int iRem = iAmt;
+    int iBuf = 0;
+    int ii = (int)iOfst;
+    while( iRem>0 && rc==SQLITE_OK ){
+      int iRealOff = pReal->nBlob - BLOCKSIZE*((ii/BLOCKSIZE)+1) + ii%BLOCKSIZE;
+      int iRealAmt = MIN(iRem, BLOCKSIZE - (iRealOff%BLOCKSIZE));
+
+      rc = pF->pMethods->xRead(pF, &((char *)zBuf)[iBuf], iRealAmt, iRealOff);
+      ii += iRealAmt;
+      iBuf += iRealAmt;
+      iRem -= iRealAmt;
+    }
+  }
+
+  return rc;
+}
+
+/*
+** Write data to an fs-file.
+*/
+static int fsWrite(
+  sqlite3_file *pFile, 
+  const void *zBuf, 
+  int iAmt, 
+  sqlite_int64 iOfst
+){
+  int rc = SQLITE_OK;
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+  sqlite3_file *pF = pReal->pFile;
+
+  if( p->eType==DATABASE_FILE ){
+    if( (iAmt+iOfst+BLOCKSIZE)>(pReal->nBlob-pReal->nJournal) ){
+      rc = SQLITE_FULL;
+    }else{
+      rc = pF->pMethods->xWrite(pF, zBuf, iAmt, iOfst+BLOCKSIZE);
+      if( rc==SQLITE_OK ){
+        pReal->nDatabase = (int)MAX(pReal->nDatabase, iAmt+iOfst);
+      }
+    }
+  }else{
+    /* Journal file. */
+    int iRem = iAmt;
+    int iBuf = 0;
+    int ii = (int)iOfst;
+    while( iRem>0 && rc==SQLITE_OK ){
+      int iRealOff = pReal->nBlob - BLOCKSIZE*((ii/BLOCKSIZE)+1) + ii%BLOCKSIZE;
+      int iRealAmt = MIN(iRem, BLOCKSIZE - (iRealOff%BLOCKSIZE));
+
+      if( iRealOff<(pReal->nDatabase+BLOCKSIZE) ){
+        rc = SQLITE_FULL;
+      }else{
+        rc = pF->pMethods->xWrite(pF, &((char *)zBuf)[iBuf], iRealAmt,iRealOff);
+        ii += iRealAmt;
+        iBuf += iRealAmt;
+        iRem -= iRealAmt;
+      }
+    }
+    if( rc==SQLITE_OK ){
+      pReal->nJournal = (int)MAX(pReal->nJournal, iAmt+iOfst);
+    }
+  }
+
+  return rc;
+}
+
+/*
+** Truncate an fs-file.
+*/
+static int fsTruncate(sqlite3_file *pFile, sqlite_int64 size){
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+  if( p->eType==DATABASE_FILE ){
+    pReal->nDatabase = (int)MIN(pReal->nDatabase, size);
+  }else{
+    pReal->nJournal = (int)MIN(pReal->nJournal, size);
+  }
+  return SQLITE_OK;
+}
+
+/*
+** Sync an fs-file.
+*/
+static int fsSync(sqlite3_file *pFile, int flags){
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+  sqlite3_file *pRealFile = pReal->pFile;
+  int rc = SQLITE_OK;
+
+  if( p->eType==DATABASE_FILE ){
+    unsigned char zSize[4];
+    zSize[0] = (pReal->nDatabase&0xFF000000)>>24;
+    zSize[1] = (pReal->nDatabase&0x00FF0000)>>16;
+    zSize[2] = (pReal->nDatabase&0x0000FF00)>>8;
+    zSize[3] = (pReal->nDatabase&0x000000FF);
+    rc = pRealFile->pMethods->xWrite(pRealFile, zSize, 4, 0);
+  }
+  if( rc==SQLITE_OK ){
+    rc = pRealFile->pMethods->xSync(pRealFile, flags&(~SQLITE_SYNC_DATAONLY));
+  }
+
+  return rc;
+}
+
+/*
+** Return the current file-size of an fs-file.
+*/
+static int fsFileSize(sqlite3_file *pFile, sqlite_int64 *pSize){
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = p->pReal;
+  if( p->eType==DATABASE_FILE ){
+    *pSize = pReal->nDatabase;
+  }else{
+    *pSize = pReal->nJournal;
+  }
+  return SQLITE_OK;
+}
+
+/*
+** Lock an fs-file.
+*/
+static int fsLock(sqlite3_file *pFile, int eLock){
+  return SQLITE_OK;
+}
+
+/*
+** Unlock an fs-file.
+*/
+static int fsUnlock(sqlite3_file *pFile, int eLock){
+  return SQLITE_OK;
+}
+
+/*
+** Check if another file-handle holds a RESERVED lock on an fs-file.
+*/
+static int fsCheckReservedLock(sqlite3_file *pFile, int *pResOut){
+  *pResOut = 0;
+  return SQLITE_OK;
+}
+
+/*
+** File control method. For custom operations on an fs-file.
+*/
+static int fsFileControl(sqlite3_file *pFile, int op, void *pArg){
+  return SQLITE_OK;
+}
+
+/*
+** Return the sector-size in bytes for an fs-file.
+*/
+static int fsSectorSize(sqlite3_file *pFile){
+  return BLOCKSIZE;
+}
+
+/*
+** Return the device characteristic flags supported by an fs-file.
+*/
+static int fsDeviceCharacteristics(sqlite3_file *pFile){
+  return 0;
+}
+
+/*
+** Open an fs file handle.
+*/
+static int fsOpen(
+  sqlite3_vfs *pVfs,
+  const char *zName,
+  sqlite3_file *pFile,
+  int flags,
+  int *pOutFlags
+){
+  fs_vfs_t *pFsVfs = (fs_vfs_t *)pVfs;
+  fs_file *p = (fs_file *)pFile;
+  fs_real_file *pReal = 0;
+  int eType;
+  int nName;
+  int rc = SQLITE_OK;
+
+  if( 0==(flags&(SQLITE_OPEN_MAIN_DB|SQLITE_OPEN_MAIN_JOURNAL)) ){
+    tmp_file *p = (tmp_file *)pFile;
+    memset(p, 0, sizeof(*p));
+    p->base.pMethods = &tmp_io_methods;
+    return SQLITE_OK;
+  }
+
+  eType = ((flags&(SQLITE_OPEN_MAIN_DB))?DATABASE_FILE:JOURNAL_FILE);
+  p->base.pMethods = &fs_io_methods;
+  p->eType = eType;
+
+  assert(strlen("-journal")==8);
+  nName = (int)strlen(zName)-((eType==JOURNAL_FILE)?8:0);
+  pReal=pFsVfs->pFileList; 
+  for(; pReal && strncmp(pReal->zName, zName, nName); pReal=pReal->pNext);
+
+  if( !pReal ){
+    int real_flags = (flags&~(SQLITE_OPEN_MAIN_DB))|SQLITE_OPEN_TEMP_DB;
+    sqlite3_int64 size;
+    sqlite3_file *pRealFile;
+    sqlite3_vfs *pParent = pFsVfs->pParent;
+    assert(eType==DATABASE_FILE);
+
+    pReal = (fs_real_file *)sqlite3_malloc(sizeof(*pReal)+pParent->szOsFile);
+    if( !pReal ){
+      rc = SQLITE_NOMEM;
+      goto open_out;
+    }
+    memset(pReal, 0, sizeof(*pReal)+pParent->szOsFile);
+    pReal->zName = zName;
+    pReal->pFile = (sqlite3_file *)(&pReal[1]);
+
+    rc = pParent->xOpen(pParent, zName, pReal->pFile, real_flags, pOutFlags);
+    if( rc!=SQLITE_OK ){
+      goto open_out;
+    }
+    pRealFile = pReal->pFile;
+
+    rc = pRealFile->pMethods->xFileSize(pRealFile, &size);
+    if( rc!=SQLITE_OK ){
+      goto open_out;
+    }
+    if( size==0 ){
+      rc = pRealFile->pMethods->xWrite(pRealFile, "\0", 1, BLOBSIZE-1);
+      pReal->nBlob = BLOBSIZE;
+    }else{
+      unsigned char zS[4];
+      pReal->nBlob = (int)size;
+      rc = pRealFile->pMethods->xRead(pRealFile, zS, 4, 0);
+      pReal->nDatabase = (zS[0]<<24)+(zS[1]<<16)+(zS[2]<<8)+zS[3];
+      if( rc==SQLITE_OK ){
+        rc = pRealFile->pMethods->xRead(pRealFile, zS, 4, pReal->nBlob-4);
+        if( zS[0] || zS[1] || zS[2] || zS[3] ){
+          pReal->nJournal = pReal->nBlob;
+        }
+      }
+    }
+
+    if( rc==SQLITE_OK ){
+      pReal->pNext = pFsVfs->pFileList;
+      if( pReal->pNext ){
+        pReal->pNext->ppThis = &pReal->pNext;
+      }
+      pReal->ppThis = &pFsVfs->pFileList;
+      pFsVfs->pFileList = pReal;
+    }
+  }
+
+open_out:
+  if( pReal ){
+    if( rc==SQLITE_OK ){
+      p->pReal = pReal;
+      pReal->nRef++;
+    }else{
+      if( pReal->pFile->pMethods ){
+        pReal->pFile->pMethods->xClose(pReal->pFile);
+      }
+      sqlite3_free(pReal);
+    }
+  }
+  return rc;
+}
+
+/*
+** Delete the file located at zPath. If the dirSync argument is true,
+** ensure the file-system modifications are synced to disk before
+** returning.
+*/
+static int fsDelete(sqlite3_vfs *pVfs, const char *zPath, int dirSync){
+  int rc = SQLITE_OK;
+  fs_vfs_t *pFsVfs = (fs_vfs_t *)pVfs;
+  fs_real_file *pReal;
+  sqlite3_file *pF;
+  int nName = (int)strlen(zPath) - 8;
+
+  assert(strlen("-journal")==8);
+  assert(strcmp("-journal", &zPath[nName])==0);
+
+  pReal = pFsVfs->pFileList; 
+  for(; pReal && strncmp(pReal->zName, zPath, nName); pReal=pReal->pNext);
+  if( pReal ){
+    pF = pReal->pFile;
+    rc = pF->pMethods->xWrite(pF, "\0\0\0\0", 4, pReal->nBlob-BLOCKSIZE);
+    if( rc==SQLITE_OK ){
+      pReal->nJournal = 0;
+    }
+  }
+  return rc;
+}
+
+/*
+** Test for access permissions. Return true if the requested permission
+** is available, or false otherwise.
+*/
+static int fsAccess(
+  sqlite3_vfs *pVfs, 
+  const char *zPath, 
+  int flags, 
+  int *pResOut
+){
+  fs_vfs_t *pFsVfs = (fs_vfs_t *)pVfs;
+  fs_real_file *pReal;
+  int isJournal = 0;
+  int nName = (int)strlen(zPath);
+
+  if( flags!=SQLITE_ACCESS_EXISTS ){
+    sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+    return pParent->xAccess(pParent, zPath, flags, pResOut);
+  }
+
+  assert(strlen("-journal")==8);
+  if( nName>8 && strcmp("-journal", &zPath[nName-8])==0 ){
+    nName -= 8;
+    isJournal = 1;
+  }
+
+  pReal = pFsVfs->pFileList; 
+  for(; pReal && strncmp(pReal->zName, zPath, nName); pReal=pReal->pNext);
+
+  *pResOut = (pReal && (!isJournal || pReal->nJournal>0));
+  return SQLITE_OK;
+}
+
+/*
+** Populate buffer zOut with the full canonical pathname corresponding
+** to the pathname in zPath. zOut is guaranteed to point to a buffer
+** of at least (FS_MAX_PATHNAME+1) bytes.
+*/
+static int fsFullPathname(
+  sqlite3_vfs *pVfs,            /* Pointer to vfs object */
+  const char *zPath,            /* Possibly relative input path */
+  int nOut,                     /* Size of output buffer in bytes */
+  char *zOut                    /* Output buffer */
+){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xFullPathname(pParent, zPath, nOut, zOut);
+}
+
+/*
+** Open the dynamic library located at zPath and return a handle.
+*/
+static void *fsDlOpen(sqlite3_vfs *pVfs, const char *zPath){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xDlOpen(pParent, zPath);
+}
+
+/*
+** Populate the buffer zErrMsg (size nByte bytes) with a human readable
+** utf-8 string describing the most recent error encountered associated 
+** with dynamic libraries.
+*/
+static void fsDlError(sqlite3_vfs *pVfs, int nByte, char *zErrMsg){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  pParent->xDlError(pParent, nByte, zErrMsg);
+}
+
+/*
+** Return a pointer to the symbol zSymbol in the dynamic library pHandle.
+*/
+static void (*fsDlSym(sqlite3_vfs *pVfs, void *pH, const char *zSym))(void){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xDlSym(pParent, pH, zSym);
+}
+
+/*
+** Close the dynamic library handle pHandle.
+*/
+static void fsDlClose(sqlite3_vfs *pVfs, void *pHandle){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  pParent->xDlClose(pParent, pHandle);
+}
+
+/*
+** Populate the buffer pointed to by zBufOut with nByte bytes of 
+** random data.
+*/
+static int fsRandomness(sqlite3_vfs *pVfs, int nByte, char *zBufOut){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xRandomness(pParent, nByte, zBufOut);
+}
+
+/*
+** Sleep for nMicro microseconds. Return the number of microseconds 
+** actually slept.
+*/
+static int fsSleep(sqlite3_vfs *pVfs, int nMicro){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xSleep(pParent, nMicro);
+}
+
+/*
+** Return the current time as a Julian Day number in *pTimeOut.
+*/
+static int fsCurrentTime(sqlite3_vfs *pVfs, double *pTimeOut){
+  sqlite3_vfs *pParent = ((fs_vfs_t *)pVfs)->pParent;
+  return pParent->xCurrentTime(pParent, pTimeOut);
+}
+
+/*
+** This procedure registers the fs vfs with SQLite. If the argument is
+** true, the fs vfs becomes the new default vfs. It is the only publicly
+** available function in this file.
+*/
+int fs_register(void){
+  if( fs_vfs.pParent ) return SQLITE_OK;
+  fs_vfs.pParent = sqlite3_vfs_find(0);
+  fs_vfs.base.mxPathname = fs_vfs.pParent->mxPathname;
+  fs_vfs.base.szOsFile = MAX(sizeof(tmp_file), sizeof(fs_file));
+  return sqlite3_vfs_register(&fs_vfs.base, 0);
+}
+
+#ifdef SQLITE_TEST
+  int SqlitetestOnefile_Init() {return fs_register();}
+#endif