SQLite

# Statically linked extensions
#
TESTSRC += \
  $(TOP)/ext/misc/amatch.c \
  $(TOP)/ext/misc/closure.c \
  $(TOP)/ext/misc/fuzzer.c \
  $(TOP)/ext/misc/ieee754.c \
  $(TOP)/ext/misc/nextchar.c \
  $(TOP)/ext/misc/regexp.c \
  $(TOP)/ext/misc/spellfix.c \
  $(TOP)/ext/misc/wholenumber.c

# Source code to the library files needed by the test fixture
#
TESTSRC2 = \

# Statically linked extensions
#
TESTEXT = \
  $(TOP)\ext\misc\amatch.c \
  $(TOP)\ext\misc\closure.c \
  $(TOP)\ext\misc\fuzzer.c \
  $(TOP)\ext\misc\ieee754.c \
  $(TOP)\ext\misc\nextchar.c \
  $(TOP)\ext\misc\regexp.c \
  $(TOP)\ext\misc\spellfix.c \
  $(TOP)\ext\misc\wholenumber.c


# Source code to the library files needed by the test fixture
#

/*
** 2013-02-28
**
** 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.
**
******************************************************************************
**
** This file contains code to implement the next_char(A,T,F,W) SQL function.
**
** The next_char(A,T,F,H) function finds all valid "next" characters for
** string A given the vocabulary in T.F.  The T.F field should be indexed.
** If the W value exists and is a non-empty string, then it is an SQL
** expression that limits the entries in T.F that will be considered.
**
** For example, suppose an application has a dictionary like this:
**
**   CREATE TABLE dictionary(word TEXT UNIQUE);
**
** Further suppose that for user keypad entry, it is desired to disable
** (gray out) keys that are not valid as the next character.  If the
** the user has previously entered (say) 'cha' then to find all allowed
** next characters (and thereby determine when keys should not be grayed
** out) run the following query:
**
**   SELECT next_char('cha','dictionary','word');
*/
#include "sqlite3ext.h"
SQLITE_EXTENSION_INIT1
#include <string.h>

/*
** A structure to hold context of the next_char() computation across
** nested function calls.
*/
typedef struct nextCharContext nextCharContext;
struct nextCharContext {
  sqlite3 *db;                      /* Database connection */
  sqlite3_stmt *pStmt;              /* Prepared statement used to query */
  const unsigned char *zPrefix;     /* Prefix to scan */
  int nPrefix;                      /* Size of zPrefix in bytes */
  int nAlloc;                       /* Space allocated to aResult */
  int nUsed;                        /* Space used in aResult */
  unsigned int *aResult;            /* Array of next characters */
  int mallocFailed;                 /* True if malloc fails */
  int otherError;                   /* True for any other failure */
};

/*
** Append a result character if the character is not already in the
** result.
*/
static void nextCharAppend(nextCharContext *p, unsigned c){
  int i;
  for(i=0; i<p->nUsed; i++){
    if( p->aResult[i]==c ) return;
  }
  if( p->nUsed+1 > p->nAlloc ){
    unsigned int *aNew;
    int n = p->nAlloc*2 + 30;
    aNew = sqlite3_realloc(p->aResult, n*sizeof(unsigned int));
    if( aNew==0 ){
      p->mallocFailed = 1;
      return;
    }else{
      p->aResult = aNew;
      p->nAlloc = n;
    }
  }
  p->aResult[p->nUsed++] = c;
}

/*
** Write a character into z[] as UTF8.  Return the number of bytes needed
** to hold the character
*/
static int writeUtf8(unsigned char *z, unsigned c){
  if( c<0x00080 ){
    z[0] = (unsigned char)(c&0xff);
    return 1;
  }
  if( c<0x00800 ){
    z[0] = 0xC0 + (unsigned char)((c>>6)&0x1F);
    z[1] = 0x80 + (unsigned char)(c & 0x3F);
    return 2;
  }
  if( c<0x10000 ){
    z[0] = 0xE0 + (unsigned char)((c>>12)&0x0F);
    z[1] = 0x80 + (unsigned char)((c>>6) & 0x3F);
    z[2] = 0x80 + (unsigned char)(c & 0x3F);
    return 3;
  }
  z[0] = 0xF0 + (unsigned char)((c>>18) & 0x07);
  z[1] = 0x80 + (unsigned char)((c>>12) & 0x3F);
  z[2] = 0x80 + (unsigned char)((c>>6) & 0x3F);
  z[3] = 0x80 + (unsigned char)(c & 0x3F);
  return 4;
}

/*
** Read a UTF8 character out of z[] and write it into *pOut.  Return
** the number of bytes in z[] that were used to construct the character.
*/
static int readUtf8(const unsigned char *z, unsigned *pOut){
  static const unsigned char validBits[] = {
    0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
    0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
    0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17,
    0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
    0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
    0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
    0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07,
    0x00, 0x01, 0x02, 0x03, 0x00, 0x01, 0x00, 0x00,
  };
  unsigned c = z[0];
  if( c<0xc0 ){
    *pOut = c;
    return 1;
  }else{
    int n = 1;
    c = validBits[c-0xc0];
    while( (z[n] & 0xc0)==0x80 ){
      c = (c<<6) + (0x3f & z[n++]);
    }
    if( c<0x80 || (c&0xFFFFF800)==0xD800 || (c&0xFFFFFFFE)==0xFFFE ){
      c = 0xFFFD;
    }
    *pOut = c;
    return n;
  }
}

/*
** The nextCharContext structure has been set up.  Add all "next" characters
** to the result set.
*/
static void findNextChars(nextCharContext *p){
  unsigned cPrev = 0;
  unsigned char zPrev[8];
  int n, rc;
  
  for(;;){
    sqlite3_bind_text(p->pStmt, 1, (char*)p->zPrefix, p->nPrefix,
                      SQLITE_STATIC);
    n = writeUtf8(zPrev, cPrev+1);
    sqlite3_bind_text(p->pStmt, 2, (char*)zPrev, n, SQLITE_STATIC);
    rc = sqlite3_step(p->pStmt);
    if( rc==SQLITE_DONE ){
      sqlite3_reset(p->pStmt);
      return;
    }else if( rc!=SQLITE_ROW ){
      p->otherError = rc;
      return;
    }else{
      const unsigned char *zOut = sqlite3_column_text(p->pStmt, 0);
      unsigned cNext;
      n = readUtf8(zOut+p->nPrefix, &cNext);
      sqlite3_reset(p->pStmt);
      nextCharAppend(p, cNext);
      cPrev = cNext;
      if( p->mallocFailed ) return;
    }
  }
}


/*
** next_character(A,T,F,W)
**
** Return a string composted of all next possible characters after
** A for elements of T.F.  If W is supplied, then it is an SQL expression
** that limits the elements in T.F that are considered.
*/
static void nextCharFunc(
  sqlite3_context *context,
  int argc,
  sqlite3_value **argv
){
  nextCharContext c;
  const unsigned char *zTable = sqlite3_value_text(argv[1]);
  const unsigned char *zField = sqlite3_value_text(argv[2]);
  const unsigned char *zWhere;
  char *zSql;
  int rc;

  memset(&c, 0, sizeof(c));
  c.db = sqlite3_context_db_handle(context);
  c.zPrefix = sqlite3_value_text(argv[0]);
  c.nPrefix = sqlite3_value_bytes(argv[0]);
  if( zTable==0 || zField==0 || c.zPrefix==0 ) return;
  if( argc<4
   || (zWhere = sqlite3_value_text(argv[3]))==0
   || zWhere[0]==0
  ){
    zSql = sqlite3_mprintf(
        "SELECT \"%w\" FROM \"%w\""
        " WHERE \"%w\">=(?1 || ?2)"
        "   AND \"%w\"<=(?1 || char(1114111))" /* 1114111 == 0x10ffff */
        " ORDER BY 1 ASC LIMIT 1",
        zField, zTable, zField, zField);
  }else{
    zSql = sqlite3_mprintf(
        "SELECT \"%w\" FROM \"%w\""
        " WHERE \"%w\">=(?1 || ?2)"
        "   AND \"%w\"<=(?1 || char(1114111))" /* 1114111 == 0x10ffff */
        "   AND (%s)"
        " ORDER BY 1 ASC LIMIT 1",
        zField, zTable, zField, zField, zWhere);
  }
  if( zSql==0 ){
    sqlite3_result_error_nomem(context);
    return;
  }

  rc = sqlite3_prepare_v2(c.db, zSql, -1, &c.pStmt, 0);
  sqlite3_free(zSql);
  if( rc ){
    sqlite3_result_error(context, sqlite3_errmsg(c.db), -1);
    return;
  }
  findNextChars(&c);
  if( c.mallocFailed ){
    sqlite3_result_error_nomem(context);
  }else{
    unsigned char *pRes;
    pRes = sqlite3_malloc( c.nUsed*4 + 1 );
    if( pRes==0 ){
      sqlite3_result_error_nomem(context);
    }else{
      int i;
      int n = 0;
      for(i=0; i<c.nUsed; i++){
        n += writeUtf8(pRes+n, c.aResult[i]);
      }
      pRes[n] = 0;
      sqlite3_result_text(context, (const char*)pRes, n, sqlite3_free);
    }
  }
  sqlite3_finalize(c.pStmt);
  sqlite3_free(c.aResult);
}

#ifdef _WIN32
__declspec(dllexport)
#endif
int sqlite3_nextchar_init(
  sqlite3 *db, 
  char **pzErrMsg, 
  const sqlite3_api_routines *pApi
){
  int rc = SQLITE_OK;
  SQLITE_EXTENSION_INIT2(pApi);
  (void)pzErrMsg;  /* Unused parameter */
  rc = sqlite3_create_function(db, "next_char", 3, SQLITE_UTF8, 0,
                               nextCharFunc, 0, 0);
  if( rc==SQLITE_OK ){
    rc = sqlite3_create_function(db, "next_char", 4, SQLITE_UTF8, 0,
                                 nextCharFunc, 0, 0);
  }
  return rc;
}

** This module implements the spellfix1 VIRTUAL TABLE that can be used
** to search a large vocabulary for close matches.  See separate
** documentation files (spellfix1.wiki and editdist3.wiki) for details.
*/
#include "sqlite3ext.h"
SQLITE_EXTENSION_INIT1

#ifndef SQLITE_AMALGAMATION
# include <string.h>
# include <stdio.h>
# include <stdlib.h>
# include <assert.h>
# define ALWAYS(X)  1
# define NEVER(X)   0
  typedef unsigned char u8;
  typedef unsigned short u16;
# include <ctype.h>
#endif

/*
** Character classes for ASCII characters:
**
**   0   ''        Silent letters:   H W
**   1   'A'       Any vowel:   A E I O U (Y)
**   2   'B'       A bilabeal stop or fricative:  B F P V W

# Extensions to be statically loaded.
#
TESTSRC += \
  $(TOP)/ext/misc/amatch.c \
  $(TOP)/ext/misc/closure.c \
  $(TOP)/ext/misc/fuzzer.c \
  $(TOP)/ext/misc/ieee754.c \
  $(TOP)/ext/misc/nextchar.c \
  $(TOP)/ext/misc/regexp.c \
  $(TOP)/ext/misc/spellfix.c \
  $(TOP)/ext/misc/wholenumber.c


#TESTSRC += $(TOP)/ext/fts2/fts2_tokenizer.c
#TESTSRC += $(TOP)/ext/fts3/fts3_tokenizer.c

  }
  sqlite3_test_control(SQLITE_TESTCTRL_OPTIMIZATIONS, db, mask);
  return TCL_OK;
}

typedef struct sqlite3_api_routines sqlite3_api_routines;
/*
**     load_static_extension DB NAME ...
**
** Load one or more statically linked extensions.
*/
static int tclLoadStaticExtensionCmd(
  void * clientData,
  Tcl_Interp *interp,
  int objc,
  Tcl_Obj *CONST objv[]
){
  extern int sqlite3_amatch_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_closure_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_fuzzer_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_ieee_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_nextchar_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_regexp_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_spellfix_init(sqlite3*,char**,const sqlite3_api_routines*);
  extern int sqlite3_wholenumber_init(sqlite3*,char**,const sqlite3_api_routines*);
  static const struct {
    const char *zExtName;
    int (*pInit)(sqlite3*,char**,const sqlite3_api_routines*);
  } aExtension[] = {
    { "amatch",                sqlite3_amatch_init               },
    { "closure",               sqlite3_closure_init              },
    { "fuzzer",                sqlite3_fuzzer_init               },
    { "ieee754",               sqlite3_ieee_init                 },
    { "nextchar",              sqlite3_nextchar_init             },
    { "regexp",                sqlite3_regexp_init               },
    { "spellfix",              sqlite3_spellfix_init             },
    { "wholenumber",           sqlite3_wholenumber_init          },
  };
  sqlite3 *db;
  const char *zName;
  int i, j, rc;
  char *zErrMsg = 0;
  if( objc<3 ){
    Tcl_WrongNumArgs(interp, 1, objv, "DB NAME ...");
    return TCL_ERROR;
  }
  if( getDbPointer(interp, Tcl_GetString(objv[1]), &db) ) return TCL_ERROR;
  for(j=2; j<objc; j++){
    zName = Tcl_GetString(objv[j]);
    for(i=0; i<ArraySize(aExtension); i++){
      if( strcmp(zName, aExtension[i].zExtName)==0 ) break;
    }
    if( i>=ArraySize(aExtension) ){
      Tcl_AppendResult(interp, "no such extension: ", zName, (char*)0);
      return TCL_ERROR;
    }
    rc = aExtension[i].pInit(db, &zErrMsg, 0);
    if( rc!=SQLITE_OK || zErrMsg ){
      Tcl_AppendResult(interp, "initialization of ", zName, " failed: ", zErrMsg,
                       (char*)0);
      sqlite3_free(zErrMsg);
      return TCL_ERROR;
    }
  }
  return TCL_OK;
}


/*
** Register commands with the TCL interpreter.

set testdir [file dirname $argv0]
source $testdir/tester.tcl
set testprefix spellfix

ifcapable !vtab { finish_test ; return }

load_static_extension db spellfix nextchar

set vocab {
rabbi rabbit rabbits rabble rabid rabies raccoon raccoons race raced racer
racers races racetrack racial racially racing rack racked racket racketeer
racketeering racketeers rackets racking racks radar radars radial radially
radian radiance radiant radiantly radiate radiated radiates radiating radiation
radiations radiator radiators radical radically radicals radices radii radio

} {
  do_execsql_test 1.2.$tn {
    SELECT word, matchlen FROM t1 WHERE word MATCH $word 
     ORDER BY score, word LIMIT 5
  } $res
}

# Tests of the next_char function.
#
do_test 1.10 {
  db eval {
    CREATE TABLE vocab(w TEXT PRIMARY KEY);
    INSERT INTO vocab SELECT word FROM t1;
  }
} {}
do_execsql_test 1.11 {
  SELECT next_char('re','vocab','w');
} {a}
do_execsql_test 1.12 {
  SELECT next_char('r','vocab','w');
} {ae}
do_execsql_test 1.13 {
  SELECT next_char('','vocab','w');
} {r}
do_test 1.14 {
  catchsql {SELECT next_char('','xyzzy','a')}
} {1 {no such table: xyzzy}}

do_execsql_test 2.1 {
  CREATE VIRTUAL TABLE t2 USING spellfix1;
  INSERT INTO t2 (word, soundslike) VALUES('school', 'skuul');
  INSERT INTO t2 (word, soundslike) VALUES('psalm', 'sarm');
  SELECT word, matchlen FROM t2 WHERE word MATCH 'sar*' LIMIT 5;
} {psalm 4}