*************************************************************************
** An tokenizer for SQL
**
** This file contains C code that splits an SQL input string up into
** individual tokens and sends those tokens one-by-one over to the
** parser for analysis.
**
** $Id: tokenize.c,v 1.88 2004/09/25 14:39:19 drh Exp $
*/
#include "sqliteInt.h"
#include "os.h"
#include <ctype.h>
#include <stdlib.h>

/*
................................................................................
  { "BEFORE",            TK_BEFORE,       },
  { "BEGIN",             TK_BEGIN,        },
  { "BETWEEN",           TK_BETWEEN,      },
  { "BY",                TK_BY,           },
  { "CASCADE",           TK_CASCADE,      },
  { "CASE",              TK_CASE,         },
  { "CHECK",             TK_CHECK,        },
  { "CLUSTER",           TK_CLUSTER,      },
  { "COLLATE",           TK_COLLATE,      },
  { "COMMIT",            TK_COMMIT,       },
  { "CONFLICT",          TK_CONFLICT,     },
  { "CONSTRAINT",        TK_CONSTRAINT,   },
  { "CREATE",            TK_CREATE,       },
  { "CROSS",             TK_JOIN_KW,      },
  { "DATABASE",          TK_DATABASE,     },
................................................................................
  static char needInit = 1;
  if( needInit ){
    /* Initialize the keyword hash table */
    sqlite3OsEnterMutex();
    if( needInit ){
      int nk;
      nk = sizeof(aKeywordTable)/sizeof(aKeywordTable[0]);
      for(i=0; i<nk; i++){
        aKeywordTable[i].len = strlen(aKeywordTable[i].zName);
        h = sqlite3HashNoCase(aKeywordTable[i].zName, aKeywordTable[i].len);
        h %= KEY_HASH_SIZE;
        aKeywordTable[i].iNext = aiHashTable[h];
        aiHashTable[h] = i+1;
      }
      needInit = 0;
    }
    sqlite3OsLeaveMutex();
  }
  h = sqlite3HashNoCase(z, n) % KEY_HASH_SIZE;
................................................................................
** alphabetic characters, digits, and "_" plus any character
** with the high-order bit set.  The latter rule means that
** any sequence of UTF-8 characters or characters taken from
** an extended ISO8859 character set can form an identifier.
*/
static const char isIdChar[] = {
/* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 0x */
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 1x */
    0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,  /* 2x */
    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,  /* 3x */
    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 4x */
    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,  /* 5x */
    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 6x */
    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,  /* 7x */
};



/*
** Return the length of the token that begins at z[0]. 
** Store the token type in *tokenType before returning.
*/
static int sqliteGetToken(const unsigned char *z, int *tokenType){
  int i;
  switch( *z ){
    case ' ': case '\t': case '\n': case '\f': case '\r': {
      for(i=1; isspace(z[i]); i++){}
      *tokenType = TK_SPACE;
      return i;
    }
    case '-': {
      if( z[1]=='-' ){
        for(i=2; z[i] && z[i]!='\n'; i++){}
        *tokenType = TK_COMMENT;
        return i;
      }
      *tokenType = TK_MINUS;
      return 1;
    }
    case '(': {
................................................................................
      return 1;
    }
    case '/': {
      if( z[1]!='*' || z[2]==0 ){
        *tokenType = TK_SLASH;
        return 1;
      }
      for(i=3; z[i] && (z[i]!='/' || z[i-1]!='*'); i++){}
      if( z[i] ) i++;
      *tokenType = TK_COMMENT;
      return i;
    }
    case '%': {
      *tokenType = TK_REM;
      return 1;
    }
    case '=': {
      *tokenType = TK_EQ;
      return 1 + (z[1]=='=');
    }
    case '<': {
      if( z[1]=='=' ){
        *tokenType = TK_LE;
        return 2;
      }else if( z[1]=='>' ){
        *tokenType = TK_NE;
        return 2;
      }else if( z[1]=='<' ){
        *tokenType = TK_LSHIFT;
        return 2;
      }else{
        *tokenType = TK_LT;
        return 1;
      }
    }
    case '>': {
      if( z[1]=='=' ){
        *tokenType = TK_GE;
        return 2;
      }else if( z[1]=='>' ){
        *tokenType = TK_RSHIFT;
        return 2;
      }else{
        *tokenType = TK_GT;
        return 1;
      }
    }
................................................................................
    }
    case '~': {
      *tokenType = TK_BITNOT;
      return 1;
    }
    case '\'': case '"': {
      int delim = z[0];
      for(i=1; z[i]; i++){
        if( z[i]==delim ){
          if( z[i+1]==delim ){
            i++;
          }else{
            break;
          }
        }
      }
      if( z[i] ) i++;
      *tokenType = TK_STRING;
      return i;
    }
    case '.': {
      *tokenType = TK_DOT;
      return 1;
    }
................................................................................
        i += 2;
        while( isdigit(z[i]) ){ i++; }
        *tokenType = TK_FLOAT;
      }
      return i;
    }
    case '[': {
      for(i=1; z[i] && z[i-1]!=']'; i++){}
      *tokenType = TK_ID;
      return i;
    }
    case '?': {
      *tokenType = TK_VARIABLE;
      for(i=1; isdigit(z[i]); i++){}
      return i;
    }
    case ':': {
      for(i=1; (z[i]&0x80)!=0 || isIdChar[z[i]]; i++){}
      *tokenType = i>1 ? TK_VARIABLE : TK_ILLEGAL;
      return i;
    }
    case '$': {
      int c;
      *tokenType = TK_VARIABLE;
      if( z[1]=='{' ){
        int nBrace = 1;
        for(i=2; (c=z[i])!=0 && nBrace; i++){
          if( c=='{' ){
            nBrace++;
          }else if( c=='}' ){
................................................................................
          }
        }
        if( n==0 ) *tokenType = TK_ILLEGAL;
      }
      return i;
    } 
    case 'x': case 'X': {
      if( z[1]=='\'' || z[1]=='"' ){
        int delim = z[1];
        *tokenType = TK_BLOB;
        for(i=2; z[i]; i++){
          if( z[i]==delim ){
            if( i%2 ) *tokenType = TK_ILLEGAL;
            break;
          }
          if( !isxdigit(z[i]) ){
            *tokenType = TK_ILLEGAL;
            return i;
          }
        }
        if( z[i] ) i++;
        return i;
      }
      /* Otherwise fall through to the next case */
    }
    default: {
      if( (*z&0x80)==0 && !isIdChar[*z] ){
        break;
      }
      for(i=1; (z[i]&0x80)!=0 || isIdChar[z[i]]; i++){}
      *tokenType = sqlite3KeywordCode((char*)z, i);
      return i;
    }
  }
  *tokenType = TK_ILLEGAL;
  return 1;
}
................................................................................
        zSql++;
        while( *zSql && *zSql!=c ){ zSql++; }
        if( *zSql==0 ) return 0;
        token = tkOTHER;
        break;
      }
      default: {
        if( isIdChar[(u8)*zSql] ){

          /* Keywords and unquoted identifiers */
          int nId;
          for(nId=1; isIdChar[(u8)zSql[nId]]; nId++){}

          switch( *zSql ){
            case 'c': case 'C': {
              if( nId==6 && sqlite3StrNICmp(zSql, "create", 6)==0 ){
                token = tkCREATE;
              }else{
                token = tkOTHER;
              }

*************************************************************************
** An tokenizer for SQL
**
** This file contains C code that splits an SQL input string up into
** individual tokens and sends those tokens one-by-one over to the
** parser for analysis.
**
** $Id: tokenize.c,v 1.89 2004/09/25 15:25:26 drh Exp $
*/
#include "sqliteInt.h"
#include "os.h"
#include <ctype.h>
#include <stdlib.h>

/*
................................................................................
  { "BEFORE",            TK_BEFORE,       },
  { "BEGIN",             TK_BEGIN,        },
  { "BETWEEN",           TK_BETWEEN,      },
  { "BY",                TK_BY,           },
  { "CASCADE",           TK_CASCADE,      },
  { "CASE",              TK_CASE,         },
  { "CHECK",             TK_CHECK,        },

  { "COLLATE",           TK_COLLATE,      },
  { "COMMIT",            TK_COMMIT,       },
  { "CONFLICT",          TK_CONFLICT,     },
  { "CONSTRAINT",        TK_CONSTRAINT,   },
  { "CREATE",            TK_CREATE,       },
  { "CROSS",             TK_JOIN_KW,      },
  { "DATABASE",          TK_DATABASE,     },
................................................................................
  static char needInit = 1;
  if( needInit ){
    /* Initialize the keyword hash table */
    sqlite3OsEnterMutex();
    if( needInit ){
      int nk;
      nk = sizeof(aKeywordTable)/sizeof(aKeywordTable[0]);
      for(i=0, p=aKeywordTable; i<nk; i++, p++){
        const char *zName = p->zName;
        int len = p->len = strlen(zName);
        h = sqlite3HashNoCase(zName, len) % KEY_HASH_SIZE;
        p->iNext = aiHashTable[h];
        aiHashTable[h] = i+1;
      }
      needInit = 0;
    }
    sqlite3OsLeaveMutex();
  }
  h = sqlite3HashNoCase(z, n) % KEY_HASH_SIZE;
................................................................................
** alphabetic characters, digits, and "_" plus any character
** with the high-order bit set.  The latter rule means that
** any sequence of UTF-8 characters or characters taken from
** an extended ISO8859 character set can form an identifier.
*/
static const char isIdChar[] = {
/* x0 x1 x2 x3 x4 x5 x6 x7 x8 x9 xA xB xC xD xE xF */



    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0,  /* 3x */
    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 4x */
    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1,  /* 5x */
    0, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,  /* 6x */
    1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,  /* 7x */
};

#define IdChar(C)  (((c=C)&0x80)!=0 || (c>0x2f && isIdChar[c-0x30]))

/*
** Return the length of the token that begins at z[0]. 
** Store the token type in *tokenType before returning.
*/
static int sqliteGetToken(const unsigned char *z, int *tokenType){
  int i, c;
  switch( *z ){
    case ' ': case '\t': case '\n': case '\f': case '\r': {
      for(i=1; isspace(z[i]); i++){}
      *tokenType = TK_SPACE;
      return i;
    }
    case '-': {
      if( z[1]=='-' ){
        for(i=2; (c=z[i])!=0 && c!='\n'; i++){}
        *tokenType = TK_COMMENT;
        return i;
      }
      *tokenType = TK_MINUS;
      return 1;
    }
    case '(': {
................................................................................
      return 1;
    }
    case '/': {
      if( z[1]!='*' || z[2]==0 ){
        *tokenType = TK_SLASH;
        return 1;
      }
      for(i=3, c=z[2]; (c!='*' || z[i]!='/') && (c=z[i])!=0; i++){}
      if( c ) i++;
      *tokenType = TK_COMMENT;
      return i;
    }
    case '%': {
      *tokenType = TK_REM;
      return 1;
    }
    case '=': {
      *tokenType = TK_EQ;
      return 1 + (z[1]=='=');
    }
    case '<': {
      if( (c=z[1])=='=' ){
        *tokenType = TK_LE;
        return 2;
      }else if( c=='>' ){
        *tokenType = TK_NE;
        return 2;
      }else if( c=='<' ){
        *tokenType = TK_LSHIFT;
        return 2;
      }else{
        *tokenType = TK_LT;
        return 1;
      }
    }
    case '>': {
      if( (c=z[1])=='=' ){
        *tokenType = TK_GE;
        return 2;
      }else if( c=='>' ){
        *tokenType = TK_RSHIFT;
        return 2;
      }else{
        *tokenType = TK_GT;
        return 1;
      }
    }
................................................................................
    }
    case '~': {
      *tokenType = TK_BITNOT;
      return 1;
    }
    case '\'': case '"': {
      int delim = z[0];
      for(i=1; (c=z[i])!=0; i++){
        if( c==delim ){
          if( z[i+1]==delim ){
            i++;
          }else{
            break;
          }
        }
      }
      if( c ) i++;
      *tokenType = TK_STRING;
      return i;
    }
    case '.': {
      *tokenType = TK_DOT;
      return 1;
    }
................................................................................
        i += 2;
        while( isdigit(z[i]) ){ i++; }
        *tokenType = TK_FLOAT;
      }
      return i;
    }
    case '[': {
      for(i=1, c=z[0]; c!=']' && (c=z[i])!=0; i++){}
      *tokenType = TK_ID;
      return i;
    }
    case '?': {
      *tokenType = TK_VARIABLE;
      for(i=1; isdigit(z[i]); i++){}
      return i;
    }
    case ':': {
      for(i=1; IdChar(z[i]); i++){}
      *tokenType = i>1 ? TK_VARIABLE : TK_ILLEGAL;
      return i;
    }
    case '$': {

      *tokenType = TK_VARIABLE;
      if( z[1]=='{' ){
        int nBrace = 1;
        for(i=2; (c=z[i])!=0 && nBrace; i++){
          if( c=='{' ){
            nBrace++;
          }else if( c=='}' ){
................................................................................
          }
        }
        if( n==0 ) *tokenType = TK_ILLEGAL;
      }
      return i;
    } 
    case 'x': case 'X': {
      if( (c=z[1])=='\'' || c=='"' ){
        int delim = c;
        *tokenType = TK_BLOB;
        for(i=2; (c=z[i])!=0; i++){
          if( c==delim ){
            if( i%2 ) *tokenType = TK_ILLEGAL;
            break;
          }
          if( !isxdigit(c) ){
            *tokenType = TK_ILLEGAL;
            return i;
          }
        }
        if( c ) i++;
        return i;
      }
      /* Otherwise fall through to the next case */
    }
    default: {
      if( !IdChar(*z) ){
        break;
      }
      for(i=1; IdChar(z[i]); i++){}
      *tokenType = sqlite3KeywordCode((char*)z, i);
      return i;
    }
  }
  *tokenType = TK_ILLEGAL;
  return 1;
}
................................................................................
        zSql++;
        while( *zSql && *zSql!=c ){ zSql++; }
        if( *zSql==0 ) return 0;
        token = tkOTHER;
        break;
      }
      default: {
        int c;
        if( IdChar((u8)*zSql) ){
          /* Keywords and unquoted identifiers */
          int nId;

          for(nId=1; IdChar(zSql[nId]); nId++){}
          switch( *zSql ){
            case 'c': case 'C': {
              if( nId==6 && sqlite3StrNICmp(zSql, "create", 6)==0 ){
                token = tkCREATE;
              }else{
                token = tkOTHER;
              }