/*
** 2001 September 15
**
** 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.
**
*************************************************************************
** 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.
*/
#include "tokenize.h"
#include <stdlib.h>
#include <tcl.h>
#include <string.h>

/* Character classes for tokenizing
**
** In the sqlite3GetToken() function, a switch() on aiClass[c] is implemented
** using a lookup table, whereas a switch() directly on c uses a binary search.
** The lookup table is much faster.  To maximize speed, and to ensure that
** a lookup table is used, all of the classes need to be small integers and
** all of them need to be used within the switch.
*/
#define CC_X          0    /* The letter 'x', or start of BLOB literal */
#define CC_KYWD       1    /* Alphabetics or '_'.  Usable in a keyword */
#define CC_ID         2    /* unicode characters usable in IDs */
#define CC_DIGIT      3    /* Digits */
#define CC_DOLLAR     4    /* '$' (returns TK_TCLVAR unless numeric var "name") */
#define CC_VARALPHA   5    /* '@', ':'.  Alphabetic SQL variables (returns TK_TCLVAR) */
#define CC_VARNUM     6    /* '?'.  Numeric SQL variables (returns TK_SQLVAR) */
#define CC_SPACE      7    /* Space characters */
#define CC_QUOTE      8    /* '"', '\'', or '`'.  String literals, quoted ids */
#define CC_QUOTE2     9    /* '['.   [...] style quoted ids */
#define CC_PIPE      10    /* '|'.   Bitwise OR or concatenate */
#define CC_MINUS     11    /* '-'.  Minus or SQL-style comment */
#define CC_LT        12    /* '<'.  Part of < or <= or <> */
#define CC_GT        13    /* '>'.  Part of > or >= */
#define CC_EQ        14    /* '='.  Part of = or == */
#define CC_BANG      15    /* '!'.  Part of != */
#define CC_SLASH     16    /* '/'.  / or c-style comment */
#define CC_LP        17    /* '(' */
#define CC_RP        18    /* ')' */
#define CC_SEMI      19    /* ';' */
#define CC_PLUS      20    /* '+' */
#define CC_STAR      21    /* '*' */
#define CC_PERCENT   22    /* '%' */
#define CC_COMMA     23    /* ',' */
#define CC_AND       24    /* '&' */
#define CC_TILDA     25    /* '~' */
#define CC_DOT       26    /* '.' */
#define CC_HASH      28    /* '#' added 20170726 PDS */
#define CC_ILLEGAL   99    /* Illegal character */

static const unsigned char aiClass[] = {
/*         x0  x1  x2  x3  x4  x5  x6  x7  x8  x9  xa  xb  xc  xd  xe  xf */
/* 0x */   99, 99, 99, 99, 99, 99, 99, 99, 99,  7,  7, 99,  7,  7, 99, 99,
/* 1x */   99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99, 99,
/* 2x */    7, 15,  8, 28,  4, 22, 24,  8, 17, 18, 21, 20, 23, 11, 26, 16,
/* 3x */    3,  3,  3,  3,  3,  3,  3,  3,  3,  3,  5, 19, 12, 14, 13,  6,
/* 4x */    5,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
/* 5x */    1,  1,  1,  1,  1,  1,  1,  1,  0,  1,  1,  9, 99, 99, 99,  1,
/* 6x */    8,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,  1,
/* 7x */    1,  1,  1,  1,  1,  1,  1,  1,  0,  1,  1, 99, 10, 99, 25, 99,
/* 8x */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* 9x */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Ax */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Bx */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Cx */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Dx */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Ex */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,
/* Fx */    2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2,  2
};

#define totoken(X) aiClass[(unsigned char)X]

// Imported from src/global.c in the sqlite distro
static const unsigned char sqlite3CtypeMap[256] = {
  0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 00..07    ........ */
  0x00, 0x01, 0x01, 0x01, 0x01, 0x01, 0x00, 0x00,  /* 08..0f    ........ */
  0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 10..17    ........ */
  0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 18..1f    ........ */
  0x01, 0x00, 0x80, 0x00, 0x40, 0x00, 0x00, 0x80,  /* 20..27     !"#$%&' */
  0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 28..2f    ()*+,-./ */
  0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c, 0x0c,  /* 30..37    01234567 */
  0x0c, 0x0c, 0x00, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 38..3f    89:;<=>? */

  0x00, 0x0a, 0x0a, 0x0a, 0x0a, 0x0a, 0x0a, 0x02,  /* 40..47    @ABCDEFG */
  0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02,  /* 48..4f    HIJKLMNO */
  0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02, 0x02,  /* 50..57    PQRSTUVW */
  0x02, 0x02, 0x02, 0x80, 0x00, 0x00, 0x00, 0x40,  /* 58..5f    XYZ[\]^_ */
  0x80, 0x2a, 0x2a, 0x2a, 0x2a, 0x2a, 0x2a, 0x22,  /* 60..67    `abcdefg */
  0x22, 0x22, 0x22, 0x22, 0x22, 0x22, 0x22, 0x22,  /* 68..6f    hijklmno */
  0x22, 0x22, 0x22, 0x22, 0x22, 0x22, 0x22, 0x22,  /* 70..77    pqrstuvw */
  0x22, 0x22, 0x22, 0x00, 0x00, 0x00, 0x00, 0x00,  /* 78..7f    xyz{|}~. */

  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* 80..87    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* 88..8f    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* 90..97    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* 98..9f    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* a0..a7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* a8..af    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* b0..b7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* b8..bf    ........ */

  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* c0..c7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* c8..cf    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* d0..d7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* d8..df    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* e0..e7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* e8..ef    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40,  /* f0..f7    ........ */
  0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40, 0x40   /* f8..ff    ........ */
};


/*
** The charMap() macro maps alphabetic characters (only) into their
** lower-case ASCII equivalent.  On ASCII machines, this is just
** an upper-to-lower case map.  On EBCDIC machines we also need
** to adjust the encoding.  The mapping is only valid for alphabetics
** which are the only characters for which this feature is used. 
**
** Used by keywordhash.h
*/
# define charMap(X) tolower((unsigned char)X)

/*
** If X is a character that can be used in an identifier then
** IdChar(X) will be true.  Otherwise it is false.
**
** For ASCII, any character with the high-order bit set is
** allowed in an identifier.  For 7-bit characters, 
** sqlite3IsIdChar[X] must be 1.
*/
#define IdChar(C)  ((sqlite3CtypeMap[(unsigned char)C]&0x46)!=0)

/*
** Return the length (in bytes) of the token that begins at z[0]. 
** Store the token type in *tokenType before returning.
*/
int Pg_sqlite3GetToken(const char *z, enum sqltoken *tokenType){
  int i, c;
  switch( totoken(*z) ){  /* Switch on the character-class of the first byte
                          ** of the token. See the comment on the CC_ defines
                          ** above. */
    case CC_SPACE: {
      for(i=1; sqlite3Isspace(z[i]); i++){}
      *tokenType = TK_SPACE;
      return i;
    }
    case CC_MINUS: {
      if( z[1]=='-' ){
        for(i=2; (c=z[i])!=0 && c!='\n'; i++){}
        *tokenType = TK_SPACE;   /* IMP: R-22934-25134 */
        return i;
      }
      *tokenType = TK_MINUS;
      return 1;
    }
    case CC_LP: {
      *tokenType = TK_LP;
      return 1;
    }
    case CC_RP: {
      *tokenType = TK_RP;
      return 1;
    }
    case CC_SEMI: {
      *tokenType = TK_SEMI;
      return 1;
    }
    case CC_PLUS: {
      *tokenType = TK_PLUS;
      return 1;
    }
    case CC_STAR: {
      *tokenType = TK_STAR;
      return 1;
    }
    case CC_HASH: {
      *tokenType = TK_HASH;
      return 1;
    }
    case CC_SLASH: {
      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_SPACE;   /* IMP: R-22934-25134 */
      return i;
    }
    case CC_PERCENT: {
      *tokenType = TK_REM;
      return 1;
    }
    case CC_EQ: {
      *tokenType = TK_EQ;
      return 1 + (z[1]=='=');
    }
    case CC_LT: {
      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 CC_GT: {
      if( (c=z[1])=='=' ){
        *tokenType = TK_GE;
        return 2;
      }else if( c=='>' ){
        *tokenType = TK_RSHIFT;
        return 2;
      }else{
        *tokenType = TK_GT;
        return 1;
      }
    }
    case CC_BANG: {
      if( z[1]!='=' ){
        *tokenType = TK_ILLEGAL;
        return 1;
      }else{
        *tokenType = TK_NE;
        return 2;
      }
    }
    case CC_PIPE: {
      if( z[1]!='|' ){
        *tokenType = TK_BITOR;
        return 1;
      }else{
        *tokenType = TK_CONCAT;
        return 2;
      }
    }
    case CC_COMMA: {
      *tokenType = TK_COMMA;
      return 1;
    }
    case CC_AND: {
      *tokenType = TK_BITAND;
      return 1;
    }
    case CC_TILDA: {
      *tokenType = TK_BITNOT;
      return 1;
    }
    case CC_QUOTE: {
      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=='\'' ){
        *tokenType = TK_STRING;
        return i+1;
      }else if( c!=0 ){
        *tokenType = TK_ID;
        return i+1;
      }else{
        *tokenType = TK_ILLEGAL;
        return i;
      }
    }
    case CC_DOT: {
#ifndef SQLITE_OMIT_FLOATING_POINT
      if( !sqlite3Isdigit(z[1]) )
#endif
      {
        *tokenType = TK_DOT;
        return 1;
      }
      /* If the next character is a digit, this is a floating point
      ** number that begins with ".".  Fall thru into the next case */
    }
    case CC_DIGIT: {
      *tokenType = TK_INTEGER;
#ifndef SQLITE_OMIT_HEX_INTEGER
      if( z[0]=='0' && (z[1]=='x' || z[1]=='X') && sqlite3Isxdigit(z[2]) ){
        for(i=3; sqlite3Isxdigit(z[i]); i++){}
        return i;
      }
#endif
      for(i=0; sqlite3Isdigit(z[i]); i++){}
#ifndef SQLITE_OMIT_FLOATING_POINT
      if( z[i]=='.' ){
        i++;
        while( sqlite3Isdigit(z[i]) ){ i++; }
        *tokenType = TK_FLOAT;
      }
      if( (z[i]=='e' || z[i]=='E') &&
           ( sqlite3Isdigit(z[i+1]) 
            || ((z[i+1]=='+' || z[i+1]=='-') && sqlite3Isdigit(z[i+2]))
           )
      ){
        i += 2;
        while( sqlite3Isdigit(z[i]) ){ i++; }
        *tokenType = TK_FLOAT;
      }
#endif
      while( IdChar(z[i]) ){
        *tokenType = TK_ILLEGAL;
        i++;
      }
      return i;
    }
    case CC_QUOTE2: {
      for(i=1, c=z[0]; c!=']' && (c=z[i])!=0; i++){}
      *tokenType = c==']' ? TK_ID : TK_ILLEGAL;
      return i;
    }
    case CC_VARNUM: {
      *tokenType = TK_SQLVAR;
      for(i=1; sqlite3Isdigit(z[i]); i++){}
      return i;
    }
// BEGIN change PDS Jul 2017
    case CC_DOLLAR: {
      // treat $NNN the same as ? because that's how PostgreSQL expects positional variables
      for(i=1; isdigit(z[i]); i++);
      if(i > 1) { // $NNN is an SQL variable
	*tokenType = TK_SQLVAR;
	return i;
      }
      // otherwise it's a dollar-quoted string (PostgreSQL special)
      *tokenType = TK_STRING;
      for(i = 1; z[i] && z[i] != '$'; i++);
      if(!z[i]) {
	*tokenType = TK_ILLEGAL;
	return i;
      }
      // len includes both leading and trailing $ signs
      int len = ++i;
      while(z[i]) {
	if(z[i] == '$') {
	  // compare &z[i] with '$token$'
	  if(strncmp(z, &z[i], len) == 0)
	    break;
	}
	i++;
      }
      if(z[i]) {
	i += len; // skip over end token
      } else {
	*tokenType = TK_ILLEGAL; // nonterminated string is illegal
      }
      return i;
    }
// END change PDS Jul 2017
    case CC_VARALPHA: {
// special case of "::" which is a cast in PostgreSQL
      if (z[0]==':' && z[1]==':' && z[2]!=':') {
	*tokenType = TK_CAST;
	return 2;
      }
      int n = 0;
      *tokenType = TK_TCLVAR;
      if(z[1] == '{') { // Handle ${...}
	i = 1;
	while((c=z[i])!=0 && c != '}')
	  i++;
	if(c == 0) { // nonterminated {...}
	  *tokenType = TK_ILLEGAL; // punt, let SQL error out
	  return 1;
	}
	return i+1;
      }
      for(i=1; (c=z[i])!=0; i++){
        if( IdChar(c) ){
          n++;
#ifndef SQLITE_OMIT_TCL_VARIABLE
        }else if( c=='(' && n>0 ){
          do{
            i++;
          }while( (c=z[i])!=0 && !sqlite3Isspace(c) && c!=')' );
          if( c==')' ){
            i++;
          }else{
            *tokenType = TK_ILLEGAL;
          }
          break;
        }else if( c==':' && z[i+1]==':' ){
          i++;
#endif
        }else{
          break;
        }
      }
      if( n==0 ) *tokenType = TK_ILLEGAL;
      return i;
    }
    case CC_KYWD: {
      for(i=1; totoken(z[i])<=CC_KYWD; i++){}
      if( IdChar(z[i]) ){
        /* This token started out using characters that can appear in keywords,
        ** but z[i] is a character not allowed within keywords, so this must
        ** be an identifier instead */
        i++;
        break;
      }
      *tokenType = TK_ID;
      return i; // was keywordCode((char*)z, i, tokenType);
    }
    case CC_X: {
#ifndef SQLITE_OMIT_BLOB_LITERAL
      if( z[1]=='\'' ){
        *tokenType = TK_BLOB;
        for(i=2; sqlite3Isxdigit(z[i]); i++){}
        if( z[i]!='\'' || i%2 ){
          *tokenType = TK_ILLEGAL;
          while( z[i] && z[i]!='\'' ){ i++; }
        }
        if( z[i] ) i++;
        return i;
      }
#endif
      /* If it is not a BLOB literal, then it must be an ID, since no
      ** SQL keywords start with the letter 'x'.  Fall through */
    }
    case CC_ID: {
      i = 1;
      break;
    }
    case CC_ILLEGAL:
    default: {
      *tokenType = TK_ILLEGAL;
      return 1;
    }
  }
  while( IdChar(z[i]) ){ i++; }
  *tokenType = TK_ID;
  return i;
}

extern int array_to_utf8(Tcl_Interp *interp, const char **paramValues, int *paramLengths, int nParams, const char **bufferPtr);

int handle_substitutions(Tcl_Interp *interp, const char *sql, char **newSqlPtr, const char ***replacementArrayPtr, int *replacementArrayLengthPtr, const char **bufferPtr)
{
	// Worst possible case, :a mapping to $99999 at the end of a really long string
	char *newSql = ckalloc(strlen(sql)*3+1);
	// Worst possible case? the sql is nothing but ":varname" and they're all one character names. This
	// will still be big enough.
	const char **replacementArray = (const char **)ckalloc((strlen(sql)/2) * (sizeof *replacementArray));
	int *lengthArray = (int *)ckalloc((strlen(sql)/2) * sizeof (int));

	const char *p;
	char *q;
	int len;
	enum sqltoken tk;
	int nextVarIndex = 0;
	int result = TCL_OK;

	p = sql;
	q = newSql;
	while(*p) {
		len = Pg_sqlite3GetToken(p, &tk);
		switch (tk) {
			case TK_SQLVAR: {
				Tcl_SetResult(interp, "Can't combine Tcl and Postgres substitutions", TCL_STATIC);
				result = TCL_ERROR;
				goto cleanup_and_exit;
			}
			case TK_TCLVAR: {
				char *nameBuf = ckalloc(len);
				int stringLength;
				Tcl_Obj *varObj = NULL;
				int i;
				int skip = 1;
				int trunc = 0;

				// if :{...} then skip ':{' and truncate '}'
				if(p[1] == '{') {
					skip = 2;
					trunc = 1;
				}

				for(i = skip; i < len; i++)
					nameBuf[i-skip] = p[i];
				nameBuf[i-skip-trunc] = 0;
				varObj = Tcl_GetVar2Ex(interp, nameBuf, NULL, 0);
				if(varObj) {
					replacementArray[nextVarIndex] = Tcl_GetStringFromObj(varObj, &stringLength);
					lengthArray[nextVarIndex] = stringLength;
				} else {
					replacementArray[nextVarIndex] = NULL;
					lengthArray[nextVarIndex] = 0;
				}
				ckfree(nameBuf);
				p += len;

				sprintf(q, "$%d", nextVarIndex+1); //1 indexed
				while(*q) q++;
				nextVarIndex++;

				break;
			}
			case TK_ILLEGAL:
			default: {
				while(len > 0) {
					*q++ = *p++;
					len--;
				}
				break;
			}
		}
	}
	*q = 0;

	if(result == TCL_OK)
		result = array_to_utf8(interp, replacementArray, lengthArray, nextVarIndex, bufferPtr);

  cleanup_and_exit:
	if(lengthArray) ckfree(lengthArray);

	if(result == TCL_OK) {
		*newSqlPtr = newSql;
		*replacementArrayPtr = replacementArray;
		*replacementArrayLengthPtr = nextVarIndex;
	} else {
		ckfree(newSql);
		ckfree((char *)replacementArray);
	}
	return result;
}