blob: 1656fd8c3dd9a3a19d6b88313ce5f52ce2ccfb9d [file] [log] [blame]
/*
** 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.
**
*************************************************************************
** A 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 <ctype.h>
#include <stdarg.h>
#include <stdlib.h>
#include "windef.h"
#include "winbase.h"
#include "wine/unicode.h"
#include "query.h"
#include "sql.tab.h"
/*
** All the keywords of the SQL language are stored as in a hash
** table composed of instances of the following structure.
*/
typedef struct Keyword Keyword;
struct Keyword {
const WCHAR *name; /* The keyword name */
unsigned int len;
int tokenType; /* The token value for this keyword */
};
#define MAX_TOKEN_LEN 11
static const WCHAR addW[] = {'A','D','D'};
static const WCHAR alterW[] = {'A','L','T','E','R'};
static const WCHAR andW[] = {'A','N','D'};
static const WCHAR byW[] = {'B','Y'};
static const WCHAR charW[] = {'C','H','A','R'};
static const WCHAR characterW[] = {'C','H','A','R','A','C','T','E','R'};
static const WCHAR createW[] = {'C','R','E','A','T','E'};
static const WCHAR deleteW[] = {'D','E','L','E','T','E'};
static const WCHAR distinctW[] = {'D','I','S','T','I','N','C','T'};
static const WCHAR dropW[] = {'D','R','O','P'};
static const WCHAR freeW[] = {'F','R','E','E'};
static const WCHAR fromW[] = {'F','R','O','M'};
static const WCHAR holdW[] = {'H','O','L','D'};
static const WCHAR insertW[] = {'I','N','S','E','R','T'};
static const WCHAR intW[] = {'I','N','T'};
static const WCHAR integerW[] = {'I','N','T','E','G','E','R'};
static const WCHAR intoW[] = {'I','N','T','O'};
static const WCHAR isW[] = {'I','S'};
static const WCHAR keyW[] = {'K','E','Y'};
static const WCHAR likeW[] = {'L','I','K','E'};
static const WCHAR localizableW[] = {'L','O','C','A','L','I','Z','A','B','L','E'};
static const WCHAR longW[] = {'L','O','N','G'};
static const WCHAR longcharW[] = {'L','O','N','G','C','H','A','R'};
static const WCHAR notW[] = {'N','O','T'};
static const WCHAR nullW[] = {'N','U','L','L'};
static const WCHAR objectW[] = {'O','B','J','E','C','T'};
static const WCHAR orW[] = {'O','R'};
static const WCHAR orderW[] = {'O','R','D','E','R'};
static const WCHAR primaryW[] = {'P','R','I','M','A','R','Y'};
static const WCHAR selectW[] = {'S','E','L','E','C','T'};
static const WCHAR setW[] = {'S','E','T'};
static const WCHAR shortW[] = {'S','H','O','R','T'};
static const WCHAR tableW[] = {'T','A','B','L','E'};
static const WCHAR temporaryW[] = {'T','E','M','P','O','R','A','R','Y'};
static const WCHAR updateW[] = {'U','P','D','A','T','E'};
static const WCHAR valuesW[] = {'V','A','L','U','E','S'};
static const WCHAR whereW[] = {'W','H','E','R','E'};
#define ARRAY_SIZE(array) (sizeof(array)/sizeof((array)[0]))
/*
** These are the keywords
** They MUST be in alphabetical order
*/
static const Keyword aKeywordTable[] = {
{ addW, ARRAY_SIZE(addW), TK_ADD },
{ alterW, ARRAY_SIZE(alterW), TK_ALTER },
{ andW, ARRAY_SIZE(andW), TK_AND },
{ byW, ARRAY_SIZE(byW), TK_BY },
{ charW, ARRAY_SIZE(charW), TK_CHAR },
{ characterW, ARRAY_SIZE(characterW), TK_CHAR },
{ createW, ARRAY_SIZE(createW), TK_CREATE },
{ deleteW, ARRAY_SIZE(deleteW), TK_DELETE },
{ distinctW, ARRAY_SIZE(distinctW), TK_DISTINCT },
{ dropW, ARRAY_SIZE(dropW), TK_DROP },
{ freeW, ARRAY_SIZE(freeW), TK_FREE },
{ fromW, ARRAY_SIZE(fromW), TK_FROM },
{ holdW, ARRAY_SIZE(holdW), TK_HOLD },
{ insertW, ARRAY_SIZE(insertW), TK_INSERT },
{ intW, ARRAY_SIZE(intW), TK_INT },
{ integerW, ARRAY_SIZE(integerW), TK_INT },
{ intoW, ARRAY_SIZE(intoW), TK_INTO },
{ isW, ARRAY_SIZE(isW), TK_IS },
{ keyW, ARRAY_SIZE(keyW), TK_KEY },
{ likeW, ARRAY_SIZE(likeW), TK_LIKE },
{ localizableW, ARRAY_SIZE(localizableW), TK_LOCALIZABLE },
{ longW, ARRAY_SIZE(longW), TK_LONG },
{ longcharW, ARRAY_SIZE(longcharW), TK_LONGCHAR },
{ notW, ARRAY_SIZE(notW), TK_NOT },
{ nullW, ARRAY_SIZE(nullW), TK_NULL },
{ objectW, ARRAY_SIZE(objectW), TK_OBJECT },
{ orW, ARRAY_SIZE(orW), TK_OR },
{ orderW, ARRAY_SIZE(orderW), TK_ORDER },
{ primaryW, ARRAY_SIZE(primaryW), TK_PRIMARY },
{ selectW, ARRAY_SIZE(selectW), TK_SELECT },
{ setW, ARRAY_SIZE(setW), TK_SET },
{ shortW, ARRAY_SIZE(shortW), TK_SHORT },
{ tableW, ARRAY_SIZE(tableW), TK_TABLE },
{ temporaryW, ARRAY_SIZE(temporaryW), TK_TEMPORARY },
{ updateW, ARRAY_SIZE(updateW), TK_UPDATE },
{ valuesW, ARRAY_SIZE(valuesW), TK_VALUES },
{ whereW, ARRAY_SIZE(whereW), TK_WHERE },
};
/*
** Comparison function for binary search.
*/
static int compKeyword(const void *m1, const void *m2){
const Keyword *k1 = m1, *k2 = m2;
int ret, len = min( k1->len, k2->len );
if ((ret = memicmpW( k1->name, k2->name, len ))) return ret;
if (k1->len < k2->len) return -1;
else if (k1->len > k2->len) return 1;
return 0;
}
/*
** This function looks up an identifier to determine if it is a
** keyword. If it is a keyword, the token code of that keyword is
** returned. If the input is not a keyword, TK_ID is returned.
*/
static int sqliteKeywordCode(const WCHAR *z, int n){
Keyword key, *r;
if( n>MAX_TOKEN_LEN )
return TK_ID;
key.tokenType = 0;
key.name = z;
key.len = n;
r = bsearch( &key, aKeywordTable, ARRAY_SIZE(aKeywordTable), sizeof(Keyword), compKeyword );
if( r )
return r->tokenType;
return TK_ID;
}
/*
** If X is a character that can be used in an identifier then
** isIdChar[X] will be 1. Otherwise isIdChar[X] will be 0.
**
** In this implementation, an identifier can be a string of
** 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, 1, 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, 1, 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 */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 8x */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* 9x */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Ax */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Bx */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Cx */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Dx */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Ex */
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, /* Fx */
};
/*
** WCHAR safe version of isdigit()
*/
static inline int isDigit(WCHAR c)
{
return c >= '0' && c <= '9';
}
/*
** WCHAR safe version of isspace(), except '\r'
*/
static inline int isSpace(WCHAR c)
{
return c == ' ' || c == '\t' || c == '\n' || c == '\f';
}
/*
** Return the length of the token that begins at z[0]. Return
** -1 if the token is (or might be) incomplete. Store the token
** type in *tokenType before returning.
*/
int sqliteGetToken(const WCHAR *z, int *tokenType, int *skip){
int i;
*skip = 0;
switch( *z ){
case ' ': case '\t': case '\n': case '\f':
for(i=1; isSpace(z[i]); i++){}
*tokenType = TK_SPACE;
return i;
case '-':
if( z[1]==0 ) return -1;
*tokenType = TK_MINUS;
return 1;
case '(':
*tokenType = TK_LP;
return 1;
case ')':
*tokenType = TK_RP;
return 1;
case '*':
*tokenType = TK_STAR;
return 1;
case '=':
*tokenType = TK_EQ;
return 1;
case '<':
if( z[1]=='=' ){
*tokenType = TK_LE;
return 2;
}else if( z[1]=='>' ){
*tokenType = TK_NE;
return 2;
}else{
*tokenType = TK_LT;
return 1;
}
case '>':
if( z[1]=='=' ){
*tokenType = TK_GE;
return 2;
}else{
*tokenType = TK_GT;
return 1;
}
case '!':
if( z[1]!='=' ){
*tokenType = TK_ILLEGAL;
return 2;
}else{
*tokenType = TK_NE;
return 2;
}
case '?':
*tokenType = TK_WILDCARD;
return 1;
case ',':
*tokenType = TK_COMMA;
return 1;
case '`': case '\'': {
int delim = z[0];
for(i=1; z[i]; i++){
if( z[i]==delim )
break;
}
if( z[i] ) i++;
if( delim == '`' )
*tokenType = TK_ID;
else
*tokenType = TK_STRING;
return i;
}
case '.':
if( !isDigit(z[1]) ){
*tokenType = TK_DOT;
return 1;
}
/* Fall through */
case '0': case '1': case '2': case '3': case '4':
case '5': case '6': case '7': case '8': case '9':
*tokenType = TK_INTEGER;
for(i=1; isDigit(z[i]); i++){}
return i;
case '[':
for(i=1; z[i] && z[i-1]!=']'; i++){}
*tokenType = TK_ID;
return i;
default:
if( !isIdChar[*z] ){
break;
}
for(i=1; isIdChar[z[i]]; i++){}
*tokenType = sqliteKeywordCode(z, i);
if( *tokenType == TK_ID && z[i] == '`' ) *skip = 1;
return i;
}
*tokenType = TK_ILLEGAL;
return 1;
}