| 2 |      1 | /*
 | 
|  |      2 | ** 2001 September 15
 | 
|  |      3 | **
 | 
|  |      4 | ** The author disclaims copyright to this source code.  In place of
 | 
|  |      5 | ** a legal notice, here is a blessing:
 | 
|  |      6 | **
 | 
|  |      7 | **    May you do good and not evil.
 | 
|  |      8 | **    May you find forgiveness for yourself and forgive others.
 | 
|  |      9 | **    May you share freely, never taking more than you give.
 | 
|  |     10 | **
 | 
|  |     11 | *************************************************************************
 | 
|  |     12 | ** An tokenizer for SQL
 | 
|  |     13 | **
 | 
|  |     14 | ** This file contains C code that implements the sqlite3_complete() API.
 | 
|  |     15 | ** This code used to be part of the tokenizer.c source file.  But by
 | 
|  |     16 | ** separating it out, the code will be automatically omitted from
 | 
|  |     17 | ** static links that do not use it.
 | 
|  |     18 | **
 | 
|  |     19 | ** $Id: complete.cpp 1282 2008-11-13 09:31:33Z LarsPson $
 | 
|  |     20 | */
 | 
|  |     21 | #include "sqliteInt.h"
 | 
|  |     22 | #ifndef SQLITE_OMIT_COMPLETE
 | 
|  |     23 | 
 | 
|  |     24 | /*
 | 
|  |     25 | ** This is defined in tokenize.c.  We just have to import the definition.
 | 
|  |     26 | */
 | 
|  |     27 | //#ifndef SQLITE_AMALGAMATION
 | 
|  |     28 | //#ifdef SQLITE_ASCII
 | 
|  |     29 | //extern const char sqlite3IsAsciiIdChar[];
 | 
|  |     30 | //#define IdChar(C)  (((c=C)&0x80)!=0 || (c>0x1f && sqlite3IsAsciiIdChar[c-0x20]))
 | 
|  |     31 | //#endif
 | 
|  |     32 | //#ifdef SQLITE_EBCDIC
 | 
|  |     33 | //extern const char sqlite3IsEbcdicIdChar[];
 | 
|  |     34 | //#define IdChar(C)  (((c=C)>=0x42 && sqlite3IsEbcdicIdChar[c-0x40]))
 | 
|  |     35 | //#endif
 | 
|  |     36 | //#endif /* SQLITE_AMALGAMATION */
 | 
|  |     37 | 
 | 
|  |     38 | 
 | 
|  |     39 | /*
 | 
|  |     40 | ** Token types used by the sqlite3_complete() routine.  See the header
 | 
|  |     41 | ** comments on that procedure for additional information.
 | 
|  |     42 | */
 | 
|  |     43 | #define tkSEMI    0
 | 
|  |     44 | #define tkWS      1
 | 
|  |     45 | #define tkOTHER   2
 | 
|  |     46 | #define tkEXPLAIN 3
 | 
|  |     47 | #define tkCREATE  4
 | 
|  |     48 | #define tkTEMP    5
 | 
|  |     49 | #define tkTRIGGER 6
 | 
|  |     50 | #define tkEND     7
 | 
|  |     51 | 
 | 
|  |     52 | /*
 | 
|  |     53 | ** Return TRUE if the given SQL string ends in a semicolon.
 | 
|  |     54 | **
 | 
|  |     55 | ** Special handling is require for CREATE TRIGGER statements.
 | 
|  |     56 | ** Whenever the CREATE TRIGGER keywords are seen, the statement
 | 
|  |     57 | ** must end with ";END;".
 | 
|  |     58 | **
 | 
|  |     59 | ** This implementation uses a state machine with 7 states:
 | 
|  |     60 | **
 | 
|  |     61 | **   (0) START     At the beginning or end of an SQL statement.  This routine
 | 
|  |     62 | **                 returns 1 if it ends in the START state and 0 if it ends
 | 
|  |     63 | **                 in any other state.
 | 
|  |     64 | **
 | 
|  |     65 | **   (1) NORMAL    We are in the middle of statement which ends with a single
 | 
|  |     66 | **                 semicolon.
 | 
|  |     67 | **
 | 
|  |     68 | **   (2) EXPLAIN   The keyword EXPLAIN has been seen at the beginning of 
 | 
|  |     69 | **                 a statement.
 | 
|  |     70 | **
 | 
|  |     71 | **   (3) CREATE    The keyword CREATE has been seen at the beginning of a
 | 
|  |     72 | **                 statement, possibly preceeded by EXPLAIN and/or followed by
 | 
|  |     73 | **                 TEMP or TEMPORARY
 | 
|  |     74 | **
 | 
|  |     75 | **   (4) TRIGGER   We are in the middle of a trigger definition that must be
 | 
|  |     76 | **                 ended by a semicolon, the keyword END, and another semicolon.
 | 
|  |     77 | **
 | 
|  |     78 | **   (5) SEMI      We've seen the first semicolon in the ";END;" that occurs at
 | 
|  |     79 | **                 the end of a trigger definition.
 | 
|  |     80 | **
 | 
|  |     81 | **   (6) END       We've seen the ";END" of the ";END;" that occurs at the end
 | 
|  |     82 | **                 of a trigger difinition.
 | 
|  |     83 | **
 | 
|  |     84 | ** Transitions between states above are determined by tokens extracted
 | 
|  |     85 | ** from the input.  The following tokens are significant:
 | 
|  |     86 | **
 | 
|  |     87 | **   (0) tkSEMI      A semicolon.
 | 
|  |     88 | **   (1) tkWS        Whitespace
 | 
|  |     89 | **   (2) tkOTHER     Any other SQL token.
 | 
|  |     90 | **   (3) tkEXPLAIN   The "explain" keyword.
 | 
|  |     91 | **   (4) tkCREATE    The "create" keyword.
 | 
|  |     92 | **   (5) tkTEMP      The "temp" or "temporary" keyword.
 | 
|  |     93 | **   (6) tkTRIGGER   The "trigger" keyword.
 | 
|  |     94 | **   (7) tkEND       The "end" keyword.
 | 
|  |     95 | **
 | 
|  |     96 | ** Whitespace never causes a state transition and is always ignored.
 | 
|  |     97 | **
 | 
|  |     98 | ** If we compile with SQLITE_OMIT_TRIGGER, all of the computation needed
 | 
|  |     99 | ** to recognize the end of a trigger can be omitted.  All we have to do
 | 
|  |    100 | ** is look for a semicolon that is not part of an string or comment.
 | 
|  |    101 | */
 | 
|  |    102 | EXPORT_C int sqlite3_complete(const char *zSql){
 | 
|  |    103 |   u8 state = 0;   /* Current state, using numbers defined in header comment */
 | 
|  |    104 |   u8 token;       /* Value of the next token */
 | 
|  |    105 | 
 | 
|  |    106 | #ifndef SQLITE_OMIT_TRIGGER
 | 
|  |    107 |   /* A complex statement machine used to detect the end of a CREATE TRIGGER
 | 
|  |    108 |   ** statement.  This is the normal case.
 | 
|  |    109 |   */
 | 
|  |    110 |   static const u8 trans[7][8] = {
 | 
|  |    111 |                      /* Token:                                                */
 | 
|  |    112 |      /* State:       **  SEMI  WS  OTHER EXPLAIN  CREATE  TEMP  TRIGGER  END  */
 | 
|  |    113 |      /* 0   START: */ {    0,  0,     1,      2,      3,    1,       1,   1,  },
 | 
|  |    114 |      /* 1  NORMAL: */ {    0,  1,     1,      1,      1,    1,       1,   1,  },
 | 
|  |    115 |      /* 2 EXPLAIN: */ {    0,  2,     1,      1,      3,    1,       1,   1,  },
 | 
|  |    116 |      /* 3  CREATE: */ {    0,  3,     1,      1,      1,    3,       4,   1,  },
 | 
|  |    117 |      /* 4 TRIGGER: */ {    5,  4,     4,      4,      4,    4,       4,   4,  },
 | 
|  |    118 |      /* 5    SEMI: */ {    5,  5,     4,      4,      4,    4,       4,   6,  },
 | 
|  |    119 |      /* 6     END: */ {    0,  6,     4,      4,      4,    4,       4,   4,  },
 | 
|  |    120 |   };
 | 
|  |    121 | #else
 | 
|  |    122 |   /* If triggers are not suppored by this compile then the statement machine
 | 
|  |    123 |   ** used to detect the end of a statement is much simplier
 | 
|  |    124 |   */
 | 
|  |    125 |   static const u8 trans[2][3] = {
 | 
|  |    126 |                      /* Token:           */
 | 
|  |    127 |      /* State:       **  SEMI  WS  OTHER */
 | 
|  |    128 |      /* 0   START: */ {    0,  0,     1, },
 | 
|  |    129 |      /* 1  NORMAL: */ {    0,  1,     1, },
 | 
|  |    130 |   };
 | 
|  |    131 | #endif /* SQLITE_OMIT_TRIGGER */
 | 
|  |    132 | 
 | 
|  |    133 |   while( *zSql ){
 | 
|  |    134 |     switch( *zSql ){
 | 
|  |    135 |       case ';': {  /* A semicolon */
 | 
|  |    136 |         token = tkSEMI;
 | 
|  |    137 |         break;
 | 
|  |    138 |       }
 | 
|  |    139 |       case ' ':
 | 
|  |    140 |       case '\r':
 | 
|  |    141 |       case '\t':
 | 
|  |    142 |       case '\n':
 | 
|  |    143 |       case '\f': {  /* White space is ignored */
 | 
|  |    144 |         token = tkWS;
 | 
|  |    145 |         break;
 | 
|  |    146 |       }
 | 
|  |    147 |       case '/': {   /* C-style comments */
 | 
|  |    148 |         if( zSql[1]!='*' ){
 | 
|  |    149 |           token = tkOTHER;
 | 
|  |    150 |           break;
 | 
|  |    151 |         }
 | 
|  |    152 |         zSql += 2;
 | 
|  |    153 |         while( zSql[0] && (zSql[0]!='*' || zSql[1]!='/') ){ zSql++; }
 | 
|  |    154 |         if( zSql[0]==0 ) return 0;
 | 
|  |    155 |         zSql++;
 | 
|  |    156 |         token = tkWS;
 | 
|  |    157 |         break;
 | 
|  |    158 |       }
 | 
|  |    159 |       case '-': {   /* SQL-style comments from "--" to end of line */
 | 
|  |    160 |         if( zSql[1]!='-' ){
 | 
|  |    161 |           token = tkOTHER;
 | 
|  |    162 |           break;
 | 
|  |    163 |         }
 | 
|  |    164 |         while( *zSql && *zSql!='\n' ){ zSql++; }
 | 
|  |    165 |         if( *zSql==0 ) return state==0;
 | 
|  |    166 |         token = tkWS;
 | 
|  |    167 |         break;
 | 
|  |    168 |       }
 | 
|  |    169 |       case '[': {   /* Microsoft-style identifiers in [...] */
 | 
|  |    170 |         zSql++;
 | 
|  |    171 |         while( *zSql && *zSql!=']' ){ zSql++; }
 | 
|  |    172 |         if( *zSql==0 ) return 0;
 | 
|  |    173 |         token = tkOTHER;
 | 
|  |    174 |         break;
 | 
|  |    175 |       }
 | 
|  |    176 |       case '`':     /* Grave-accent quoted symbols used by MySQL */
 | 
|  |    177 |       case '"':     /* single- and double-quoted strings */
 | 
|  |    178 |       case '\'': {
 | 
|  |    179 |         int c = *zSql;
 | 
|  |    180 |         zSql++;
 | 
|  |    181 |         while( *zSql && *zSql!=c ){ zSql++; }
 | 
|  |    182 |         if( *zSql==0 ) return 0;
 | 
|  |    183 |         token = tkOTHER;
 | 
|  |    184 |         break;
 | 
|  |    185 |       }
 | 
|  |    186 |       default: {
 | 
|  |    187 |         int c;
 | 
|  |    188 |         if( IdChar((u8)*zSql) ){
 | 
|  |    189 |           /* Keywords and unquoted identifiers */
 | 
|  |    190 |           int nId;
 | 
|  |    191 |           for(nId=1; IdChar(zSql[nId]); nId++){}
 | 
|  |    192 | #ifdef SQLITE_OMIT_TRIGGER
 | 
|  |    193 |           token = tkOTHER;
 | 
|  |    194 | #else
 | 
|  |    195 |           switch( *zSql ){
 | 
|  |    196 |             case 'c': case 'C': {
 | 
|  |    197 |               if( nId==6 && sqlite3StrNICmp(zSql, "create", 6)==0 ){
 | 
|  |    198 |                 token = tkCREATE;
 | 
|  |    199 |               }else{
 | 
|  |    200 |                 token = tkOTHER;
 | 
|  |    201 |               }
 | 
|  |    202 |               break;
 | 
|  |    203 |             }
 | 
|  |    204 |             case 't': case 'T': {
 | 
|  |    205 |               if( nId==7 && sqlite3StrNICmp(zSql, "trigger", 7)==0 ){
 | 
|  |    206 |                 token = tkTRIGGER;
 | 
|  |    207 |               }else if( nId==4 && sqlite3StrNICmp(zSql, "temp", 4)==0 ){
 | 
|  |    208 |                 token = tkTEMP;
 | 
|  |    209 |               }else if( nId==9 && sqlite3StrNICmp(zSql, "temporary", 9)==0 ){
 | 
|  |    210 |                 token = tkTEMP;
 | 
|  |    211 |               }else{
 | 
|  |    212 |                 token = tkOTHER;
 | 
|  |    213 |               }
 | 
|  |    214 |               break;
 | 
|  |    215 |             }
 | 
|  |    216 |             case 'e':  case 'E': {
 | 
|  |    217 |               if( nId==3 && sqlite3StrNICmp(zSql, "end", 3)==0 ){
 | 
|  |    218 |                 token = tkEND;
 | 
|  |    219 |               }else
 | 
|  |    220 | #ifndef SQLITE_OMIT_EXPLAIN
 | 
|  |    221 |               if( nId==7 && sqlite3StrNICmp(zSql, "explain", 7)==0 ){
 | 
|  |    222 |                 token = tkEXPLAIN;
 | 
|  |    223 |               }else
 | 
|  |    224 | #endif
 | 
|  |    225 |               {
 | 
|  |    226 |                 token = tkOTHER;
 | 
|  |    227 |               }
 | 
|  |    228 |               break;
 | 
|  |    229 |             }
 | 
|  |    230 |             default: {
 | 
|  |    231 |               token = tkOTHER;
 | 
|  |    232 |               break;
 | 
|  |    233 |             }
 | 
|  |    234 |           }
 | 
|  |    235 | #endif /* SQLITE_OMIT_TRIGGER */
 | 
|  |    236 |           zSql += nId-1;
 | 
|  |    237 |         }else{
 | 
|  |    238 |           /* Operators and special symbols */
 | 
|  |    239 |           token = tkOTHER;
 | 
|  |    240 |         }
 | 
|  |    241 |         break;
 | 
|  |    242 |       }
 | 
|  |    243 |     }
 | 
|  |    244 |     state = trans[state][token];
 | 
|  |    245 |     zSql++;
 | 
|  |    246 |   }
 | 
|  |    247 |   return state==0;
 | 
|  |    248 | }
 | 
|  |    249 | 
 | 
|  |    250 | #ifndef SQLITE_OMIT_UTF16
 | 
|  |    251 | /*
 | 
|  |    252 | ** This routine is the same as the sqlite3_complete() routine described
 | 
|  |    253 | ** above, except that the parameter is required to be UTF-16 encoded, not
 | 
|  |    254 | ** UTF-8.
 | 
|  |    255 | */
 | 
|  |    256 | EXPORT_C int sqlite3_complete16(const void *zSql){
 | 
|  |    257 |   sqlite3_value *pVal;
 | 
|  |    258 |   char const *zSql8;
 | 
|  |    259 |   int rc = SQLITE_NOMEM;
 | 
|  |    260 | 
 | 
|  |    261 |   pVal = sqlite3ValueNew(0);
 | 
|  |    262 |   sqlite3ValueSetStr(pVal, -1, zSql, SQLITE_UTF16NATIVE, SQLITE_STATIC);
 | 
|  |    263 |   zSql8 = (const char*)sqlite3ValueText(pVal, SQLITE_UTF8);
 | 
|  |    264 |   if( zSql8 ){
 | 
|  |    265 |     rc = sqlite3_complete(zSql8);
 | 
|  |    266 |   }
 | 
|  |    267 |   sqlite3ValueFree(pVal);
 | 
|  |    268 |   return sqlite3ApiExit(0, rc);
 | 
|  |    269 | }
 | 
|  |    270 | #endif /* SQLITE_OMIT_UTF16 */
 | 
|  |    271 | #endif /* SQLITE_OMIT_COMPLETE */
 |