granicus.if.org Git - postgresql/blob - src/backend/parser/keywords.c

   1 /*-------------------------------------------------------------------------
   2  *
   3  * keywords.c
   4  *        lexical token lookup for reserved words in PostgreSQL
   5  *
   6  * Portions Copyright (c) 1996-2001, PostgreSQL Global Development Group
   7  * Portions Copyright (c) 1994, Regents of the University of California
   8  *
   9  *
  10  * IDENTIFICATION
  11  *        $Header: /cvsroot/pgsql/src/backend/parser/keywords.c,v 1.101 2002/03/05 05:33:15 momjian Exp $
  12  *
  13  *-------------------------------------------------------------------------
  14  */
  15 #include "postgres.h"
  16
  17 #include <ctype.h>
  18
  19 #include "nodes/parsenodes.h"
  20 #include "parser/keywords.h"
  21 #include "parser/parse.h"
  22
  23 /*
  24  * List of (keyword-name, keyword-token-value) pairs.
  25  *
  26  * !!WARNING!!: This list must be sorted, because binary
  27  *               search is used to locate entries.
  28  */
  29 static ScanKeyword ScanKeywords[] = {
  30         /* name, value */
  31         {"abort", ABORT_TRANS},
  32         {"absolute", ABSOLUTE},
  33         {"access", ACCESS},
  34         {"action", ACTION},
  35         {"add", ADD},
  36         {"after", AFTER},
  37         {"aggregate", AGGREGATE},
  38         {"all", ALL},
  39         {"alter", ALTER},
  40         {"analyse", ANALYSE},           /* British spelling */
  41         {"analyze", ANALYZE},
  42         {"and", AND},
  43         {"any", ANY},
  44         {"as", AS},
  45         {"asc", ASC},
  46         {"at", AT},
  47         {"authorization", AUTHORIZATION},
  48         {"backward", BACKWARD},
  49         {"before", BEFORE},
  50         {"begin", BEGIN_TRANS},
  51         {"between", BETWEEN},
  52         {"binary", BINARY},
  53         {"bit", BIT},
  54         {"both", BOTH},
  55         {"by", BY},
  56         {"cache", CACHE},
  57         {"cascade", CASCADE},
  58         {"case", CASE},
  59         {"cast", CAST},
  60         {"chain", CHAIN},
  61         {"char", CHAR},
  62         {"character", CHARACTER},
  63         {"characteristics", CHARACTERISTICS},
  64         {"check", CHECK},
  65         {"checkpoint", CHECKPOINT},
  66         {"close", CLOSE},
  67         {"cluster", CLUSTER},
  68         {"coalesce", COALESCE},
  69         {"collate", COLLATE},
  70         {"column", COLUMN},
  71         {"comment", COMMENT},
  72         {"commit", COMMIT},
  73         {"committed", COMMITTED},
  74         {"constraint", CONSTRAINT},
  75         {"constraints", CONSTRAINTS},
  76         {"copy", COPY},
  77         {"create", CREATE},
  78         {"createdb", CREATEDB},
  79         {"createuser", CREATEUSER},
  80         {"cross", CROSS},
  81         {"current_date", CURRENT_DATE},
  82         {"current_time", CURRENT_TIME},
  83         {"current_timestamp", CURRENT_TIMESTAMP},
  84         {"current_user", CURRENT_USER},
  85         {"cursor", CURSOR},
  86         {"cycle", CYCLE},
  87         {"database", DATABASE},
  88         {"day", DAY_P},
  89         {"dec", DEC},
  90         {"decimal", DECIMAL},
  91         {"declare", DECLARE},
  92         {"default", DEFAULT},
  93         {"deferrable", DEFERRABLE},
  94         {"deferred", DEFERRED},
  95         {"delete", DELETE},
  96         {"delimiters", DELIMITERS},
  97         {"desc", DESC},
  98         {"distinct", DISTINCT},
  99         {"do", DO},
 100         {"double", DOUBLE},
 101         {"drop", DROP},
 102         {"each", EACH},
 103         {"else", ELSE},
 104         {"encoding", ENCODING},
 105         {"encrypted", ENCRYPTED},
 106         {"end", END_TRANS},
 107         {"escape", ESCAPE},
 108         {"except", EXCEPT},
 109         {"exclusive", EXCLUSIVE},
 110         {"execute", EXECUTE},
 111         {"exists", EXISTS},
 112         {"explain", EXPLAIN},
 113         {"extract", EXTRACT},
 114         {"false", FALSE_P},
 115         {"fetch", FETCH},
 116         {"float", FLOAT},
 117         {"for", FOR},
 118         {"force", FORCE},
 119         {"foreign", FOREIGN},
 120         {"forward", FORWARD},
 121         {"freeze", FREEZE},
 122         {"from", FROM},
 123         {"full", FULL},
 124         {"function", FUNCTION},
 125         {"global", GLOBAL},
 126         {"grant", GRANT},
 127         {"group", GROUP},
 128         {"handler", HANDLER},
 129         {"having", HAVING},
 130         {"hour", HOUR_P},
 131         {"ilike", ILIKE},
 132         {"immediate", IMMEDIATE},
 133         {"in", IN},
 134         {"increment", INCREMENT},
 135         {"index", INDEX},
 136         {"inherits", INHERITS},
 137         {"initially", INITIALLY},
 138         {"inner", INNER_P},
 139         {"inout", INOUT},
 140         {"insensitive", INSENSITIVE},
 141         {"insert", INSERT},
 142         {"instead", INSTEAD},
 143         {"intersect", INTERSECT},
 144         {"interval", INTERVAL},
 145         {"into", INTO},
 146         {"is", IS},
 147         {"isnull", ISNULL},
 148         {"isolation", ISOLATION},
 149         {"join", JOIN},
 150         {"key", KEY},
 151         {"lancompiler", LANCOMPILER},
 152         {"language", LANGUAGE},
 153         {"leading", LEADING},
 154         {"left", LEFT},
 155         {"level", LEVEL},
 156         {"like", LIKE},
 157         {"limit", LIMIT},
 158         {"listen", LISTEN},
 159         {"load", LOAD},
 160         {"local", LOCAL},
 161         {"location", LOCATION},
 162         {"lock", LOCK_P},
 163         {"match", MATCH},
 164         {"maxvalue", MAXVALUE},
 165         {"minute", MINUTE_P},
 166         {"minvalue", MINVALUE},
 167         {"mode", MODE},
 168         {"month", MONTH_P},
 169         {"move", MOVE},
 170         {"names", NAMES},
 171         {"national", NATIONAL},
 172         {"natural", NATURAL},
 173         {"nchar", NCHAR},
 174         {"new", NEW},
 175         {"next", NEXT},
 176         {"no", NO},
 177         {"nocreatedb", NOCREATEDB},
 178         {"nocreateuser", NOCREATEUSER},
 179         {"none", NONE},
 180         {"not", NOT},
 181         {"nothing", NOTHING},
 182         {"notify", NOTIFY},
 183         {"notnull", NOTNULL},
 184         {"null", NULL_P},
 185         {"nullif", NULLIF},
 186         {"numeric", NUMERIC},
 187         {"of", OF},
 188         {"off", OFF},
 189         {"offset", OFFSET},
 190         {"oids", OIDS},
 191         {"old", OLD},
 192         {"on", ON},
 193         {"only", ONLY},
 194         {"operator", OPERATOR},
 195         {"option", OPTION},
 196         {"or", OR},
 197         {"order", ORDER},
 198         {"out", OUT},
 199         {"outer", OUTER_P},
 200         {"overlaps", OVERLAPS},
 201         {"owner", OWNER},
 202         {"partial", PARTIAL},
 203         {"password", PASSWORD},
 204         {"path", PATH_P},
 205         {"pendant", PENDANT},
 206         {"position", POSITION},
 207         {"precision", PRECISION},
 208         {"primary", PRIMARY},
 209         {"prior", PRIOR},
 210         {"privileges", PRIVILEGES},
 211         {"procedural", PROCEDURAL},
 212         {"procedure", PROCEDURE},
 213         {"public", PUBLIC},
 214         {"read", READ},
 215         {"references", REFERENCES},
 216         {"reindex", REINDEX},
 217         {"relative", RELATIVE},
 218         {"rename", RENAME},
 219         {"replace", REPLACE},
 220         {"reset", RESET},
 221         {"restrict", RESTRICT},
 222         {"returns", RETURNS},
 223         {"revoke", REVOKE},
 224         {"right", RIGHT},
 225         {"rollback", ROLLBACK},
 226         {"row", ROW},
 227         {"rule", RULE},
 228         {"schema", SCHEMA},
 229         {"scroll", SCROLL},
 230         {"second", SECOND_P},
 231         {"select", SELECT},
 232         {"sequence", SEQUENCE},
 233         {"serializable", SERIALIZABLE},
 234         {"session", SESSION},
 235         {"session_user", SESSION_USER},
 236         {"set", SET},
 237         {"setof", SETOF},
 238         {"share", SHARE},
 239         {"show", SHOW},
 240         {"some", SOME},
 241         {"start", START},
 242         {"statement", STATEMENT},
 243         {"statistics", STATISTICS},
 244         {"stdin", STDIN},
 245         {"stdout", STDOUT},
 246         {"storage", STORAGE},
 247         {"substring", SUBSTRING},
 248         {"sysid", SYSID},
 249         {"table", TABLE},
 250         {"temp", TEMP},
 251         {"template", TEMPLATE},
 252         {"temporary", TEMPORARY},
 253         {"then", THEN},
 254         {"time", TIME},
 255         {"timestamp", TIMESTAMP},
 256         {"to", TO},
 257         {"toast", TOAST},
 258         {"trailing", TRAILING},
 259         {"transaction", TRANSACTION},
 260         {"trigger", TRIGGER},
 261         {"trim", TRIM},
 262         {"true", TRUE_P},
 263         {"truncate", TRUNCATE},
 264         {"trusted", TRUSTED},
 265         {"type", TYPE_P},
 266         {"unencrypted", UNENCRYPTED},
 267         {"union", UNION},
 268         {"unique", UNIQUE},
 269         {"unknown", UNKNOWN},
 270         {"unlisten", UNLISTEN},
 271         {"until", UNTIL},
 272         {"update", UPDATE},
 273         {"usage", USAGE},
 274         {"user", USER},
 275         {"using", USING},
 276         {"vacuum", VACUUM},
 277         {"valid", VALID},
 278         {"values", VALUES},
 279         {"varchar", VARCHAR},
 280         {"varying", VARYING},
 281         {"verbose", VERBOSE},
 282         {"version", VERSION},
 283         {"view", VIEW},
 284         {"when", WHEN},
 285         {"where", WHERE},
 286         {"with", WITH},
 287         {"without", WITHOUT},
 288         {"work", WORK},
 289         {"year", YEAR_P},
 290         {"zone", ZONE},
 291 };
 292
 293 /*
 294  * ScanKeywordLookup - see if a given word is a keyword
 295  *
 296  * Returns a pointer to the ScanKeyword table entry, or NULL if no match.
 297  *
 298  * The match is done case-insensitively.  Note that we deliberately use a
 299  * dumbed-down case conversion that will only translate 'A'-'Z' into 'a'-'z',
 300  * even if we are in a locale where tolower() would produce more or different
 301  * translations.  This is to conform to the SQL99 spec, which says that
 302  * keywords are to be matched in this way even though non-keyword identifiers
 303  * receive a different case-normalization mapping.
 304  */
 305 ScanKeyword *
 306 ScanKeywordLookup(char *text)
 307 {
 308         int                     len,
 309                                 i;
 310         char            word[NAMEDATALEN];
 311         ScanKeyword *low;
 312         ScanKeyword *high;
 313
 314         len = strlen(text);
 315         /* We assume all keywords are shorter than NAMEDATALEN. */
 316         if (len >= NAMEDATALEN)
 317                 return NULL;
 318
 319         /*
 320          * Apply an ASCII-only downcasing.      We must not use tolower() since it
 321          * may produce the wrong translation in some locales (eg, Turkish),
 322          * and we don't trust isupper() very much either.  In an ASCII-based
 323          * encoding the tests against A and Z are sufficient, but we also
 324          * check isupper() so that we will work correctly under EBCDIC.  The
 325          * actual case conversion step should work for either ASCII or EBCDIC.
 326          */
 327         for (i = 0; i < len; i++)
 328         {
 329                 char            ch = text[i];
 330
 331                 if (ch >= 'A' && ch <= 'Z' && isupper((unsigned char) ch))
 332                         ch += 'a' - 'A';
 333                 word[i] = ch;
 334         }
 335         word[len] = '\0';
 336
 337         /*
 338          * Now do a binary search using plain strcmp() comparison.
 339          */
 340         low = &ScanKeywords[0];
 341         high = endof(ScanKeywords) - 1;
 342         while (low <= high)
 343         {
 344                 ScanKeyword *middle;
 345                 int                     difference;
 346
 347                 middle = low + (high - low) / 2;
 348                 difference = strcmp(middle->name, word);
 349                 if (difference == 0)
 350                         return middle;
 351                 else if (difference < 0)
 352                         low = middle + 1;
 353                 else
 354                         high = middle - 1;
 355         }
 356
 357         return NULL;
 358 }
 359
 360
 361 /*
 362  * This does the reverse mapping from token number to string.
 363  */
 364 const char *
 365 TokenString(int token)
 366 {
 367     int i = 0;
 368     static char buf[NAMEDATALEN];
 369
 370     while (i < sizeof(ScanKeywords))
 371     {
 372         if (ScanKeywords[i].value == token)
 373         {
 374             int k;
 375
 376             /* uppercase */
 377             for (k = 0; k < NAMEDATALEN; k++)
 378                 if (ScanKeywords[i].name[k] >= 'a'
 379                     && ScanKeywords[i].name[k] <= 'z')
 380                     buf[k] = ScanKeywords[i].name[k] + ('A' - 'a');
 381                 else
 382                     buf[k] = ScanKeywords[i].name[k];
 383
 384             return buf;
 385         }
 386
 387         i++;
 388     }
 389
 390     return NULL;
 391 }