gcc/java/lex.h

   1 /* Language lexer definitions for the GNU compiler for the Java(TM) language.
   2    Copyright (C) 1997, 1998, 1999, 2000, 2001, 2002, 2003
   3    Free Software Foundation, Inc.
   4    Contributed by Alexandre Petit-Bianco (apbianco@cygnus.com)
   5
   6 This file is part of GCC.
   7
   8 GCC is free software; you can redistribute it and/or modify
   9 it under the terms of the GNU General Public License as published by
  10 the Free Software Foundation; either version 2, or (at your option)
  11 any later version.
  12
  13 GCC is distributed in the hope that it will be useful,
  14 but WITHOUT ANY WARRANTY; without even the implied warranty of
  15 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
  16 GNU General Public License for more details.
  17
  18 You should have received a copy of the GNU General Public License
  19 along with GCC; see the file COPYING.  If not, write to
  20 the Free Software Foundation, 59 Temple Place - Suite 330,
  21 Boston, MA 02111-1307, USA.
  22
  23 Java and all Java-based marks are trademarks or registered trademarks
  24 of Sun Microsystems, Inc. in the United States and other countries.
  25 The Free Software Foundation is independent of Sun Microsystems, Inc.  */
  26
  27 #ifndef GCC_JAVA_LEX_H
  28 #define GCC_JAVA_LEX_H
  29
  30 /* Extern global variables declarations  */
  31 extern FILE *finput;
  32 extern int   lineno;
  33
  34 /* A Unicode character, as read from the input file  */
  35 typedef unsigned short unicode_t;
  36
  37 #ifdef HAVE_ICONV
  38 #include <iconv.h>
  39 #endif /* HAVE_ICONV */
  40
  41 /* Default encoding to use if no encoding is specified.  */
  42 #define DEFAULT_ENCODING "UTF-8"
  43
  44 /* Debug macro to print-out what we match  */
  45 #ifdef JAVA_LEX_DEBUG
  46 #ifdef JAVA_LEX_DEBUG_CHAR
  47 #define JAVA_LEX_CHAR(c)      printf ("java_lex:%d: char '%c'.%d\n",    \
  48                                       lineno, (c < 128 ? c : '.'), c);
  49 #else
  50 #define JAVA_LEX_CHAR(c)
  51 #endif
  52 #define JAVA_LEX_KW(c)        printf ("java_lex:%d: keyword: '%s'\n", lineno,c)
  53 #define JAVA_LEX_ID(s)        printf ("java_lex:%d: ID: '%s'\n",        \
  54                                       lineno,                           \
  55                                       (all_ascii ? s : "<U>"))
  56 #define JAVA_LEX_LIT(s, r)    printf ("java_lex:%d: literal '%s'_%d\n", \
  57                                       lineno, s, r)
  58 #define JAVA_LEX_CHAR_LIT(s)  printf ("java_lex:%d: literal '%d'\n", lineno, s)
  59 #define JAVA_LEX_STR_LIT(s)   {                                          \
  60                                  int i;                                  \
  61                                  printf ("java_lex:%d: literal '%s'\n",  \
  62                                          lineno, s);                     \
  63                                }
  64 #define JAVA_LEX_SEP(c)       printf ("java_lex:%d: separator '%c'\n",lineno,c)
  65 #define JAVA_LEX_OP(c)        printf ("java_lex:%d: operator '%s'\n", lineno,c)
  66 #else
  67 #define JAVA_LEX_CHAR(c)
  68 #define JAVA_LEX_KW(c)
  69 #define JAVA_LEX_ID(s)
  70 #define JAVA_LEX_LIT(s,r)
  71 #define JAVA_LEX_CHAR_LIT(s)
  72 #define JAVA_LEX_STR_LIT(s)
  73 #define JAVA_LEX_SEP(c)
  74 #define JAVA_LEX_OP(s)
  75 #endif
  76
  77 /* Line information containers  */
  78 struct java_line {
  79   unicode_t *line;              /* The line's unicode */
  80   char      *unicode_escape_p;  /* The matching char was a unicode escape */
  81   unicode_t ahead[1];           /* Character ahead */
  82   char unicode_escape_ahead_p;  /* Character ahead is a unicode escape */
  83   int max;                      /* buffer's max size */
  84   int size;                     /* number of unicodes */
  85   int current;                  /* Current position, unicode based */
  86   int char_col;                 /* Current position, input char based */
  87   int lineno;                   /* Its line number */
  88   int white_space_only;         /* If it contains only white spaces */
  89 };
  90 #define JAVA_COLUMN_DELTA(p)                                            \
  91   (ctxp->c_line->unicode_escape_p [ctxp->c_line->current+(p)] ? 6 :     \
  92    (ctxp->c_line->line [ctxp->c_line->current+(p)] == '\t' ? 8 : 1))
  93
  94 struct java_error {
  95   struct java_line *line;
  96   int error;
  97 };
  98
  99 typedef struct java_lc_s GTY(()) {
 100   int line;
 101   int prev_col;
 102   int col;
 103 } java_lc;
 104
 105 struct java_lexer
 106 {
 107   /* The file from which we're reading.  */
 108   FILE *finput;
 109
 110   /* Number of consecutive backslashes we've read.  */
 111   int bs_count;
 112
 113   /* If nonzero, a value that was pushed back.  */
 114   unicode_t unget_value;
 115
 116   /* If nonzero, we've hit EOF.  Used only by java_get_unicode().  */
 117   int hit_eof : 1;
 118
 119 #ifdef HAVE_ICONV
 120   /* Nonzero if we've read any bytes.  We only recognize the
 121      byte-order-marker (BOM) as the first word.  */
 122   int read_anything : 1;
 123
 124   /* Nonzero if we have to byte swap.  */
 125   int byte_swap : 1;
 126
 127   /* Nonzero if we're using the fallback decoder.  */
 128   int use_fallback : 1;
 129
 130   /* The handle for the iconv converter we're using.  */
 131   iconv_t handle;
 132
 133   /* Bytes we've read from the file but have not sent to iconv.  */
 134   char buffer[1024];
 135
 136   /* Index of first valid character in buffer, -1 if no valid
 137      characters.  */
 138   int first;
 139
 140   /* Index of last valid character in buffer, plus one.  -1 if no
 141      valid characters in buffer.  */
 142   int last;
 143
 144   /* This is a buffer of characters already converted by iconv.  We
 145      use `char' here because we're assuming that iconv() converts to
 146      UCS-2, and then we convert it ourselves.  */
 147   unsigned char out_buffer[1024];
 148
 149   /* Index of first valid output character.  -1 if no valid
 150      characters.  */
 151   int out_first;
 152
 153   /* Index of last valid output character, plus one.  -1 if no valid
 154      characters.  */
 155   int out_last;
 156
 157 #endif /* HAVE_ICONV */
 158 };
 159 typedef struct java_lexer java_lexer;
 160
 161 /* Destroy a lexer object.  */
 162 extern void java_destroy_lexer (java_lexer *);
 163
 164 #define JAVA_LINE_MAX 80
 165
 166 /* Build a location compound integer */
 167 #define BUILD_LOCATION() ((ctxp->elc.line << 12) | (ctxp->elc.col & 0xfff))
 168
 169 /* Those macros are defined differently if we compile jc1-lite
 170    (JC1_LITE defined) or jc1.  */
 171 #ifdef JC1_LITE
 172
 173 #define DCONST0 0
 174 #define REAL_VALUE_TYPE int
 175 #define GET_IDENTIFIER(S) xstrdup ((S))
 176 #define REAL_VALUE_ATOF(LIT,MODE) 0
 177 #define REAL_VALUE_ISINF(VALUE)   0
 178 #define REAL_VALUE_ISNAN(VALUE)   0
 179 #define SET_REAL_VALUE_ATOF(TARGET,SOURCE)
 180 #define FLOAT_TYPE_NODE 0
 181 #define DOUBLE_TYPE_NODE 0
 182 #define SET_MODIFIER_CTX(TOKEN) java_lval->value = (TOKEN)
 183 #define GET_TYPE_PRECISION(NODE) 4
 184 #define BUILD_OPERATOR(TOKEN)   return TOKEN
 185 #define BUILD_OPERATOR2(TOKEN)  return ASSIGN_ANY_TK
 186 #define SET_LVAL_NODE(NODE)
 187 #define SET_LVAL_NODE_TYPE(NODE, TYPE)
 188 #define BUILD_ID_WFL(EXP) (EXP)
 189 #define JAVA_FLOAT_RANGE_ERROR(S) {}
 190 #define JAVA_INTEGRAL_RANGE_ERROR(S) do { } while (0)
 191
 192 #else
 193
 194 #define DCONST0 dconst0
 195 #define GET_IDENTIFIER(S) get_identifier ((S))
 196 #define SET_REAL_VALUE_ATOF(TARGET,SOURCE) (TARGET) = (SOURCE)
 197 #define FLOAT_TYPE_NODE float_type_node
 198 #define DOUBLE_TYPE_NODE double_type_node
 199 /* Set modifier_ctx according to TOKEN */
 200 #define SET_MODIFIER_CTX(TOKEN)                                            \
 201   {                                                                        \
 202     ctxp->modifier_ctx [(TOKEN)-PUBLIC_TK] = build_wfl_node (NULL_TREE); \
 203     java_lval->value = (TOKEN)-PUBLIC_TK;                                  \
 204   }
 205 /* Type precision for long */
 206 #define GET_TYPE_PRECISION(NODE) TYPE_PRECISION (long_type_node) / 8;
 207 /* Build an operator tree node and return TOKEN */
 208 #define BUILD_OPERATOR(TOKEN)                           \
 209   {                                                     \
 210     java_lval->operator.token = (TOKEN);                \
 211     java_lval->operator.location = BUILD_LOCATION();    \
 212     return (TOKEN);                                     \
 213   }
 214
 215 /* Build an operator tree node but return ASSIGN_ANY_TK */
 216 #define BUILD_OPERATOR2(TOKEN)                          \
 217   {                                                     \
 218     java_lval->operator.token = (TOKEN);                \
 219     java_lval->operator.location = BUILD_LOCATION();    \
 220     return ASSIGN_ANY_TK;                               \
 221   }
 222 /* Set java_lval->node and TREE_TYPE(java_lval->node) in macros */
 223 #define SET_LVAL_NODE(NODE) java_lval->node = (NODE)
 224 #define SET_LVAL_NODE_TYPE(NODE,TYPE)           \
 225   {                                             \
 226     java_lval->node = (NODE);                   \
 227     TREE_TYPE (java_lval->node) = (TYPE);       \
 228   }
 229 /* Wrap identifier around a wfl */
 230 #define BUILD_ID_WFL(EXP) build_wfl_node ((EXP))
 231 /* Special ways to report error on numeric literals  */
 232 #define JAVA_FLOAT_RANGE_ERROR(m)                                         \
 233   {                                                                       \
 234     char msg [1024];                                                      \
 235     int i = ctxp->c_line->current;                                        \
 236     ctxp->c_line->current = number_beginning;                             \
 237     sprintf (msg, "Floating point literal exceeds range of `%s'", (m)); \
 238     java_lex_error (msg, 0);                                              \
 239     ctxp->c_line->current = i;                                            \
 240   }
 241 #define JAVA_INTEGRAL_RANGE_ERROR(m)            \
 242   do {                                          \
 243     int i = ctxp->c_line->current;              \
 244     ctxp->c_line->current = number_beginning;   \
 245     java_lex_error (m, 0);                      \
 246     ctxp->c_line->current = i;                  \
 247   } while (0)
 248
 249 #endif /* Definitions for jc1 compilation only */
 250
 251 /* Macros to decode character ranges */
 252 #define RANGE(c, l, h)           (((c) >= l && (c) <= h))
 253 #define JAVA_WHITE_SPACE_P(c) (c == ' ' || c == '\t' || c == '\f')
 254 #define JAVA_START_CHAR_P(c) ((c < 128                                        \
 255                                && (ISIDST (c) || c == '$'))                   \
 256                               || (c >= 128 && java_start_char_p (c)))
 257 #define JAVA_PART_CHAR_P(c) ((c < 128                                         \
 258                                && (ISIDNUM (c)                                \
 259                                    || c == '$'                                \
 260                                    || c == 0x0000                             \
 261                                    || RANGE (c, 0x01, 0x08)                   \
 262                                    || RANGE (c, 0x0e, 0x1b)                   \
 263                                    || c == 0x7f))                             \
 264                               || (c >= 128 && java_part_char_p (c)))
 265 #define JAVA_ASCII_DIGIT(c)    ISDIGIT (c)
 266 #define JAVA_ASCII_OCTDIGIT(c) RANGE (c, '0', '7')
 267 #define JAVA_ASCII_HEXDIGIT(c) ISXDIGIT (c)
 268 #define JAVA_ASCII_FPCHAR(c)   (RANGE (c, 'd', 'f') || RANGE (c, 'D', 'F') || \
 269                                 c == '.' || JAVA_ASCII_DIGIT (c))
 270 #define JAVA_FP_SUFFIX(c)      (c == 'D' || c == 'd' || c == 'f' || c == 'F')
 271 #define JAVA_FP_EXP(c)         (c == 'E' || c == 'F')
 272 #define JAVA_FP_PM(c)          (c == '-' || c == '+')
 273 #define JAVA_ASCII_LETTER(c)   ISALPHA (c)
 274
 275 /* Constants  */
 276 #define JAVA_READ_BUFFER 256
 277 #define JAVA_CHAR_ERROR -2
 278 #define UEOF -1
 279
 280 #endif /* ! GCC_JAVA_LEX_H */