?? c-lex.c
字號:
/* Lexical analyzer for C and Objective C. Copyright (C) 1987, 88, 89, 92, 94, 1995 Free Software Foundation, Inc.This file is part of GNU CC.GNU CC is free software; you can redistribute it and/or modifyit under the terms of the GNU General Public License as published bythe Free Software Foundation; either version 2, or (at your option)any later version.GNU CC is distributed in the hope that it will be useful,but WITHOUT ANY WARRANTY; without even the implied warranty ofMERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See theGNU General Public License for more details.You should have received a copy of the GNU General Public Licensealong with GNU CC; see the file COPYING. If not, write tothe Free Software Foundation, 59 Temple Place - Suite 330,Boston, MA 02111-1307, USA. */#include <stdio.h>#include <errno.h>#include <setjmp.h>#include "config.h"#include "rtl.h"#include "tree.h"#include "input.h"#include "c-lex.h"#include "c-tree.h"#include "flags.h"#include "c-parse.h"#include "c-pragma.h"#include <ctype.h>#ifdef MULTIBYTE_CHARS#include <stdlib.h>#include <locale.h>#endif#ifndef errnoextern int errno;#endif/* The elements of `ridpointers' are identifier nodes for the reserved type names and storage classes. It is indexed by a RID_... value. */tree ridpointers[(int) RID_MAX];/* Cause the `yydebug' variable to be defined. */#define YYDEBUG 1/* the declaration found for the last IDENTIFIER token read in. yylex must look this up to detect typedefs, which get token type TYPENAME, so it is left around in case the identifier is not a typedef but is used in a context which makes it a reference to a variable. */tree lastiddecl;/* Nonzero enables objc features. */int doing_objc_thang;extern tree is_class_name ();extern int yydebug;/* File used for outputting assembler code. */extern FILE *asm_out_file;#ifndef WCHAR_TYPE_SIZE#ifdef INT_TYPE_SIZE#define WCHAR_TYPE_SIZE INT_TYPE_SIZE#else#define WCHAR_TYPE_SIZE BITS_PER_WORD#endif#endif/* Number of bytes in a wide character. */#define WCHAR_BYTES (WCHAR_TYPE_SIZE / BITS_PER_UNIT)static int maxtoken; /* Current nominal length of token buffer. */char *token_buffer; /* Pointer to token buffer. Actual allocated length is maxtoken + 2. This is not static because objc-parse.y uses it. *//* Nonzero if end-of-file has been seen on input. */static int end_of_file;/* Buffered-back input character; faster than using ungetc. */static int nextchar = -1;int check_newline ();/* Do not insert generated code into the source, instead, include it. This allows us to build gcc automatically even for targets that need to add or modify the reserved keyword lists. */#include "c-gperf.h"/* Return something to represent absolute declarators containing a *. TARGET is the absolute declarator that the * contains. TYPE_QUALS is a list of modifiers such as const or volatile to apply to the pointer type, represented as identifiers. We return an INDIRECT_REF whose "contents" are TARGET and whose type is the modifier list. */treemake_pointer_declarator (type_quals, target) tree type_quals, target;{ return build1 (INDIRECT_REF, type_quals, target);}voidforget_protocol_qualifiers (){ int i, n = sizeof wordlist / sizeof (struct resword); for (i = 0; i < n; i++) if ((int) wordlist[i].rid >= (int) RID_IN && (int) wordlist[i].rid <= (int) RID_ONEWAY) wordlist[i].name = "";}voidremember_protocol_qualifiers (){ int i, n = sizeof wordlist / sizeof (struct resword); for (i = 0; i < n; i++) if (wordlist[i].rid == RID_IN) wordlist[i].name = "in"; else if (wordlist[i].rid == RID_OUT) wordlist[i].name = "out"; else if (wordlist[i].rid == RID_INOUT) wordlist[i].name = "inout"; else if (wordlist[i].rid == RID_BYCOPY) wordlist[i].name = "bycopy"; else if (wordlist[i].rid == RID_ONEWAY) wordlist[i].name = "oneway"; }voidinit_lex (){ /* Make identifier nodes long enough for the language-specific slots. */ set_identifier_size (sizeof (struct lang_identifier)); /* Start it at 0, because check_newline is called at the very beginning and will increment it to 1. */ lineno = 0;#ifdef MULTIBYTE_CHARS /* Change to the native locale for multibyte conversions. */ setlocale (LC_CTYPE, "");#endif maxtoken = 40; token_buffer = (char *) xmalloc (maxtoken + 2); ridpointers[(int) RID_INT] = get_identifier ("int"); ridpointers[(int) RID_CHAR] = get_identifier ("char"); ridpointers[(int) RID_VOID] = get_identifier ("void"); ridpointers[(int) RID_FLOAT] = get_identifier ("float"); ridpointers[(int) RID_DOUBLE] = get_identifier ("double"); ridpointers[(int) RID_SHORT] = get_identifier ("short"); ridpointers[(int) RID_LONG] = get_identifier ("long"); ridpointers[(int) RID_UNSIGNED] = get_identifier ("unsigned"); ridpointers[(int) RID_SIGNED] = get_identifier ("signed"); ridpointers[(int) RID_INLINE] = get_identifier ("inline"); ridpointers[(int) RID_CONST] = get_identifier ("const"); ridpointers[(int) RID_VOLATILE] = get_identifier ("volatile"); ridpointers[(int) RID_AUTO] = get_identifier ("auto"); ridpointers[(int) RID_STATIC] = get_identifier ("static"); ridpointers[(int) RID_EXTERN] = get_identifier ("extern"); ridpointers[(int) RID_TYPEDEF] = get_identifier ("typedef"); ridpointers[(int) RID_REGISTER] = get_identifier ("register"); ridpointers[(int) RID_ITERATOR] = get_identifier ("iterator"); ridpointers[(int) RID_COMPLEX] = get_identifier ("complex"); ridpointers[(int) RID_ID] = get_identifier ("id"); ridpointers[(int) RID_IN] = get_identifier ("in"); ridpointers[(int) RID_OUT] = get_identifier ("out"); ridpointers[(int) RID_INOUT] = get_identifier ("inout"); ridpointers[(int) RID_BYCOPY] = get_identifier ("bycopy"); ridpointers[(int) RID_ONEWAY] = get_identifier ("oneway"); forget_protocol_qualifiers(); /* Some options inhibit certain reserved words. Clear those words out of the hash table so they won't be recognized. */#define UNSET_RESERVED_WORD(STRING) \ do { struct resword *s = is_reserved_word (STRING, sizeof (STRING) - 1); \ if (s) s->name = ""; } while (0) if (! doing_objc_thang) UNSET_RESERVED_WORD ("id"); if (flag_traditional) { UNSET_RESERVED_WORD ("const"); UNSET_RESERVED_WORD ("volatile"); UNSET_RESERVED_WORD ("typeof"); UNSET_RESERVED_WORD ("signed"); UNSET_RESERVED_WORD ("inline"); UNSET_RESERVED_WORD ("iterator"); UNSET_RESERVED_WORD ("complex"); } if (flag_no_asm) { UNSET_RESERVED_WORD ("asm"); UNSET_RESERVED_WORD ("typeof"); UNSET_RESERVED_WORD ("inline"); UNSET_RESERVED_WORD ("iterator"); UNSET_RESERVED_WORD ("complex"); }}voidreinit_parse_for_function (){}/* Function used when yydebug is set, to print a token in more detail. */voidyyprint (file, yychar, yylval) FILE *file; int yychar; YYSTYPE yylval;{ tree t; switch (yychar) { case IDENTIFIER: case TYPENAME: case OBJECTNAME: t = yylval.ttype; if (IDENTIFIER_POINTER (t)) fprintf (file, " `%s'", IDENTIFIER_POINTER (t)); break; case CONSTANT: t = yylval.ttype; if (TREE_CODE (t) == INTEGER_CST) fprintf (file,#if HOST_BITS_PER_WIDE_INT == 64#if HOST_BITS_PER_WIDE_INT != HOST_BITS_PER_INT " 0x%lx%016lx",#else " 0x%x%016x",#endif#else#if HOST_BITS_PER_WIDE_INT != HOST_BITS_PER_INT " 0x%lx%08lx",#else " 0x%x%08x",#endif#endif TREE_INT_CST_HIGH (t), TREE_INT_CST_LOW (t)); break; }}/* If C is not whitespace, return C. Otherwise skip whitespace and return first nonwhite char read. */static intskip_white_space (c) register int c;{ static int newline_warning = 0; for (;;) { switch (c) { /* We don't recognize comments here, because cpp output can include / and * consecutively as operators. Also, there's no need, since cpp removes all comments. */ case '\n': c = check_newline (); break; case ' ': case '\t': case '\f': case '\v': case '\b': c = getc (finput); break; case '\r': /* ANSI C says the effects of a carriage return in a source file are undefined. */ if (pedantic && !newline_warning) { warning ("carriage return in source file"); warning ("(we only warn about the first carriage return)"); newline_warning = 1; } c = getc (finput); break; case '\\': c = getc (finput); if (c == '\n') lineno++; else error ("stray '\\' in program"); c = getc (finput); break; default: return (c); } }}/* Skips all of the white space at the current location in the input file. Must use and reset nextchar if it has the next character. */voidposition_after_white_space (){ register int c; if (nextchar != -1) c = nextchar, nextchar = -1; else c = getc (finput); ungetc (skip_white_space (c), finput);}/* Make the token buffer longer, preserving the data in it. P should point to just beyond the last valid character in the old buffer. The value we return is a pointer to the new buffer at a place corresponding to P. */static char *extend_token_buffer (p) char *p;{ int offset = p - token_buffer; maxtoken = maxtoken * 2 + 10; token_buffer = (char *) xrealloc (token_buffer, maxtoken + 2); return token_buffer + offset;}/* At the beginning of a line, increment the line number and process any #-directive on this line. If the line is a #-directive, read the entire line and return a newline. Otherwise, return the line's first non-whitespace character. */intcheck_newline (){ register int c; register int token; lineno++; /* Read first nonwhite char on the line. */ c = getc (finput); while (c == ' ' || c == '\t') c = getc (finput); if (c != '#') { /* If not #, return it so caller will use it. */ return c; } /* Read first nonwhite char after the `#'. */ c = getc (finput); while (c == ' ' || c == '\t') c = getc (finput); /* If a letter follows, then if the word here is `line', skip it and ignore it; otherwise, ignore the line, with an error if the word isn't `pragma', `ident', `define', or `undef'. */ if ((c >= 'a' && c <= 'z') || (c >= 'A' && c <= 'Z')) { if (c == 'p') { if (getc (finput) == 'r' && getc (finput) == 'a' && getc (finput) == 'g' && getc (finput) == 'm' && getc (finput) == 'a' && ((c = getc (finput)) == ' ' || c == '\t' || c == '\n')) {#ifdef HANDLE_SYSV_PRAGMA return handle_sysv_pragma (finput, c);#else /* !HANDLE_SYSV_PRAGMA */#ifdef HANDLE_PRAGMA HANDLE_PRAGMA (finput);#endif /* HANDLE_PRAGMA */ goto skipline;#endif /* !HANDLE_SYSV_PRAGMA */ } } else if (c == 'd') { if (getc (finput) == 'e' && getc (finput) == 'f' && getc (finput) == 'i' && getc (finput) == 'n' && getc (finput) == 'e' && ((c = getc (finput)) == ' ' || c == '\t' || c == '\n')) {#ifdef DWARF_DEBUGGING_INFO if ((debug_info_level == DINFO_LEVEL_VERBOSE) && (write_symbols == DWARF_DEBUG)) dwarfout_define (lineno, get_directive_line (finput));#endif /* DWARF_DEBUGGING_INFO */ goto skipline; } } else if (c == 'u') { if (getc (finput) == 'n' && getc (finput) == 'd' && getc (finput) == 'e' && getc (finput) == 'f' && ((c = getc (finput)) == ' ' || c == '\t' || c == '\n')) {#ifdef DWARF_DEBUGGING_INFO if ((debug_info_level == DINFO_LEVEL_VERBOSE) && (write_symbols == DWARF_DEBUG)) dwarfout_undef (lineno, get_directive_line (finput));#endif /* DWARF_DEBUGGING_INFO */ goto skipline; } } else if (c == 'l') { if (getc (finput) == 'i' && getc (finput) == 'n' && getc (finput) == 'e' && ((c = getc (finput)) == ' ' || c == '\t')) goto linenum; } else if (c == 'i') { if (getc (finput) == 'd' && getc (finput) == 'e' && getc (finput) == 'n' && getc (finput) == 't' && ((c = getc (finput)) == ' ' || c == '\t')) { /* #ident. The pedantic warning is now in cccp.c. */ /* Here we have just seen `#ident '. A string constant should follow. */ while (c == ' ' || c == '\t') c = getc (finput); /* If no argument, ignore the line. */ if (c == '\n') return c; ungetc (c, finput); token = yylex (); if (token != STRING || TREE_CODE (yylval.ttype) != STRING_CST) { error ("invalid #ident"); goto skipline; } if (!flag_no_ident) {#ifdef ASM_OUTPUT_IDENT ASM_OUTPUT_IDENT (asm_out_file, TREE_STRING_POINTER (yylval.ttype));#endif } /* Skip the rest of this line. */ goto skipline; } } error ("undefined or invalid # directive"); goto skipline; }linenum: /* Here we have either `#line' or `# <nonletter>'. In either case, it should be a line number; a digit should follow. */ while (c == ' ' || c == '\t') c = getc (finput); /* If the # is the only nonwhite char on the line, just ignore it. Check the new newline. */
?? 快捷鍵說明
復制代碼
Ctrl + C
搜索代碼
Ctrl + F
全屏模式
F11
切換主題
Ctrl + Shift + D
顯示快捷鍵
?
增大字號
Ctrl + =
減小字號
Ctrl + -