Subversion Repositories tendra.SVN

/*

 * Copyright (c) 2002-2005 The TenDRA Project <http://www.tendra.org/>.

 * All rights reserved.

 *

 * Redistribution and use in source and binary forms, with or without

 * modification, are permitted provided that the following conditions are met:

 *

 * 1. Redistributions of source code must retain the above copyright notice,

 *    this list of conditions and the following disclaimer.

 * 2. Redistributions in binary form must reproduce the above copyright notice,

 *    this list of conditions and the following disclaimer in the documentation

 *    and/or other materials provided with the distribution.

 * 3. Neither the name of The TenDRA Project nor the names of its contributors

 *    may be used to endorse or promote products derived from this software

 *    without specific, prior written permission.

 *

 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS ``AS

 * IS'' AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO,

 * THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR

 * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL THE COPYRIGHT HOLDERS OR

 * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,

 * EXEMPLARY OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,

 * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;

 * OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,

 * WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR

 * OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF

 * ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

 *

 * $Id$

 */

/*

    		 Crown Copyright (c) 1997

    This TenDRA(r) Computer Program is subject to Copyright

    owned by the United Kingdom Secretary of State for Defence

    acting through the Defence Evaluation and Research Agency

    (DERA).  It is made available to Recipients with a

    royalty-free licence for its use, reproduction, transfer

    to other parties and amendment for any purpose not excluding

    product development provided that any such use et cetera

    shall be deemed to be acceptance of the following conditions:-

        (1) Its Recipients shall ensure that this Notice is

        reproduced upon any copies or amended versions of it;

        (2) Any amended version of it shall be clearly marked to

        show both the nature of and the organisation responsible

        for the relevant amendment or amendments;

        (3) Its onward transfer from a recipient to another

        party shall be deemed to be that party's acceptance of

        these conditions;

        (4) DERA gives no warranty or assurance as to its

        quality or suitability for any purpose and DERA accepts

        no liability whatsoever in relation to any use to which

        it may be put.

*/

#include "config.h"

#include "system.h"

#include "version.h"

#include "c_types.h"

#include "hashid_ops.h"

#include "id_ops.h"

#include "nspace_ops.h"

#include "error.h"

#include "catalog.h"

#include "option.h"

#include "buffer.h"

#include "char.h"

#include "dump.h"

#include "file.h"

#include "hash.h"

#include "lex.h"

#include "literal.h"

#include "macro.h"

#include "namespace.h"

#include "predict.h"

#include "preproc.h"

#include "print.h"

#include "symbols.h"

#include "syntax.h"

#include "ustring.h"

#include "xalloc.h"

/*

    LIST OF FREE LEXICAL TOKENS

    All the free lexical tokens are formed into a list.

*/

PPTOKEN *free_tokens = NULL;

static LIST(PPTOKEN_P)alloc_tokens = NULL_list(PPTOKEN_P);

/*

    ALLOCATE A NEW TOKEN

    This routine allocates a new token from the list free_tokens.

*/

PPTOKEN *

new_pptok(void)

{

	PPTOKEN *p = free_tokens;

	if (p == NULL) {

		PPTOKEN *q;

		int i, n = 2000;

		p = xmalloc_nof(PPTOKEN, n);

		CONS_pptok(p, alloc_tokens, alloc_tokens);

		q = p;

		for (i = 1; i < n; i++) {

			q->next = q + 1;

			q++;

		}

		q->next = NULL;

	}

	free_tokens = p->next;

	p->pp_opts = real_opts;

	return(p);

}

/*

    FREE A SINGLE TOKEN

    This macro frees the single token P by adding it to the list of all

    free tokens.

*/

#define free_pptok(P)			\

    {					\

	(P)->next = free_tokens;	\

	free_tokens = (P);		\

    }

/*

    FREE A LIST OF TOKENS

    This routine adds the list of tokens p to the list of all free tokens.

*/

void

free_tok_list(PPTOKEN *p)

{

	PPTOKEN *q = p;

	if (q == NULL) {

		return;

	}

	while (q->next) {

		q = q->next;

	}

	q->next = free_tokens;

	free_tokens = p;

	return;

}

/*

    FREE ALL ALLOCATED PREPROCESSING TOKENS

    This routine frees all the space allocated for preprocessing tokens.

    It should only be called after the input has been completely processed.

*/

void

term_macros(void)

{

	LIST(PPTOKEN_P)p = alloc_tokens;

	while (!IS_NULL_list(p)) {

		PPTOKEN *q;

		DESTROY_CONS_pptok(destroy, q, p, p);

		xfree_nof(q);

	}

	alloc_tokens = p;

	free_tokens = NULL;

	return;

}

/*

    COPY A TOKEN

    This macro copies the contents of the token with token value T and data

    Q into P.

*/

#define copy_pptok(P, T, Q)		\

    {					\

	(P)->tok = (T);			\

	(P)->pp_data = (Q)->pp_data;	\

	(P)->pp_opts = (Q)->pp_opts;	\

	(P)->pp_space = (Q)->pp_space;	\

    }

/*

    ASSIGN TOKEN COMPONENTS

    This routine assigns the token components for the token t, which has

    just been read from the input file (or faked on occasions - these are

    indicated) into p.  It is only necessary to call this routine is T is

    less than or equal to LAST_COMPLEX_TOKEN (defined in symbols.h).  If any

    cases are added to this routine then it may be necessary to change the

    value of this macro.

*/

void

token_parts(int t, PPTOKEN *p)

{

	switch (t) {

	case lex_identifier: {

		/* Identifiers */

		HASHID nm = token_hashid;

		IDENTIFIER id = DEREF_id(hashid_id(nm));

		p->pp_data.id.hash = nm;

		p->pp_data.id.use = id;

		break;

	}

	case lex_char_Hlit:

	case lex_string_Hlit:

	case lex_wchar_Hlit:

	case lex_wstring_Hlit: {

		/* String and character literals */

		string s1 = token_buff.start;

		gen_size n = (gen_size)(token_buff.posn - s1);

		string s2;

		if (n < 2) {

			/* Optimise for small strings */

			s2 = xustrcpy(s1);

		} else {

			s2 = xustr(n + 1);

			xumemcpy(s2, s1, n);

			s2[n] = 0;

		}

		p->pp_data.str.start = s2;

		p->pp_data.str.end = s2 + n;

		break;

	}

	case lex_integer_Hlit:

		/* Integer and floating-point literals */

		p->pp_data.text = xustrcpy(token_buff.start);

		break;

	case lex_hash_Hif:

	case lex_hash_Helif:

		/* Target dependent conditionals */

		p->pp_data.exp = crt_hash_if_exp;

		break;

	case lex_unknown: {

		/* Unknown characters */

		int i;

		string s1 = token_buff.start;

		string s2 = p->pp_data.buff;

		ASSERT(MULTI_WIDTH <= sizeof(p->pp_data.buff));

		for (i = 0; i < MULTI_WIDTH; i++)s2[i] = s1[i];

		break;

	}

	}

	return;

}

/*

    REMOVE ANY IGNORED TOKENS FROM A LIST

    This routine removes any ignored tokens from the list tok, returning

    the result.

*/

PPTOKEN *

clean_tok_list(PPTOKEN *toks)

{

	unsigned long sp = 0;

	PPTOKEN p0, *p = &p0;

	PPTOKEN *q;

	p->next = toks;

	while (q = p->next, q != NULL) {

		if (q->tok == lex_ignore_token) {

			sp |= q->pp_space;

			p->next = q->next;

			free_pptok(q);

			q = p->next;

			if (q == NULL) {

				break;

			}

		} else {

			if (sp) {

				q->pp_space |= sp;

				sp = 0;

			}

		}

		p = q;

	}

	return(p0.next);

}

/*

    READ A LINE OF TOKENS

    This routine reads the sequence of preprocessing tokens comprising a

    preprocessing directive (for example, a macro definition).  If t1 is

    not lex_ignore_token then it is taken to be the first token in the

    definition, similarly tn gives the last token.

*/

PPTOKEN *

read_line(int t1, int tn)

{

	int t = t1;

	unsigned long sp = 0;

	PPTOKEN dummy_tok, *this_tok = &dummy_tok;

	if (t == lex_ignore_token) {

		t = read_token();

		update_column();

		if (in_preproc_dir) {

			preproc_loc = crt_loc;

		}

	}

	while (t != lex_newline && t != lex_eof) {

		this_tok->next = new_pptok();

		this_tok = this_tok->next;

		this_tok->tok = t;

		if (t <= LAST_COMPLEX_TOKEN) {

			token_parts(t, this_tok);

		}

		this_tok->pp_space = (sp & WHITE_MASK);

		sp = skip_white(0);

		t = read_token();

		update_column();

		if (in_preproc_dir) {

			preproc_loc = crt_loc;

		}

	}

	if (tn != lex_ignore_token) {

		this_tok->next = new_pptok();

		this_tok = this_tok->next;

		this_tok->tok = tn;

		token_parts(tn, this_tok);

		this_tok->pp_space = (sp & WHITE_MASK);

	}

	this_tok->next = NULL;

	if (in_preproc_dir) {

		IGNORE skip_to_end();

	}

	return(dummy_tok.next);

}

/*

    COPY A LIST OF TOKENS

    This routine copies the list of tokens toks, excluding any ignored

    tokens.

*/

static PPTOKEN *

copy_tok_list(PPTOKEN *toks)

{

	PPTOKEN *ptr_tok;

	PPTOKEN dummy_tok, *this_tok = &dummy_tok;

	for (ptr_tok = toks; ptr_tok != NULL; ptr_tok = ptr_tok->next) {

		int t = ptr_tok->tok;

		if (t != lex_ignore_token) {

			this_tok->next = new_pptok();

			this_tok = this_tok->next;

			copy_pptok(this_tok, t, ptr_tok);

		}

	}

	this_tok->next = NULL;

	return(dummy_tok.next);

}

/*

    STRINGISE A LIST OF TOKENS

    This routine turns the list of tokens toks into a string.  The result

    is built up in token_buff.  If esc is true then any '"' (or whatever

    the value of quote is) and '\' characters in string and character

    literals (including the initial and terminating quotes) are preceded

    by a '\'.  This routine is used in the implementation of the # operator,

    in macro #include directives and a couple of other preprocessing

    directives.  It returns 1 to indicate a valid string.

*/

int

quote_tok_list(PPTOKEN *toks, int esc, int quote)

{

	int res = 1;

	string st, se;

	int started = 0;

	int escaped = 0;

	PPTOKEN *ptr_tok;

	character qo = (character)quote;

	BUFFER *bf = clear_buffer(&token_buff, NIL(FILE));

	/* Scan through tokens */

	for (ptr_tok = toks; ptr_tok != NULL; ptr_tok = ptr_tok->next) {

		character p, q;

		int t = ptr_tok->tok;

		if (t == lex_ignore_token) {

			continue;

		}

		/* Print initial space if necessary */

		if (ptr_tok->pp_space && started) {

			bfputc(bf, char_space);

		}

		/* Find the token name */

		switch (t) {

		case lex_identifier: {

			/* Identifiers */

			HASHID nm = ptr_tok->pp_data.id.hash;

			st = DEREF_string(hashid_name_etc_text(nm));

			bfputs(bf, st);

			break;

		}

		case lex_integer_Hlit:

			/* Integer and floating-point literals */

			st = ptr_tok->pp_data.text;

			bfputs(bf, st);

			break;

		case lex_char_Hlit:

			/* Character literals */

			p = 0;

			q = char_single_quote;

string_label:

			st = ptr_tok->pp_data.str.start;

			se = ptr_tok->pp_data.str.end;

			/* Prefix and opening quote */

			if (p) {

				bfputc(bf, (int)p);

			}

			if (esc && q == qo) {

				bfputc(bf, char_backslash);

			}

			bfputc(bf, (int)q);

			/* Copy string */

			while (st != se) {

				character c = *(st++);

				if (c == qo || c == char_backslash) {

					/* Escaped characters */

					if (esc) {

						bfputc(bf, char_backslash);

					}

				}

				bfputc(bf, (int)c);

			}

			/* Closing quote */

			if (esc && q == qo) {

				bfputc(bf, char_backslash);

			}

			bfputc(bf, (int)q);

			break;

		case lex_wchar_Hlit:

			/* Wide character literals */

			p = char_L;

			q = char_single_quote;

			goto string_label;

		case lex_string_Hlit:

			/* String literals */

			p = 0;

			q = char_quote;

			goto string_label;

		case lex_wstring_Hlit:

			/* Wide string literals */

			p = char_L;

			q = char_quote;

			goto string_label;

		case lex_unknown: {

			/* Unknown characters */

			unsigned long u;

			int ch = CHAR_SIMPLE;

			u = get_multi_char(ptr_tok->pp_data.buff, &ch);

			if (ch == CHAR_SIMPLE) {

				bfputc(bf, (int)u);

			} else {

				print_char(u, ch, 0, bf);

			}

			break;

		}

		case lex_macro_Harg: {

			/* Macro parameters */

			HASHID nm = ptr_tok->pp_data.par.hash;

			st = DEREF_string(hashid_name_etc_text(nm));

			bfputs(bf, st);

			break;

		}

		default:

			/* Symbols */

			st = token_name(t);

			bfputs(bf, st);

			break;

		}

		started = 1;

	}

	/* End of string */

	bfputc(bf, 0);

	bf->posn--;

	/* Check for legal strings */

	st = bf->start;

	se = bf->posn;

	while (st != se) {

		if (escaped) {

			escaped = 0;

		} else {

			character c = *st;

			if (c == qo) {

				res = 0;

			}

			if (c == char_backslash) {

				escaped = 1;

			}

		}

		st++;

	}

	if (escaped) {

		res = 0;

	}

	return(res);

}

/*

    CONCATENATE TWO TOKENS

    This routine concatenates the two tokens p and q into a single token.

    This is used to implement the ## operator.  If the result is a valid

    preprocessing token then p is overwritten by the result and 1 is

    returned.  Otherwise p and q are unchanged and 0 is returned.

*/

static int

concat_pptoks(PPTOKEN *p, PPTOKEN *q)

{

	int a = p->tok;

	int b = q->tok;

	unsigned long sa = p->pp_space;

	unsigned long sb = q->pp_space;

	p->pp_space = (sa | sb);

	q->pp_space = 0;

	if (a >= FIRST_SYMBOL && a <= LAST_SYMBOL) {

		if (b >= FIRST_SYMBOL && b <= LAST_SYMBOL) {

			/* Two symbols may combine to give another symbol */

			int c;

			string s = token_buff.start;

			ustrcpy_v(s, token_name(a));

			ustrcpy_v(s + ustrlen(s), token_name(b));

			for (c = FIRST_SYMBOL; c <= LAST_SYMBOL; c++) {

				if (ustreq(s, token_name(c))) {

					/* Token found - check options */

					p->tok = c;

					if (c >= FIRST_C_SYMBOL &&

					    c <= LAST_C_SYMBOL) {

						return(1);

					}

#if LANGUAGE_CPP

					if (c >= FIRST_CPP_SYMBOL &&

					    c <= LAST_CPP_SYMBOL) {

						return(1);

					}

#endif

					if (c >= FIRST_EXTRA_SYMBOL &&

					    c <= LAST_EXTRA_SYMBOL) {

						if (allow_extra_symbols) {

							return(1);

						}

					}

					if (c >= FIRST_DIGRAPH &&

					    c <= LAST_DIGRAPH) {

						if (allow_digraphs) {

							return(1);

						}

					}

					p->tok = a;

				}

			}

			return(0);

		} else if (a == lex_dot && b == lex_integer_Hlit) {

			/* A dot may start a number */

			string s = q->pp_data.text;

			if (s[0] == char_dot) {

				return(0);

			}

			p->tok = lex_integer_Hlit;

			p->pp_data.text = xustrcat(token_name(a), s);

			return(1);

		} else if (a == lex_backslash && b == lex_identifier) {

			/* A backslash may start a universal character */

			/* NOT YET IMPLEMENTED */

			/* EMPTY */

		}

	} else if (a == lex_identifier) {

		HASHID nm = p->pp_data.id.hash;

		string s = DEREF_string(hashid_name_etc_text(nm));

		if (b == lex_identifier) {

			/* Two identifiers give another identifier */

			HASHID nm2 = q->pp_data.id.hash;

			string s2 = DEREF_string(hashid_name_etc_text(nm2));

			s = xustrcat(s, s2);

			nm = lookup_name(s, hash(s), 2, lex_identifier);

			p->pp_data.id.hash = nm;

			p->pp_data.id.use = DEREF_id(hashid_id(nm));

			return(1);

		} else if (b == lex_integer_Hlit) {

			/* An identifier and a number may give an identifier */

			character c;

			string n = q->pp_data.text;

			while (c = *(n++), c != 0) {

				if (c == char_dot || c == char_plus ||

				    c == char_minus) {

					/* The number must be entirely

					 * alphanumeric */

					return(0);

				}

			}

			s = xustrcat(s, q->pp_data.text);

			nm = lookup_name(s, hash(s), 2, lex_identifier);

			p->pp_data.id.hash = nm;

			p->pp_data.id.use = DEREF_id(hashid_id(nm));

			return(1);

		} else if (s[0] == char_L && s[1] == 0) {

			/* An L may start a wide character or string */

			if (b == lex_char_Hlit) {

				p->tok = lex_wchar_Hlit;

				p->pp_data.str.start = q->pp_data.str.start;

				p->pp_data.str.end = q->pp_data.str.end;

				return(1);

			} else if (b == lex_string_Hlit) {

				p->tok = lex_wstring_Hlit;

				p->pp_data.str.start = q->pp_data.str.start;

				p->pp_data.str.end = q->pp_data.str.end;

				return(1);

			}

		}

	} else if (a == lex_integer_Hlit) {

		string s = p->pp_data.text;

		if (b == lex_identifier) {

			/* A number followed by an identifier is a number */

			HASHID nm = q->pp_data.id.hash;

			string s2 = DEREF_string(hashid_name_etc_text(nm));

			p->pp_data.text = xustrcat(s, s2);

			return(1);

		} else if (b == lex_integer_Hlit) {

			/* Two numbers form another number */

			string s2 = q->pp_data.text;

			p->pp_data.text = xustrcat(s, s2);

			return(1);

		} else if (b == lex_dot || b == lex_ellipsis) {

			/* A number followed by a sequence of dots is a

			 * number */

			p->pp_data.text = xustrcat(s, token_name(b));

			return(1);

		} else if (b == lex_plus || b == lex_minus) {

			/* A sign may terminate a number after e or E */

			unsigned n = (unsigned)ustrlen(s) - 1;

			if (s[n] == char_e || s[n] == char_E) {

				p->pp_data.text = xustrcat(s, token_name(b));

				return(1);

			}

		}

	}

	return(0);

}

/*

    DUMMY LOCATION FOR INPUT FILE

    This dummy location represents tokens read directly from the input file.

    If present, it will always be the last element of a list of token

    locations.

*/

static PPTOKEN *dummy_loc_toks = NULL;

static TOKEN_LOC dummy_loc = { &dummy_loc_toks, NULL };

TOKEN_LOC *file_loc = &dummy_loc;

/*

    FORWARD DECLARATION

    The functions expand_macro, expand_toks and expand_tok_list are defined

    recursively.  This gives the necessary forward declarations.

*/

static PPTOKEN *expand_toks(PPTOKEN *, TOKEN_LOC *, int);

/*

    HANDLE OLD STYLE STRINGISING

    This routine handles the old style stringising for the definition defn

    for the given macro.  Argument replacement has already been performed

    on defn.  If this facility is enabled then in macro definitions of the

    form:

		#define f( X )	"X"

    quotes are classified as unknown characters rather than string

    terminators.  This means that the X is recognised as a macro parameter

    and is replaced during argument replacement.  The job of this routine

    is to spot these unrecognised quotes and turn them into proper strings.

*/

PPTOKEN *

recognise_strings(PPTOKEN *defn, HASHID macro, int act)

{

	PPTOKEN *this_tok = defn;

	PPTOKEN *last_tok = defn;

	while (this_tok != NULL) {

		if (this_tok->tok == lex_unknown) {

			unsigned long u;

			int ch = CHAR_SIMPLE;

			character qo = char_question;

			u = get_multi_char(this_tok->pp_data.buff, &ch);

			if (ch == CHAR_SIMPLE) {

				qo = (character)u;

			}

			if (qo == char_quote || qo == char_single_quote) {

				/* Start of string */

				int t;

				int escaped = 0;

				PPTOKEN *next_tok = this_tok->next;

				PPTOKEN *ptr_tok = next_tok;

				while (ptr_tok != NULL) {

					t = ptr_tok->tok;

					if (t == lex_macro_Harg) {

						HASHID nm =

						    ptr_tok->pp_data.par.hash;

						ERROR err =

						    ERR_cpp_stringize_old(nm,

									 macro);

						report(preproc_loc, err);

					}

					if (escaped) {

						escaped = 0;

					} else if (t == lex_unknown) {

						character qc = char_question;

						u = get_multi_char(ptr_tok->pp_data.buff, &ch);

						if (ch == CHAR_SIMPLE) {

							qc = (character)u;

						}

						if (qc == qo) {

							break;

						}

						if (qc == char_backslash) {

							escaped = 1;

						}

					}

					ptr_tok = ptr_tok->next;

				}

				if (act) {

					if (ptr_tok == NULL) {

						/* No closing quote */

						report(crt_loc,

						  ERR_cpp_stringize_bad(macro));

						this_tok->next = NULL;

					} else {

						ptr_tok->tok = lex_ignore_token;

						this_tok->next = ptr_tok->next;

						ptr_tok->next = NULL;

					}

					/* Form the string */

					if (!quote_tok_list(next_tok, 0,

							    (int)qo)) {

						report(crt_loc,

						  ERR_cpp_stringize_bad(macro));

					}

					t = (qo == char_quote ?

					     lex_string_Hlit : lex_char_Hlit);

					this_tok->tok = t;

					token_parts(t, this_tok);

					free_tok_list(next_tok);

					/* Check for wide strings */

					if (last_tok->tok == lex_identifier) {

						string s;

						HASHID nm =

						    last_tok->pp_data.id.hash;

						s = DEREF_string(hashid_name_etc_text(nm));

						if (s[0] == char_L &&

						    s[1] == 0) {

							if (t ==

							    lex_string_Hlit) {

								t = lex_wstring_Hlit;

							} else {

								t = lex_wchar_Hlit;

							}

							copy_pptok(last_tok, t,

								   this_tok);

							last_tok->next =

							    this_tok->next;

							free_pptok(this_tok);

							this_tok = last_tok;

						}

					}

				}

			}

		}

		last_tok = this_tok;

		this_tok = this_tok->next;

	}

	return(defn);

}

/*

    HANDLE TOKEN CONCATENATION

    This routine handles any ## operators in the definition defn of the

    given macro.  Note that any initial or terminal ## operators have

    already been reported.

*/

static PPTOKEN *

process_concat(PPTOKEN *defn, HASHID macro)

{

	PPTOKEN *this_tok;

	while (defn && defn->tok == lex_hash_Hhash_Hop) {

		/* Check for initial ## */

		this_tok = defn;

		defn = defn->next;

		free_pptok(this_tok);

	}

	this_tok = defn;

	while (this_tok != NULL) {

		PPTOKEN *next_tok = this_tok->next;

		if (next_tok == NULL) {

			break;

		}

		if (next_tok->tok == lex_hash_Hhash_Hop) {

			/* Delete the ## */

			this_tok->next = next_tok->next;