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 "c_types.h"

#include "hashid_ops.h"

#include "id_ops.h"

#include "tok_ops.h"

#include "error.h"

#include "catalog.h"

#include "option.h"

#include "char.h"

#include "declare.h"

#include "dump.h"

#include "file.h"

#include "lex.h"

#include "hash.h"

#include "macro.h"

#include "parse.h"

#include "pragma.h"

#include "predict.h"

#include "preproc.h"

#include "psyntax.h"

#include "redeclare.h"

#include "statement.h"

#include "symbols.h"

#include "tok.h"

#include "token.h"

#include "ustring.h"

#include "variable.h"

/*

    CHECK THAT PRAGMA SYNTAX AND MAIN SYNTAX ARE IN STEP

    By including both psyntax.h and syntax.h, the compiler will check

    that the lex_* macros defined in each are consistent.  Note that

    because the #pragma syntax is the same for both C and C++, this

    also checks the consistency of the C and C++ syntaxes.  This is only

    done in development mode.

*/

#ifdef DEBUG

#include "syntax.h"

#endif

/*

    READ A PREPROCESSING TOKEN

    This routine reads the next preprocessing token from the input file.

*/

static PPTOKEN *

get_token(void)

{

	PPTOKEN *p = new_pptok();

	int t = read_token();

	update_column();

	p->tok = t;

	p->pp_space = WHITE_SPACE;

	p->next = NULL;

	if (t <= LAST_COMPLEX_TOKEN) {

		token_parts(t, p);

	}

	return(p);

}

/*

    FIND A KEYWORD LEXICAL TOKEN NUMBER

    This routine finds the lexical token corresponding to the identifier id.

    If id does not represent an underlying keyword then an error is reported

    and lex_identifier is returned.

*/

int

find_keyword(IDENTIFIER id)

{

	HASHID nm = DEREF_hashid(id_name(id));

	int t = find_hashid(nm);

	if (t == lex_identifier) {

		report(preproc_loc, ERR_pragma_keyword_bad(nm));

	}

	return(t);

}

/*

    DEFINE A KEYWORD

    This routine defines id to be a keyword with lexical token number t.

    Certain language extensions are implemented by tokens disguised as

    keywords.

*/

void

define_keyword(IDENTIFIER id, int t)

{

	switch (t) {

	case lex_identifier: {

		break;

	}

	case lex_representation: {

		TOKEN tok = make_sort("ZZZ", 1);

		id = make_token_decl(tok, 0, id, id);

		token_interface(id, lex_no_Hdef);

		COPY_int(tok_proc_key(tok), t);

		break;

	}

	case lex_typeof: {

		TOKEN tok = make_sort("SE", 1);

		id = make_token_decl(tok, 0, id, id);

		token_interface(id, lex_no_Hdef);

		COPY_int(tok_proc_key(tok), t);

		break;

	}

	default: {

		HASHID nm = DEREF_hashid(id_name(id));

		IGNORE make_keyword(nm, t, NULL_id);

		break;

	}

	}

	return;

}

/*

    UNDEFINE A KEYWORD

    This routine undefines the keyword id.

*/

void

undef_keyword(IDENTIFIER id)

{

	unsigned tag;

	HASHID nm = DEREF_hashid(id_name(id));

	PTR(IDENTIFIER)ptr = hashid_id(nm);

	do {

		IDENTIFIER pid = DEREF_id(ptr);

		tag = TAG_id(pid);

		switch (tag) {

		case id_keyword_tag:

		case id_iso_keyword_tag:

		case id_reserved_tag:

			/* Undefine a keyword */

			if (do_keyword) {

				dump_undefine(pid, &preproc_loc, 1);

			}

			pid = DEREF_id(id_alias(pid));

			COPY_id(ptr, pid);

			COPY_id(hashid_cache(nm), NULL_id);

			return;

		}

		ptr = id_alias(pid);

	} while (tag != id_dummy_tag);

	return;

}

/*

    RESCAN A PRAGMA STATEMENT

    The routine read_tendra replaces all identifier tokens within a '#pragma'

    command by their corresponding underlying keywords.  This routine restores

    these keywords, except that given by s, to identifiers.

*/

void

rescan_pragma(int s)

{

	PPTOKEN *p;

	for (p = crt_token; p != NULL; p = p->next) {

		int t = p->tok;

		if (t >= FIRST_KEYWORD && t <= LAST_KEYWORD && t != s) {

			p->tok = lex_identifier;

		}

	}

	return;

}

/*

    SET A TOKEN TO A KEYWORD

    This routine sets the preprocessing token p to the keyword corresponding

    to the lexical token number t.

*/

static void

set_token(PPTOKEN *p, int t)

{

	HASHID nm = KEYWORD(t);

	p->tok = t;

	p->pp_data.id.hash = nm;

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

	return;

}

/*

    PATCH A PRAGMA STATEMENT

    This routine is used by the preprocessor to preserve a '#pragma'

    statement.  The arguments are as in parse_pragma.

*/

static void

patch_pragma(PPTOKEN *p, int tendra)

{

	p = clean_tok_list(p);

	switch (p->tok) {

	case lex_interface: {

		int t = crt_interface;

		if (t == lex_ignore && tendra) {

			t = lex_reject;

		}

		set_token(p, t);

		break;

	}

	case lex_member:

		if (!tendra) {

			if (p->next->tok == lex_definition) {

				set_token(p, lex_define_Hcap);

				set_token(p->next, lex_member_Hcap);

			}

		}

		break;

	case lex_promoted:

		if (!tendra) {

			set_token(p, lex_promote);

		}

		break;

	case lex_reject:

		if (!tendra) {

			set_token(p, lex_ignore);

		}

		break;

	}

	patch_preproc_dir(p);

	if (tendra) {

		p = patch_tokens(tendra);

		set_token(p, lex_tendra);

		if (tendra == 2)p->next->tok = lex_plus_Hplus;

	}

	return;

}

/*

    PARSE A PRAGMA STATEMENT

    This routine parses the '#pragma' statement given by the preprocessing

    tokens p.  tendra is 1 for '#pragma TenDRA' statements, 2 for '#pragma

    TenDRA++' statements, and 0 otherwise.

*/

static int

parse_pragma(PPTOKEN *p, int tendra)

{

	int nt;

	PPTOKEN *pt;

	PARSE_STATE s;

	int pp = preproc_only;

	int tok = lex_ignore_token;

	/* Parsing action */

	new_linkage = crt_linkage;

	save_state(&s, 0);

	init_parser(p);

	crt_loc = preproc_loc;

	crt_line_changed = 1;

	ADVANCE_LEXER;

	if (pp) {

		parse_preproc(&tok);

		if (tok != lex_ignore_token) {

			tok = lex_ignore_token;

			have_syntax_error = 1;

		}

	} else {

		parse_tendra(&tok);

	}

	nt = crt_lex_token;

	pt = crt_token;

	p = restore_parser();

	/* Check for end of input */

	switch (tok) {

	case lex_set:

	case lex_unused: {

		/* Patch inset pragma statements */

		PPTOKEN *q = clean_tok_list(p);

		PPTOKEN *r = new_pptok();

		p = q;

		p->tok = tok;

		while (q->tok != lex_newline) {

			q = q->next;

		}

		q->tok = lex_inset_Hend;

		r->tok = lex_newline;

		r->next = q->next;

		q->next = r;

		patch_preproc_dir(p);

		tok = lex_inset_Hstart;

		p = NULL;

		break;

	}

	default:

		/* Should have reached the end of the line */

		if (nt != lex_newline && !have_syntax_error) {

			ERROR err = ERR_lex_parse(pt);

			err = concat_error(err, ERR_cpp_end(lex_pragma));

			report(preproc_loc, err);

		}

		break;

	}

	/* Preprocessing action */

	if (pp) {

		patch_pragma(p, tendra);

		tok = lex_hash_Hpragma;

	} else {

		free_tok_list(p);

	}

	restore_state(&s);

	crt_linkage = new_linkage;

	return(tok);

}

/*

    SKIP TO NEXT COLON

    This routine scans along the list of preprocessing tokens p until it

    finds the first colon.  It then returns the following token.  The null

    token is returned if there is no colon in p.

*/

static PPTOKEN *

skip_to_colon(PPTOKEN *p)

{

	while (p) {

		PPTOKEN *q = p->next;

		if (p->tok == lex_colon) {

			return(q);

		}

		p = q;

	}

	return(NULL);

}

/*

    MARK A TOKEN PARAMETER

    This routine marks the TDF token parameter given by the preprocessing

    tokens p.  This consists of an optional 'TAG' followed by an identifier,

    which is optional if n is false.  Macro expansion of this identifier

    is inhibited.  Macro is true if the identifier is declared in the

    token namespace.

*/

static PPTOKEN *

mark_tdf_param(PPTOKEN *p, int n, int macro)

{

	if (p && p->tok == lex_identifier) {

		int t = find_hashid(p->pp_data.id.hash);

		if (t == lex_tag_Hcap) {

			if (p->next && p->next->tok == lex_identifier) {

				/* Have 'TAG id' */

				p->tok = t;

				p = p->next;

				p->pp_data.id.use = NULL_id;

			} else {

				/* Have 'TAG' */

				if (n == 0) {

					/* Interpret as 'TAG' */

					p->tok = t;

				} else {

					/* Interpret as 'id' where id = TAG */

					p->pp_data.id.use = NULL_id;

				}

			}

		} else {

			/* Have 'id' */

			if (macro && preproc_only) {

				/* Mark macro names when preprocessing */

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

				unsigned c = check_macro(nm, 0);

				IDENTIFIER id =

				    underlying_id(p->pp_data.id.use);

				DECL_SPEC ds = DEREF_dspec(id_storage(id));

				ds |= dspec_token;

				COPY_dspec(id_storage(id), ds);

				if (c == PP_TRUE) {

					/* Token previously defined as macro */

					PPTOKEN *q;

					token_macro = DEREF_id(hashid_id(nm));

					IGNORE patch_cond(lex_hash_Hop,

							  lex_define);

					q = patch_tokens(2);

					q->tok = lex_builtin_Hfile;

					q->next->tok = lex_hash_Hop;

				}

			}

			p->pp_data.id.use = NULL_id;

		}

		return(p->next);

	}

	if (n) {

		return(NULL);

	}

	return(p);

}

/*

    MARK A TOKEN PROGRAM PARAMETER

    This routine marks the TDF token program parameter given by the

    preprocessing tokens p.

*/

static PPTOKEN *

mark_prog_param(PPTOKEN *p)

{

	if (p && p->tok == lex_identifier) {

		int t = find_hashid(p->pp_data.id.hash);

		switch (t) {

		case lex_exp_Hcap:

		case lex_nat_Hcap:

		case lex_int_Hcap:

		case lex_stmt_Hcap:

		case lex_proc_Hcap:

			/* Have the form 'EXP id' etc */

			p->tok = t;

			p = p->next;

			if (p) {

				p = p->next;

			}

			return(p);

		case lex_type_Hcap: {

			/* Have the form 'TYPE t' */

			int depth = 0;

			p->tok = t;

			while (p) {

				/* Step over type */

				switch (p->tok) {

				case lex_open_Hround:

				case lex_open_Hsquare_H1:

				case lex_open_Hsquare_H2:

					depth++;

					break;

				case lex_close_Hround:

				case lex_close_Hsquare_H1:

				case lex_close_Hsquare_H2:

					if (depth > 0) {

						depth--;

					}

					break;

				case lex_comma:

					if (depth == 0) {

						return(p);

					}

					break;

				case lex_close_Hbrace_H1:

				case lex_close_Hbrace_H2:

					return(p);

				}

				p = p->next;

			}

			break;

		}

		case lex_member_Hcap:

			/* Have the form 'MEMBER t : id' */

			p->tok = t;

			p = skip_to_colon(p->next);

			if (p) {

				p = p->next;

			}

			return(p);

		}

	}

	return(NULL);

}

/*

    MARK A PRAGMA TOKEN STATEMENT

    The macro expansion of '#pragma token' statements is rather complex.

    The token syntax skeleton is not subject to macro expansion, while

    any types etc. involved in the specification are.  This routine marks

    the skeleton keywords in the list of preprocessing tokens p, returning

    the token immediately following the token specification.  It also sets

    macro to true if the declared token lies in the macro namespace.

*/

static PPTOKEN *

mark_tdf_token(PPTOKEN *p, int *macro)

{

	if (p == NULL) {

		return(NULL);

	}

	if (p->tok == lex_identifier) {

		int t = find_hashid(p->pp_data.id.hash);

		switch (t) {

		case lex_exp_Hcap:

			/* Expression token specifiers */

			*macro = 1;

			p->tok = t;

			p = p->next;

			if (p->tok == lex_identifier) {

				/* Check for lvalue or rvalue qualifier */

				t = find_hashid(p->pp_data.id.hash);

				if (t == lex_lvalue || t == lex_rvalue ||

				    t == lex_const) {

					p->tok = t;

				}

				p = p->next;

			}

			if (p->tok == lex_colon) {

				/* Step over type name */

				p = skip_to_colon(p->next);

			}

			return(p);

		case lex_func_Hcap:

			/* Function token specifiers */

			*macro = 1;

			p->tok = t;

			p = skip_to_colon(p->next);

			return(p);

		case lex_member_Hcap:

			/* Member token specifiers */

			p->tok = t;

			p = p->next;

			if (p->tok == lex_identifier) {

				/* Check for access specifier */

				t = find_hashid(p->pp_data.id.hash);

				if (t == lex_public || t == lex_protected ||

				    t == lex_private) {

					p->tok = t;

					p = p->next;

				}

			}

			p = skip_to_colon(p);

			p = skip_to_colon(p);

			return(p);

		case lex_proc_Hcap:

			/* Procedure token specifiers */

			*macro = 1;

			p->tok = t;

			p = p->next;

			if (p == NULL) {

				return(NULL);

			}

			t = p->tok;

			if (t == lex_open_Hbrace_H1 ||

			    t == lex_open_Hbrace_H2) {

				/* General procedure parameters */

				p = p->next;

				for (;;) {

					if (p == NULL) {

						break;

					}

					t = p->tok;

					if (t == lex_or_H1) {

						p = p->next;

						break;

					}

					p = mark_tdf_token(p, macro);

					p = mark_tdf_param(p, 1, 0);

					if (p == NULL) {

						break;

					}

					if (p->tok == lex_comma)p = p->next;

				}

				for (;;) {

					if (p == NULL) {

						break;

					}

					t = p->tok;

					if (t == lex_close_Hbrace_H1 ||

					    t == lex_close_Hbrace_H2) {

						p = p->next;

						break;

					}

					p = mark_prog_param(p);

					if (p == NULL) {

						break;

					}

					if (p->tok == lex_comma)p = p->next;

				}

			} else if (t == lex_open_Hround) {

				/* Simple procedure parameters */

				p = p->next;

				for (;;) {

					if (p == NULL) {

						break;

					}

					t = p->tok;

					if (t == lex_close_Hround) {

						p = p->next;

						break;

					}

					p = mark_tdf_token(p, macro);

					p = mark_tdf_param(p, 0, 0);

					if (p == NULL) {

						break;

					}

					if (p->tok == lex_comma) {

						p = p->next;

					}

				}

			}

			p = mark_tdf_token(p, macro);

			return(p);

		case lex_variety_Hcap:

			/* Integral type token specifiers */

			p->tok = t;

			p = p->next;

			if (p->tok == lex_identifier) {

				/* Check for signed or unsigned qualifier */

				t = find_hashid(p->pp_data.id.hash);

				if (t == lex_signed || t == lex_unsigned) {

					p->tok = t;

					p = p->next;

				}

			}

			return(p);

		case lex_nat_Hcap:

		case lex_int_Hcap:

		case lex_stmt_Hcap:

			/* Simple token specifiers */

			*macro = 1;

			p->tok = t;

			return(p->next);

		case lex_arith_Hcap:

		case lex_class_Hcap:

		case lex_float_Hcap:

		case lex_scalar_Hcap:

		case lex_struct_Hcap:

		case lex_type_Hcap:

		case lex_union_Hcap:

			/* Type token specifiers */

			p->tok = t;

			return(p->next);

		}

	}

	return(p->next);

}

/*

    READ AN EXTERNAL TOKEN NAME

    This routine processes an external token name.  This consists of a

    hash symbol, pointed to by p, followed a list of preprocessing

    tokens.  Note that the result is an extended identifier, unless it

    is a simple identifier.

*/

static PPTOKEN *

quote_token_name(PPTOKEN *p)

{

	PPTOKEN *q = p->next;

	if (q != NULL) {

		/* All following tokens are quoted */

		string s;

		unsigned long sp = q->pp_space;

		IGNORE quote_tok_list(q, 0, char_quote);

		free_tok_list(q);

		q = new_pptok();

		q->pp_space = sp;

		q->next = NULL;

		s = token_buff.start;

		if (ustrseq(s, "-")) {

			/* Special form '-' */

			q->tok = lex_minus;

		} else {

			/* Create an identifier */

			unsigned long h = hash(s);

			token_hashid = lookup_name(s, h, 1, lex_unknown);

			q->tok = lex_identifier;

			token_parts(lex_identifier, q);

		}

		p->next = q;

		p = q;

	}

	return(p);

}

/*

    READ A PRAGMA TOKEN STATEMENT

    This routine processes a '#pragma token' statement.  p gives the first

    preprocessing token (i.e. 'token').  tendra is as in parse_pragma.

*/

static int

read_tdf_token(PPTOKEN *p, int tendra)

{

	/* Read and macro expand the rest of the line */

	int t;

	PPTOKEN *q = p;

	while (q->next)q = q->next;

	if (q->tok != lex_newline) {

		int macro = 0;

		PPTOKEN *r = read_line(lex_ignore_token, lex_ignore_token);

		if (r) {

			r->pp_space = WHITE_SPACE;

		}

		q->next = r;

		r = mark_tdf_token(r, &macro);

		r = mark_tdf_param(r, 1, macro);

		while (r) {

			/* Find token name */

			t = r->tok;

			if (t == lex_hash_H1 || t == lex_hash_H2) {

				q = quote_token_name(r);

				break;

			}

			q = r;

			r = r->next;

		}

		/* Add newline token */

		while (q->next) {

			q = q->next;

		}

		q->next = new_pptok();

		q->next->tok = lex_newline;

		q->next->next = NULL;

	}

	q = expand_tok_list(p);

	free_tok_list(p);

	/* Parse the line */

	in_token_decl = 1;

	decl_loc = preproc_loc;

	t = parse_pragma(q, tendra);

	in_token_decl = 0;

	return(t);

}

/*

    READ A PRAGMA TENDRA STATEMENT

    This routine processes a '#pragma TenDRA' statement, returning the

    corresponding lexical token.  One or two tokens from the line will

    have already been read into p.  tendra is as in parse_pragma.

*/

static int

read_tendra(PPTOKEN *p, int tendra)

{

	/* Read and macro expand the rest of the line */

	PPTOKEN *q = p;

	while (q->next) {

		q = q->next;

	}

	if (q->tok != lex_newline) {

		PPTOKEN *r = read_line(lex_ignore_token, lex_newline);

		if (r) {

			r->pp_space = WHITE_SPACE;

		}

		q->next = r;

	}

	q = expand_tok_list(p);

	free_tok_list(p);

	p = q;

	/* Replace any identifiers by keywords */

	for (q = p; q != NULL; q = q->next) {

		int t = q->tok;

		if (t == lex_identifier) {

			t = find_hashid(q->pp_data.id.hash);