X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/7685e2f7ba6d5b3cc8db5862c14c1772d2a39872..faff3bef5b72ac85093799f997596631c3202c12:/src/reader.c diff --git a/src/reader.c b/src/reader.c index ec560aa5..7dd0aaa4 100644 --- a/src/reader.c +++ b/src/reader.c @@ -1,7 +1,7 @@ /* Input parser for Bison - Copyright (C) 1984, 1986, 1989, 1992, 1998, 2000, 2001, 2002, 2003, - 2005, 2006, 2007, 2009 Free Software Foundation, Inc. + Copyright (C) 1984, 1986, 1989, 1992, 1998, 2000-2003, 2005-2007, + 2009-2012 Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -22,20 +22,20 @@ #include "system.h" #include -#include #include "complain.h" #include "conflicts.h" #include "files.h" #include "getargs.h" #include "gram.h" -#include "muscle_tab.h" +#include "muscle-tab.h" #include "reader.h" #include "symlist.h" #include "symtab.h" #include "scan-gram.h" #include "scan-code.h" +static void prepare_percent_define_front_end_variables (void); static void check_and_convert_grammar (void); static symbol_list *grammar = NULL; @@ -129,8 +129,8 @@ record_merge_function_type (int merger, uniqstr type, location declaration_loc) if (merge_function->type != NULL && !UNIQSTR_EQ (merge_function->type, type)) { complain_at (declaration_loc, - _("result type clash on merge function `%s': <%s> != <%s>"), - merge_function->name, type, merge_function->type); + _("result type clash on merge function %s: <%s> != <%s>"), + quote (merge_function->name), type, merge_function->type); complain_at (merge_function->type_declaration_location, _("previous declaration")); } @@ -190,6 +190,23 @@ grammar_symbol_append (symbol *sym, location loc) return p; } +static void +assign_named_ref (symbol_list *p, named_ref *name) +{ + symbol *sym = p->content.sym; + + if (name->id == sym->tag) + { + warn_at (name->loc, + _("duplicated symbol name for %s ignored"), + quote (sym->tag)); + named_ref_free (name); + } + else + p->named_ref = name; +} + + /* The rule currently being defined, and the previous rule. CURRENT_RULE points to the first LHS of the current rule, while PREVIOUS_RULE_END points to the *end* of the previous rule (NULL). */ @@ -203,7 +220,7 @@ static symbol_list *previous_rule_end = NULL; void grammar_current_rule_begin (symbol *lhs, location loc, - named_ref *lhs_named_ref) + named_ref *lhs_name) { symbol_list* p; @@ -212,18 +229,8 @@ grammar_current_rule_begin (symbol *lhs, location loc, previous_rule_end = grammar_end; p = grammar_symbol_append (lhs, loc); - if (lhs_named_ref) - { - if (lhs_named_ref->id == lhs->tag) - { - warn_at (lhs_named_ref->loc, - _("duplicated symbol name for %s ignored"), - quote (lhs->tag)); - named_ref_free (lhs_named_ref); - } - else - p->named_ref = lhs_named_ref; - } + if (lhs_name) + assign_named_ref (p, named_ref_copy (lhs_name)); current_rule = grammar_end; @@ -242,22 +249,25 @@ grammar_current_rule_begin (symbol *lhs, location loc, /*----------------------------------------------------------------------. | A symbol should be used if either: | | 1. It has a destructor. | -| 2. --warnings=midrule-values and the symbol is a mid-rule symbol | -| (i.e., the generated LHS replacing a mid-rule action) that was | -| assigned to or used, as in "exp: { $$ = 1; } { $$ = $1; }". | +| 2. The symbol is a mid-rule symbol (i.e., the generated LHS | +| replacing a mid-rule action) that was assigned to or used, as in | +| "exp: { $$ = 1; } { $$ = $1; }". | `----------------------------------------------------------------------*/ static bool -symbol_should_be_used (symbol_list const *s) +symbol_should_be_used (symbol_list const *s, bool *midrule_warning) { if (symbol_destructor_get (s->content.sym)->code) return true; - if (warnings_flag & warnings_midrule_values) - return ((s->midrule && s->midrule->action_props.is_value_used) - || (s->midrule_parent_rule - && symbol_list_n_get (s->midrule_parent_rule, - s->midrule_parent_rhs_index) - ->action_props.is_value_used)); + if ((s->midrule && s->midrule->action_props.is_value_used) + || (s->midrule_parent_rule + && symbol_list_n_get (s->midrule_parent_rule, + s->midrule_parent_rhs_index) + ->action_props.is_value_used)) + { + *midrule_warning = true; + return true; + } return false; } @@ -301,17 +311,31 @@ grammar_rule_check (const symbol_list *r) symbol_list const *l = r; int n = 0; for (; l && l->content.sym; l = l->next, ++n) - if (! (l->action_props.is_value_used - || !symbol_should_be_used (l) - /* The default action, $$ = $1, `uses' both. */ - || (!r->action_props.code && (n == 0 || n == 1)))) - { - if (n) - warn_at (r->location, _("unused value: $%d"), n); - else - warn_at (r->location, _("unset value: $$")); - } + { + bool midrule_warning = false; + if (!l->action_props.is_value_used + && symbol_should_be_used (l, &midrule_warning) + /* The default action, $$ = $1, `uses' both. */ + && (r->action_props.code || (n != 0 && n != 1))) + { + void (*warn_at_ptr)(location, char const*, ...) = + midrule_warning ? midrule_value_at : warn_at; + if (n) + warn_at_ptr (r->location, _("unused value: $%d"), n); + else + warn_at_ptr (r->location, _("unset value: $$")); + } + } } + + /* See comments in grammar_current_rule_prec_set for how POSIX + mandates this complaint. It's only for identifiers, so skip + it for char literals and strings, which are always tokens. */ + if (r->ruleprec + && r->ruleprec->tag[0] != '\'' && r->ruleprec->tag[0] != '"' + && !r->ruleprec->declared && !r->ruleprec->prec) + warn_at (r->location, _("token for %%prec is not defined: %s"), + r->ruleprec->tag); } @@ -348,8 +372,8 @@ grammar_midrule_action (void) symbol *dummy = dummy_symbol_get (dummy_location); symbol_list *midrule = symbol_list_sym_new (dummy, dummy_location); - /* Remember named_ref of previous action */ - named_ref *named_ref = current_rule->action_props.named_ref; + /* Remember named_ref of previous action. */ + named_ref *action_name = current_rule->action_props.named_ref; /* Make a new rule, whose body is empty, before the current one, so that the action just read can belong to it. */ @@ -376,7 +400,8 @@ grammar_midrule_action (void) /* Insert the dummy nonterminal replacing the midrule action into the current rule. Bind it to its dedicated rule. */ - grammar_current_rule_symbol_append (dummy, dummy_location, named_ref); + grammar_current_rule_symbol_append (dummy, dummy_location, + action_name); grammar_end->midrule = midrule; midrule->midrule_parent_rule = current_rule; midrule->midrule_parent_rhs_index = symbol_list_length (current_rule->next); @@ -387,6 +412,16 @@ grammar_midrule_action (void) void grammar_current_rule_prec_set (symbol *precsym, location loc) { + /* POSIX says that any identifier is a nonterminal if it does not + appear on the LHS of a grammar rule and is not defined by %token + or by one of the directives that assigns precedence to a token. We + ignore this here because the only kind of identifier that POSIX + allows to follow a %prec is a token and because assuming it's a + token now can produce more logical error messages. Nevertheless, + grammar_rule_check does obey what we believe is the real intent of + POSIX here: that an error be reported for any identifier that + appears after %prec but that is not defined separately as a + token. */ symbol_class_set (precsym, token_sym, loc, false); if (current_rule->ruleprec) complain_at (loc, _("only one %s allowed per rule"), "%prec"); @@ -426,39 +461,28 @@ grammar_current_rule_merge_set (uniqstr name, location loc) void grammar_current_rule_symbol_append (symbol *sym, location loc, - named_ref *named_ref) + named_ref *name) { symbol_list *p; if (current_rule->action_props.code) grammar_midrule_action (); p = grammar_symbol_append (sym, loc); - - if (named_ref) - { - if (named_ref->id == sym->tag) - { - warn_at (named_ref->loc, - _("duplicated symbol name for %s ignored"), - quote (sym->tag)); - named_ref_free (named_ref); - } - else - p->named_ref = named_ref; - } + if (name) + assign_named_ref(p, name); } /* Attach an ACTION to the current rule. */ void grammar_current_rule_action_append (const char *action, location loc, - named_ref *named_ref) + named_ref *name) { if (current_rule->action_props.code) grammar_midrule_action (); /* After all symbol declarations have been parsed, packgram invokes code_props_translate_code. */ code_props_rule_action_init (¤t_rule->action_props, action, loc, - current_rule, named_ref); + current_rule, name); } @@ -593,36 +617,41 @@ reader (void) gram_debug = trace_flag & trace_parse; gram_scanner_initialize (); gram_parse (); + prepare_percent_define_front_end_variables (); + + if (! complaint_issued) + check_and_convert_grammar (); - /* Set front-end %define variable defaults. */ + xfclose (gram_in); +} + +static void +prepare_percent_define_front_end_variables (void) +{ + /* Set %define front-end variable defaults. */ muscle_percent_define_default ("lr.keep-unreachable-states", "false"); { char *lr_type; /* IELR would be a better default, but LALR is historically the default. */ - muscle_percent_define_default ("lr.type", "LALR"); + muscle_percent_define_default ("lr.type", "lalr"); lr_type = muscle_percent_define_get ("lr.type"); - if (0 != strcmp (lr_type, "canonical LR")) - muscle_percent_define_default ("lr.default-reductions", "all"); + if (0 != strcmp (lr_type, "canonical-lr")) + muscle_percent_define_default ("lr.default-reductions", "most"); else muscle_percent_define_default ("lr.default-reductions", "accepting"); free (lr_type); } - /* Check front-end %define variables. */ + /* Check %define front-end variables. */ { static char const * const values[] = { - "lr.type", "LALR", "IELR", "canonical LR", NULL, - "lr.default-reductions", "all", "consistent", "accepting", NULL, + "lr.type", "lalr", "ielr", "canonical-lr", NULL, + "lr.default-reductions", "most", "consistent", "accepting", NULL, NULL }; muscle_percent_define_check_values (values); } - - if (! complaint_issued) - check_and_convert_grammar (); - - xfclose (gram_in); } @@ -638,9 +667,6 @@ check_and_convert_grammar (void) if (nrules == 0) fatal (_("no rules in the input grammar")); - /* Report any undefined symbols and consider them nonterminals. */ - symbols_check_defined (); - /* If the user did not define her ENDTOKEN, do it now. */ if (!endtoken) { @@ -651,6 +677,9 @@ check_and_convert_grammar (void) endtoken->user_token_number = 0; } + /* Report any undefined symbols and consider them nonterminals. */ + symbols_check_defined (); + /* Find the start symbol if no %start. */ if (!start_flag) {