X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/f64e406b80ddaadf25c5b64bbe1d65097568db09..3eb4f1a3e62807125a815dc572da9204b5b51865:/src/scan-code.l diff --git a/src/scan-code.l b/src/scan-code.l index b2077160..24ef5d92 100644 --- a/src/scan-code.l +++ b/src/scan-code.l @@ -1,6 +1,6 @@ /* Bison Action Scanner -*- C -*- - Copyright (C) 2006, 2007, 2008, 2009 Free Software Foundation, Inc. + Copyright (C) 2006-2012 Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -48,7 +48,7 @@ YY_DECL; #define YY_USER_ACTION location_compute (loc, &loc->end, yytext, yyleng); static void handle_action_dollar (symbol_list *rule, char *cp, - location dollar_loc); + location dollar_loc); static void handle_action_at (symbol_list *rule, char *cp, location at_loc); /* A string to be pushed to obstack after dollar/at has been handled. */ @@ -76,17 +76,17 @@ static bool untyped_var_seen; /* POSIX says that a tag must be both an id and a C union member, but historically almost any character is allowed in a tag. We disallow NUL and newline, as this simplifies our implementation. */ -tag [^\0\n>]+ +tag [^\0\n>]+ /* Zero or more instances of backslash-newline. Following GCC, allow white space between the backslash and the newline. */ -splice (\\[ \f\t\v]*\n)* +splice (\\[ \f\t\v]*\n)* /* C style identifier. Must start with letter. Will be used for named symbol references. Shall be kept synchronized with scan-gram.l "letter" and "id". */ -letter [-.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] -id {letter}({letter}|[0-9])* +letter [.abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ_] +id {letter}({letter}|[-0-9])* ref -?[0-9]+|{id}|"["{id}"]"|"$" %% @@ -96,10 +96,10 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" int braces_level = 0; /* Whether a semicolon is probably needed. - The heuristic is that a semicolon is not needed after `{', `}', `;', + The heuristic is that a semicolon is not needed after '{', '}', ';', or a C preprocessor directive, and that whitespaces and comments do not affect this flag. - Note that `{' does not need a semicolon because of `{}'. + Note that '{' does not need a semicolon because of '{}'. A semicolon may be needed before a cpp direcive, but don't bother. */ bool need_semicolon = false; @@ -112,13 +112,13 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" is expected to return only once. This initialization is therefore done once per action to translate. */ aver (sc_context == SC_SYMBOL_ACTION - || sc_context == SC_RULE_ACTION - || sc_context == INITIAL); + || sc_context == SC_RULE_ACTION + || sc_context == INITIAL); BEGIN sc_context; %} /*------------------------------------------------------------. - | Scanning a C comment. The initial `/ *' is already eaten. | + | Scanning a C comment. The initial '/ *' is already eaten. | `------------------------------------------------------------*/ @@ -128,13 +128,13 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" /*--------------------------------------------------------------. - | Scanning a line comment. The initial `//' is already eaten. | + | Scanning a line comment. The initial '//' is already eaten. | `--------------------------------------------------------------*/ { - "\n" STRING_GROW; BEGIN sc_context; - {splice} STRING_GROW; + "\n" STRING_GROW; BEGIN sc_context; + {splice} STRING_GROW; } @@ -144,17 +144,17 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" { - {splice}|\\{splice}. STRING_GROW; + {splice}|\\{splice}. STRING_GROW; } { - "'" STRING_GROW; BEGIN sc_context; + "'" STRING_GROW; BEGIN sc_context; } { - "\"" STRING_GROW; BEGIN sc_context; + "\"" STRING_GROW; BEGIN sc_context; } @@ -198,12 +198,12 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" need_semicolon = true; } "$" { - warn_at (*loc, _("stray `$'")); + complain_at (*loc, Wother, _("stray '$'")); obstack_sgrow (&obstack_for_string, "$]["); need_semicolon = true; } "@" { - warn_at (*loc, _("stray `@'")); + complain_at (*loc, Wother, _("stray '@'")); obstack_sgrow (&obstack_for_string, "@@"); need_semicolon = true; } @@ -221,16 +221,18 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" "}" { bool outer_brace = --braces_level == 0; - /* As an undocumented Bison extension, append `;' before the last + /* As an undocumented Bison extension, append ';' before the last brace in braced code, so that the user code can omit trailing - `;'. But do not append `;' if emulating Yacc, since Yacc does + ';'. But do not append ';' if emulating Yacc, since Yacc does not append one. */ if (outer_brace && !yacc_flag && language_prio == default_prio && skeleton_prio == default_prio && need_semicolon && ! in_cpp) { - warn_at (*loc, _("a `;' might be needed at the end of action code")); - warn_at (*loc, _("future versions of Bison will not add the `;'")); - obstack_1grow (&obstack_for_string, ';'); + complain_at (*loc, Wother, + _("a ';' might be needed at the end of action code")); + complain_at (*loc, Wother, + _("future versions of Bison will not add the ';'")); + obstack_1grow (&obstack_for_string, ';'); } STRING_GROW; @@ -239,14 +241,27 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" /* Preprocessing directives should only be recognized at the beginning of lines, allowing whitespace including comments, but in C/C++, - `#' can only be the start of preprocessor directives or within - `#define' directives anyway, so don't bother with begin of line. */ + '#' can only be the start of preprocessor directives or within + '#define' directives anyway, so don't bother with begin of line. */ "#" STRING_GROW; in_cpp = true; {splice} STRING_GROW; [\n\r] STRING_GROW; if (in_cpp) in_cpp = need_semicolon = false; [ \t\f] STRING_GROW; - . STRING_GROW; need_semicolon = true; + + /* YYFAIL is undocumented and was formally deprecated in Bison + 2.4.2. */ + YYFAIL { + STRING_GROW; need_semicolon = true; + complain_at (*loc, Wother, + _("use of YYFAIL, which is deprecated and will be removed")); + } + + /* The sole purpose of this is to make sure identifiers that merely + contain YYFAIL don't produce the above warning. */ + [A-Za-z_][0-9A-Za-z_]* STRING_GROW; need_semicolon = true; + + . STRING_GROW; need_semicolon = true; } @@ -268,20 +283,20 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" <*> { - \$ obstack_sgrow (&obstack_for_string, "$]["); - \@ obstack_sgrow (&obstack_for_string, "@@"); - \[ obstack_sgrow (&obstack_for_string, "@{"); - \] obstack_sgrow (&obstack_for_string, "@}"); + \$ obstack_sgrow (&obstack_for_string, "$]["); + \@ obstack_sgrow (&obstack_for_string, "@@"); + \[ obstack_sgrow (&obstack_for_string, "@{"); + \] obstack_sgrow (&obstack_for_string, "@}"); } /*-----------------------------------------------------. | By default, grow the string obstack with the input. | `-----------------------------------------------------*/ -<*>.|\n STRING_GROW; +<*>.|\n STRING_GROW; /* End of processing. */ -<*><> { +<*><> { STRING_FINISH; return last_string; } @@ -344,9 +359,9 @@ variant_table_grow (void) if (variant_count > variant_table_size) { while (variant_count > variant_table_size) - variant_table_size = 2 * variant_table_size + 3; + variant_table_size = 2 * variant_table_size + 3; variant_table = xnrealloc (variant_table, variant_table_size, - sizeof *variant_table); + sizeof *variant_table); } return &variant_table[variant_count - 1]; } @@ -376,7 +391,7 @@ find_prefix_end (const char *prefix, char *begin, char *end) static variant * variant_add (uniqstr id, location id_loc, unsigned symbol_index, - char *cp, char *cp_end, bool explicit_bracketing) + char *cp, char *cp_end, bool explicit_bracketing) { char *prefix_end; @@ -410,7 +425,8 @@ get_at_spec(unsigned symbol_index) static void show_sub_messages (const char* cp, bool explicit_bracketing, - int midrule_rhs_index, char dollar_or_at, bool is_warning) + int midrule_rhs_index, char dollar_or_at, + bool is_warning, unsigned indent) { unsigned i; @@ -422,55 +438,62 @@ show_sub_messages (const char* cp, bool explicit_bracketing, if (var->err == 0) { if (is_warning) - warn_at (var->loc, _(" refers to: %c%s at %s"), - dollar_or_at, var->id, at_spec); + complain_at_indent (var->loc, Wother, &indent, + _("refers to: %c%s at %s"), dollar_or_at, + var->id, at_spec); else - complain_at (var->loc, _(" refers to: %c%s at %s"), - dollar_or_at, var->id, at_spec); + complain_at_indent (var->loc, complaint, &indent, + _("refers to: %c%s at %s"), dollar_or_at, + var->id, at_spec); } else - { - static struct obstack msg_buf; - const char *tail = explicit_bracketing ? "" : - cp + strlen (var->id); - const char *id = var->hidden_by ? var->hidden_by->id : - var->id; - location id_loc = var->hidden_by ? var->hidden_by->loc : - var->loc; - - /* Create the explanation message. */ - obstack_init (&msg_buf); - - obstack_fgrow1 (&msg_buf, " possibly meant: %c", dollar_or_at); - if (contains_dot_or_dash (id)) - obstack_fgrow1 (&msg_buf, "[%s]", id); - else - obstack_sgrow (&msg_buf, id); - obstack_sgrow (&msg_buf, tail); - - if (var->err & VARIANT_HIDDEN) - { - obstack_fgrow1 (&msg_buf, ", hiding %c", dollar_or_at); - if (contains_dot_or_dash (var->id)) - obstack_fgrow1 (&msg_buf, "[%s]", var->id); - else - obstack_sgrow (&msg_buf, var->id); - obstack_sgrow (&msg_buf, tail); - } - - obstack_fgrow1 (&msg_buf, " at %s", at_spec); - - if (var->err & VARIANT_NOT_VISIBLE_FROM_MIDRULE) - obstack_fgrow1 (&msg_buf, ", cannot be accessed from " - "mid-rule action at $%d", midrule_rhs_index); - - obstack_1grow (&msg_buf, '\0'); + { + static struct obstack msg_buf; + const char *tail = explicit_bracketing ? "" : + cp + strlen (var->id); + const char *id = var->hidden_by ? var->hidden_by->id : + var->id; + location id_loc = var->hidden_by ? var->hidden_by->loc : + var->loc; + + /* Create the explanation message. */ + obstack_init (&msg_buf); + + obstack_fgrow1 (&msg_buf, _("possibly meant: %c"), dollar_or_at); + if (contains_dot_or_dash (id)) + obstack_fgrow1 (&msg_buf, "[%s]", id); + else + obstack_sgrow (&msg_buf, id); + obstack_sgrow (&msg_buf, tail); + + if (var->err & VARIANT_HIDDEN) + { + obstack_fgrow1 (&msg_buf, _(", hiding %c"), dollar_or_at); + if (contains_dot_or_dash (var->id)) + obstack_fgrow1 (&msg_buf, "[%s]", var->id); + else + obstack_sgrow (&msg_buf, var->id); + obstack_sgrow (&msg_buf, tail); + } + + obstack_fgrow1 (&msg_buf, _(" at %s"), at_spec); + + if (var->err & VARIANT_NOT_VISIBLE_FROM_MIDRULE) + { + const char *format = + _(", cannot be accessed from mid-rule action at $%d"); + obstack_fgrow1 (&msg_buf, format, midrule_rhs_index); + } + + obstack_1grow (&msg_buf, '\0'); if (is_warning) - warn_at (id_loc, _("%s"), (char *) obstack_finish (&msg_buf)); + complain_at_indent (id_loc, Wother, &indent, "%s", + (char *) obstack_finish (&msg_buf)); else - complain_at (id_loc, _("%s"), (char *) obstack_finish (&msg_buf)); - obstack_free (&msg_buf, 0); - } + complain_at_indent (id_loc, complaint, &indent, "%s", + (char *) obstack_finish (&msg_buf)); + obstack_free (&msg_buf, 0); + } } } @@ -482,13 +505,16 @@ show_sub_messages (const char* cp, bool explicit_bracketing, points to LHS ($$) of the current rule or midrule. */ #define LHS_REF (INT_MIN + 1) +/* Sub-messages indent. */ +#define SUB_INDENT (4) + /* Parse named or positional reference. In case of positional references, can return negative values for $-n "deep" stack accesses. */ static long int parse_ref (char *cp, symbol_list *rule, int rule_length, - int midrule_rhs_index, char *text, location text_loc, - char dollar_or_at) + int midrule_rhs_index, char *text, location text_loc, + char dollar_or_at) { symbol_list *l; char *cp_end; @@ -504,13 +530,13 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, { long int num = strtol (cp, &cp, 10); if (1 - INT_MAX + rule_length <= num && num <= rule_length) - return num; + return num; else - { - complain_at (text_loc, _("integer out of range: %s"), + { + complain_at (text_loc, complaint, _("integer out of range: %s"), quote (text)); - return INVALID_REF; - } + return INVALID_REF; + } } if ('[' == *cp) @@ -518,7 +544,7 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, /* Ignore the brackets. */ char *p; for (p = ++cp; *p != ']'; ++p) - continue; + continue; cp_end = p; explicit_bracketing = true; @@ -528,13 +554,13 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, /* Take all characters of the name. */ char* p; for (p = cp; *p; ++p) - if (is_dot_or_dash (*p)) - { - ref_tail_fields = p; - break; - } + if (is_dot_or_dash (*p)) + { + ref_tail_fields = p; + break; + } for (p = cp; *p; ++p) - continue; + continue; cp_end = p; explicit_bracketing = false; @@ -547,17 +573,17 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, for (symbol_index = 0, l = rule; !symbol_list_null (l); ++symbol_index, l = l->next) { - variant *var; - if (l->content_type != SYMLIST_SYMBOL) - continue; + variant *var; + if (l->content_type != SYMLIST_SYMBOL) + continue; - var = variant_add (l->content.sym->tag, l->sym_loc, + var = variant_add (l->content.sym->tag, l->sym_loc, symbol_index, cp, cp_end, explicit_bracketing); - if (var && l->named_ref) - var->hidden_by = l->named_ref; + if (var && l->named_ref) + var->hidden_by = l->named_ref; - if (l->named_ref) - variant_add (l->named_ref->id, l->named_ref->loc, + if (l->named_ref) + variant_add (l->named_ref->id, l->named_ref->loc, symbol_index, cp, cp_end, explicit_bracketing); } } @@ -570,7 +596,7 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, /* Check visibility from mid-rule actions. */ if (midrule_rhs_index != 0 - && (symbol_index == 0 || midrule_rhs_index < symbol_index)) + && (symbol_index == 0 || midrule_rhs_index < symbol_index)) var->err |= VARIANT_NOT_VISIBLE_FROM_MIDRULE; /* Check correct bracketing. */ @@ -594,29 +620,51 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, { unsigned len = (explicit_bracketing || !ref_tail_fields) ? cp_end - cp : ref_tail_fields - cp; - const char *message = "symbol not found in production"; + unsigned indent = 0; - complain_at (text_loc, _("invalid reference: %s"), quote (text)); - if (midrule_rhs_index) - complain_at (rule->location, _(" %s before $%d: %.*s"), - message, midrule_rhs_index, len, cp); + complain_at_indent (text_loc, complaint, &indent, + _("invalid reference: %s"), quote (text)); + indent += SUB_INDENT; + if (len == 0) + { + location sym_loc = text_loc; + sym_loc.start.column += 1; + sym_loc.end = sym_loc.start; + const char *format = + _("syntax error after '%c', expecting integer, letter," + " '_', '[', or '$'"); + complain_at_indent (sym_loc, complaint, &indent, format, + dollar_or_at); + } + else if (midrule_rhs_index) + { + const char *format = + _("symbol not found in production before $%d: %.*s"); + complain_at_indent (rule->location, complaint, &indent, format, + midrule_rhs_index, len, cp); + } else - complain_at (rule->location, _(" %s: %.*s"), - message, len, cp); + { + const char *format = + _("symbol not found in production: %.*s"); + complain_at_indent (rule->location, complaint, &indent, format, + len, cp); + } if (variant_count > 0) show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, false); + dollar_or_at, false, indent); return INVALID_REF; } case 1: { + unsigned indent = 0; if (variant_count > 1) { - warn_at (text_loc, _("misleading reference: %s"), - quote (text)); + complain_at_indent (text_loc, Wother, &indent, + _("misleading reference: %s"), quote (text)); show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, true); + dollar_or_at, true, indent + SUB_INDENT); } { unsigned symbol_index = @@ -626,11 +674,14 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, } case 2: default: - complain_at (text_loc, _("ambiguous reference: %s"), - quote (text)); - show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, false); - return INVALID_REF; + { + unsigned indent = 0; + complain_at_indent (text_loc, complaint, &indent, + _("ambiguous reference: %s"), quote (text)); + show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, + dollar_or_at, false, indent + SUB_INDENT); + return INVALID_REF; + } } /* Not reachable. */ @@ -644,7 +695,7 @@ int max_left_semantic_context = 0; /*------------------------------------------------------------------. -| TEXT is pointing to a wannabee semantic value (i.e., a `$'). | +| TEXT is pointing to a wannabee semantic value (i.e., a '$'). | | | | Possible inputs: $[]($|integer) | | | @@ -677,19 +728,20 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) { type_name = ++cp; while (*cp != '>') - ++cp; + ++cp; /* The '>' symbol will be later replaced by '\0'. Original - 'text' is needed for error messages. */ + 'text' is needed for error messages. */ gt_ptr = cp; ++cp; if (untyped_var_seen) - complain_at (dollar_loc, _("explicit type given in untyped grammar")); + complain_at (dollar_loc, complaint, + _("explicit type given in untyped grammar")); tag_seen = true; } n = parse_ref (cp, effective_rule, effective_rule_length, - rule->midrule_parent_rhs_index, text, dollar_loc, '$'); + rule->midrule_parent_rhs_index, text, dollar_loc, '$'); if (gt_ptr) *gt_ptr = '\0'; @@ -701,61 +753,63 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) case LHS_REF: if (!type_name) - type_name = symbol_list_n_type_name_get (rule, dollar_loc, 0); + type_name = symbol_list_n_type_name_get (rule, dollar_loc, 0); if (!type_name) - { - if (union_seen | tag_seen) - { - if (rule->midrule_parent_rule) - complain_at (dollar_loc, - _("$$ for the midrule at $%d of `%s'" - " has no declared type"), - rule->midrule_parent_rhs_index, - effective_rule->content.sym->tag); - else - complain_at (dollar_loc, _("$$ of `%s' has no declared type"), - rule->content.sym->tag); - } - else - untyped_var_seen = true; - type_name = ""; - } + { + if (union_seen | tag_seen) + { + if (rule->midrule_parent_rule) + complain_at (dollar_loc, complaint, + _("$$ for the midrule at $%d of %s" + " has no declared type"), + rule->midrule_parent_rhs_index, + quote (effective_rule->content.sym->tag)); + else + complain_at (dollar_loc, complaint, + _("$$ of %s has no declared type"), + quote (rule->content.sym->tag)); + } + else + untyped_var_seen = true; + type_name = ""; + } obstack_fgrow1 (&obstack_for_string, - "]b4_lhs_value([%s])[", type_name); + "]b4_lhs_value([%s])[", type_name); rule->action_props.is_value_used = true; break; default: if (max_left_semantic_context < 1 - n) - max_left_semantic_context = 1 - n; + max_left_semantic_context = 1 - n; if (!type_name && 0 < n) - type_name = - symbol_list_n_type_name_get (effective_rule, dollar_loc, n); + type_name = + symbol_list_n_type_name_get (effective_rule, dollar_loc, n); if (!type_name) - { - if (union_seen | tag_seen) - complain_at (dollar_loc, _("$%s of `%s' has no declared type"), - cp, effective_rule->content.sym->tag); - else - untyped_var_seen = true; - type_name = ""; - } + { + if (union_seen | tag_seen) + complain_at (dollar_loc, complaint, + _("$%s of %s has no declared type"), cp, + quote (effective_rule->content.sym->tag)); + else + untyped_var_seen = true; + type_name = ""; + } obstack_fgrow3 (&obstack_for_string, - "]b4_rhs_value(%d, %d, [%s])[", - effective_rule_length, n, type_name); + "]b4_rhs_value(%d, %d, [%s])[", + effective_rule_length, n, type_name); if (n > 0) - symbol_list_n_get (effective_rule, n)->action_props.is_value_used = - true; + symbol_list_n_get (effective_rule, n)->action_props.is_value_used = + true; break; } } /*------------------------------------------------------. -| TEXT is a location token (i.e., a `@...'). Output to | +| TEXT is a location token (i.e., a '@...'). Output to | | OBSTACK_FOR_STRING a reference to this location. | `------------------------------------------------------*/ @@ -781,7 +835,7 @@ handle_action_at (symbol_list *rule, char *text, location at_loc) muscle_percent_define_ensure("locations", at_loc, true); n = parse_ref (cp, effective_rule, effective_rule_length, - rule->midrule_parent_rhs_index, text, at_loc, '@'); + rule->midrule_parent_rhs_index, text, at_loc, '@'); switch (n) { case INVALID_REF: @@ -793,7 +847,7 @@ handle_action_at (symbol_list *rule, char *text, location at_loc) default: obstack_fgrow2 (&obstack_for_string, "]b4_rhs_location(%d, %d)[", - effective_rule_length, n); + effective_rule_length, n); break; } } @@ -840,39 +894,36 @@ code_props const code_props_none = CODE_PROPS_NONE_INIT; void code_props_plain_init (code_props *self, char const *code, - location code_loc) + location code_loc) { + code_props_none_init (self); self->kind = CODE_PROPS_PLAIN; self->code = code; self->location = code_loc; - self->is_value_used = false; - self->rule = NULL; - self->named_ref = NULL; } void code_props_symbol_action_init (code_props *self, char const *code, location code_loc) { + code_props_none_init (self); self->kind = CODE_PROPS_SYMBOL_ACTION; self->code = code; self->location = code_loc; - self->is_value_used = false; - self->rule = NULL; - self->named_ref = NULL; } void code_props_rule_action_init (code_props *self, char const *code, location code_loc, symbol_list *rule, - named_ref *name) + named_ref *name, bool is_predicate) { + code_props_none_init (self); self->kind = CODE_PROPS_RULE_ACTION; self->code = code; self->location = code_loc; - self->is_value_used = false; self->rule = rule; self->named_ref = name; + self->is_predicate = is_predicate; } void