X-Git-Url: https://git.saurik.com/bison.git/blobdiff_plain/68f91d58f15abe21c65a8b4d7022ccb854c857dc..975bb564319aa4f4204c48aba265757ba207a80f:/src/scan-code.l diff --git a/src/scan-code.l b/src/scan-code.l index f1329e0b..1eb17eed 100644 --- a/src/scan-code.l +++ b/src/scan-code.l @@ -1,6 +1,6 @@ /* Bison Action Scanner -*- C -*- - Copyright (C) 2006-2013 Free Software Foundation, Inc. + Copyright (C) 2006-2014 Free Software Foundation, Inc. This file is part of Bison, the GNU Compiler Compiler. @@ -27,13 +27,14 @@ #define code_wrap() 1 #define FLEX_PREFIX(Id) code_ ## Id -#include "flex-scanner.h" +#include -#include "complain.h" -#include "reader.h" -#include "getargs.h" -#include "scan-code.h" -#include "symlist.h" +#include +#include +#include +#include +#include +#include #include #include @@ -50,7 +51,7 @@ static char *fetch_type_name (char *cp, char const **type_name, location dollar_loc); 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. */ @@ -77,49 +78,30 @@ 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>]+ + NUL and newline, as this simplifies our implementation. We allow + "->" as a means to dereference a pointer. */ +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}"]"|"$" %% %{ - /* Nesting level of the current code in braces. */ - int braces_level = 0; - - /* Whether a semicolon is probably needed. - - The heuristic is that a semicolon is not needed after '{', '}', - ';', or a C preprocessor directive, and that whitespaces and - comments do not affect this flag. Note that '{' does not need a - semicolon because of '{}'. A semicolon may be needed before a - cpp directive, but don't bother. - - While it is maintained in several start-conditions (factoring - opportunities), it is meaningful only for SC_RULE_ACTION. */ - bool need_semicolon = false; - - /* Whether in a C preprocessor directive. Don't use a start condition - for this because, at the end of strings and comments, we still need - to know whether we're in a directive. */ - bool in_cpp = false; - /* This scanner is special: it is invoked only once, henceforth is expected to return only once. This initialization is therefore done once per action to translate. */ aver (sc_context == SC_SYMBOL_ACTION - || sc_context == SC_RULE_ACTION - || sc_context == INITIAL); + || sc_context == SC_RULE_ACTION + || sc_context == INITIAL); BEGIN sc_context; %} @@ -139,8 +121,8 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" { - "\n" STRING_GROW; BEGIN sc_context; - {splice} STRING_GROW; + "\n" STRING_GROW; BEGIN sc_context; + {splice} STRING_GROW; } @@ -150,48 +132,30 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" { - {splice}|\\{splice}. STRING_GROW; + {splice}|\\{splice}. STRING_GROW; } { - "'" STRING_GROW; BEGIN sc_context; + "'" STRING_GROW; BEGIN sc_context; } { - "\"" STRING_GROW; BEGIN sc_context; + "\"" STRING_GROW; BEGIN sc_context; } { - "'" { - STRING_GROW; - BEGIN SC_CHARACTER; - need_semicolon = true; - } - "\"" { - STRING_GROW; - BEGIN SC_STRING; - need_semicolon = true; - } - "/"{splice}"*" { - STRING_GROW; - BEGIN SC_COMMENT; - } - "/"{splice}"/" { - STRING_GROW; - BEGIN SC_LINE_COMMENT; - } + "'" STRING_GROW; BEGIN SC_CHARACTER; + "\"" STRING_GROW; BEGIN SC_STRING; + "/"{splice}"*" STRING_GROW; BEGIN SC_COMMENT; + "/"{splice}"/" STRING_GROW; BEGIN SC_LINE_COMMENT; + [$@] { - warn_at (*loc, _("stray '%s'"), yytext); - obstack_escape (&obstack_for_string, yytext); - need_semicolon = true; - } - [\[\]] { + complain (loc, Wother, _("stray '%s'"), yytext); obstack_escape (&obstack_for_string, yytext); - need_semicolon = true; } } @@ -202,64 +166,13 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" handle_action_dollar (self->rule, yytext, *loc); if (ref_tail_fields) obstack_sgrow (&obstack_for_string, ref_tail_fields); - need_semicolon = true; } "@"{ref} { ref_tail_fields = NULL; handle_action_at (self->rule, yytext, *loc); if (ref_tail_fields) obstack_sgrow (&obstack_for_string, ref_tail_fields); - need_semicolon = true; } - - ";" STRING_GROW; need_semicolon = false; - "{" STRING_GROW; ++braces_level; need_semicolon = false; - "}" { - bool outer_brace = --braces_level == 0; - - /* 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 - not append one. This is deprecated since release 2.4.1. */ - if (outer_brace && !yacc_flag && language_prio == default_prio - && skeleton_prio == default_prio && need_semicolon && ! in_cpp) - { - unsigned int indent = 0; - warn_at_indent (*loc, &indent, - _("a ';' might be needed at the end of action code")); - indent += SUB_INDENT; - warn_at_indent (*loc, &indent, - _("future versions of Bison will not add the ';'")); - obstack_1grow (&obstack_for_string, ';'); - } - - STRING_GROW; - need_semicolon = false; - } - - /* 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. */ - "#" STRING_GROW; in_cpp = true; - - {splice} STRING_GROW; - [\n\r] STRING_GROW; if (in_cpp) in_cpp = need_semicolon = false; - [ \t\f] STRING_GROW; - - /* YYFAIL is undocumented and was formally deprecated in Bison - 2.4.2. */ - YYFAIL { - STRING_GROW; need_semicolon = true; - warn_at (*loc, _("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; } @@ -274,7 +187,7 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" } "@$" { obstack_sgrow (&obstack_for_string, "]b4_at_dollar["); - locations_flag = true; + muscle_percent_define_ensure("locations", the_location, true); } } @@ -287,7 +200,7 @@ ref -?[0-9]+|{id}|"["{id}"]"|"$" /* By default, grow the string obstack with the input. */ .|\n STRING_GROW; - /* End of processing. */ + /* End of processing. */ <> STRING_FINISH; return last_string; } @@ -349,9 +262,9 @@ 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]; } @@ -381,7 +294,7 @@ 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; @@ -403,7 +316,7 @@ variant_add (uniqstr id, location id_loc, unsigned symbol_index, } static const char * -get_at_spec (unsigned symbol_index) +get_at_spec(unsigned symbol_index) { static char at_buf[20]; if (symbol_index == 0) @@ -414,77 +327,83 @@ 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, unsigned indent) +show_sub_message (warnings warning, + const char* cp, bool explicit_bracketing, + int midrule_rhs_index, char dollar_or_at, + unsigned indent, const variant *var) { - unsigned i; + const char *at_spec = get_at_spec (var->symbol_index); - for (i = 0; i < variant_count; ++i) + if (var->err == 0) + complain_indent (&var->loc, warning, &indent, + _("refers to: %c%s at %s"), dollar_or_at, + var->id, at_spec); + else { - const variant *var = &variant_table[i]; - const char *at_spec = get_at_spec (var->symbol_index); + static struct obstack msg_buf; + const char *tail = explicit_bracketing ? "" : cp + strlen (var->id); + const char *id; + location id_loc; - if (var->err == 0) + if (var->hidden_by) { - if (is_warning) - warn_at_indent (var->loc, &indent, _("refers to: %c%s at %s"), - dollar_or_at, var->id, at_spec); - else - complain_at_indent (var->loc, &indent, _("refers to: %c%s at %s"), - dollar_or_at, var->id, at_spec); + id = var->hidden_by->id; + id_loc = var->hidden_by->loc; } 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_printf (&msg_buf, _("possibly meant: %c"), dollar_or_at); - if (contains_dot_or_dash (id)) - obstack_printf (&msg_buf, "[%s]", id); - else - obstack_sgrow (&msg_buf, id); - obstack_sgrow (&msg_buf, tail); - - if (var->err & VARIANT_HIDDEN) - { - obstack_printf (&msg_buf, _(", hiding %c"), dollar_or_at); - if (contains_dot_or_dash (var->id)) - obstack_printf (&msg_buf, "[%s]", var->id); - else - obstack_sgrow (&msg_buf, var->id); - obstack_sgrow (&msg_buf, tail); - } - - obstack_printf (&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_printf (&msg_buf, format, midrule_rhs_index); - } + { + id = var->id; + id_loc = var->loc; + } + + /* Create the explanation message. */ + obstack_init (&msg_buf); - obstack_1grow (&msg_buf, '\0'); - if (is_warning) - warn_at_indent (id_loc, &indent, "%s", - (char *) obstack_finish (&msg_buf)); + obstack_printf (&msg_buf, _("possibly meant: %c"), dollar_or_at); + if (contains_dot_or_dash (id)) + obstack_printf (&msg_buf, "[%s]", id); + else + obstack_sgrow (&msg_buf, id); + obstack_sgrow (&msg_buf, tail); + + if (var->err & VARIANT_HIDDEN) + { + obstack_printf (&msg_buf, _(", hiding %c"), dollar_or_at); + if (contains_dot_or_dash (var->id)) + obstack_printf (&msg_buf, "[%s]", var->id); else - complain_at_indent (id_loc, &indent, "%s", - (char *) obstack_finish (&msg_buf)); - obstack_free (&msg_buf, 0); - } + obstack_sgrow (&msg_buf, var->id); + obstack_sgrow (&msg_buf, tail); + } + + obstack_printf (&msg_buf, _(" at %s"), at_spec); + + if (var->err & VARIANT_NOT_VISIBLE_FROM_MIDRULE) + obstack_printf (&msg_buf, + _(", cannot be accessed from mid-rule action at $%d"), + midrule_rhs_index); + + complain_indent (&id_loc, warning, &indent, "%s", + obstack_finish0 (&msg_buf)); + obstack_free (&msg_buf, 0); } } +static void +show_sub_messages (warnings warning, + const char* cp, bool explicit_bracketing, + int midrule_rhs_index, char dollar_or_at, + unsigned indent) +{ + unsigned i; + + for (i = 0; i < variant_count; ++i) + show_sub_message (warning | silent, + cp, explicit_bracketing, + midrule_rhs_index, dollar_or_at, + indent, &variant_table[i]); +} + /* Returned from "parse_ref" when the reference is inappropriate. */ #define INVALID_REF (INT_MIN) @@ -498,8 +417,8 @@ show_sub_messages (const char* cp, bool explicit_bracketing, 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; @@ -515,13 +434,13 @@ 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"), - quote (text)); - return INVALID_REF; - } + { + complain (&text_loc, complaint, _("integer out of range: %s"), + quote (text)); + return INVALID_REF; + } } if ('[' == *cp) @@ -529,7 +448,7 @@ 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; @@ -539,13 +458,13 @@ 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; @@ -558,17 +477,17 @@ 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); } } @@ -581,7 +500,7 @@ 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. */ @@ -607,37 +526,33 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, cp_end - cp : ref_tail_fields - cp; unsigned indent = 0; - complain_at_indent (text_loc, &indent, _("invalid reference: %s"), - quote (text)); + complain_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, &indent, format, dollar_or_at); + complain_indent (&sym_loc, complaint, &indent, + _("syntax error after '%c', expecting integer, " + "letter, '_', '[', or '$'"), + dollar_or_at); } else if (midrule_rhs_index) - { - const char *format = - _("symbol not found in production before $%d: %.*s"); - complain_at_indent (rule->location, &indent, format, - midrule_rhs_index, len, cp); - } + complain_indent (&rule->location, complaint, &indent, + _("symbol not found in production before $%d: " + "%.*s"), + midrule_rhs_index, len, cp); else - { - const char *format = - _("symbol not found in production: %.*s"); - complain_at_indent (rule->location, &indent, format, - len, cp); - } + complain_indent (&rule->location, complaint, &indent, + _("symbol not found in production: %.*s"), + len, cp); if (variant_count > 0) - show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, false, indent); + show_sub_messages (complaint, + cp, explicit_bracketing, midrule_rhs_index, + dollar_or_at, indent); return INVALID_REF; } case 1: @@ -645,10 +560,11 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, unsigned indent = 0; if (variant_count > 1) { - warn_at_indent (text_loc, &indent, _("misleading reference: %s"), - quote (text)); - show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, true, indent + SUB_INDENT); + complain_indent (&text_loc, Wother, &indent, + _("misleading reference: %s"), quote (text)); + show_sub_messages (Wother, + cp, explicit_bracketing, midrule_rhs_index, + dollar_or_at, indent + SUB_INDENT); } { unsigned symbol_index = @@ -660,16 +576,14 @@ parse_ref (char *cp, symbol_list *rule, int rule_length, default: { unsigned indent = 0; - complain_at_indent (text_loc, &indent, _("ambiguous reference: %s"), - quote (text)); - show_sub_messages (cp, explicit_bracketing, midrule_rhs_index, - dollar_or_at, false, indent + SUB_INDENT); + complain_indent (&text_loc, complaint, &indent, + _("ambiguous reference: %s"), quote (text)); + show_sub_messages (complaint, + cp, explicit_bracketing, midrule_rhs_index, + dollar_or_at, indent + SUB_INDENT); return INVALID_REF; } } - - /* Not reachable. */ - return INVALID_REF; } /* Keeps track of the maximum number of semantic values to the left of @@ -690,14 +604,16 @@ fetch_type_name (char *cp, char const **type_name, if (*cp == '<') { *type_name = ++cp; - while (*cp != '>') - ++cp; + /* Series of non-'>' or "->". */ + while (*cp != '>' || cp[-1] == '-') + ++cp; /* The '>' symbol will be later replaced by '\0'. Original - 'text' is needed for error messages. */ + 'text' is needed for error messages. */ ++cp; if (untyped_var_seen) - complain_at (dollar_loc, _("explicit type given in untyped grammar")); + complain (&dollar_loc, complaint, + _("explicit type given in untyped grammar")); tag_seen = true; } return cp; @@ -735,7 +651,7 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) cp = fetch_type_name (cp, &type_name, dollar_loc); 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, '$'); /* End type_name. */ if (type_name) @@ -748,21 +664,22 @@ 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, + complain (&dollar_loc, complaint, _("$$ for the midrule at $%d of %s" " has no declared type"), rule->midrule_parent_rhs_index, quote (effective_rule->content.sym->tag)); else - complain_at (dollar_loc, _("$$ of %s has no declared type"), - quote (rule->content.sym->tag)); + complain (&dollar_loc, complaint, + _("$$ of %s has no declared type"), + quote (rule->content.sym->tag)); } else untyped_var_seen = true; @@ -776,26 +693,27 @@ handle_action_dollar (symbol_list *rule, char *text, location dollar_loc) 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, _("$%s of %s has no declared type"), - cp, quote (effective_rule->content.sym->tag)); + complain (&dollar_loc, complaint, + _("$%s of %s has no declared type"), cp, + quote (effective_rule->content.sym->tag)); else untyped_var_seen = true; } obstack_printf (&obstack_for_string, - "]b4_rhs_value(%d, %d, ", effective_rule_length, n); + "]b4_rhs_value(%d, %d, ", effective_rule_length, n); obstack_quote (&obstack_for_string, type_name); obstack_sgrow (&obstack_for_string, ")["); 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; } } @@ -825,10 +743,10 @@ handle_action_at (symbol_list *rule, char *text, location at_loc) effective_rule_length = symbol_list_length (rule->next); } - locations_flag = true; + 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: @@ -840,7 +758,7 @@ handle_action_at (symbol_list *rule, char *text, location at_loc) default: obstack_printf (&obstack_for_string, "]b4_rhs_location(%d, %d)[", - effective_rule_length, n); + effective_rule_length, n); break; } } @@ -883,43 +801,40 @@ code_props_none_init (code_props *self) *self = code_props_none; } -code_props const code_props_none = CODE_PROPS_NONE_INIT; +code_props 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