/********************************************************************** parse.y - $Author$ created at: Fri May 28 18:02:42 JST 1993 Copyright (C) 1993-2007 Yukihiro Matsumoto **********************************************************************/ %{ #ifndef PARSER_DEBUG #define PARSER_DEBUG 0 #endif #define YYDEBUG 1 #define YYERROR_VERBOSE 1 #define YYSTACK_USE_ALLOCA 0 #include "ruby/ruby.h" #include "ruby/st.h" #include "ruby/encoding.h" #include "internal.h" #include "node.h" #include "parse.h" #include "id.h" #include "regenc.h" #include #include #include #include "probes.h" #define YYMALLOC(size) rb_parser_malloc(parser, (size)) #define YYREALLOC(ptr, size) rb_parser_realloc(parser, (ptr), (size)) #define YYCALLOC(nelem, size) rb_parser_calloc(parser, (nelem), (size)) #define YYFREE(ptr) rb_parser_free(parser, (ptr)) #define malloc YYMALLOC #define realloc YYREALLOC #define calloc YYCALLOC #define free YYFREE #ifndef RIPPER static ID register_symid(ID, const char *, long, rb_encoding *); static ID register_symid_str(ID, VALUE); #define REGISTER_SYMID(id, name) register_symid((id), (name), strlen(name), enc) #include "id.c" #endif #define is_notop_id(id) ((id)>tLAST_OP_ID) #define is_local_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_LOCAL) #define is_global_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_GLOBAL) #define is_instance_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_INSTANCE) #define is_attrset_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_ATTRSET) #define is_const_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_CONST) #define is_class_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_CLASS) #define is_junk_id(id) (is_notop_id(id)&&((id)&ID_SCOPE_MASK)==ID_JUNK) #define id_type(id) (is_notop_id(id) ? (int)((id)&ID_SCOPE_MASK) : -1) #define is_asgn_or_id(id) ((is_notop_id(id)) && \ (((id)&ID_SCOPE_MASK) == ID_GLOBAL || \ ((id)&ID_SCOPE_MASK) == ID_INSTANCE || \ ((id)&ID_SCOPE_MASK) == ID_CLASS)) enum lex_state_bits { EXPR_BEG_bit, /* ignore newline, +/- is a sign. */ EXPR_END_bit, /* newline significant, +/- is an operator. */ EXPR_ENDARG_bit, /* ditto, and unbound braces. */ EXPR_ENDFN_bit, /* ditto, and unbound braces. */ EXPR_ARG_bit, /* newline significant, +/- is an operator. */ EXPR_CMDARG_bit, /* newline significant, +/- is an operator. */ EXPR_MID_bit, /* newline significant, +/- is an operator. */ EXPR_FNAME_bit, /* ignore newline, no reserved words. */ EXPR_DOT_bit, /* right after `.' or `::', no reserved words. */ EXPR_CLASS_bit, /* immediate after `class', no here document. */ EXPR_VALUE_bit, /* alike EXPR_BEG but label is disallowed. */ EXPR_MAX_STATE }; /* examine combinations */ enum lex_state_e { #define DEF_EXPR(n) EXPR_##n = (1 << EXPR_##n##_bit) DEF_EXPR(BEG), DEF_EXPR(END), DEF_EXPR(ENDARG), DEF_EXPR(ENDFN), DEF_EXPR(ARG), DEF_EXPR(CMDARG), DEF_EXPR(MID), DEF_EXPR(FNAME), DEF_EXPR(DOT), DEF_EXPR(CLASS), DEF_EXPR(VALUE), EXPR_BEG_ANY = (EXPR_BEG | EXPR_VALUE | EXPR_MID | EXPR_CLASS), EXPR_ARG_ANY = (EXPR_ARG | EXPR_CMDARG), EXPR_END_ANY = (EXPR_END | EXPR_ENDARG | EXPR_ENDFN) }; #define IS_lex_state_for(x, ls) ((x) & (ls)) #define IS_lex_state(ls) IS_lex_state_for(lex_state, (ls)) #if PARSER_DEBUG static const char *lex_state_name(enum lex_state_e state); #endif typedef VALUE stack_type; # define BITSTACK_PUSH(stack, n) ((stack) = ((stack)<<1)|((n)&1)) # define BITSTACK_POP(stack) ((stack) = (stack) >> 1) # define BITSTACK_LEXPOP(stack) ((stack) = ((stack) >> 1) | ((stack) & 1)) # define BITSTACK_SET_P(stack) ((stack)&1) #define COND_PUSH(n) BITSTACK_PUSH(cond_stack, (n)) #define COND_POP() BITSTACK_POP(cond_stack) #define COND_LEXPOP() BITSTACK_LEXPOP(cond_stack) #define COND_P() BITSTACK_SET_P(cond_stack) #define CMDARG_PUSH(n) BITSTACK_PUSH(cmdarg_stack, (n)) #define CMDARG_POP() BITSTACK_POP(cmdarg_stack) #define CMDARG_LEXPOP() BITSTACK_LEXPOP(cmdarg_stack) #define CMDARG_P() BITSTACK_SET_P(cmdarg_stack) struct vtable { ID *tbl; int pos; int capa; struct vtable *prev; }; struct local_vars { struct vtable *args; struct vtable *vars; struct vtable *used; struct local_vars *prev; }; #define DVARS_INHERIT ((void*)1) #define DVARS_TOPSCOPE NULL #define DVARS_SPECIAL_P(tbl) (!POINTER_P(tbl)) #define POINTER_P(val) ((VALUE)(val) & ~(VALUE)3) static int vtable_size(const struct vtable *tbl) { if (POINTER_P(tbl)) { return tbl->pos; } else { return 0; } } #define VTBL_DEBUG 0 static struct vtable * vtable_alloc(struct vtable *prev) { struct vtable *tbl = ALLOC(struct vtable); tbl->pos = 0; tbl->capa = 8; tbl->tbl = ALLOC_N(ID, tbl->capa); tbl->prev = prev; if (VTBL_DEBUG) printf("vtable_alloc: %p\n", (void *)tbl); return tbl; } static void vtable_free(struct vtable *tbl) { if (VTBL_DEBUG)printf("vtable_free: %p\n", (void *)tbl); if (POINTER_P(tbl)) { if (tbl->tbl) { xfree(tbl->tbl); } xfree(tbl); } } static void vtable_add(struct vtable *tbl, ID id) { if (!POINTER_P(tbl)) { rb_bug("vtable_add: vtable is not allocated (%p)", (void *)tbl); } if (VTBL_DEBUG) printf("vtable_add: %p, %s\n", (void *)tbl, rb_id2name(id)); if (tbl->pos == tbl->capa) { tbl->capa = tbl->capa * 2; REALLOC_N(tbl->tbl, ID, tbl->capa); } tbl->tbl[tbl->pos++] = id; } static int vtable_included(const struct vtable * tbl, ID id) { int i; if (POINTER_P(tbl)) { for (i = 0; i < tbl->pos; i++) { if (tbl->tbl[i] == id) { return i+1; } } } return 0; } #ifndef RIPPER typedef struct token_info { const char *token; int linenum; int column; int nonspc; struct token_info *next; } token_info; #endif /* Structure of Lexer Buffer: lex_pbeg tokp lex_p lex_pend | | | | |-----------+--------------+------------| |<------------>| token */ struct parser_params { int is_ripper; NODE *heap; YYSTYPE *parser_yylval; VALUE eofp; NODE *parser_lex_strterm; enum lex_state_e parser_lex_state; stack_type parser_cond_stack; stack_type parser_cmdarg_stack; int parser_class_nest; int parser_paren_nest; int parser_lpar_beg; int parser_in_single; int parser_in_def; int parser_brace_nest; int parser_compile_for_eval; VALUE parser_cur_mid; int parser_in_defined; char *parser_tokenbuf; int parser_tokidx; int parser_toksiz; int parser_tokline; VALUE parser_lex_input; VALUE parser_lex_lastline; VALUE parser_lex_nextline; const char *parser_lex_pbeg; const char *parser_lex_p; const char *parser_lex_pend; int parser_heredoc_end; int parser_command_start; NODE *parser_deferred_nodes; long parser_lex_gets_ptr; VALUE (*parser_lex_gets)(struct parser_params*,VALUE); struct local_vars *parser_lvtbl; int parser_ruby__end__seen; int line_count; int has_shebang; char *parser_ruby_sourcefile; /* current source file */ int parser_ruby_sourceline; /* current line no. */ VALUE parser_ruby_sourcefile_string; rb_encoding *enc; int parser_yydebug; int last_cr_line; #ifndef RIPPER /* Ruby core only */ NODE *parser_eval_tree_begin; NODE *parser_eval_tree; VALUE debug_lines; VALUE coverage; int nerr; int parser_token_info_enabled; token_info *parser_token_info; #else /* Ripper only */ const char *tokp; VALUE delayed; int delayed_line; int delayed_col; VALUE value; VALUE result; VALUE parsing_thread; int toplevel_p; #endif }; #define STR_NEW(p,n) rb_enc_str_new((p),(n),current_enc) #define STR_NEW0() rb_enc_str_new(0,0,current_enc) #define STR_NEW2(p) rb_enc_str_new((p),strlen(p),current_enc) #define STR_NEW3(p,n,e,func) parser_str_new((p),(n),(e),(func),current_enc) #define ENC_SINGLE(cr) ((cr)==ENC_CODERANGE_7BIT) #define TOK_INTERN(mb) rb_intern3(tok(), toklen(), current_enc) static int parser_yyerror(struct parser_params*, const char*); #define yyerror(msg) parser_yyerror(parser, (msg)) #define lex_strterm (parser->parser_lex_strterm) #define lex_state (parser->parser_lex_state) #define cond_stack (parser->parser_cond_stack) #define cmdarg_stack (parser->parser_cmdarg_stack) #define class_nest (parser->parser_class_nest) #define paren_nest (parser->parser_paren_nest) #define lpar_beg (parser->parser_lpar_beg) #define brace_nest (parser->parser_brace_nest) #define in_single (parser->parser_in_single) #define in_def (parser->parser_in_def) #define compile_for_eval (parser->parser_compile_for_eval) #define cur_mid (parser->parser_cur_mid) #define in_defined (parser->parser_in_defined) #define tokenbuf (parser->parser_tokenbuf) #define tokidx (parser->parser_tokidx) #define toksiz (parser->parser_toksiz) #define tokline (parser->parser_tokline) #define lex_input (parser->parser_lex_input) #define lex_lastline (parser->parser_lex_lastline) #define lex_nextline (parser->parser_lex_nextline) #define lex_pbeg (parser->parser_lex_pbeg) #define lex_p (parser->parser_lex_p) #define lex_pend (parser->parser_lex_pend) #define heredoc_end (parser->parser_heredoc_end) #define command_start (parser->parser_command_start) #define deferred_nodes (parser->parser_deferred_nodes) #define lex_gets_ptr (parser->parser_lex_gets_ptr) #define lex_gets (parser->parser_lex_gets) #define lvtbl (parser->parser_lvtbl) #define ruby__end__seen (parser->parser_ruby__end__seen) #define ruby_sourceline (parser->parser_ruby_sourceline) #define ruby_sourcefile (parser->parser_ruby_sourcefile) #define ruby_sourcefile_string (parser->parser_ruby_sourcefile_string) #define current_enc (parser->enc) #define yydebug (parser->parser_yydebug) #ifdef RIPPER #else #define ruby_eval_tree (parser->parser_eval_tree) #define ruby_eval_tree_begin (parser->parser_eval_tree_begin) #define ruby_debug_lines (parser->debug_lines) #define ruby_coverage (parser->coverage) #endif #if YYPURE static int yylex(void*, void*); #else static int yylex(void*); #endif #ifndef RIPPER #define yyparse ruby_yyparse static NODE* node_newnode(struct parser_params *, enum node_type, VALUE, VALUE, VALUE); #define rb_node_newnode(type, a1, a2, a3) node_newnode(parser, (type), (a1), (a2), (a3)) static NODE *cond_gen(struct parser_params*,NODE*); #define cond(node) cond_gen(parser, (node)) static NODE *logop_gen(struct parser_params*,enum node_type,NODE*,NODE*); #define logop(type,node1,node2) logop_gen(parser, (type), (node1), (node2)) static NODE *newline_node(NODE*); static void fixpos(NODE*,NODE*); static int value_expr_gen(struct parser_params*,NODE*); static void void_expr_gen(struct parser_params*,NODE*); static NODE *remove_begin(NODE*); static NODE *remove_begin_all(NODE*); #define value_expr(node) value_expr_gen(parser, (node) = remove_begin(node)) #define void_expr0(node) void_expr_gen(parser, (node)) #define void_expr(node) void_expr0((node) = remove_begin(node)) static void void_stmts_gen(struct parser_params*,NODE*); #define void_stmts(node) void_stmts_gen(parser, (node)) static void reduce_nodes_gen(struct parser_params*,NODE**); #define reduce_nodes(n) reduce_nodes_gen(parser,(n)) static void block_dup_check_gen(struct parser_params*,NODE*,NODE*); #define block_dup_check(n1,n2) block_dup_check_gen(parser,(n1),(n2)) static NODE *block_append_gen(struct parser_params*,NODE*,NODE*); #define block_append(h,t) block_append_gen(parser,(h),(t)) static NODE *list_append_gen(struct parser_params*,NODE*,NODE*); #define list_append(l,i) list_append_gen(parser,(l),(i)) static NODE *list_concat_gen(struct parser_params*,NODE*,NODE*); #define list_concat(h,t) list_concat_gen(parser,(h),(t)) static NODE *arg_append_gen(struct parser_params*,NODE*,NODE*); #define arg_append(h,t) arg_append_gen(parser,(h),(t)) static NODE *arg_concat_gen(struct parser_params*,NODE*,NODE*); #define arg_concat(h,t) arg_concat_gen(parser,(h),(t)) static NODE *literal_concat_gen(struct parser_params*,NODE*,NODE*); #define literal_concat(h,t) literal_concat_gen(parser,(h),(t)) static int literal_concat0(struct parser_params *, VALUE, VALUE); static NODE *new_evstr_gen(struct parser_params*,NODE*); #define new_evstr(n) new_evstr_gen(parser,(n)) static NODE *evstr2dstr_gen(struct parser_params*,NODE*); #define evstr2dstr(n) evstr2dstr_gen(parser,(n)) static NODE *str_suffix_gen(struct parser_params*, NODE*, long); #define str_suffix(n,o) str_suffix_gen(parser,(n),(o)) static NODE *splat_array(NODE*); static NODE *call_bin_op_gen(struct parser_params*,NODE*,ID,NODE*); #define call_bin_op(recv,id,arg1) call_bin_op_gen(parser, (recv),(id),(arg1)) static NODE *call_uni_op_gen(struct parser_params*,NODE*,ID); #define call_uni_op(recv,id) call_uni_op_gen(parser, (recv),(id)) static NODE *new_args_gen(struct parser_params*,NODE*,NODE*,ID,NODE*,NODE*); #define new_args(f,o,r,p,t) new_args_gen(parser, (f),(o),(r),(p),(t)) static NODE *new_args_tail_gen(struct parser_params*,NODE*,ID,ID); #define new_args_tail(k,kr,b) new_args_tail_gen(parser, (k),(kr),(b)) static NODE *negate_lit(NODE*); static NODE *ret_args_gen(struct parser_params*,NODE*); #define ret_args(node) ret_args_gen(parser, (node)) static NODE *arg_blk_pass(NODE*,NODE*); static NODE *new_yield_gen(struct parser_params*,NODE*); #define new_yield(node) new_yield_gen(parser, (node)) static NODE *dsym_node_gen(struct parser_params*,NODE*); #define dsym_node(node) dsym_node_gen(parser, (node)) static NODE *gettable_gen(struct parser_params*,ID); #define gettable(id) gettable_gen(parser,(id)) static NODE *assignable_gen(struct parser_params*,ID,NODE*); #define assignable(id,node) assignable_gen(parser, (id), (node)) static NODE *aryset_gen(struct parser_params*,NODE*,NODE*); #define aryset(node1,node2) aryset_gen(parser, (node1), (node2)) static NODE *attrset_gen(struct parser_params*,NODE*,ID); #define attrset(node,id) attrset_gen(parser, (node), (id)) static void rb_backref_error_gen(struct parser_params*,NODE*); #define rb_backref_error(n) rb_backref_error_gen(parser,(n)) static NODE *node_assign_gen(struct parser_params*,NODE*,NODE*); #define node_assign(node1, node2) node_assign_gen(parser, (node1), (node2)) static NODE *new_op_assign_gen(struct parser_params *parser, NODE *lhs, ID op, NODE *rhs); static NODE *new_attr_op_assign_gen(struct parser_params *parser, NODE *lhs, ID attr, ID op, NODE *rhs); #define new_attr_op_assign(lhs, type, attr, op, rhs) new_attr_op_assign_gen(parser, (lhs), (attr), (op), (rhs)) static NODE *new_const_op_assign_gen(struct parser_params *parser, NODE *lhs, ID op, NODE *rhs); #define new_const_op_assign(lhs, op, rhs) new_const_op_assign_gen(parser, (lhs), (op), (rhs)) #define new_defined(expr) NEW_DEFINED(remove_begin_all(expr)) static NODE *match_op_gen(struct parser_params*,NODE*,NODE*); #define match_op(node1,node2) match_op_gen(parser, (node1), (node2)) static ID *local_tbl_gen(struct parser_params*); #define local_tbl() local_tbl_gen(parser) static void fixup_nodes(NODE **); static VALUE reg_compile_gen(struct parser_params*, VALUE, int); #define reg_compile(str,options) reg_compile_gen(parser, (str), (options)) static void reg_fragment_setenc_gen(struct parser_params*, VALUE, int); #define reg_fragment_setenc(str,options) reg_fragment_setenc_gen(parser, (str), (options)) static int reg_fragment_check_gen(struct parser_params*, VALUE, int); #define reg_fragment_check(str,options) reg_fragment_check_gen(parser, (str), (options)) static NODE *reg_named_capture_assign_gen(struct parser_params* parser, VALUE regexp, NODE *match); #define reg_named_capture_assign(regexp,match) reg_named_capture_assign_gen(parser,(regexp),(match)) #define get_id(id) (id) #define get_value(val) (val) #else #define value_expr(node) ((void)(node)) #define remove_begin(node) (node) #define rb_dvar_defined(id) 0 #define rb_local_defined(id) 0 static ID ripper_get_id(VALUE); #define get_id(id) ripper_get_id(id) static VALUE ripper_get_value(VALUE); #define get_value(val) ripper_get_value(val) static VALUE assignable_gen(struct parser_params*,VALUE); #define assignable(lhs,node) assignable_gen(parser, (lhs)) static int id_is_var_gen(struct parser_params *parser, ID id); #define id_is_var(id) id_is_var_gen(parser, (id)) #define node_assign(node1, node2) dispatch2(assign, (node1), (node2)) static VALUE new_op_assign_gen(struct parser_params *parser, VALUE lhs, VALUE op, VALUE rhs); static VALUE new_attr_op_assign_gen(struct parser_params *parser, VALUE lhs, VALUE type, VALUE attr, VALUE op, VALUE rhs); #define new_attr_op_assign(lhs, type, attr, op, rhs) new_attr_op_assign_gen(parser, (lhs), (type), (attr), (op), (rhs)) #endif /* !RIPPER */ #define new_op_assign(lhs, op, rhs) new_op_assign_gen(parser, (lhs), (op), (rhs)) static ID formal_argument_gen(struct parser_params*, ID); #define formal_argument(id) formal_argument_gen(parser, (id)) static ID shadowing_lvar_gen(struct parser_params*,ID); #define shadowing_lvar(name) shadowing_lvar_gen(parser, (name)) static void new_bv_gen(struct parser_params*,ID); #define new_bv(id) new_bv_gen(parser, (id)) static void local_push_gen(struct parser_params*,int); #define local_push(top) local_push_gen(parser,(top)) static void local_pop_gen(struct parser_params*); #define local_pop() local_pop_gen(parser) static int local_var_gen(struct parser_params*, ID); #define local_var(id) local_var_gen(parser, (id)) static int arg_var_gen(struct parser_params*, ID); #define arg_var(id) arg_var_gen(parser, (id)) static int local_id_gen(struct parser_params*, ID); #define local_id(id) local_id_gen(parser, (id)) static ID internal_id_gen(struct parser_params*); #define internal_id() internal_id_gen(parser) static const struct vtable *dyna_push_gen(struct parser_params *); #define dyna_push() dyna_push_gen(parser) static void dyna_pop_gen(struct parser_params*, const struct vtable *); #define dyna_pop(node) dyna_pop_gen(parser, (node)) static int dyna_in_block_gen(struct parser_params*); #define dyna_in_block() dyna_in_block_gen(parser) #define dyna_var(id) local_var(id) static int dvar_defined_gen(struct parser_params*,ID,int); #define dvar_defined(id) dvar_defined_gen(parser, (id), 0) #define dvar_defined_get(id) dvar_defined_gen(parser, (id), 1) static int dvar_curr_gen(struct parser_params*,ID); #define dvar_curr(id) dvar_curr_gen(parser, (id)) static int lvar_defined_gen(struct parser_params*, ID); #define lvar_defined(id) lvar_defined_gen(parser, (id)) #define RE_OPTION_ONCE (1<<16) #define RE_OPTION_ENCODING_SHIFT 8 #define RE_OPTION_ENCODING(e) (((e)&0xff)<>RE_OPTION_ENCODING_SHIFT)&0xff) #define RE_OPTION_ENCODING_NONE(o) ((o)&RE_OPTION_ARG_ENCODING_NONE) #define RE_OPTION_MASK 0xff #define RE_OPTION_ARG_ENCODING_NONE 32 #define STR_OPTION_FROZEN 1 #define STR_OPTION_BINARY 0 /* disabled */ #define NODE_STRTERM NODE_ZARRAY /* nothing to gc */ #define NODE_HEREDOC NODE_ARRAY /* 1, 3 to gc */ #define SIGN_EXTEND(x,n) (((1<<(n)-1)^((x)&~(~0<<(n))))-(1<<(n)-1)) #define nd_func u1.id #if SIZEOF_SHORT == 2 #define nd_term(node) ((signed short)(node)->u2.id) #else #define nd_term(node) SIGN_EXTEND((node)->u2.id, CHAR_BIT*2) #endif #define nd_paren(node) (char)((node)->u2.id >> CHAR_BIT*2) #define nd_nest u3.cnt /****** Ripper *******/ #ifdef RIPPER #define RIPPER_VERSION "0.1.0" #include "eventids1.c" #include "eventids2.c" static VALUE ripper_dispatch0(struct parser_params*,ID); static VALUE ripper_dispatch1(struct parser_params*,ID,VALUE); static VALUE ripper_dispatch2(struct parser_params*,ID,VALUE,VALUE); static VALUE ripper_dispatch3(struct parser_params*,ID,VALUE,VALUE,VALUE); static VALUE ripper_dispatch4(struct parser_params*,ID,VALUE,VALUE,VALUE,VALUE); static VALUE ripper_dispatch5(struct parser_params*,ID,VALUE,VALUE,VALUE,VALUE,VALUE); static VALUE ripper_dispatch7(struct parser_params*,ID,VALUE,VALUE,VALUE,VALUE,VALUE,VALUE,VALUE); #define dispatch0(n) ripper_dispatch0(parser, TOKEN_PASTE(ripper_id_, n)) #define dispatch1(n,a) ripper_dispatch1(parser, TOKEN_PASTE(ripper_id_, n), (a)) #define dispatch2(n,a,b) ripper_dispatch2(parser, TOKEN_PASTE(ripper_id_, n), (a), (b)) #define dispatch3(n,a,b,c) ripper_dispatch3(parser, TOKEN_PASTE(ripper_id_, n), (a), (b), (c)) #define dispatch4(n,a,b,c,d) ripper_dispatch4(parser, TOKEN_PASTE(ripper_id_, n), (a), (b), (c), (d)) #define dispatch5(n,a,b,c,d,e) ripper_dispatch5(parser, TOKEN_PASTE(ripper_id_, n), (a), (b), (c), (d), (e)) #define dispatch7(n,a,b,c,d,e,f,g) ripper_dispatch7(parser, TOKEN_PASTE(ripper_id_, n), (a), (b), (c), (d), (e), (f), (g)) #define yyparse ripper_yyparse #define ripper_intern(s) ID2SYM(rb_intern(s)) static VALUE ripper_id2sym(ID); #ifdef __GNUC__ #define ripper_id2sym(id) ((id) < 256 && rb_ispunct(id) ? \ ID2SYM(id) : ripper_id2sym(id)) #endif #define arg_new() dispatch0(args_new) #define arg_add(l,a) dispatch2(args_add, (l), (a)) #define arg_add_star(l,a) dispatch2(args_add_star, (l), (a)) #define arg_add_block(l,b) dispatch2(args_add_block, (l), (b)) #define arg_add_optblock(l,b) ((b)==Qundef? (l) : dispatch2(args_add_block, (l), (b))) #define bare_assoc(v) dispatch1(bare_assoc_hash, (v)) #define arg_add_assocs(l,b) arg_add((l), bare_assoc(b)) #define args2mrhs(a) dispatch1(mrhs_new_from_args, (a)) #define mrhs_new() dispatch0(mrhs_new) #define mrhs_add(l,a) dispatch2(mrhs_add, (l), (a)) #define mrhs_add_star(l,a) dispatch2(mrhs_add_star, (l), (a)) #define mlhs_new() dispatch0(mlhs_new) #define mlhs_add(l,a) dispatch2(mlhs_add, (l), (a)) #define mlhs_add_star(l,a) dispatch2(mlhs_add_star, (l), (a)) #define params_new(pars, opts, rest, pars2, kws, kwrest, blk) \ dispatch7(params, (pars), (opts), (rest), (pars2), (kws), (kwrest), (blk)) #define blockvar_new(p,v) dispatch2(block_var, (p), (v)) #define blockvar_add_star(l,a) dispatch2(block_var_add_star, (l), (a)) #define blockvar_add_block(l,a) dispatch2(block_var_add_block, (l), (a)) #define method_optarg(m,a) ((a)==Qundef ? (m) : dispatch2(method_add_arg,(m),(a))) #define method_arg(m,a) dispatch2(method_add_arg,(m),(a)) #define method_add_block(m,b) dispatch2(method_add_block, (m), (b)) #define escape_Qundef(x) ((x)==Qundef ? Qnil : (x)) static inline VALUE new_args_gen(struct parser_params *parser, VALUE f, VALUE o, VALUE r, VALUE p, VALUE tail) { NODE *t = (NODE *)tail; VALUE k = t->u1.value, kr = t->u2.value, b = t->u3.value; return params_new(f, o, r, p, k, kr, escape_Qundef(b)); } #define new_args(f,o,r,p,t) new_args_gen(parser, (f),(o),(r),(p),(t)) static inline VALUE new_args_tail_gen(struct parser_params *parser, VALUE k, VALUE kr, VALUE b) { return (VALUE)rb_node_newnode(NODE_MEMO, k, kr, b); } #define new_args_tail(k,kr,b) new_args_tail_gen(parser, (k),(kr),(b)) #define new_defined(expr) dispatch1(defined, (expr)) #define FIXME 0 #endif /* RIPPER */ #ifndef RIPPER # define Qnone 0 # define ifndef_ripper(x) (x) #else # define Qnone Qnil # define ifndef_ripper(x) #endif #ifndef RIPPER # define rb_warn0(fmt) rb_compile_warn(ruby_sourcefile, ruby_sourceline, (fmt)) # define rb_warnI(fmt,a) rb_compile_warn(ruby_sourcefile, ruby_sourceline, (fmt), (a)) # define rb_warnS(fmt,a) rb_compile_warn(ruby_sourcefile, ruby_sourceline, (fmt), (a)) # define rb_warn4S(file,line,fmt,a) rb_compile_warn((file), (line), (fmt), (a)) # define rb_warning0(fmt) rb_compile_warning(ruby_sourcefile, ruby_sourceline, (fmt)) # define rb_warningS(fmt,a) rb_compile_warning(ruby_sourcefile, ruby_sourceline, (fmt), (a)) #else # define rb_warn0(fmt) ripper_warn0(parser, (fmt)) # define rb_warnI(fmt,a) ripper_warnI(parser, (fmt), (a)) # define rb_warnS(fmt,a) ripper_warnS(parser, (fmt), (a)) # define rb_warn4S(file,line,fmt,a) ripper_warnS(parser, (fmt), (a)) # define rb_warning0(fmt) ripper_warning0(parser, (fmt)) # define rb_warningS(fmt,a) ripper_warningS(parser, (fmt), (a)) static void ripper_warn0(struct parser_params*, const char*); static void ripper_warnI(struct parser_params*, const char*, int); static void ripper_warnS(struct parser_params*, const char*, const char*); static void ripper_warning0(struct parser_params*, const char*); static void ripper_warningS(struct parser_params*, const char*, const char*); #endif #ifdef RIPPER static void ripper_compile_error(struct parser_params*, const char *fmt, ...); # define rb_compile_error ripper_compile_error # define compile_error ripper_compile_error # define PARSER_ARG parser, #else # define rb_compile_error rb_compile_error_with_enc # define compile_error parser->nerr++,rb_compile_error_with_enc # define PARSER_ARG ruby_sourcefile, ruby_sourceline, current_enc, #endif /* Older versions of Yacc set YYMAXDEPTH to a very low value by default (150, for instance). This is too low for Ruby to parse some files, such as date/format.rb, therefore bump the value up to at least Bison's default. */ #ifdef OLD_YACC #ifndef YYMAXDEPTH #define YYMAXDEPTH 10000 #endif #endif #ifndef RIPPER static void token_info_push(struct parser_params*, const char *token); static void token_info_pop(struct parser_params*, const char *token); #define token_info_push(token) (RTEST(ruby_verbose) ? token_info_push(parser, (token)) : (void)0) #define token_info_pop(token) (RTEST(ruby_verbose) ? token_info_pop(parser, (token)) : (void)0) #else #define token_info_push(token) /* nothing */ #define token_info_pop(token) /* nothing */ #endif %} %pure-parser %lex-param {struct parser_params *parser} %parse-param {struct parser_params *parser} %union { VALUE val; NODE *node; ID id; int num; const struct vtable *vars; } /*%%%*/ %token /*% %token %*/ keyword_class keyword_module keyword_def keyword_undef keyword_begin keyword_rescue keyword_ensure keyword_end keyword_if keyword_unless keyword_then keyword_elsif keyword_else keyword_case keyword_when keyword_while keyword_until keyword_for keyword_break keyword_next keyword_redo keyword_retry keyword_in keyword_do keyword_do_cond keyword_do_block keyword_do_LAMBDA keyword_return keyword_yield keyword_super keyword_self keyword_nil keyword_true keyword_false keyword_and keyword_or keyword_not modifier_if modifier_unless modifier_while modifier_until modifier_rescue keyword_alias keyword_defined keyword_BEGIN keyword_END keyword__LINE__ keyword__FILE__ keyword__ENCODING__ %token tIDENTIFIER tFID tGVAR tIVAR tCONSTANT tCVAR tLABEL %token tINTEGER tFLOAT tRATIONAL tIMAGINARY tSTRING_CONTENT tCHAR %token tNTH_REF tBACK_REF %token tREGEXP_END tSTRING_SUFFIX %type singleton strings string string1 xstring regexp %type string_contents xstring_contents regexp_contents string_content %type words symbols symbol_list qwords qsymbols word_list qword_list qsym_list word %type literal numeric simple_numeric dsym cpath %type top_compstmt top_stmts top_stmt %type bodystmt compstmt stmts stmt_or_begin stmt expr arg primary command command_call method_call %type expr_value arg_value primary_value fcall %type if_tail opt_else case_body cases opt_rescue exc_list exc_var opt_ensure %type args call_args opt_call_args %type paren_args opt_paren_args args_tail opt_args_tail block_args_tail opt_block_args_tail %type command_args aref_args opt_block_arg block_arg var_ref var_lhs %type command_asgn mrhs mrhs_arg superclass block_call block_command %type f_block_optarg f_block_opt %type f_arglist f_args f_arg f_arg_item f_optarg f_marg f_marg_list f_margs %type assoc_list assocs assoc undef_list backref string_dvar for_var %type block_param opt_block_param block_param_def f_opt %type f_kwarg f_kw f_block_kwarg f_block_kw %type bv_decls opt_bv_decl bvar %type lambda f_larglist lambda_body %type brace_block cmd_brace_block do_block lhs none fitem %type mlhs mlhs_head mlhs_basic mlhs_item mlhs_node mlhs_post mlhs_inner %type fsym keyword_variable user_variable sym symbol operation operation2 operation3 %type cname fname op f_rest_arg f_block_arg opt_f_block_arg f_norm_arg f_bad_arg %type f_kwrest f_label %type opt_string_sfx /*%%%*/ /*% %type program reswords then do dot_or_colon %*/ %token END_OF_INPUT 0 "end-of-input" %token tUPLUS RUBY_TOKEN(UPLUS) "unary+" %token tUMINUS RUBY_TOKEN(UMINUS) "unary-" %token tPOW RUBY_TOKEN(POW) "**" %token tCMP RUBY_TOKEN(CMP) "<=>" %token tEQ RUBY_TOKEN(EQ) "==" %token tEQQ RUBY_TOKEN(EQQ) "===" %token tNEQ RUBY_TOKEN(NEQ) "!=" %token tGEQ RUBY_TOKEN(GEQ) ">=" %token tLEQ RUBY_TOKEN(LEQ) "<=" %token tANDOP "&&" %token tOROP "||" %token tMATCH RUBY_TOKEN(MATCH) "=~" %token tNMATCH RUBY_TOKEN(NMATCH) "!~" %token tDOT2 RUBY_TOKEN(DOT2) ".." %token tDOT3 RUBY_TOKEN(DOT3) "..." %token tAREF RUBY_TOKEN(AREF) "[]" %token tASET RUBY_TOKEN(ASET) "[]=" %token tLSHFT RUBY_TOKEN(LSHFT) "<<" %token tRSHFT RUBY_TOKEN(RSHFT) ">>" %token tCOLON2 "::" %token tCOLON3 ":: at EXPR_BEG" %token tOP_ASGN /* +=, -= etc. */ %token tASSOC "=>" %token tLPAREN "(" %token tLPAREN_ARG "( arg" %token tRPAREN ")" %token tLBRACK "[" %token tLBRACE "{" %token tLBRACE_ARG "{ arg" %token tSTAR "*" %token tDSTAR "**arg" %token tAMPER "&" %token tLAMBDA "->" %token tSYMBEG tSTRING_BEG tXSTRING_BEG tREGEXP_BEG tWORDS_BEG tQWORDS_BEG tSYMBOLS_BEG tQSYMBOLS_BEG %token tSTRING_DBEG tSTRING_DEND tSTRING_DVAR tSTRING_END tLAMBEG /* * precedence table */ %nonassoc tLOWEST %nonassoc tLBRACE_ARG %nonassoc modifier_if modifier_unless modifier_while modifier_until %left keyword_or keyword_and %right keyword_not %nonassoc keyword_defined %right '=' tOP_ASGN %left modifier_rescue %right '?' ':' %nonassoc tDOT2 tDOT3 %left tOROP %left tANDOP %nonassoc tCMP tEQ tEQQ tNEQ tMATCH tNMATCH %left '>' tGEQ '<' tLEQ %left '|' '^' %left '&' %left tLSHFT tRSHFT %left '+' '-' %left '*' '/' '%' %right tUMINUS_NUM tUMINUS %right tPOW %right '!' '~' tUPLUS %token tLAST_TOKEN %% program : { lex_state = EXPR_BEG; /*%%%*/ local_push(compile_for_eval || rb_parse_in_main()); /*% local_push(0); %*/ } top_compstmt { /*%%%*/ if ($2 && !compile_for_eval) { /* last expression should not be void */ if (nd_type($2) != NODE_BLOCK) void_expr($2); else { NODE *node = $2; while (node->nd_next) { node = node->nd_next; } void_expr(node->nd_head); } } ruby_eval_tree = NEW_SCOPE(0, block_append(ruby_eval_tree, $2)); /*% $$ = $2; parser->result = dispatch1(program, $$); %*/ local_pop(); } ; top_compstmt : top_stmts opt_terms { /*%%%*/ void_stmts($1); fixup_nodes(&deferred_nodes); /*% %*/ $$ = $1; } ; top_stmts : none { /*%%%*/ $$ = NEW_BEGIN(0); /*% $$ = dispatch2(stmts_add, dispatch0(stmts_new), dispatch0(void_stmt)); %*/ } | top_stmt { /*%%%*/ $$ = newline_node($1); /*% $$ = dispatch2(stmts_add, dispatch0(stmts_new), $1); %*/ } | top_stmts terms top_stmt { /*%%%*/ $$ = block_append($1, newline_node($3)); /*% $$ = dispatch2(stmts_add, $1, $3); %*/ } | error top_stmt { $$ = remove_begin($2); } ; top_stmt : stmt | keyword_BEGIN { /*%%%*/ /* local_push(0); */ /*% %*/ } '{' top_compstmt '}' { /*%%%*/ ruby_eval_tree_begin = block_append(ruby_eval_tree_begin, $4); /* NEW_PREEXE($4)); */ /* local_pop(); */ $$ = NEW_BEGIN(0); /*% $$ = dispatch1(BEGIN, $4); %*/ } ; bodystmt : compstmt opt_rescue opt_else opt_ensure { /*%%%*/ $$ = $1; if ($2) { $$ = NEW_RESCUE($1, $2, $3); } else if ($3) { rb_warn0("else without rescue is useless"); $$ = block_append($$, $3); } if ($4) { if ($$) { $$ = NEW_ENSURE($$, $4); } else { $$ = block_append($4, NEW_NIL()); } } fixpos($$, $1); /*% $$ = dispatch4(bodystmt, escape_Qundef($1), escape_Qundef($2), escape_Qundef($3), escape_Qundef($4)); %*/ } ; compstmt : stmts opt_terms { /*%%%*/ void_stmts($1); fixup_nodes(&deferred_nodes); /*% %*/ $$ = $1; } ; stmts : none { /*%%%*/ $$ = NEW_BEGIN(0); /*% $$ = dispatch2(stmts_add, dispatch0(stmts_new), dispatch0(void_stmt)); %*/ } | stmt_or_begin { /*%%%*/ $$ = newline_node($1); /*% $$ = dispatch2(stmts_add, dispatch0(stmts_new), $1); %*/ } | stmts terms stmt_or_begin { /*%%%*/ $$ = block_append($1, newline_node($3)); /*% $$ = dispatch2(stmts_add, $1, $3); %*/ } | error stmt { $$ = remove_begin($2); } ; stmt_or_begin : stmt { $$ = $1; } | keyword_BEGIN { yyerror("BEGIN is permitted only at toplevel"); /*%%%*/ /* local_push(0); */ /*% %*/ } '{' top_compstmt '}' { /*%%%*/ ruby_eval_tree_begin = block_append(ruby_eval_tree_begin, $4); /* NEW_PREEXE($4)); */ /* local_pop(); */ $$ = NEW_BEGIN(0); /*% $$ = dispatch1(BEGIN, $4); %*/ } stmt : keyword_alias fitem {lex_state = EXPR_FNAME;} fitem { /*%%%*/ $$ = NEW_ALIAS($2, $4); /*% $$ = dispatch2(alias, $2, $4); %*/ } | keyword_alias tGVAR tGVAR { /*%%%*/ $$ = NEW_VALIAS($2, $3); /*% $$ = dispatch2(var_alias, $2, $3); %*/ } | keyword_alias tGVAR tBACK_REF { /*%%%*/ char buf[2]; buf[0] = '$'; buf[1] = (char)$3->nd_nth; $$ = NEW_VALIAS($2, rb_intern2(buf, 2)); /*% $$ = dispatch2(var_alias, $2, $3); %*/ } | keyword_alias tGVAR tNTH_REF { /*%%%*/ yyerror("can't make alias for the number variables"); $$ = NEW_BEGIN(0); /*% $$ = dispatch2(var_alias, $2, $3); $$ = dispatch1(alias_error, $$); %*/ } | keyword_undef undef_list { /*%%%*/ $$ = $2; /*% $$ = dispatch1(undef, $2); %*/ } | stmt modifier_if expr_value { /*%%%*/ $$ = NEW_IF(cond($3), remove_begin($1), 0); fixpos($$, $3); /*% $$ = dispatch2(if_mod, $3, $1); %*/ } | stmt modifier_unless expr_value { /*%%%*/ $$ = NEW_UNLESS(cond($3), remove_begin($1), 0); fixpos($$, $3); /*% $$ = dispatch2(unless_mod, $3, $1); %*/ } | stmt modifier_while expr_value { /*%%%*/ if ($1 && nd_type($1) == NODE_BEGIN) { $$ = NEW_WHILE(cond($3), $1->nd_body, 0); } else { $$ = NEW_WHILE(cond($3), $1, 1); } /*% $$ = dispatch2(while_mod, $3, $1); %*/ } | stmt modifier_until expr_value { /*%%%*/ if ($1 && nd_type($1) == NODE_BEGIN) { $$ = NEW_UNTIL(cond($3), $1->nd_body, 0); } else { $$ = NEW_UNTIL(cond($3), $1, 1); } /*% $$ = dispatch2(until_mod, $3, $1); %*/ } | stmt modifier_rescue stmt { /*%%%*/ NODE *resq = NEW_RESBODY(0, remove_begin($3), 0); $$ = NEW_RESCUE(remove_begin($1), resq, 0); /*% $$ = dispatch2(rescue_mod, $1, $3); %*/ } | keyword_END '{' compstmt '}' { if (in_def || in_single) { rb_warn0("END in method; use at_exit"); } /*%%%*/ $$ = NEW_POSTEXE(NEW_NODE( NODE_SCOPE, 0 /* tbl */, $3 /* body */, 0 /* args */)); /*% $$ = dispatch1(END, $3); %*/ } | command_asgn | mlhs '=' command_call { /*%%%*/ value_expr($3); $1->nd_value = $3; $$ = $1; /*% $$ = dispatch2(massign, $1, $3); %*/ } | var_lhs tOP_ASGN command_call { value_expr($3); $$ = new_op_assign($1, $2, $3); } | primary_value '[' opt_call_args rbracket tOP_ASGN command_call { /*%%%*/ NODE *args; value_expr($6); if (!$3) $3 = NEW_ZARRAY(); args = arg_concat($3, $6); if ($5 == tOROP) { $5 = 0; } else if ($5 == tANDOP) { $5 = 1; } $$ = NEW_OP_ASGN1($1, $5, args); fixpos($$, $1); /*% $$ = dispatch2(aref_field, $1, escape_Qundef($3)); $$ = dispatch3(opassign, $$, $5, $6); %*/ } | primary_value '.' tIDENTIFIER tOP_ASGN command_call { value_expr($5); $$ = new_attr_op_assign($1, ripper_id2sym('.'), $3, $4, $5); } | primary_value '.' tCONSTANT tOP_ASGN command_call { value_expr($5); $$ = new_attr_op_assign($1, ripper_id2sym('.'), $3, $4, $5); } | primary_value tCOLON2 tCONSTANT tOP_ASGN command_call { /*%%%*/ $$ = NEW_COLON2($1, $3); $$ = new_const_op_assign($$, $4, $5); /*% $$ = dispatch2(const_path_field, $1, $3); $$ = dispatch3(opassign, $$, $4, $5); %*/ } | primary_value tCOLON2 tIDENTIFIER tOP_ASGN command_call { value_expr($5); $$ = new_attr_op_assign($1, ripper_intern("::"), $3, $4, $5); } | backref tOP_ASGN command_call { /*%%%*/ rb_backref_error($1); $$ = NEW_BEGIN(0); /*% $$ = dispatch2(assign, dispatch1(var_field, $1), $3); $$ = dispatch1(assign_error, $$); %*/ } | lhs '=' mrhs { /*%%%*/ value_expr($3); $$ = node_assign($1, $3); /*% $$ = dispatch2(assign, $1, $3); %*/ } | mlhs '=' mrhs_arg { /*%%%*/ $1->nd_value = $3; $$ = $1; /*% $$ = dispatch2(massign, $1, $3); %*/ } | expr ; command_asgn : lhs '=' command_call { /*%%%*/ value_expr($3); $$ = node_assign($1, $3); /*% $$ = dispatch2(assign, $1, $3); %*/ } | lhs '=' command_asgn { /*%%%*/ value_expr($3); $$ = node_assign($1, $3); /*% $$ = dispatch2(assign, $1, $3); %*/ } ; expr : command_call | expr keyword_and expr { /*%%%*/ $$ = logop(NODE_AND, $1, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("and"), $3); %*/ } | expr keyword_or expr { /*%%%*/ $$ = logop(NODE_OR, $1, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("or"), $3); %*/ } | keyword_not opt_nl expr { /*%%%*/ $$ = call_uni_op(cond($3), '!'); /*% $$ = dispatch2(unary, ripper_intern("not"), $3); %*/ } | '!' command_call { /*%%%*/ $$ = call_uni_op(cond($2), '!'); /*% $$ = dispatch2(unary, ripper_id2sym('!'), $2); %*/ } | arg ; expr_value : expr { /*%%%*/ value_expr($1); $$ = $1; if (!$$) $$ = NEW_NIL(); /*% $$ = $1; %*/ } ; command_call : command | block_command ; block_command : block_call | block_call dot_or_colon operation2 command_args { /*%%%*/ $$ = NEW_CALL($1, $3, $4); /*% $$ = dispatch3(call, $1, $2, $3); $$ = method_arg($$, $4); %*/ } ; cmd_brace_block : tLBRACE_ARG { $1 = dyna_push(); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } opt_block_param compstmt '}' { /*%%%*/ $$ = NEW_ITER($3,$4); nd_set_line($$, $2); /*% $$ = dispatch2(brace_block, escape_Qundef($3), $4); %*/ dyna_pop($1); } ; fcall : operation { /*%%%*/ $$ = NEW_FCALL($1, 0); nd_set_line($$, tokline); /*% %*/ } ; command : fcall command_args %prec tLOWEST { /*%%%*/ $$ = $1; $$->nd_args = $2; /*% $$ = dispatch2(command, $1, $2); %*/ } | fcall command_args cmd_brace_block { /*%%%*/ block_dup_check($2,$3); $1->nd_args = $2; $3->nd_iter = $1; $$ = $3; fixpos($$, $1); /*% $$ = dispatch2(command, $1, $2); $$ = method_add_block($$, $3); %*/ } | primary_value '.' operation2 command_args %prec tLOWEST { /*%%%*/ $$ = NEW_CALL($1, $3, $4); fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, ripper_id2sym('.'), $3, $4); %*/ } | primary_value '.' operation2 command_args cmd_brace_block { /*%%%*/ block_dup_check($4,$5); $5->nd_iter = NEW_CALL($1, $3, $4); $$ = $5; fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, ripper_id2sym('.'), $3, $4); $$ = method_add_block($$, $5); %*/ } | primary_value tCOLON2 operation2 command_args %prec tLOWEST { /*%%%*/ $$ = NEW_CALL($1, $3, $4); fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, ripper_intern("::"), $3, $4); %*/ } | primary_value tCOLON2 operation2 command_args cmd_brace_block { /*%%%*/ block_dup_check($4,$5); $5->nd_iter = NEW_CALL($1, $3, $4); $$ = $5; fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, ripper_intern("::"), $3, $4); $$ = method_add_block($$, $5); %*/ } | keyword_super command_args { /*%%%*/ $$ = NEW_SUPER($2); fixpos($$, $2); /*% $$ = dispatch1(super, $2); %*/ } | keyword_yield command_args { /*%%%*/ $$ = new_yield($2); fixpos($$, $2); /*% $$ = dispatch1(yield, $2); %*/ } | keyword_return call_args { /*%%%*/ $$ = NEW_RETURN(ret_args($2)); /*% $$ = dispatch1(return, $2); %*/ } | keyword_break call_args { /*%%%*/ $$ = NEW_BREAK(ret_args($2)); /*% $$ = dispatch1(break, $2); %*/ } | keyword_next call_args { /*%%%*/ $$ = NEW_NEXT(ret_args($2)); /*% $$ = dispatch1(next, $2); %*/ } ; mlhs : mlhs_basic | tLPAREN mlhs_inner rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(mlhs_paren, $2); %*/ } ; mlhs_inner : mlhs_basic | tLPAREN mlhs_inner rparen { /*%%%*/ $$ = NEW_MASGN(NEW_LIST($2), 0); /*% $$ = dispatch1(mlhs_paren, $2); %*/ } ; mlhs_basic : mlhs_head { /*%%%*/ $$ = NEW_MASGN($1, 0); /*% $$ = $1; %*/ } | mlhs_head mlhs_item { /*%%%*/ $$ = NEW_MASGN(list_append($1,$2), 0); /*% $$ = mlhs_add($1, $2); %*/ } | mlhs_head tSTAR mlhs_node { /*%%%*/ $$ = NEW_MASGN($1, $3); /*% $$ = mlhs_add_star($1, $3); %*/ } | mlhs_head tSTAR mlhs_node ',' mlhs_post { /*%%%*/ $$ = NEW_MASGN($1, NEW_POSTARG($3,$5)); /*% $1 = mlhs_add_star($1, $3); $$ = mlhs_add($1, $5); %*/ } | mlhs_head tSTAR { /*%%%*/ $$ = NEW_MASGN($1, -1); /*% $$ = mlhs_add_star($1, Qnil); %*/ } | mlhs_head tSTAR ',' mlhs_post { /*%%%*/ $$ = NEW_MASGN($1, NEW_POSTARG(-1, $4)); /*% $1 = mlhs_add_star($1, Qnil); $$ = mlhs_add($1, $4); %*/ } | tSTAR mlhs_node { /*%%%*/ $$ = NEW_MASGN(0, $2); /*% $$ = mlhs_add_star(mlhs_new(), $2); %*/ } | tSTAR mlhs_node ',' mlhs_post { /*%%%*/ $$ = NEW_MASGN(0, NEW_POSTARG($2,$4)); /*% $2 = mlhs_add_star(mlhs_new(), $2); $$ = mlhs_add($2, $4); %*/ } | tSTAR { /*%%%*/ $$ = NEW_MASGN(0, -1); /*% $$ = mlhs_add_star(mlhs_new(), Qnil); %*/ } | tSTAR ',' mlhs_post { /*%%%*/ $$ = NEW_MASGN(0, NEW_POSTARG(-1, $3)); /*% $$ = mlhs_add_star(mlhs_new(), Qnil); $$ = mlhs_add($$, $3); %*/ } ; mlhs_item : mlhs_node | tLPAREN mlhs_inner rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(mlhs_paren, $2); %*/ } ; mlhs_head : mlhs_item ',' { /*%%%*/ $$ = NEW_LIST($1); /*% $$ = mlhs_add(mlhs_new(), $1); %*/ } | mlhs_head mlhs_item ',' { /*%%%*/ $$ = list_append($1, $2); /*% $$ = mlhs_add($1, $2); %*/ } ; mlhs_post : mlhs_item { /*%%%*/ $$ = NEW_LIST($1); /*% $$ = mlhs_add(mlhs_new(), $1); %*/ } | mlhs_post ',' mlhs_item { /*%%%*/ $$ = list_append($1, $3); /*% $$ = mlhs_add($1, $3); %*/ } ; mlhs_node : user_variable { $$ = assignable($1, 0); } | keyword_variable { $$ = assignable($1, 0); } | primary_value '[' opt_call_args rbracket { /*%%%*/ $$ = aryset($1, $3); /*% $$ = dispatch2(aref_field, $1, escape_Qundef($3)); %*/ } | primary_value '.' tIDENTIFIER { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch3(field, $1, ripper_id2sym('.'), $3); %*/ } | primary_value tCOLON2 tIDENTIFIER { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch2(const_path_field, $1, $3); %*/ } | primary_value '.' tCONSTANT { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch3(field, $1, ripper_id2sym('.'), $3); %*/ } | primary_value tCOLON2 tCONSTANT { /*%%%*/ if (in_def || in_single) yyerror("dynamic constant assignment"); $$ = NEW_CDECL(0, 0, NEW_COLON2($1, $3)); /*% if (in_def || in_single) yyerror("dynamic constant assignment"); $$ = dispatch2(const_path_field, $1, $3); %*/ } | tCOLON3 tCONSTANT { /*%%%*/ if (in_def || in_single) yyerror("dynamic constant assignment"); $$ = NEW_CDECL(0, 0, NEW_COLON3($2)); /*% $$ = dispatch1(top_const_field, $2); %*/ } | backref { /*%%%*/ rb_backref_error($1); $$ = NEW_BEGIN(0); /*% $$ = dispatch1(var_field, $1); $$ = dispatch1(assign_error, $$); %*/ } ; lhs : user_variable { $$ = assignable($1, 0); /*%%%*/ if (!$$) $$ = NEW_BEGIN(0); /*% $$ = dispatch1(var_field, $$); %*/ } | keyword_variable { $$ = assignable($1, 0); /*%%%*/ if (!$$) $$ = NEW_BEGIN(0); /*% $$ = dispatch1(var_field, $$); %*/ } | primary_value '[' opt_call_args rbracket { /*%%%*/ $$ = aryset($1, $3); /*% $$ = dispatch2(aref_field, $1, escape_Qundef($3)); %*/ } | primary_value '.' tIDENTIFIER { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch3(field, $1, ripper_id2sym('.'), $3); %*/ } | primary_value tCOLON2 tIDENTIFIER { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch3(field, $1, ripper_intern("::"), $3); %*/ } | primary_value '.' tCONSTANT { /*%%%*/ $$ = attrset($1, $3); /*% $$ = dispatch3(field, $1, ripper_id2sym('.'), $3); %*/ } | primary_value tCOLON2 tCONSTANT { /*%%%*/ if (in_def || in_single) yyerror("dynamic constant assignment"); $$ = NEW_CDECL(0, 0, NEW_COLON2($1, $3)); /*% $$ = dispatch2(const_path_field, $1, $3); if (in_def || in_single) { $$ = dispatch1(assign_error, $$); } %*/ } | tCOLON3 tCONSTANT { /*%%%*/ if (in_def || in_single) yyerror("dynamic constant assignment"); $$ = NEW_CDECL(0, 0, NEW_COLON3($2)); /*% $$ = dispatch1(top_const_field, $2); if (in_def || in_single) { $$ = dispatch1(assign_error, $$); } %*/ } | backref { /*%%%*/ rb_backref_error($1); $$ = NEW_BEGIN(0); /*% $$ = dispatch1(assign_error, $1); %*/ } ; cname : tIDENTIFIER { /*%%%*/ yyerror("class/module name must be CONSTANT"); /*% $$ = dispatch1(class_name_error, $1); %*/ } | tCONSTANT ; cpath : tCOLON3 cname { /*%%%*/ $$ = NEW_COLON3($2); /*% $$ = dispatch1(top_const_ref, $2); %*/ } | cname { /*%%%*/ $$ = NEW_COLON2(0, $$); /*% $$ = dispatch1(const_ref, $1); %*/ } | primary_value tCOLON2 cname { /*%%%*/ $$ = NEW_COLON2($1, $3); /*% $$ = dispatch2(const_path_ref, $1, $3); %*/ } ; fname : tIDENTIFIER | tCONSTANT | tFID | op { lex_state = EXPR_ENDFN; $$ = $1; } | reswords { lex_state = EXPR_ENDFN; /*%%%*/ $$ = $1; /*% $$ = $1; %*/ } ; fsym : fname | symbol ; fitem : fsym { /*%%%*/ $$ = NEW_LIT(ID2SYM($1)); /*% $$ = dispatch1(symbol_literal, $1); %*/ } | dsym ; undef_list : fitem { /*%%%*/ $$ = NEW_UNDEF($1); /*% $$ = rb_ary_new3(1, $1); %*/ } | undef_list ',' {lex_state = EXPR_FNAME;} fitem { /*%%%*/ $$ = block_append($1, NEW_UNDEF($4)); /*% rb_ary_push($1, $4); %*/ } ; op : '|' { ifndef_ripper($$ = '|'); } | '^' { ifndef_ripper($$ = '^'); } | '&' { ifndef_ripper($$ = '&'); } | tCMP { ifndef_ripper($$ = tCMP); } | tEQ { ifndef_ripper($$ = tEQ); } | tEQQ { ifndef_ripper($$ = tEQQ); } | tMATCH { ifndef_ripper($$ = tMATCH); } | tNMATCH { ifndef_ripper($$ = tNMATCH); } | '>' { ifndef_ripper($$ = '>'); } | tGEQ { ifndef_ripper($$ = tGEQ); } | '<' { ifndef_ripper($$ = '<'); } | tLEQ { ifndef_ripper($$ = tLEQ); } | tNEQ { ifndef_ripper($$ = tNEQ); } | tLSHFT { ifndef_ripper($$ = tLSHFT); } | tRSHFT { ifndef_ripper($$ = tRSHFT); } | '+' { ifndef_ripper($$ = '+'); } | '-' { ifndef_ripper($$ = '-'); } | '*' { ifndef_ripper($$ = '*'); } | tSTAR { ifndef_ripper($$ = '*'); } | '/' { ifndef_ripper($$ = '/'); } | '%' { ifndef_ripper($$ = '%'); } | tPOW { ifndef_ripper($$ = tPOW); } | tDSTAR { ifndef_ripper($$ = tDSTAR); } | '!' { ifndef_ripper($$ = '!'); } | '~' { ifndef_ripper($$ = '~'); } | tUPLUS { ifndef_ripper($$ = tUPLUS); } | tUMINUS { ifndef_ripper($$ = tUMINUS); } | tAREF { ifndef_ripper($$ = tAREF); } | tASET { ifndef_ripper($$ = tASET); } | '`' { ifndef_ripper($$ = '`'); } ; reswords : keyword__LINE__ | keyword__FILE__ | keyword__ENCODING__ | keyword_BEGIN | keyword_END | keyword_alias | keyword_and | keyword_begin | keyword_break | keyword_case | keyword_class | keyword_def | keyword_defined | keyword_do | keyword_else | keyword_elsif | keyword_end | keyword_ensure | keyword_false | keyword_for | keyword_in | keyword_module | keyword_next | keyword_nil | keyword_not | keyword_or | keyword_redo | keyword_rescue | keyword_retry | keyword_return | keyword_self | keyword_super | keyword_then | keyword_true | keyword_undef | keyword_when | keyword_yield | keyword_if | keyword_unless | keyword_while | keyword_until ; arg : lhs '=' arg { /*%%%*/ value_expr($3); $$ = node_assign($1, $3); /*% $$ = dispatch2(assign, $1, $3); %*/ } | lhs '=' arg modifier_rescue arg { /*%%%*/ value_expr($3); $3 = NEW_RESCUE($3, NEW_RESBODY(0,$5,0), 0); $$ = node_assign($1, $3); /*% $$ = dispatch2(assign, $1, dispatch2(rescue_mod, $3, $5)); %*/ } | var_lhs tOP_ASGN arg { value_expr($3); $$ = new_op_assign($1, $2, $3); } | var_lhs tOP_ASGN arg modifier_rescue arg { /*%%%*/ value_expr($3); $3 = NEW_RESCUE($3, NEW_RESBODY(0,$5,0), 0); /*% $3 = dispatch2(rescue_mod, $3, $5); %*/ $$ = new_op_assign($1, $2, $3); } | primary_value '[' opt_call_args rbracket tOP_ASGN arg { /*%%%*/ NODE *args; value_expr($6); if (!$3) $3 = NEW_ZARRAY(); if (nd_type($3) == NODE_BLOCK_PASS) { args = NEW_ARGSCAT($3, $6); } else { args = arg_concat($3, $6); } if ($5 == tOROP) { $5 = 0; } else if ($5 == tANDOP) { $5 = 1; } $$ = NEW_OP_ASGN1($1, $5, args); fixpos($$, $1); /*% $1 = dispatch2(aref_field, $1, escape_Qundef($3)); $$ = dispatch3(opassign, $1, $5, $6); %*/ } | primary_value '.' tIDENTIFIER tOP_ASGN arg { value_expr($5); $$ = new_attr_op_assign($1, ripper_id2sym('.'), $3, $4, $5); } | primary_value '.' tCONSTANT tOP_ASGN arg { value_expr($5); $$ = new_attr_op_assign($1, ripper_id2sym('.'), $3, $4, $5); } | primary_value tCOLON2 tIDENTIFIER tOP_ASGN arg { value_expr($5); $$ = new_attr_op_assign($1, ripper_intern("::"), $3, $4, $5); } | primary_value tCOLON2 tCONSTANT tOP_ASGN arg { /*%%%*/ $$ = NEW_COLON2($1, $3); $$ = new_const_op_assign($$, $4, $5); /*% $$ = dispatch2(const_path_field, $1, $3); $$ = dispatch3(opassign, $$, $4, $5); %*/ } | tCOLON3 tCONSTANT tOP_ASGN arg { /*%%%*/ $$ = NEW_COLON3($2); $$ = new_const_op_assign($$, $3, $4); /*% $$ = dispatch1(top_const_field, $2); $$ = dispatch3(opassign, $$, $3, $4); %*/ } | backref tOP_ASGN arg { /*%%%*/ rb_backref_error($1); $$ = NEW_BEGIN(0); /*% $$ = dispatch1(var_field, $1); $$ = dispatch3(opassign, $$, $2, $3); $$ = dispatch1(assign_error, $$); %*/ } | arg tDOT2 arg { /*%%%*/ value_expr($1); value_expr($3); $$ = NEW_DOT2($1, $3); if (nd_type($1) == NODE_LIT && FIXNUM_P($1->nd_lit) && nd_type($3) == NODE_LIT && FIXNUM_P($3->nd_lit)) { deferred_nodes = list_append(deferred_nodes, $$); } /*% $$ = dispatch2(dot2, $1, $3); %*/ } | arg tDOT3 arg { /*%%%*/ value_expr($1); value_expr($3); $$ = NEW_DOT3($1, $3); if (nd_type($1) == NODE_LIT && FIXNUM_P($1->nd_lit) && nd_type($3) == NODE_LIT && FIXNUM_P($3->nd_lit)) { deferred_nodes = list_append(deferred_nodes, $$); } /*% $$ = dispatch2(dot3, $1, $3); %*/ } | arg '+' arg { /*%%%*/ $$ = call_bin_op($1, '+', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('+'), $3); %*/ } | arg '-' arg { /*%%%*/ $$ = call_bin_op($1, '-', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('-'), $3); %*/ } | arg '*' arg { /*%%%*/ $$ = call_bin_op($1, '*', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('*'), $3); %*/ } | arg '/' arg { /*%%%*/ $$ = call_bin_op($1, '/', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('/'), $3); %*/ } | arg '%' arg { /*%%%*/ $$ = call_bin_op($1, '%', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('%'), $3); %*/ } | arg tPOW arg { /*%%%*/ $$ = call_bin_op($1, tPOW, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("**"), $3); %*/ } | tUMINUS_NUM simple_numeric tPOW arg { /*%%%*/ $$ = NEW_CALL(call_bin_op($2, tPOW, $4), tUMINUS, 0); /*% $$ = dispatch3(binary, $2, ripper_intern("**"), $4); $$ = dispatch2(unary, ripper_intern("-@"), $$); %*/ } | tUPLUS arg { /*%%%*/ $$ = call_uni_op($2, tUPLUS); /*% $$ = dispatch2(unary, ripper_intern("+@"), $2); %*/ } | tUMINUS arg { /*%%%*/ $$ = call_uni_op($2, tUMINUS); /*% $$ = dispatch2(unary, ripper_intern("-@"), $2); %*/ } | arg '|' arg { /*%%%*/ $$ = call_bin_op($1, '|', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('|'), $3); %*/ } | arg '^' arg { /*%%%*/ $$ = call_bin_op($1, '^', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('^'), $3); %*/ } | arg '&' arg { /*%%%*/ $$ = call_bin_op($1, '&', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('&'), $3); %*/ } | arg tCMP arg { /*%%%*/ $$ = call_bin_op($1, tCMP, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("<=>"), $3); %*/ } | arg '>' arg { /*%%%*/ $$ = call_bin_op($1, '>', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('>'), $3); %*/ } | arg tGEQ arg { /*%%%*/ $$ = call_bin_op($1, tGEQ, $3); /*% $$ = dispatch3(binary, $1, ripper_intern(">="), $3); %*/ } | arg '<' arg { /*%%%*/ $$ = call_bin_op($1, '<', $3); /*% $$ = dispatch3(binary, $1, ID2SYM('<'), $3); %*/ } | arg tLEQ arg { /*%%%*/ $$ = call_bin_op($1, tLEQ, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("<="), $3); %*/ } | arg tEQ arg { /*%%%*/ $$ = call_bin_op($1, tEQ, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("=="), $3); %*/ } | arg tEQQ arg { /*%%%*/ $$ = call_bin_op($1, tEQQ, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("==="), $3); %*/ } | arg tNEQ arg { /*%%%*/ $$ = call_bin_op($1, tNEQ, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("!="), $3); %*/ } | arg tMATCH arg { /*%%%*/ $$ = match_op($1, $3); if (nd_type($1) == NODE_LIT && RB_TYPE_P($1->nd_lit, T_REGEXP)) { $$ = reg_named_capture_assign($1->nd_lit, $$); } /*% $$ = dispatch3(binary, $1, ripper_intern("=~"), $3); %*/ } | arg tNMATCH arg { /*%%%*/ $$ = call_bin_op($1, tNMATCH, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("!~"), $3); %*/ } | '!' arg { /*%%%*/ $$ = call_uni_op(cond($2), '!'); /*% $$ = dispatch2(unary, ID2SYM('!'), $2); %*/ } | '~' arg { /*%%%*/ $$ = call_uni_op($2, '~'); /*% $$ = dispatch2(unary, ID2SYM('~'), $2); %*/ } | arg tLSHFT arg { /*%%%*/ $$ = call_bin_op($1, tLSHFT, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("<<"), $3); %*/ } | arg tRSHFT arg { /*%%%*/ $$ = call_bin_op($1, tRSHFT, $3); /*% $$ = dispatch3(binary, $1, ripper_intern(">>"), $3); %*/ } | arg tANDOP arg { /*%%%*/ $$ = logop(NODE_AND, $1, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("&&"), $3); %*/ } | arg tOROP arg { /*%%%*/ $$ = logop(NODE_OR, $1, $3); /*% $$ = dispatch3(binary, $1, ripper_intern("||"), $3); %*/ } | keyword_defined opt_nl {in_defined = 1;} arg { /*%%%*/ in_defined = 0; $$ = new_defined($4); /*% in_defined = 0; $$ = dispatch1(defined, $4); %*/ } | arg '?' arg opt_nl ':' arg { /*%%%*/ value_expr($1); $$ = NEW_IF(cond($1), $3, $6); fixpos($$, $1); /*% $$ = dispatch3(ifop, $1, $3, $6); %*/ } | primary { $$ = $1; } ; arg_value : arg { /*%%%*/ value_expr($1); $$ = $1; if (!$$) $$ = NEW_NIL(); /*% $$ = $1; %*/ } ; aref_args : none | args trailer { $$ = $1; } | args ',' assocs trailer { /*%%%*/ $$ = arg_append($1, NEW_HASH($3)); /*% $$ = arg_add_assocs($1, $3); %*/ } | assocs trailer { /*%%%*/ $$ = NEW_LIST(NEW_HASH($1)); /*% $$ = arg_add_assocs(arg_new(), $1); %*/ } ; paren_args : '(' opt_call_args rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(arg_paren, escape_Qundef($2)); %*/ } ; opt_paren_args : none | paren_args ; opt_call_args : none | call_args | args ',' { $$ = $1; } | args ',' assocs ',' { /*%%%*/ $$ = arg_append($1, NEW_HASH($3)); /*% $$ = arg_add_assocs($1, $3); %*/ } | assocs ',' { /*%%%*/ $$ = NEW_LIST(NEW_HASH($1)); /*% $$ = arg_add_assocs(arg_new(), $1); %*/ } ; call_args : command { /*%%%*/ value_expr($1); $$ = NEW_LIST($1); /*% $$ = arg_add(arg_new(), $1); %*/ } | args opt_block_arg { /*%%%*/ $$ = arg_blk_pass($1, $2); /*% $$ = arg_add_optblock($1, $2); %*/ } | assocs opt_block_arg { /*%%%*/ $$ = NEW_LIST(NEW_HASH($1)); $$ = arg_blk_pass($$, $2); /*% $$ = arg_add_assocs(arg_new(), $1); $$ = arg_add_optblock($$, $2); %*/ } | args ',' assocs opt_block_arg { /*%%%*/ $$ = arg_append($1, NEW_HASH($3)); $$ = arg_blk_pass($$, $4); /*% $$ = arg_add_optblock(arg_add_assocs($1, $3), $4); %*/ } | block_arg /*%c%*/ /*%c { $$ = arg_add_block(arg_new(), $1); } %*/ ; command_args : { $$ = cmdarg_stack; CMDARG_PUSH(1); } call_args { /* CMDARG_POP() */ cmdarg_stack = $1; $$ = $2; } ; block_arg : tAMPER arg_value { /*%%%*/ $$ = NEW_BLOCK_PASS($2); /*% $$ = $2; %*/ } ; opt_block_arg : ',' block_arg { $$ = $2; } | none { $$ = 0; } ; args : arg_value { /*%%%*/ $$ = NEW_LIST($1); /*% $$ = arg_add(arg_new(), $1); %*/ } | tSTAR arg_value { /*%%%*/ $$ = NEW_SPLAT($2); /*% $$ = arg_add_star(arg_new(), $2); %*/ } | args ',' arg_value { /*%%%*/ NODE *n1; if ((n1 = splat_array($1)) != 0) { $$ = list_append(n1, $3); } else { $$ = arg_append($1, $3); } /*% $$ = arg_add($1, $3); %*/ } | args ',' tSTAR arg_value { /*%%%*/ NODE *n1; if ((nd_type($4) == NODE_ARRAY) && (n1 = splat_array($1)) != 0) { $$ = list_concat(n1, $4); } else { $$ = arg_concat($1, $4); } /*% $$ = arg_add_star($1, $4); %*/ } ; mrhs_arg : mrhs | arg_value ; mrhs : args ',' arg_value { /*%%%*/ NODE *n1; if ((n1 = splat_array($1)) != 0) { $$ = list_append(n1, $3); } else { $$ = arg_append($1, $3); } /*% $$ = mrhs_add(args2mrhs($1), $3); %*/ } | args ',' tSTAR arg_value { /*%%%*/ NODE *n1; if (nd_type($4) == NODE_ARRAY && (n1 = splat_array($1)) != 0) { $$ = list_concat(n1, $4); } else { $$ = arg_concat($1, $4); } /*% $$ = mrhs_add_star(args2mrhs($1), $4); %*/ } | tSTAR arg_value { /*%%%*/ $$ = NEW_SPLAT($2); /*% $$ = mrhs_add_star(mrhs_new(), $2); %*/ } ; primary : literal | strings | xstring | regexp | words | qwords | symbols | qsymbols | var_ref | backref | tFID { /*%%%*/ $$ = NEW_FCALL($1, 0); /*% $$ = method_arg(dispatch1(fcall, $1), arg_new()); %*/ } | k_begin { $1 = cmdarg_stack; cmdarg_stack = 0; /*%%%*/ $$ = ruby_sourceline; /*% %*/ } bodystmt k_end { cmdarg_stack = $1; /*%%%*/ if ($3 == NULL) { $$ = NEW_NIL(); } else { if (nd_type($3) == NODE_RESCUE || nd_type($3) == NODE_ENSURE) nd_set_line($3, $2); $$ = NEW_BEGIN($3); } nd_set_line($$, $2); /*% $$ = dispatch1(begin, $3); %*/ } | tLPAREN_ARG {lex_state = EXPR_ENDARG;} rparen { /*%%%*/ $$ = 0; /*% $$ = dispatch1(paren, 0); %*/ } | tLPAREN_ARG expr {lex_state = EXPR_ENDARG;} rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(paren, $2); %*/ } | tLPAREN compstmt ')' { /*%%%*/ $$ = $2; /*% $$ = dispatch1(paren, $2); %*/ } | primary_value tCOLON2 tCONSTANT { /*%%%*/ $$ = NEW_COLON2($1, $3); /*% $$ = dispatch2(const_path_ref, $1, $3); %*/ } | tCOLON3 tCONSTANT { /*%%%*/ $$ = NEW_COLON3($2); /*% $$ = dispatch1(top_const_ref, $2); %*/ } | tLBRACK aref_args ']' { /*%%%*/ if ($2 == 0) { $$ = NEW_ZARRAY(); /* zero length array*/ } else { $$ = $2; } /*% $$ = dispatch1(array, escape_Qundef($2)); %*/ } | tLBRACE assoc_list '}' { /*%%%*/ $$ = NEW_HASH($2); /*% $$ = dispatch1(hash, escape_Qundef($2)); %*/ } | keyword_return { /*%%%*/ $$ = NEW_RETURN(0); /*% $$ = dispatch0(return0); %*/ } | keyword_yield '(' call_args rparen { /*%%%*/ $$ = new_yield($3); /*% $$ = dispatch1(yield, dispatch1(paren, $3)); %*/ } | keyword_yield '(' rparen { /*%%%*/ $$ = NEW_YIELD(0); /*% $$ = dispatch1(yield, dispatch1(paren, arg_new())); %*/ } | keyword_yield { /*%%%*/ $$ = NEW_YIELD(0); /*% $$ = dispatch0(yield0); %*/ } | keyword_defined opt_nl '(' {in_defined = 1;} expr rparen { /*%%%*/ in_defined = 0; $$ = new_defined($5); /*% in_defined = 0; $$ = dispatch1(defined, $5); %*/ } | keyword_not '(' expr rparen { /*%%%*/ $$ = call_uni_op(cond($3), '!'); /*% $$ = dispatch2(unary, ripper_intern("not"), $3); %*/ } | keyword_not '(' rparen { /*%%%*/ $$ = call_uni_op(cond(NEW_NIL()), '!'); /*% $$ = dispatch2(unary, ripper_intern("not"), Qnil); %*/ } | fcall brace_block { /*%%%*/ $2->nd_iter = $1; $$ = $2; /*% $$ = method_arg(dispatch1(fcall, $1), arg_new()); $$ = method_add_block($$, $2); %*/ } | method_call | method_call brace_block { /*%%%*/ block_dup_check($1->nd_args, $2); $2->nd_iter = $1; $$ = $2; /*% $$ = method_add_block($1, $2); %*/ } | tLAMBDA lambda { $$ = $2; } | k_if expr_value then compstmt if_tail k_end { /*%%%*/ $$ = NEW_IF(cond($2), $4, $5); fixpos($$, $2); /*% $$ = dispatch3(if, $2, $4, escape_Qundef($5)); %*/ } | k_unless expr_value then compstmt opt_else k_end { /*%%%*/ $$ = NEW_UNLESS(cond($2), $4, $5); fixpos($$, $2); /*% $$ = dispatch3(unless, $2, $4, escape_Qundef($5)); %*/ } | k_while {COND_PUSH(1);} expr_value do {COND_POP();} compstmt k_end { /*%%%*/ $$ = NEW_WHILE(cond($3), $6, 1); fixpos($$, $3); /*% $$ = dispatch2(while, $3, $6); %*/ } | k_until {COND_PUSH(1);} expr_value do {COND_POP();} compstmt k_end { /*%%%*/ $$ = NEW_UNTIL(cond($3), $6, 1); fixpos($$, $3); /*% $$ = dispatch2(until, $3, $6); %*/ } | k_case expr_value opt_terms case_body k_end { /*%%%*/ $$ = NEW_CASE($2, $4); fixpos($$, $2); /*% $$ = dispatch2(case, $2, $4); %*/ } | k_case opt_terms case_body k_end { /*%%%*/ $$ = NEW_CASE(0, $3); /*% $$ = dispatch2(case, Qnil, $3); %*/ } | k_for for_var keyword_in {COND_PUSH(1);} expr_value do {COND_POP();} compstmt k_end { /*%%%*/ /* * for a, b, c in e * #=> * e.each{|*x| a, b, c = x * * for a in e * #=> * e.each{|x| a, = x} */ ID id = internal_id(); ID *tbl = ALLOC_N(ID, 2); NODE *m = NEW_ARGS_AUX(0, 0); NODE *args, *scope; if (nd_type($2) == NODE_MASGN) { /* if args.length == 1 && args[0].kind_of?(Array) * args = args[0] * end */ NODE *one = NEW_LIST(NEW_LIT(INT2FIX(1))); NODE *zero = NEW_LIST(NEW_LIT(INT2FIX(0))); m->nd_next = block_append( NEW_IF( NEW_NODE(NODE_AND, NEW_CALL(NEW_CALL(NEW_DVAR(id), idLength, 0), idEq, one), NEW_CALL(NEW_CALL(NEW_DVAR(id), idAREF, zero), rb_intern("kind_of?"), NEW_LIST(NEW_LIT(rb_cArray))), 0), NEW_DASGN_CURR(id, NEW_CALL(NEW_DVAR(id), idAREF, zero)), 0), node_assign($2, NEW_DVAR(id))); args = new_args(m, 0, id, 0, new_args_tail(0, 0, 0)); } else { if (nd_type($2) == NODE_LASGN || nd_type($2) == NODE_DASGN || nd_type($2) == NODE_DASGN_CURR) { $2->nd_value = NEW_DVAR(id); m->nd_plen = 1; m->nd_next = $2; args = new_args(m, 0, 0, 0, new_args_tail(0, 0, 0)); } else { m->nd_next = node_assign(NEW_MASGN(NEW_LIST($2), 0), NEW_DVAR(id)); args = new_args(m, 0, id, 0, new_args_tail(0, 0, 0)); } } scope = NEW_NODE(NODE_SCOPE, tbl, $8, args); tbl[0] = 1; tbl[1] = id; $$ = NEW_FOR(0, $5, scope); fixpos($$, $2); /*% $$ = dispatch3(for, $2, $5, $8); %*/ } | k_class cpath superclass { if (in_def || in_single) yyerror("class definition in method body"); local_push(0); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } bodystmt k_end { /*%%%*/ $$ = NEW_CLASS($2, $5, $3); nd_set_line($$, $4); /*% $$ = dispatch3(class, $2, $3, $5); %*/ local_pop(); } | k_class tLSHFT expr { $$ = in_def; in_def = 0; } term { $$ = in_single; in_single = 0; local_push(0); } bodystmt k_end { /*%%%*/ $$ = NEW_SCLASS($3, $7); fixpos($$, $3); /*% $$ = dispatch2(sclass, $3, $7); %*/ local_pop(); in_def = $4; in_single = $6; } | k_module cpath { if (in_def || in_single) yyerror("module definition in method body"); local_push(0); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } bodystmt k_end { /*%%%*/ $$ = NEW_MODULE($2, $4); nd_set_line($$, $3); /*% $$ = dispatch2(module, $2, $4); %*/ local_pop(); } | k_def fname { $$ = cur_mid; cur_mid = $2; in_def++; local_push(0); } f_arglist bodystmt k_end { /*%%%*/ NODE *body = remove_begin($5); reduce_nodes(&body); $$ = NEW_DEFN($2, $4, body, NOEX_PRIVATE); nd_set_line($$, $1); /*% $$ = dispatch3(def, $2, $4, $5); %*/ local_pop(); in_def--; cur_mid = $3; } | k_def singleton dot_or_colon {lex_state = EXPR_FNAME;} fname { in_single++; lex_state = EXPR_ENDFN; /* force for args */ local_push(0); } f_arglist bodystmt k_end { /*%%%*/ NODE *body = remove_begin($8); reduce_nodes(&body); $$ = NEW_DEFS($2, $5, $7, body); nd_set_line($$, $1); /*% $$ = dispatch5(defs, $2, $3, $5, $7, $8); %*/ local_pop(); in_single--; } | keyword_break { /*%%%*/ $$ = NEW_BREAK(0); /*% $$ = dispatch1(break, arg_new()); %*/ } | keyword_next { /*%%%*/ $$ = NEW_NEXT(0); /*% $$ = dispatch1(next, arg_new()); %*/ } | keyword_redo { /*%%%*/ $$ = NEW_REDO(); /*% $$ = dispatch0(redo); %*/ } | keyword_retry { /*%%%*/ $$ = NEW_RETRY(); /*% $$ = dispatch0(retry); %*/ } ; primary_value : primary { /*%%%*/ value_expr($1); $$ = $1; if (!$$) $$ = NEW_NIL(); /*% $$ = $1; %*/ } ; k_begin : keyword_begin { token_info_push("begin"); } ; k_if : keyword_if { token_info_push("if"); } ; k_unless : keyword_unless { token_info_push("unless"); } ; k_while : keyword_while { token_info_push("while"); } ; k_until : keyword_until { token_info_push("until"); } ; k_case : keyword_case { token_info_push("case"); } ; k_for : keyword_for { token_info_push("for"); } ; k_class : keyword_class { token_info_push("class"); } ; k_module : keyword_module { token_info_push("module"); } ; k_def : keyword_def { token_info_push("def"); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } ; k_end : keyword_end { token_info_pop("end"); } ; then : term /*%c%*/ /*%c { $$ = Qnil; } %*/ | keyword_then | term keyword_then /*%c%*/ /*%c { $$ = $2; } %*/ ; do : term /*%c%*/ /*%c { $$ = Qnil; } %*/ | keyword_do_cond ; if_tail : opt_else | keyword_elsif expr_value then compstmt if_tail { /*%%%*/ $$ = NEW_IF(cond($2), $4, $5); fixpos($$, $2); /*% $$ = dispatch3(elsif, $2, $4, escape_Qundef($5)); %*/ } ; opt_else : none | keyword_else compstmt { /*%%%*/ $$ = $2; /*% $$ = dispatch1(else, $2); %*/ } ; for_var : lhs | mlhs ; f_marg : f_norm_arg { $$ = assignable($1, 0); /*%%%*/ /*% $$ = dispatch1(mlhs_paren, $$); %*/ } | tLPAREN f_margs rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(mlhs_paren, $2); %*/ } ; f_marg_list : f_marg { /*%%%*/ $$ = NEW_LIST($1); /*% $$ = mlhs_add(mlhs_new(), $1); %*/ } | f_marg_list ',' f_marg { /*%%%*/ $$ = list_append($1, $3); /*% $$ = mlhs_add($1, $3); %*/ } ; f_margs : f_marg_list { /*%%%*/ $$ = NEW_MASGN($1, 0); /*% $$ = $1; %*/ } | f_marg_list ',' tSTAR f_norm_arg { $$ = assignable($4, 0); /*%%%*/ $$ = NEW_MASGN($1, $$); /*% $$ = mlhs_add_star($1, $$); %*/ } | f_marg_list ',' tSTAR f_norm_arg ',' f_marg_list { $$ = assignable($4, 0); /*%%%*/ $$ = NEW_MASGN($1, NEW_POSTARG($$, $6)); /*% $$ = mlhs_add_star($1, $$); %*/ } | f_marg_list ',' tSTAR { /*%%%*/ $$ = NEW_MASGN($1, -1); /*% $$ = mlhs_add_star($1, Qnil); %*/ } | f_marg_list ',' tSTAR ',' f_marg_list { /*%%%*/ $$ = NEW_MASGN($1, NEW_POSTARG(-1, $5)); /*% $$ = mlhs_add_star($1, $5); %*/ } | tSTAR f_norm_arg { $$ = assignable($2, 0); /*%%%*/ $$ = NEW_MASGN(0, $$); /*% $$ = mlhs_add_star(mlhs_new(), $$); %*/ } | tSTAR f_norm_arg ',' f_marg_list { $$ = assignable($2, 0); /*%%%*/ $$ = NEW_MASGN(0, NEW_POSTARG($$, $4)); /*% #if 0 TODO: Check me #endif $$ = mlhs_add_star($$, $4); %*/ } | tSTAR { /*%%%*/ $$ = NEW_MASGN(0, -1); /*% $$ = mlhs_add_star(mlhs_new(), Qnil); %*/ } | tSTAR ',' f_marg_list { /*%%%*/ $$ = NEW_MASGN(0, NEW_POSTARG(-1, $3)); /*% $$ = mlhs_add_star(mlhs_new(), Qnil); %*/ } ; block_args_tail : f_block_kwarg ',' f_kwrest opt_f_block_arg { $$ = new_args_tail($1, $3, $4); } | f_block_kwarg opt_f_block_arg { $$ = new_args_tail($1, Qnone, $2); } | f_kwrest opt_f_block_arg { $$ = new_args_tail(Qnone, $1, $2); } | f_block_arg { $$ = new_args_tail(Qnone, Qnone, $1); } ; opt_block_args_tail : ',' block_args_tail { $$ = $2; } | /* none */ { $$ = new_args_tail(Qnone, Qnone, Qnone); } ; block_param : f_arg ',' f_block_optarg ',' f_rest_arg opt_block_args_tail { $$ = new_args($1, $3, $5, Qnone, $6); } | f_arg ',' f_block_optarg ',' f_rest_arg ',' f_arg opt_block_args_tail { $$ = new_args($1, $3, $5, $7, $8); } | f_arg ',' f_block_optarg opt_block_args_tail { $$ = new_args($1, $3, Qnone, Qnone, $4); } | f_arg ',' f_block_optarg ',' f_arg opt_block_args_tail { $$ = new_args($1, $3, Qnone, $5, $6); } | f_arg ',' f_rest_arg opt_block_args_tail { $$ = new_args($1, Qnone, $3, Qnone, $4); } | f_arg ',' { $$ = new_args($1, Qnone, 1, Qnone, new_args_tail(Qnone, Qnone, Qnone)); /*%%%*/ /*% dispatch1(excessed_comma, $$); %*/ } | f_arg ',' f_rest_arg ',' f_arg opt_block_args_tail { $$ = new_args($1, Qnone, $3, $5, $6); } | f_arg opt_block_args_tail { $$ = new_args($1, Qnone, Qnone, Qnone, $2); } | f_block_optarg ',' f_rest_arg opt_block_args_tail { $$ = new_args(Qnone, $1, $3, Qnone, $4); } | f_block_optarg ',' f_rest_arg ',' f_arg opt_block_args_tail { $$ = new_args(Qnone, $1, $3, $5, $6); } | f_block_optarg opt_block_args_tail { $$ = new_args(Qnone, $1, Qnone, Qnone, $2); } | f_block_optarg ',' f_arg opt_block_args_tail { $$ = new_args(Qnone, $1, Qnone, $3, $4); } | f_rest_arg opt_block_args_tail { $$ = new_args(Qnone, Qnone, $1, Qnone, $2); } | f_rest_arg ',' f_arg opt_block_args_tail { $$ = new_args(Qnone, Qnone, $1, $3, $4); } | block_args_tail { $$ = new_args(Qnone, Qnone, Qnone, Qnone, $1); } ; opt_block_param : none | block_param_def { command_start = TRUE; } ; block_param_def : '|' opt_bv_decl '|' { /*%%%*/ $$ = 0; /*% $$ = blockvar_new(params_new(Qnil,Qnil,Qnil,Qnil,Qnil,Qnil,Qnil), escape_Qundef($2)); %*/ } | tOROP { /*%%%*/ $$ = 0; /*% $$ = blockvar_new(params_new(Qnil,Qnil,Qnil,Qnil,Qnil,Qnil,Qnil), Qnil); %*/ } | '|' block_param opt_bv_decl '|' { /*%%%*/ $$ = $2; /*% $$ = blockvar_new(escape_Qundef($2), escape_Qundef($3)); %*/ } ; opt_bv_decl : opt_nl { $$ = 0; } | opt_nl ';' bv_decls opt_nl { /*%%%*/ $$ = 0; /*% $$ = $3; %*/ } ; bv_decls : bvar /*%c%*/ /*%c { $$ = rb_ary_new3(1, $1); } %*/ | bv_decls ',' bvar /*%c%*/ /*%c { rb_ary_push($1, $3); } %*/ ; bvar : tIDENTIFIER { new_bv(get_id($1)); /*%%%*/ /*% $$ = get_value($1); %*/ } | f_bad_arg { $$ = 0; } ; lambda : { $$ = dyna_push(); } { $$ = lpar_beg; lpar_beg = ++paren_nest; } f_larglist lambda_body { lpar_beg = $2; /*%%%*/ $$ = NEW_LAMBDA($3, $4); /*% $$ = dispatch2(lambda, $3, $4); %*/ dyna_pop($1); } ; f_larglist : '(' f_args opt_bv_decl ')' { /*%%%*/ $$ = $2; /*% $$ = dispatch1(paren, $2); %*/ } | f_args { /*%%%*/ $$ = $1; /*% $$ = $1; %*/ } ; lambda_body : tLAMBEG compstmt '}' { $$ = $2; } | keyword_do_LAMBDA compstmt keyword_end { $$ = $2; } ; do_block : keyword_do_block { $1 = dyna_push(); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } opt_block_param compstmt keyword_end { /*%%%*/ $$ = NEW_ITER($3,$4); nd_set_line($$, $2); /*% $$ = dispatch2(do_block, escape_Qundef($3), $4); %*/ dyna_pop($1); } ; block_call : command do_block { /*%%%*/ if (nd_type($1) == NODE_YIELD) { compile_error(PARSER_ARG "block given to yield"); } else { block_dup_check($1->nd_args, $2); } $2->nd_iter = $1; $$ = $2; fixpos($$, $1); /*% $$ = method_add_block($1, $2); %*/ } | block_call dot_or_colon operation2 opt_paren_args { /*%%%*/ $$ = NEW_CALL($1, $3, $4); /*% $$ = dispatch3(call, $1, $2, $3); $$ = method_optarg($$, $4); %*/ } | block_call dot_or_colon operation2 opt_paren_args brace_block { /*%%%*/ block_dup_check($4, $5); $5->nd_iter = NEW_CALL($1, $3, $4); $$ = $5; fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, $2, $3, $4); $$ = method_add_block($$, $5); %*/ } | block_call dot_or_colon operation2 command_args do_block { /*%%%*/ block_dup_check($4, $5); $5->nd_iter = NEW_CALL($1, $3, $4); $$ = $5; fixpos($$, $1); /*% $$ = dispatch4(command_call, $1, $2, $3, $4); $$ = method_add_block($$, $5); %*/ } ; method_call : fcall paren_args { /*%%%*/ $$ = $1; $$->nd_args = $2; /*% $$ = method_arg(dispatch1(fcall, $1), $2); %*/ } | primary_value '.' operation2 { /*%%%*/ $$ = ruby_sourceline; /*% %*/ } opt_paren_args { /*%%%*/ $$ = NEW_CALL($1, $3, $5); nd_set_line($$, $4); /*% $$ = dispatch3(call, $1, ripper_id2sym('.'), $3); $$ = method_optarg($$, $5); %*/ } | primary_value tCOLON2 operation2 { /*%%%*/ $$ = ruby_sourceline; /*% %*/ } paren_args { /*%%%*/ $$ = NEW_CALL($1, $3, $5); nd_set_line($$, $4); /*% $$ = dispatch3(call, $1, ripper_id2sym('.'), $3); $$ = method_optarg($$, $5); %*/ } | primary_value tCOLON2 operation3 { /*%%%*/ $$ = NEW_CALL($1, $3, 0); /*% $$ = dispatch3(call, $1, ripper_intern("::"), $3); %*/ } | primary_value '.' { /*%%%*/ $$ = ruby_sourceline; /*% %*/ } paren_args { /*%%%*/ $$ = NEW_CALL($1, rb_intern("call"), $4); nd_set_line($$, $3); /*% $$ = dispatch3(call, $1, ripper_id2sym('.'), ripper_intern("call")); $$ = method_optarg($$, $4); %*/ } | primary_value tCOLON2 { /*%%%*/ $$ = ruby_sourceline; /*% %*/ } paren_args { /*%%%*/ $$ = NEW_CALL($1, rb_intern("call"), $4); nd_set_line($$, $3); /*% $$ = dispatch3(call, $1, ripper_intern("::"), ripper_intern("call")); $$ = method_optarg($$, $4); %*/ } | keyword_super paren_args { /*%%%*/ $$ = NEW_SUPER($2); /*% $$ = dispatch1(super, $2); %*/ } | keyword_super { /*%%%*/ $$ = NEW_ZSUPER(); /*% $$ = dispatch0(zsuper); %*/ } | primary_value '[' opt_call_args rbracket { /*%%%*/ if ($1 && nd_type($1) == NODE_SELF) $$ = NEW_FCALL(tAREF, $3); else $$ = NEW_CALL($1, tAREF, $3); fixpos($$, $1); /*% $$ = dispatch2(aref, $1, escape_Qundef($3)); %*/ } ; brace_block : '{' { $1 = dyna_push(); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } opt_block_param compstmt '}' { /*%%%*/ $$ = NEW_ITER($3,$4); nd_set_line($$, $2); /*% $$ = dispatch2(brace_block, escape_Qundef($3), $4); %*/ dyna_pop($1); } | keyword_do { $1 = dyna_push(); /*%%%*/ $$ = ruby_sourceline; /*% %*/ } opt_block_param compstmt keyword_end { /*%%%*/ $$ = NEW_ITER($3,$4); nd_set_line($$, $2); /*% $$ = dispatch2(do_block, escape_Qundef($3), $4); %*/ dyna_pop($1); } ; case_body : keyword_when args then compstmt cases { /*%%%*/ $$ = NEW_WHEN($2, $4, $5); /*% $$ = dispatch3(when, $2, $4, escape_Qundef($5)); %*/ } ; cases : opt_else | case_body ; opt_rescue : keyword_rescue exc_list exc_var then compstmt opt_rescue { /*%%%*/ if ($3) { $3 = node_assign($3, NEW_ERRINFO()); $5 = block_append($3, $5); } $$ = NEW_RESBODY($2, $5, $6); fixpos($$, $2?$2:$5); /*% $$ = dispatch4(rescue, escape_Qundef($2), escape_Qundef($3), escape_Qundef($5), escape_Qundef($6)); %*/ } | none ; exc_list : arg_value { /*%%%*/ $$ = NEW_LIST($1); /*% $$ = rb_ary_new3(1, $1); %*/ } | mrhs { /*%%%*/ if (!($$ = splat_array($1))) $$ = $1; /*% $$ = $1; %*/ } | none ; exc_var : tASSOC lhs { $$ = $2; } | none ; opt_ensure : keyword_ensure compstmt { /*%%%*/ $$ = $2; /*% $$ = dispatch1(ensure, $2); %*/ } | none ; literal : numeric | symbol { /*%%%*/ $$ = NEW_LIT(ID2SYM($1)); /*% $$ = dispatch1(symbol_literal, $1); %*/ } | dsym ; strings : string opt_string_sfx { /*%%%*/ NODE *node = $1; if (!node) { node = NEW_STR(STR_NEW0()); } else { node = evstr2dstr(node); } node = str_suffix(node, $2); $$ = node; /*% $$ = $1; %*/ } ; string : tCHAR | string1 | string string1 { /*%%%*/ $$ = literal_concat($1, $2); /*% $$ = dispatch2(string_concat, $1, $2); %*/ } ; string1 : tSTRING_BEG string_contents tSTRING_END { /*%%%*/ $$ = $2; /*% $$ = dispatch1(string_literal, $2); %*/ } ; opt_string_sfx : tSTRING_SUFFIX | /* none */ {$$ = 0;} ; xstring : tXSTRING_BEG xstring_contents tSTRING_END { /*%%%*/ NODE *node = $2; if (!node) { node = NEW_XSTR(STR_NEW0()); } else { switch (nd_type(node)) { case NODE_STR: nd_set_type(node, NODE_XSTR); break; case NODE_DSTR: nd_set_type(node, NODE_DXSTR); break; default: node = NEW_NODE(NODE_DXSTR, Qnil, 1, NEW_LIST(node)); break; } } $$ = node; /*% $$ = dispatch1(xstring_literal, $2); %*/ } ; regexp : tREGEXP_BEG regexp_contents tREGEXP_END { /*%%%*/ int options = $3; NODE *node = $2; NODE *list, *prev; if (!node) { node = NEW_LIT(reg_compile(STR_NEW0(), options)); } else switch (nd_type(node)) { case NODE_STR: { VALUE src = node->nd_lit; nd_set_type(node, NODE_LIT); node->nd_lit = reg_compile(src, options); } break; default: node = NEW_NODE(NODE_DSTR, STR_NEW0(), 1, NEW_LIST(node)); case NODE_DSTR: if (options & RE_OPTION_ONCE) { nd_set_type(node, NODE_DREGX_ONCE); } else { nd_set_type(node, NODE_DREGX); } node->nd_cflag = options & RE_OPTION_MASK; if (!NIL_P(node->nd_lit)) reg_fragment_check(node->nd_lit, options); for (list = (prev = node)->nd_next; list; list = list->nd_next) { if (nd_type(list->nd_head) == NODE_STR) { VALUE tail = list->nd_head->nd_lit; if (reg_fragment_check(tail, options) && prev && !NIL_P(prev->nd_lit)) { VALUE lit = prev == node ? prev->nd_lit : prev->nd_head->nd_lit; if (!literal_concat0(parser, lit, tail)) { node = 0; break; } rb_str_resize(tail, 0); prev->nd_next = list->nd_next; rb_gc_force_recycle((VALUE)list->nd_head); rb_gc_force_recycle((VALUE)list); list = prev; } else { prev = list; } } else { prev = 0; } } if (!node->nd_next) { VALUE src = node->nd_lit; nd_set_type(node, NODE_LIT); node->nd_lit = reg_compile(src, options); } break; } $$ = node; /*% $$ = dispatch2(regexp_literal, $2, $3); %*/ } ; words : tWORDS_BEG ' ' tSTRING_END { /*%%%*/ $$ = NEW_ZARRAY(); /*% $$ = dispatch0(words_new); $$ = dispatch1(array, $$); %*/ } | tWORDS_BEG word_list tSTRING_END { /*%%%*/ $$ = $2; /*% $$ = dispatch1(array, $2); %*/ } ; word_list : /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(words_new); %*/ } | word_list word ' ' { /*%%%*/ $$ = list_append($1, evstr2dstr($2)); /*% $$ = dispatch2(words_add, $1, $2); %*/ } ; word : string_content /*%c%*/ /*%c { $$ = dispatch0(word_new); $$ = dispatch2(word_add, $$, $1); } %*/ | word string_content { /*%%%*/ $$ = literal_concat($1, $2); /*% $$ = dispatch2(word_add, $1, $2); %*/ } ; symbols : tSYMBOLS_BEG ' ' tSTRING_END { /*%%%*/ $$ = NEW_ZARRAY(); /*% $$ = dispatch0(symbols_new); $$ = dispatch1(array, $$); %*/ } | tSYMBOLS_BEG symbol_list tSTRING_END { /*%%%*/ $$ = $2; /*% $$ = dispatch1(array, $2); %*/ } ; symbol_list : /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(symbols_new); %*/ } | symbol_list word ' ' { /*%%%*/ $2 = evstr2dstr($2); nd_set_type($2, NODE_DSYM); $$ = list_append($1, $2); /*% $$ = dispatch2(symbols_add, $1, $2); %*/ } ; qwords : tQWORDS_BEG ' ' tSTRING_END { /*%%%*/ $$ = NEW_ZARRAY(); /*% $$ = dispatch0(qwords_new); $$ = dispatch1(array, $$); %*/ } | tQWORDS_BEG qword_list tSTRING_END { /*%%%*/ $$ = $2; /*% $$ = dispatch1(array, $2); %*/ } ; qsymbols : tQSYMBOLS_BEG ' ' tSTRING_END { /*%%%*/ $$ = NEW_ZARRAY(); /*% $$ = dispatch0(qsymbols_new); $$ = dispatch1(array, $$); %*/ } | tQSYMBOLS_BEG qsym_list tSTRING_END { /*%%%*/ $$ = $2; /*% $$ = dispatch1(array, $2); %*/ } ; qword_list : /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(qwords_new); %*/ } | qword_list tSTRING_CONTENT ' ' { /*%%%*/ $$ = list_append($1, $2); /*% $$ = dispatch2(qwords_add, $1, $2); %*/ } ; qsym_list : /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(qsymbols_new); %*/ } | qsym_list tSTRING_CONTENT ' ' { /*%%%*/ VALUE lit; lit = $2->nd_lit; $2->nd_lit = ID2SYM(rb_intern_str(lit)); nd_set_type($2, NODE_LIT); $$ = list_append($1, $2); /*% $$ = dispatch2(qsymbols_add, $1, $2); %*/ } ; string_contents : /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(string_content); %*/ } | string_contents string_content { /*%%%*/ $$ = literal_concat($1, $2); /*% $$ = dispatch2(string_add, $1, $2); %*/ } ; xstring_contents: /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(xstring_new); %*/ } | xstring_contents string_content { /*%%%*/ $$ = literal_concat($1, $2); /*% $$ = dispatch2(xstring_add, $1, $2); %*/ } ; regexp_contents: /* none */ { /*%%%*/ $$ = 0; /*% $$ = dispatch0(regexp_new); %*/ } | regexp_contents string_content { /*%%%*/ NODE *head = $1, *tail = $2; if (!head) { $$ = tail; } else if (!tail) { $$ = head; } else { switch (nd_type(head)) { case NODE_STR: nd_set_type(head, NODE_DSTR); break; case NODE_DSTR: break; default: head = list_append(NEW_DSTR(Qnil), head); break; } $$ = list_append(head, tail); } /*% $$ = dispatch2(regexp_add, $1, $2); %*/ } ; string_content : tSTRING_CONTENT | tSTRING_DVAR { $$ = lex_strterm; lex_strterm = 0; lex_state = EXPR_BEG; } string_dvar { /*%%%*/ lex_strterm = $2; $$ = NEW_EVSTR($3); /*% lex_strterm = $2; $$ = dispatch1(string_dvar, $3); %*/ } | tSTRING_DBEG { $1 = cond_stack; $$ = cmdarg_stack; cond_stack = 0; cmdarg_stack = 0; } { $$ = lex_strterm; lex_strterm = 0; lex_state = EXPR_BEG; } { $$ = brace_nest; brace_nest = 0; } compstmt tSTRING_DEND { cond_stack = $1; cmdarg_stack = $2; lex_strterm = $3; brace_nest = $4; /*%%%*/ if ($5) $5->flags &= ~NODE_FL_NEWLINE; $$ = new_evstr($5); /*% $$ = dispatch1(string_embexpr, $5); %*/ } ; string_dvar : tGVAR { /*%%%*/ $$ = NEW_GVAR($1); /*% $$ = dispatch1(var_ref, $1); %*/ } | tIVAR { /*%%%*/ $$ = NEW_IVAR($1); /*% $$ = dispatch1(var_ref, $1); %*/ } | tCVAR { /*%%%*/ $$ = NEW_CVAR($1); /*% $$ = dispatch1(var_ref, $1); %*/ } | backref ; symbol : tSYMBEG sym { lex_state = EXPR_END; /*%%%*/ $$ = $2; /*% $$ = dispatch1(symbol, $2); %*/ } ; sym : fname | tIVAR | tGVAR | tCVAR ; dsym : tSYMBEG xstring_contents tSTRING_END { lex_state = EXPR_END; /*%%%*/ $$ = dsym_node($2); /*% $$ = dispatch1(dyna_symbol, $2); %*/ } ; numeric : simple_numeric | tUMINUS_NUM simple_numeric %prec tLOWEST { /*%%%*/ $$ = negate_lit($2); /*% $$ = dispatch2(unary, ripper_intern("-@"), $2); %*/ } ; simple_numeric : tINTEGER | tFLOAT | tRATIONAL | tIMAGINARY ; user_variable : tIDENTIFIER | tIVAR | tGVAR | tCONSTANT | tCVAR ; keyword_variable: keyword_nil {ifndef_ripper($$ = keyword_nil);} | keyword_self {ifndef_ripper($$ = keyword_self);} | keyword_true {ifndef_ripper($$ = keyword_true);} | keyword_false {ifndef_ripper($$ = keyword_false);} | keyword__FILE__ {ifndef_ripper($$ = keyword__FILE__);} | keyword__LINE__ {ifndef_ripper($$ = keyword__LINE__);} | keyword__ENCODING__ {ifndef_ripper($$ = keyword__ENCODING__);} ; var_ref : user_variable { /*%%%*/ if (!($$ = gettable($1))) $$ = NEW_BEGIN(0); /*% if (id_is_var(get_id($1))) { $$ = dispatch1(var_ref, $1); } else { $$ = dispatch1(vcall, $1); } %*/ } | keyword_variable { /*%%%*/ if (!($$ = gettable($1))) $$ = NEW_BEGIN(0); /*% $$ = dispatch1(var_ref, $1); %*/ } ; var_lhs : user_variable { $$ = assignable($1, 0); /*%%%*/ /*% $$ = dispatch1(var_field, $$); %*/ } | keyword_variable { $$ = assignable($1, 0); /*%%%*/ /*% $$ = dispatch1(var_field, $$); %*/ } ; backref : tNTH_REF | tBACK_REF ; superclass : term { /*%%%*/ $$ = 0; /*% $$ = Qnil; %*/ } | '<' { lex_state = EXPR_BEG; command_start = TRUE; } expr_value term { $$ = $3; } | error term { /*%%%*/ yyerrok; $$ = 0; /*% yyerrok; $$ = Qnil; %*/ } ; f_arglist : '(' f_args rparen { /*%%%*/ $$ = $2; /*% $$ = dispatch1(paren, $2); %*/ lex_state = EXPR_BEG; command_start = TRUE; } | f_args term { $$ = $1; lex_state = EXPR_BEG; command_start = TRUE; } ; args_tail : f_kwarg ',' f_kwrest opt_f_block_arg { $$ = new_args_tail($1, $3, $4); } | f_kwarg opt_f_block_arg { $$ = new_args_tail($1, Qnone, $2); } | f_kwrest opt_f_block_arg { $$ = new_args_tail(Qnone, $1, $2); } | f_block_arg { $$ = new_args_tail(Qnone, Qnone, $1); } ; opt_args_tail : ',' args_tail { $$ = $2; } | /* none */ { $$ = new_args_tail(Qnone, Qnone, Qnone); } ; f_args : f_arg ',' f_optarg ',' f_rest_arg opt_args_tail { $$ = new_args($1, $3, $5, Qnone, $6); } | f_arg ',' f_optarg ',' f_rest_arg ',' f_arg opt_args_tail { $$ = new_args($1, $3, $5, $7, $8); } | f_arg ',' f_optarg opt_args_tail { $$ = new_args($1, $3, Qnone, Qnone, $4); } | f_arg ',' f_optarg ',' f_arg opt_args_tail { $$ = new_args($1, $3, Qnone, $5, $6); } | f_arg ',' f_rest_arg opt_args_tail { $$ = new_args($1, Qnone, $3, Qnone, $4); } | f_arg ',' f_rest_arg ',' f_arg opt_args_tail { $$ = new_args($1, Qnone, $3, $5, $6); } | f_arg opt_args_tail { $$ = new_args($1, Qnone, Qnone, Qnone, $2); } | f_optarg ',' f_rest_arg opt_args_tail { $$ = new_args(Qnone, $1, $3, Qnone, $4); } | f_optarg ',' f_rest_arg ',' f_arg opt_args_tail { $$ = new_args(Qnone, $1, $3, $5, $6); } | f_optarg opt_args_tail { $$ = new_args(Qnone, $1, Qnone, Qnone, $2); } | f_optarg ',' f_arg opt_args_tail { $$ = new_args(Qnone, $1, Qnone, $3, $4); } | f_rest_arg opt_args_tail { $$ = new_args(Qnone, Qnone, $1, Qnone, $2); } | f_rest_arg ',' f_arg opt_args_tail { $$ = new_args(Qnone, Qnone, $1, $3, $4); } | args_tail { $$ = new_args(Qnone, Qnone, Qnone, Qnone, $1); } | /* none */ { $$ = new_args_tail(Qnone, Qnone, Qnone); $$ = new_args(Qnone, Qnone, Qnone, Qnone, $$); } ; f_bad_arg : tCONSTANT { /*%%%*/ yyerror("formal argument cannot be a constant"); $$ = 0; /*% $$ = dispatch1(param_error, $1); %*/ } | tIVAR { /*%%%*/ yyerror("formal argument cannot be an instance variable"); $$ = 0; /*% $$ = dispatch1(param_error, $1); %*/ } | tGVAR { /*%%%*/ yyerror("formal argument cannot be a global variable"); $$ = 0; /*% $$ = dispatch1(param_error, $1); %*/ } | tCVAR { /*%%%*/ yyerror("formal argument cannot be a class variable"); $$ = 0; /*% $$ = dispatch1(param_error, $1); %*/ } ; f_norm_arg : f_bad_arg | tIDENTIFIER { formal_argument(get_id($1)); $$ = $1; } ; f_arg_item : f_norm_arg { arg_var(get_id($1)); /*%%%*/ $$ = NEW_ARGS_AUX($1, 1); /*% $$ = get_value($1); %*/ } | tLPAREN f_margs rparen { ID tid = internal_id(); arg_var(tid); /*%%%*/ if (dyna_in_block()) { $2->nd_value = NEW_DVAR(tid); } else { $2->nd_value = NEW_LVAR(tid); } $$ = NEW_ARGS_AUX(tid, 1); $$->nd_next = $2; /*% $$ = dispatch1(mlhs_paren, $2); %*/ } ; f_arg : f_arg_item /*%c%*/ /*%c { $$ = rb_ary_new3(1, $1); } c%*/ | f_arg ',' f_arg_item { /*%%%*/ $$ = $1; $$->nd_plen++; $$->nd_next = block_append($$->nd_next, $3->nd_next); rb_gc_force_recycle((VALUE)$3); /*% $$ = rb_ary_push($1, $3); %*/ } ; f_label : tLABEL { arg_var(formal_argument(get_id($1))); $$ = $1; } ; f_kw : f_label arg_value { $$ = assignable($1, $2); /*%%%*/ $$ = NEW_KW_ARG(0, $$); /*% $$ = rb_assoc_new($$, $2); %*/ } | f_label { $$ = assignable($1, (NODE *)-1); /*%%%*/ $$ = NEW_KW_ARG(0, $$); /*% $$ = rb_assoc_new($$, 0); %*/ } ; f_block_kw : f_label primary_value { $$ = assignable($1, $2); /*%%%*/ $$ = NEW_KW_ARG(0, $$); /*% $$ = rb_assoc_new($$, $2); %*/ } | f_label { $$ = assignable($1, (NODE *)-1); /*%%%*/ $$ = NEW_KW_ARG(0, $$); /*% $$ = rb_assoc_new($$, 0); %*/ } ; f_block_kwarg : f_block_kw { /*%%%*/ $$ = $1; /*% $$ = rb_ary_new3(1, $1); %*/ } | f_block_kwarg ',' f_block_kw { /*%%%*/ NODE *kws = $1; while (kws->nd_next) { kws = kws->nd_next; } kws->nd_next = $3; $$ = $1; /*% $$ = rb_ary_push($1, $3); %*/ } ; f_kwarg : f_kw { /*%%%*/ $$ = $1; /*% $$ = rb_ary_new3(1, $1); %*/ } | f_kwarg ',' f_kw { /*%%%*/ NODE *kws = $1; while (kws->nd_next) { kws = kws->nd_next; } kws->nd_next = $3; $$ = $1; /*% $$ = rb_ary_push($1, $3); %*/ } ; kwrest_mark : tPOW | tDSTAR ; f_kwrest : kwrest_mark tIDENTIFIER { shadowing_lvar(get_id($2)); $$ = $2; } | kwrest_mark { $$ = internal_id(); } ; f_opt : f_norm_arg '=' arg_value { arg_var(get_id($1)); $$ = assignable($1, $3); /*%%%*/ $$ = NEW_OPT_ARG(0, $$); /*% $$ = rb_assoc_new($$, $3); %*/ } ; f_block_opt : f_norm_arg '=' primary_value { arg_var(get_id($1)); $$ = assignable($1, $3); /*%%%*/ $$ = NEW_OPT_ARG(0, $$); /*% $$ = rb_assoc_new($$, $3); %*/ } ; f_block_optarg : f_block_opt { /*%%%*/ $$ = $1; /*% $$ = rb_ary_new3(1, $1); %*/ } | f_block_optarg ',' f_block_opt { /*%%%*/ NODE *opts = $1; while (opts->nd_next) { opts = opts->nd_next; } opts->nd_next = $3; $$ = $1; /*% $$ = rb_ary_push($1, $3); %*/ } ; f_optarg : f_opt { /*%%%*/ $$ = $1; /*% $$ = rb_ary_new3(1, $1); %*/ } | f_optarg ',' f_opt { /*%%%*/ NODE *opts = $1; while (opts->nd_next) { opts = opts->nd_next; } opts->nd_next = $3; $$ = $1; /*% $$ = rb_ary_push($1, $3); %*/ } ; restarg_mark : '*' | tSTAR ; f_rest_arg : restarg_mark tIDENTIFIER { /*%%%*/ if (!is_local_id($2)) yyerror("rest argument must be local variable"); /*% %*/ arg_var(shadowing_lvar(get_id($2))); /*%%%*/ $$ = $2; /*% $$ = dispatch1(rest_param, $2); %*/ } | restarg_mark { /*%%%*/ $$ = internal_id(); arg_var($$); /*% $$ = dispatch1(rest_param, Qnil); %*/ } ; blkarg_mark : '&' | tAMPER ; f_block_arg : blkarg_mark tIDENTIFIER { /*%%%*/ if (!is_local_id($2)) yyerror("block argument must be local variable"); else if (!dyna_in_block() && local_id($2)) yyerror("duplicated block argument name"); /*% %*/ arg_var(shadowing_lvar(get_id($2))); /*%%%*/ $$ = $2; /*% $$ = dispatch1(blockarg, $2); %*/ } ; opt_f_block_arg : ',' f_block_arg { $$ = $2; } | none { /*%%%*/ $$ = 0; /*% $$ = Qundef; %*/ } ; singleton : var_ref { /*%%%*/ value_expr($1); $$ = $1; if (!$$) $$ = NEW_NIL(); /*% $$ = $1; %*/ } | '(' {lex_state = EXPR_BEG;} expr rparen { /*%%%*/ if ($3 == 0) { yyerror("can't define singleton method for ()."); } else { switch (nd_type($3)) { case NODE_STR: case NODE_DSTR: case NODE_XSTR: case NODE_DXSTR: case NODE_DREGX: case NODE_LIT: case NODE_ARRAY: case NODE_ZARRAY: yyerror("can't define singleton method for literals"); default: value_expr($3); break; } } $$ = $3; /*% $$ = dispatch1(paren, $3); %*/ } ; assoc_list : none | assocs trailer { /*%%%*/ $$ = $1; /*% $$ = dispatch1(assoclist_from_args, $1); %*/ } ; assocs : assoc /*%c%*/ /*%c { $$ = rb_ary_new3(1, $1); } %*/ | assocs ',' assoc { /*%%%*/ $$ = list_concat($1, $3); /*% $$ = rb_ary_push($1, $3); %*/ } ; assoc : arg_value tASSOC arg_value { /*%%%*/ $$ = list_append(NEW_LIST($1), $3); /*% $$ = dispatch2(assoc_new, $1, $3); %*/ } | tLABEL arg_value { /*%%%*/ $$ = list_append(NEW_LIST(NEW_LIT(ID2SYM($1))), $2); /*% $$ = dispatch2(assoc_new, $1, $2); %*/ } | tDSTAR arg_value { /*%%%*/ $$ = list_append(NEW_LIST(0), $2); /*% $$ = dispatch1(assoc_splat, $2); %*/ } ; ; operation : tIDENTIFIER | tCONSTANT | tFID ; operation2 : tIDENTIFIER | tCONSTANT | tFID | op ; operation3 : tIDENTIFIER | tFID | op ; dot_or_colon : '.' /*%c%*/ /*%c { $$ = $1; } %*/ | tCOLON2 /*%c%*/ /*%c { $$ = $1; } %*/ ; opt_terms : /* none */ | terms ; opt_nl : /* none */ | '\n' ; rparen : opt_nl ')' ; rbracket : opt_nl ']' ; trailer : /* none */ | '\n' | ',' ; term : ';' {yyerrok;} | '\n' ; terms : term | terms ';' {yyerrok;} ; none : /* none */ { /*%%%*/ $$ = 0; /*% $$ = Qundef; %*/ } ; %% # undef parser # undef yylex # undef yylval # define yylval (*((YYSTYPE*)(parser->parser_yylval))) static int parser_regx_options(struct parser_params*); static int parser_str_options(struct parser_params*); static int parser_tokadd_string(struct parser_params*,int,int,int,long*,rb_encoding**); static void parser_tokaddmbc(struct parser_params *parser, int c, rb_encoding *enc); static int parser_parse_string(struct parser_params*,NODE*); static int parser_here_document(struct parser_params*,NODE*); # define nextc() parser_nextc(parser) # define pushback(c) parser_pushback(parser, (c)) # define newtok() parser_newtok(parser) # define tokspace(n) parser_tokspace(parser, (n)) # define tokadd(c) parser_tokadd(parser, (c)) # define tok_hex(numlen) parser_tok_hex(parser, (numlen)) # define read_escape(flags,e) parser_read_escape(parser, (flags), (e)) # define tokadd_escape(e) parser_tokadd_escape(parser, (e)) # define regx_options() parser_regx_options(parser) # define str_options() parser_str_options(parser) # define tokadd_string(f,t,p,n,e) parser_tokadd_string(parser,(f),(t),(p),(n),(e)) # define parse_string(n) parser_parse_string(parser,(n)) # define tokaddmbc(c, enc) parser_tokaddmbc(parser, (c), (enc)) # define here_document(n) parser_here_document(parser,(n)) # define heredoc_identifier() parser_heredoc_identifier(parser) # define heredoc_restore(n) parser_heredoc_restore(parser,(n)) # define whole_match_p(e,l,i) parser_whole_match_p(parser,(e),(l),(i)) # define number_literal_suffix(f) parser_number_literal_suffix(parser, (f)) # define set_number_literal(v, t, f) parser_set_number_literal(parser, (v), (t), (f)) # define set_integer_literal(v, f) parser_set_integer_literal(parser, (v), (f)) #ifndef RIPPER # define set_yylval_str(x) (yylval.node = NEW_STR(x)) # define set_yylval_num(x) (yylval.num = (x)) # define set_yylval_id(x) (yylval.id = (x)) # define set_yylval_name(x) (yylval.id = (x)) # define set_yylval_literal(x) (yylval.node = NEW_LIT(x)) # define set_yylval_node(x) (yylval.node = (x)) # define yylval_id() (yylval.id) #else static inline VALUE ripper_yylval_id(ID x) { return (VALUE)NEW_LASGN(x, ID2SYM(x)); } # define set_yylval_str(x) (void)(x) # define set_yylval_num(x) (void)(x) # define set_yylval_id(x) (void)(x) # define set_yylval_name(x) (void)(yylval.val = ripper_yylval_id(x)) # define set_yylval_literal(x) (void)(x) # define set_yylval_node(x) (void)(x) # define yylval_id() yylval.id #endif #ifndef RIPPER #define ripper_flush(p) (void)(p) #else #define ripper_flush(p) ((p)->tokp = (p)->parser_lex_p) #define yylval_rval (*(RB_TYPE_P(yylval.val, T_NODE) ? &yylval.node->nd_rval : &yylval.val)) static int ripper_has_scan_event(struct parser_params *parser) { if (lex_p < parser->tokp) rb_raise(rb_eRuntimeError, "lex_p < tokp"); return lex_p > parser->tokp; } static VALUE ripper_scan_event_val(struct parser_params *parser, int t) { VALUE str = STR_NEW(parser->tokp, lex_p - parser->tokp); VALUE rval = ripper_dispatch1(parser, ripper_token2eventid(t), str); ripper_flush(parser); return rval; } static void ripper_dispatch_scan_event(struct parser_params *parser, int t) { if (!ripper_has_scan_event(parser)) return; yylval_rval = ripper_scan_event_val(parser, t); } static void ripper_dispatch_ignored_scan_event(struct parser_params *parser, int t) { if (!ripper_has_scan_event(parser)) return; (void)ripper_scan_event_val(parser, t); } static void ripper_dispatch_delayed_token(struct parser_params *parser, int t) { int saved_line = ruby_sourceline; const char *saved_tokp = parser->tokp; ruby_sourceline = parser->delayed_line; parser->tokp = lex_pbeg + parser->delayed_col; yylval_rval = ripper_dispatch1(parser, ripper_token2eventid(t), parser->delayed); parser->delayed = Qnil; ruby_sourceline = saved_line; parser->tokp = saved_tokp; } #endif /* RIPPER */ #include "ruby/regex.h" #include "ruby/util.h" /* We remove any previous definition of `SIGN_EXTEND_CHAR', since ours (we hope) works properly with all combinations of machines, compilers, `char' and `unsigned char' argument types. (Per Bothner suggested the basic approach.) */ #undef SIGN_EXTEND_CHAR #if __STDC__ # define SIGN_EXTEND_CHAR(c) ((signed char)(c)) #else /* not __STDC__ */ /* As in Harbison and Steele. */ # define SIGN_EXTEND_CHAR(c) ((((unsigned char)(c)) ^ 128) - 128) #endif #define parser_encoding_name() (current_enc->name) #define parser_mbclen() mbclen((lex_p-1),lex_pend,current_enc) #define parser_precise_mbclen() rb_enc_precise_mbclen((lex_p-1),lex_pend,current_enc) #define is_identchar(p,e,enc) (rb_enc_isalnum((unsigned char)(*(p)),(enc)) || (*(p)) == '_' || !ISASCII(*(p))) #define parser_is_identchar() (!parser->eofp && is_identchar((lex_p-1),lex_pend,current_enc)) #define parser_isascii() ISASCII(*(lex_p-1)) #ifndef RIPPER static int token_info_get_column(struct parser_params *parser, const char *token) { int column = 1; const char *p, *pend = lex_p - strlen(token); for (p = lex_pbeg; p < pend; p++) { if (*p == '\t') { column = (((column - 1) / 8) + 1) * 8; } column++; } return column; } static int token_info_has_nonspaces(struct parser_params *parser, const char *token) { const char *p, *pend = lex_p - strlen(token); for (p = lex_pbeg; p < pend; p++) { if (*p != ' ' && *p != '\t') { return 1; } } return 0; } #undef token_info_push static void token_info_push(struct parser_params *parser, const char *token) { token_info *ptinfo; if (!parser->parser_token_info_enabled) return; ptinfo = ALLOC(token_info); ptinfo->token = token; ptinfo->linenum = ruby_sourceline; ptinfo->column = token_info_get_column(parser, token); ptinfo->nonspc = token_info_has_nonspaces(parser, token); ptinfo->next = parser->parser_token_info; parser->parser_token_info = ptinfo; } #undef token_info_pop static void token_info_pop(struct parser_params *parser, const char *token) { int linenum; token_info *ptinfo = parser->parser_token_info; if (!ptinfo) return; parser->parser_token_info = ptinfo->next; if (token_info_get_column(parser, token) == ptinfo->column) { /* OK */ goto finish; } linenum = ruby_sourceline; if (linenum == ptinfo->linenum) { /* SKIP */ goto finish; } if (token_info_has_nonspaces(parser, token) || ptinfo->nonspc) { /* SKIP */ goto finish; } if (parser->parser_token_info_enabled) { rb_compile_warn(ruby_sourcefile, linenum, "mismatched indentations at '%s' with '%s' at %d", token, ptinfo->token, ptinfo->linenum); } finish: xfree(ptinfo); } #endif /* RIPPER */ static int parser_yyerror(struct parser_params *parser, const char *msg) { #ifndef RIPPER const int max_line_margin = 30; const char *p, *pe; char *buf; long len; int i; compile_error(PARSER_ARG "%s", msg); p = lex_p; while (lex_pbeg <= p) { if (*p == '\n') break; p--; } p++; pe = lex_p; while (pe < lex_pend) { if (*pe == '\n') break; pe++; } len = pe - p; if (len > 4) { char *p2; const char *pre = "", *post = ""; if (len > max_line_margin * 2 + 10) { if (lex_p - p > max_line_margin) { p = rb_enc_prev_char(p, lex_p - max_line_margin, pe, rb_enc_get(lex_lastline)); pre = "..."; } if (pe - lex_p > max_line_margin) { pe = rb_enc_prev_char(lex_p, lex_p + max_line_margin, pe, rb_enc_get(lex_lastline)); post = "..."; } len = pe - p; } buf = ALLOCA_N(char, len+2); MEMCPY(buf, p, char, len); buf[len] = '\0'; rb_compile_error_append("%s%s%s", pre, buf, post); i = (int)(lex_p - p); p2 = buf; pe = buf + len; while (p2 < pe) { if (*p2 != '\t') *p2 = ' '; p2++; } buf[i] = '^'; buf[i+1] = '\0'; rb_compile_error_append("%s%s", pre, buf); } #else dispatch1(parse_error, STR_NEW2(msg)); #endif /* !RIPPER */ return 0; } static void parser_prepare(struct parser_params *parser); #ifndef RIPPER static VALUE debug_lines(VALUE fname) { ID script_lines; CONST_ID(script_lines, "SCRIPT_LINES__"); if (rb_const_defined_at(rb_cObject, script_lines)) { VALUE hash = rb_const_get_at(rb_cObject, script_lines); if (RB_TYPE_P(hash, T_HASH)) { VALUE lines = rb_ary_new(); rb_hash_aset(hash, fname, lines); return lines; } } return 0; } static VALUE coverage(VALUE fname, int n) { VALUE coverages = rb_get_coverages(); if (RTEST(coverages) && RBASIC(coverages)->klass == 0) { VALUE lines = rb_ary_new2(n); int i; RBASIC_CLEAR_CLASS(lines); for (i = 0; i < n; i++) RARRAY_ASET(lines, i, Qnil); RARRAY(lines)->as.heap.len = n; rb_hash_aset(coverages, fname, lines); return lines; } return 0; } static int e_option_supplied(struct parser_params *parser) { return strcmp(ruby_sourcefile, "-e") == 0; } static VALUE yycompile0(VALUE arg) { int n; NODE *tree; struct parser_params *parser = (struct parser_params *)arg; if (!compile_for_eval && rb_safe_level() == 0) { ruby_debug_lines = debug_lines(ruby_sourcefile_string); if (ruby_debug_lines && ruby_sourceline > 0) { VALUE str = STR_NEW0(); n = ruby_sourceline; do { rb_ary_push(ruby_debug_lines, str); } while (--n); } if (!e_option_supplied(parser)) { ruby_coverage = coverage(ruby_sourcefile_string, ruby_sourceline); } } parser->last_cr_line = ruby_sourceline - 1; parser_prepare(parser); deferred_nodes = 0; #ifndef RIPPER parser->parser_token_info_enabled = !compile_for_eval && RTEST(ruby_verbose); #endif #ifndef RIPPER if (RUBY_DTRACE_PARSE_BEGIN_ENABLED()) { RUBY_DTRACE_PARSE_BEGIN(parser->parser_ruby_sourcefile, parser->parser_ruby_sourceline); } #endif n = yyparse((void*)parser); #ifndef RIPPER if (RUBY_DTRACE_PARSE_END_ENABLED()) { RUBY_DTRACE_PARSE_END(parser->parser_ruby_sourcefile, parser->parser_ruby_sourceline); } #endif ruby_debug_lines = 0; ruby_coverage = 0; compile_for_eval = 0; lex_strterm = 0; lex_p = lex_pbeg = lex_pend = 0; lex_lastline = lex_nextline = 0; if (parser->nerr) { return 0; } tree = ruby_eval_tree; if (!tree) { tree = NEW_NIL(); } else if (ruby_eval_tree_begin) { tree->nd_body = NEW_PRELUDE(ruby_eval_tree_begin, tree->nd_body); } return (VALUE)tree; } static NODE* yycompile(struct parser_params *parser, VALUE fname, int line) { ruby_sourcefile_string = rb_str_new_frozen(fname); ruby_sourcefile = RSTRING_PTR(fname); ruby_sourceline = line - 1; return (NODE *)rb_suppress_tracing(yycompile0, (VALUE)parser); } #endif /* !RIPPER */ static rb_encoding * must_be_ascii_compatible(VALUE s) { rb_encoding *enc = rb_enc_get(s); if (!rb_enc_asciicompat(enc)) { rb_raise(rb_eArgError, "invalid source encoding"); } return enc; } static VALUE lex_get_str(struct parser_params *parser, VALUE s) { char *beg, *end, *pend; rb_encoding *enc = must_be_ascii_compatible(s); beg = RSTRING_PTR(s); if (lex_gets_ptr) { if (RSTRING_LEN(s) == lex_gets_ptr) return Qnil; beg += lex_gets_ptr; } pend = RSTRING_PTR(s) + RSTRING_LEN(s); end = beg; while (end < pend) { if (*end++ == '\n') break; } lex_gets_ptr = end - RSTRING_PTR(s); return rb_enc_str_new(beg, end - beg, enc); } static VALUE lex_getline(struct parser_params *parser) { VALUE line = (*parser->parser_lex_gets)(parser, parser->parser_lex_input); if (NIL_P(line)) return line; must_be_ascii_compatible(line); #ifndef RIPPER if (ruby_debug_lines) { rb_enc_associate(line, current_enc); rb_ary_push(ruby_debug_lines, line); } if (ruby_coverage) { rb_ary_push(ruby_coverage, Qnil); } #endif return line; } #ifdef RIPPER static rb_data_type_t parser_data_type; #else static const rb_data_type_t parser_data_type; static NODE* parser_compile_string(volatile VALUE vparser, VALUE fname, VALUE s, int line) { struct parser_params *parser; NODE *node; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); lex_gets = lex_get_str; lex_gets_ptr = 0; lex_input = s; lex_pbeg = lex_p = lex_pend = 0; compile_for_eval = rb_parse_in_eval(); node = yycompile(parser, fname, line); RB_GC_GUARD(vparser); /* prohibit tail call optimization */ return node; } NODE* rb_compile_string(const char *f, VALUE s, int line) { must_be_ascii_compatible(s); return parser_compile_string(rb_parser_new(), rb_filesystem_str_new_cstr(f), s, line); } NODE* rb_parser_compile_string(volatile VALUE vparser, const char *f, VALUE s, int line) { return rb_parser_compile_string_path(vparser, rb_filesystem_str_new_cstr(f), s, line); } NODE* rb_parser_compile_string_path(volatile VALUE vparser, VALUE f, VALUE s, int line) { must_be_ascii_compatible(s); return parser_compile_string(vparser, f, s, line); } NODE* rb_compile_cstr(const char *f, const char *s, int len, int line) { VALUE str = rb_str_new(s, len); return parser_compile_string(rb_parser_new(), rb_filesystem_str_new_cstr(f), str, line); } NODE* rb_parser_compile_cstr(volatile VALUE vparser, const char *f, const char *s, int len, int line) { VALUE str = rb_str_new(s, len); return parser_compile_string(vparser, rb_filesystem_str_new_cstr(f), str, line); } static VALUE lex_io_gets(struct parser_params *parser, VALUE io) { return rb_io_gets(io); } NODE* rb_compile_file(const char *f, VALUE file, int start) { VALUE volatile vparser = rb_parser_new(); return rb_parser_compile_file(vparser, f, file, start); } NODE* rb_parser_compile_file(volatile VALUE vparser, const char *f, VALUE file, int start) { return rb_parser_compile_file_path(vparser, rb_filesystem_str_new_cstr(f), file, start); } NODE* rb_parser_compile_file_path(volatile VALUE vparser, VALUE fname, VALUE file, int start) { struct parser_params *parser; NODE *node; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); lex_gets = lex_io_gets; lex_input = file; lex_pbeg = lex_p = lex_pend = 0; compile_for_eval = rb_parse_in_eval(); node = yycompile(parser, fname, start); RB_GC_GUARD(vparser); /* prohibit tail call optimization */ return node; } #endif /* !RIPPER */ #define STR_FUNC_ESCAPE 0x01 #define STR_FUNC_EXPAND 0x02 #define STR_FUNC_REGEXP 0x04 #define STR_FUNC_QWORDS 0x08 #define STR_FUNC_SYMBOL 0x10 #define STR_FUNC_INDENT 0x20 #define STR_FUNC_OPTION 0x40 enum string_type { str_squote = (STR_FUNC_OPTION), str_dquote = (STR_FUNC_EXPAND|STR_FUNC_OPTION), str_xquote = (STR_FUNC_EXPAND), str_regexp = (STR_FUNC_REGEXP|STR_FUNC_ESCAPE|STR_FUNC_EXPAND), str_sword = (STR_FUNC_QWORDS), str_dword = (STR_FUNC_QWORDS|STR_FUNC_EXPAND), str_ssym = (STR_FUNC_SYMBOL), str_dsym = (STR_FUNC_SYMBOL|STR_FUNC_EXPAND) }; static VALUE parser_str_new(const char *p, long n, rb_encoding *enc, int func, rb_encoding *enc0) { VALUE str; str = rb_enc_str_new(p, n, enc); if (!(func & STR_FUNC_REGEXP) && rb_enc_asciicompat(enc)) { if (rb_enc_str_coderange(str) == ENC_CODERANGE_7BIT) { } else if (enc0 == rb_usascii_encoding() && enc != rb_utf8_encoding()) { rb_enc_associate(str, rb_ascii8bit_encoding()); } } return str; } #define lex_goto_eol(parser) ((parser)->parser_lex_p = (parser)->parser_lex_pend) #define lex_eol_p() (lex_p >= lex_pend) #define peek(c) peek_n((c), 0) #define peek_n(c,n) (lex_p+(n) < lex_pend && (c) == (unsigned char)lex_p[n]) static inline int parser_nextc(struct parser_params *parser) { int c; if (lex_p == lex_pend) { VALUE v = lex_nextline; lex_nextline = 0; if (!v) { if (parser->eofp) return -1; if (!lex_input || NIL_P(v = lex_getline(parser))) { parser->eofp = Qtrue; lex_goto_eol(parser); return -1; } } { #ifdef RIPPER if (parser->tokp < lex_pend) { if (NIL_P(parser->delayed)) { parser->delayed = rb_str_buf_new(1024); rb_enc_associate(parser->delayed, current_enc); rb_str_buf_cat(parser->delayed, parser->tokp, lex_pend - parser->tokp); parser->delayed_line = ruby_sourceline; parser->delayed_col = (int)(parser->tokp - lex_pbeg); } else { rb_str_buf_cat(parser->delayed, parser->tokp, lex_pend - parser->tokp); } } #endif if (heredoc_end > 0) { ruby_sourceline = heredoc_end; heredoc_end = 0; } ruby_sourceline++; parser->line_count++; lex_pbeg = lex_p = RSTRING_PTR(v); lex_pend = lex_p + RSTRING_LEN(v); ripper_flush(parser); lex_lastline = v; } } c = (unsigned char)*lex_p++; if (c == '\r') { if (peek('\n')) { lex_p++; c = '\n'; } else if (ruby_sourceline > parser->last_cr_line) { parser->last_cr_line = ruby_sourceline; rb_compile_warn(ruby_sourcefile, ruby_sourceline, "encountered \\r in middle of line, treated as a mere space"); } } return c; } static void parser_pushback(struct parser_params *parser, int c) { if (c == -1) return; lex_p--; if (lex_p > lex_pbeg && lex_p[0] == '\n' && lex_p[-1] == '\r') { lex_p--; } } #define was_bol() (lex_p == lex_pbeg + 1) #define tokfix() (tokenbuf[tokidx]='\0') #define tok() tokenbuf #define toklen() tokidx #define toklast() (tokidx>0?tokenbuf[tokidx-1]:0) static char* parser_newtok(struct parser_params *parser) { tokidx = 0; tokline = ruby_sourceline; if (!tokenbuf) { toksiz = 60; tokenbuf = ALLOC_N(char, 60); } if (toksiz > 4096) { toksiz = 60; REALLOC_N(tokenbuf, char, 60); } return tokenbuf; } static char * parser_tokspace(struct parser_params *parser, int n) { tokidx += n; if (tokidx >= toksiz) { do {toksiz *= 2;} while (toksiz < tokidx); REALLOC_N(tokenbuf, char, toksiz); } return &tokenbuf[tokidx-n]; } static void parser_tokadd(struct parser_params *parser, int c) { tokenbuf[tokidx++] = (char)c; if (tokidx >= toksiz) { toksiz *= 2; REALLOC_N(tokenbuf, char, toksiz); } } static int parser_tok_hex(struct parser_params *parser, size_t *numlen) { int c; c = scan_hex(lex_p, 2, numlen); if (!*numlen) { yyerror("invalid hex escape"); return 0; } lex_p += *numlen; return c; } #define tokcopy(n) memcpy(tokspace(n), lex_p - (n), (n)) /* return value is for ?\u3042 */ static int parser_tokadd_utf8(struct parser_params *parser, rb_encoding **encp, int string_literal, int symbol_literal, int regexp_literal) { /* * If string_literal is true, then we allow multiple codepoints * in \u{}, and add the codepoints to the current token. * Otherwise we're parsing a character literal and return a single * codepoint without adding it */ int codepoint; size_t numlen; if (regexp_literal) { tokadd('\\'); tokadd('u'); } if (peek('{')) { /* handle \u{...} form */ do { if (regexp_literal) { tokadd(*lex_p); } nextc(); codepoint = scan_hex(lex_p, 6, &numlen); if (numlen == 0) { yyerror("invalid Unicode escape"); return 0; } if (codepoint > 0x10ffff) { yyerror("invalid Unicode codepoint (too large)"); return 0; } lex_p += numlen; if (regexp_literal) { tokcopy((int)numlen); } else if (codepoint >= 0x80) { *encp = rb_utf8_encoding(); if (string_literal) tokaddmbc(codepoint, *encp); } else if (string_literal) { tokadd(codepoint); } } while (string_literal && (peek(' ') || peek('\t'))); if (!peek('}')) { yyerror("unterminated Unicode escape"); return 0; } if (regexp_literal) { tokadd('}'); } nextc(); } else { /* handle \uxxxx form */ codepoint = scan_hex(lex_p, 4, &numlen); if (numlen < 4) { yyerror("invalid Unicode escape"); return 0; } lex_p += 4; if (regexp_literal) { tokcopy(4); } else if (codepoint >= 0x80) { *encp = rb_utf8_encoding(); if (string_literal) tokaddmbc(codepoint, *encp); } else if (string_literal) { tokadd(codepoint); } } return codepoint; } #define ESCAPE_CONTROL 1 #define ESCAPE_META 2 static int parser_read_escape(struct parser_params *parser, int flags, rb_encoding **encp) { int c; size_t numlen; switch (c = nextc()) { case '\\': /* Backslash */ return c; case 'n': /* newline */ return '\n'; case 't': /* horizontal tab */ return '\t'; case 'r': /* carriage-return */ return '\r'; case 'f': /* form-feed */ return '\f'; case 'v': /* vertical tab */ return '\13'; case 'a': /* alarm(bell) */ return '\007'; case 'e': /* escape */ return 033; case '0': case '1': case '2': case '3': /* octal constant */ case '4': case '5': case '6': case '7': pushback(c); c = scan_oct(lex_p, 3, &numlen); lex_p += numlen; return c; case 'x': /* hex constant */ c = tok_hex(&numlen); if (numlen == 0) return 0; return c; case 'b': /* backspace */ return '\010'; case 's': /* space */ return ' '; case 'M': if (flags & ESCAPE_META) goto eof; if ((c = nextc()) != '-') { pushback(c); goto eof; } if ((c = nextc()) == '\\') { if (peek('u')) goto eof; return read_escape(flags|ESCAPE_META, encp) | 0x80; } else if (c == -1 || !ISASCII(c)) goto eof; else { return ((c & 0xff) | 0x80); } case 'C': if ((c = nextc()) != '-') { pushback(c); goto eof; } case 'c': if (flags & ESCAPE_CONTROL) goto eof; if ((c = nextc())== '\\') { if (peek('u')) goto eof; c = read_escape(flags|ESCAPE_CONTROL, encp); } else if (c == '?') return 0177; else if (c == -1 || !ISASCII(c)) goto eof; return c & 0x9f; eof: case -1: yyerror("Invalid escape character syntax"); return '\0'; default: return c; } } static void parser_tokaddmbc(struct parser_params *parser, int c, rb_encoding *enc) { int len = rb_enc_codelen(c, enc); rb_enc_mbcput(c, tokspace(len), enc); } static int parser_tokadd_escape(struct parser_params *parser, rb_encoding **encp) { int c; int flags = 0; size_t numlen; first: switch (c = nextc()) { case '\n': return 0; /* just ignore */ case '0': case '1': case '2': case '3': /* octal constant */ case '4': case '5': case '6': case '7': { ruby_scan_oct(--lex_p, 3, &numlen); if (numlen == 0) goto eof; lex_p += numlen; tokcopy((int)numlen + 1); } return 0; case 'x': /* hex constant */ { tok_hex(&numlen); if (numlen == 0) return -1; tokcopy((int)numlen + 2); } return 0; case 'M': if (flags & ESCAPE_META) goto eof; if ((c = nextc()) != '-') { pushback(c); goto eof; } tokcopy(3); flags |= ESCAPE_META; goto escaped; case 'C': if (flags & ESCAPE_CONTROL) goto eof; if ((c = nextc()) != '-') { pushback(c); goto eof; } tokcopy(3); goto escaped; case 'c': if (flags & ESCAPE_CONTROL) goto eof; tokcopy(2); flags |= ESCAPE_CONTROL; escaped: if ((c = nextc()) == '\\') { goto first; } else if (c == -1) goto eof; tokadd(c); return 0; eof: case -1: yyerror("Invalid escape character syntax"); return -1; default: tokadd('\\'); tokadd(c); } return 0; } static int parser_regx_options(struct parser_params *parser) { int kcode = 0; int kopt = 0; int options = 0; int c, opt, kc; newtok(); while (c = nextc(), ISALPHA(c)) { if (c == 'o') { options |= RE_OPTION_ONCE; } else if (rb_char_to_option_kcode(c, &opt, &kc)) { if (kc >= 0) { if (kc != rb_ascii8bit_encindex()) kcode = c; kopt = opt; } else { options |= opt; } } else { tokadd(c); } } options |= kopt; pushback(c); if (toklen()) { tokfix(); compile_error(PARSER_ARG "unknown regexp option%s - %s", toklen() > 1 ? "s" : "", tok()); } return options | RE_OPTION_ENCODING(kcode); } static int parser_str_options(struct parser_params *parser) { int c, options = 0; const char *save_p = lex_p; while (c = nextc(), ISALPHA(c)) { switch (c) { #if STR_OPTION_FROZEN case 'f': options |= STR_OPTION_FROZEN; break; #endif #if STR_OPTION_BINARY case 'b': options |= STR_OPTION_BINARY; break; #endif default: lex_p = save_p; return 0; } } pushback(c); return options; } static void dispose_string(VALUE str) { rb_str_free(str); rb_gc_force_recycle(str); } static int parser_tokadd_mbchar(struct parser_params *parser, int c) { int len = parser_precise_mbclen(); if (!MBCLEN_CHARFOUND_P(len)) { compile_error(PARSER_ARG "invalid multibyte char (%s)", parser_encoding_name()); return -1; } tokadd(c); lex_p += --len; if (len > 0) tokcopy(len); return c; } #define tokadd_mbchar(c) parser_tokadd_mbchar(parser, (c)) static inline int simple_re_meta(int c) { switch (c) { case '$': case '*': case '+': case '.': case '?': case '^': case '|': case ')': case ']': case '}': case '>': return TRUE; default: return FALSE; } } static int parser_tokadd_string(struct parser_params *parser, int func, int term, int paren, long *nest, rb_encoding **encp) { int c; int has_nonascii = 0; rb_encoding *enc = *encp; char *errbuf = 0; static const char mixed_msg[] = "%s mixed within %s source"; #define mixed_error(enc1, enc2) if (!errbuf) { \ size_t len = sizeof(mixed_msg) - 4; \ len += strlen(rb_enc_name(enc1)); \ len += strlen(rb_enc_name(enc2)); \ errbuf = ALLOCA_N(char, len); \ snprintf(errbuf, len, mixed_msg, \ rb_enc_name(enc1), \ rb_enc_name(enc2)); \ yyerror(errbuf); \ } #define mixed_escape(beg, enc1, enc2) do { \ const char *pos = lex_p; \ lex_p = (beg); \ mixed_error((enc1), (enc2)); \ lex_p = pos; \ } while (0) while ((c = nextc()) != -1) { if (paren && c == paren) { ++*nest; } else if (c == term) { if (!nest || !*nest) { pushback(c); break; } --*nest; } else if ((func & STR_FUNC_EXPAND) && c == '#' && lex_p < lex_pend) { int c2 = *lex_p; if (c2 == '$' || c2 == '@' || c2 == '{') { pushback(c); break; } } else if (c == '\\') { const char *beg = lex_p - 1; c = nextc(); switch (c) { case '\n': if (func & STR_FUNC_QWORDS) break; if (func & STR_FUNC_EXPAND) continue; tokadd('\\'); break; case '\\': if (func & STR_FUNC_ESCAPE) tokadd(c); break; case 'u': if ((func & STR_FUNC_EXPAND) == 0) { tokadd('\\'); break; } parser_tokadd_utf8(parser, &enc, 1, func & STR_FUNC_SYMBOL, func & STR_FUNC_REGEXP); if (has_nonascii && enc != *encp) { mixed_escape(beg, enc, *encp); } continue; default: if (c == -1) return -1; if (!ISASCII(c)) { if ((func & STR_FUNC_EXPAND) == 0) tokadd('\\'); goto non_ascii; } if (func & STR_FUNC_REGEXP) { if (c == term && !simple_re_meta(c)) { tokadd(c); continue; } pushback(c); if ((c = tokadd_escape(&enc)) < 0) return -1; if (has_nonascii && enc != *encp) { mixed_escape(beg, enc, *encp); } continue; } else if (func & STR_FUNC_EXPAND) { pushback(c); if (func & STR_FUNC_ESCAPE) tokadd('\\'); c = read_escape(0, &enc); } else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) { /* ignore backslashed spaces in %w */ } else if (c != term && !(paren && c == paren)) { tokadd('\\'); pushback(c); continue; } } } else if (!parser_isascii()) { non_ascii: has_nonascii = 1; if (enc != *encp) { mixed_error(enc, *encp); continue; } if (tokadd_mbchar(c) == -1) return -1; continue; } else if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) { pushback(c); break; } if (c & 0x80) { has_nonascii = 1; if (enc != *encp) { mixed_error(enc, *encp); continue; } } tokadd(c); } *encp = enc; return c; } #define NEW_STRTERM(func, term, paren) \ rb_node_newnode(NODE_STRTERM, (func), (term) | ((paren) << (CHAR_BIT * 2)), 0) #ifdef RIPPER static void ripper_flush_string_content(struct parser_params *parser, rb_encoding *enc) { if (!NIL_P(parser->delayed)) { ptrdiff_t len = lex_p - parser->tokp; if (len > 0) { rb_enc_str_buf_cat(parser->delayed, parser->tokp, len, enc); } ripper_dispatch_delayed_token(parser, tSTRING_CONTENT); parser->tokp = lex_p; } } #define flush_string_content(enc) ripper_flush_string_content(parser, (enc)) #else #define flush_string_content(enc) ((void)(enc)) #endif RUBY_FUNC_EXPORTED const unsigned int ruby_global_name_punct_bits[(0x7e - 0x20 + 31) / 32]; /* this can be shared with ripper, since it's independent from struct * parser_params. */ #ifndef RIPPER #define BIT(c, idx) (((c) / 32 - 1 == idx) ? (1U << ((c) % 32)) : 0) #define SPECIAL_PUNCT(idx) ( \ BIT('~', idx) | BIT('*', idx) | BIT('$', idx) | BIT('?', idx) | \ BIT('!', idx) | BIT('@', idx) | BIT('/', idx) | BIT('\\', idx) | \ BIT(';', idx) | BIT(',', idx) | BIT('.', idx) | BIT('=', idx) | \ BIT(':', idx) | BIT('<', idx) | BIT('>', idx) | BIT('\"', idx) | \ BIT('&', idx) | BIT('`', idx) | BIT('\'', idx) | BIT('+', idx) | \ BIT('0', idx)) const unsigned int ruby_global_name_punct_bits[] = { SPECIAL_PUNCT(0), SPECIAL_PUNCT(1), SPECIAL_PUNCT(2), }; #undef BIT #undef SPECIAL_PUNCT #endif static inline int is_global_name_punct(const int c) { if (c <= 0x20 || 0x7e < c) return 0; return (ruby_global_name_punct_bits[(c - 0x20) / 32] >> (c % 32)) & 1; } static int parser_peek_variable_name(struct parser_params *parser) { int c; const char *p = lex_p; if (p + 1 >= lex_pend) return 0; c = *p++; switch (c) { case '$': if ((c = *p) == '-') { if (++p >= lex_pend) return 0; c = *p; } else if (is_global_name_punct(c) || ISDIGIT(c)) { return tSTRING_DVAR; } break; case '@': if ((c = *p) == '@') { if (++p >= lex_pend) return 0; c = *p; } break; case '{': lex_p = p; command_start = TRUE; return tSTRING_DBEG; default: return 0; } if (!ISASCII(c) || c == '_' || ISALPHA(c)) return tSTRING_DVAR; return 0; } static int parser_parse_string(struct parser_params *parser, NODE *quote) { int func = (int)quote->nd_func; int term = nd_term(quote); int paren = nd_paren(quote); int c, space = 0; rb_encoding *enc = current_enc; if (func == -1) return tSTRING_END; if (func == 0) { set_yylval_num(term); return tSTRING_SUFFIX; } c = nextc(); if ((func & STR_FUNC_QWORDS) && ISSPACE(c)) { do {c = nextc();} while (ISSPACE(c)); space = 1; } if (c == term && !quote->nd_nest) { if (func & STR_FUNC_QWORDS) { quote->nd_func = -1; quote->u2.id = 0; return ' '; } if (func & STR_FUNC_REGEXP) { set_yylval_num(regx_options()); return tREGEXP_END; } if ((func & STR_FUNC_OPTION) && (func = str_options()) != 0) { quote->nd_func = 0; quote->u2.id = func; } return tSTRING_END; } if (space) { pushback(c); return ' '; } newtok(); if ((func & STR_FUNC_EXPAND) && c == '#') { int t = parser_peek_variable_name(parser); if (t) return t; tokadd('#'); c = nextc(); } pushback(c); if (tokadd_string(func, term, paren, "e->nd_nest, &enc) == -1) { ruby_sourceline = nd_line(quote); if (func & STR_FUNC_REGEXP) { if (parser->eofp) compile_error(PARSER_ARG "unterminated regexp meets end of file"); return tREGEXP_END; } else { if (parser->eofp) compile_error(PARSER_ARG "unterminated string meets end of file"); return tSTRING_END; } } tokfix(); set_yylval_str(STR_NEW3(tok(), toklen(), enc, func)); flush_string_content(enc); return tSTRING_CONTENT; } static int parser_heredoc_identifier(struct parser_params *parser) { int c = nextc(), term, func = 0; long len; if (c == '-') { c = nextc(); func = STR_FUNC_INDENT; } switch (c) { case '\'': func |= str_squote; goto quoted; case '"': func |= str_dquote; goto quoted; case '`': func |= str_xquote; quoted: newtok(); tokadd(func); term = c; while ((c = nextc()) != -1 && c != term) { if (tokadd_mbchar(c) == -1) return 0; } if (c == -1) { compile_error(PARSER_ARG "unterminated here document identifier"); return 0; } break; default: if (!parser_is_identchar()) { pushback(c); if (func & STR_FUNC_INDENT) { pushback('-'); } return 0; } newtok(); term = '"'; tokadd(func |= str_dquote); do { if (tokadd_mbchar(c) == -1) return 0; } while ((c = nextc()) != -1 && parser_is_identchar()); pushback(c); break; } tokfix(); #ifdef RIPPER ripper_dispatch_scan_event(parser, tHEREDOC_BEG); #endif len = lex_p - lex_pbeg; lex_goto_eol(parser); lex_strterm = rb_node_newnode(NODE_HEREDOC, STR_NEW(tok(), toklen()), /* nd_lit */ len, /* nd_nth */ lex_lastline); /* nd_orig */ nd_set_line(lex_strterm, ruby_sourceline); ripper_flush(parser); return term == '`' ? tXSTRING_BEG : tSTRING_BEG; } static void parser_heredoc_restore(struct parser_params *parser, NODE *here) { VALUE line; lex_strterm = 0; line = here->nd_orig; lex_lastline = line; lex_pbeg = RSTRING_PTR(line); lex_pend = lex_pbeg + RSTRING_LEN(line); lex_p = lex_pbeg + here->nd_nth; heredoc_end = ruby_sourceline; ruby_sourceline = nd_line(here); dispose_string(here->nd_lit); rb_gc_force_recycle((VALUE)here); ripper_flush(parser); } static int parser_whole_match_p(struct parser_params *parser, const char *eos, long len, int indent) { const char *p = lex_pbeg; long n; if (indent) { while (*p && ISSPACE(*p)) p++; } n = lex_pend - (p + len); if (n < 0) return FALSE; if (n > 0 && p[len] != '\n') { if (p[len] != '\r') return FALSE; if (n <= 1 || p[len+1] != '\n') return FALSE; } return strncmp(eos, p, len) == 0; } #define NUM_SUFFIX_R (1<<0) #define NUM_SUFFIX_I (1<<1) #define NUM_SUFFIX_ALL 3 static int parser_number_literal_suffix(struct parser_params *parser, int mask) { int c, result = 0; const char *lastp = lex_p; while ((c = nextc()) != -1) { if ((mask & NUM_SUFFIX_I) && c == 'i') { result |= (mask & NUM_SUFFIX_I); mask &= ~NUM_SUFFIX_I; /* r after i, rational of complex is disallowed */ mask &= ~NUM_SUFFIX_R; continue; } if ((mask & NUM_SUFFIX_R) && c == 'r') { result |= (mask & NUM_SUFFIX_R); mask &= ~NUM_SUFFIX_R; continue; } if (!ISASCII(c) || ISALPHA(c) || c == '_') { lex_p = lastp; return 0; } pushback(c); break; } return result; } static int parser_set_number_literal(struct parser_params *parser, VALUE v, int type, int suffix) { if (suffix & NUM_SUFFIX_I) { v = rb_complex_raw(INT2FIX(0), v); type = tIMAGINARY; } set_yylval_literal(v); return type; } static int parser_set_integer_literal(struct parser_params *parser, VALUE v, int suffix) { int type = tINTEGER; if (suffix & NUM_SUFFIX_R) { v = rb_rational_raw1(v); type = tRATIONAL; } return set_number_literal(v, type, suffix); } #ifdef RIPPER static void ripper_dispatch_heredoc_end(struct parser_params *parser) { if (!NIL_P(parser->delayed)) ripper_dispatch_delayed_token(parser, tSTRING_CONTENT); lex_goto_eol(parser); ripper_dispatch_ignored_scan_event(parser, tHEREDOC_END); } #define dispatch_heredoc_end() ripper_dispatch_heredoc_end(parser) #else #define dispatch_heredoc_end() ((void)0) #endif static int parser_here_document(struct parser_params *parser, NODE *here) { int c, func, indent = 0; const char *eos, *p, *pend; long len; VALUE str = 0; rb_encoding *enc = current_enc; eos = RSTRING_PTR(here->nd_lit); len = RSTRING_LEN(here->nd_lit) - 1; indent = (func = *eos++) & STR_FUNC_INDENT; if ((c = nextc()) == -1) { error: compile_error(PARSER_ARG "can't find string \"%s\" anywhere before EOF", eos); #ifdef RIPPER if (NIL_P(parser->delayed)) { ripper_dispatch_scan_event(parser, tSTRING_CONTENT); } else { if (str || ((len = lex_p - parser->tokp) > 0 && (str = STR_NEW3(parser->tokp, len, enc, func), 1))) { rb_str_append(parser->delayed, str); } ripper_dispatch_delayed_token(parser, tSTRING_CONTENT); } lex_goto_eol(parser); #endif restore: heredoc_restore(lex_strterm); return 0; } if (was_bol() && whole_match_p(eos, len, indent)) { dispatch_heredoc_end(); heredoc_restore(lex_strterm); return tSTRING_END; } if (!(func & STR_FUNC_EXPAND)) { do { p = RSTRING_PTR(lex_lastline); pend = lex_pend; if (pend > p) { switch (pend[-1]) { case '\n': if (--pend == p || pend[-1] != '\r') { pend++; break; } case '\r': --pend; } } if (str) rb_str_cat(str, p, pend - p); else str = STR_NEW(p, pend - p); if (pend < lex_pend) rb_str_cat(str, "\n", 1); lex_goto_eol(parser); if (nextc() == -1) { if (str) dispose_string(str); goto error; } } while (!whole_match_p(eos, len, indent)); } else { /* int mb = ENC_CODERANGE_7BIT, *mbp = &mb;*/ newtok(); if (c == '#') { int t = parser_peek_variable_name(parser); if (t) return t; tokadd('#'); c = nextc(); } do { pushback(c); if ((c = tokadd_string(func, '\n', 0, NULL, &enc)) == -1) { if (parser->eofp) goto error; goto restore; } if (c != '\n') { set_yylval_str(STR_NEW3(tok(), toklen(), enc, func)); flush_string_content(enc); return tSTRING_CONTENT; } tokadd(nextc()); /* if (mbp && mb == ENC_CODERANGE_UNKNOWN) mbp = 0;*/ if ((c = nextc()) == -1) goto error; } while (!whole_match_p(eos, len, indent)); str = STR_NEW3(tok(), toklen(), enc, func); } dispatch_heredoc_end(); heredoc_restore(lex_strterm); lex_strterm = NEW_STRTERM(-1, 0, 0); set_yylval_str(str); return tSTRING_CONTENT; } #include "lex.c" static void arg_ambiguous_gen(struct parser_params *parser) { #ifndef RIPPER rb_warning0("ambiguous first argument; put parentheses or even spaces"); #else dispatch0(arg_ambiguous); #endif } #define arg_ambiguous() (arg_ambiguous_gen(parser), 1) static ID formal_argument_gen(struct parser_params *parser, ID lhs) { #ifndef RIPPER if (!is_local_id(lhs)) yyerror("formal argument must be local variable"); #endif shadowing_lvar(lhs); return lhs; } static int lvar_defined_gen(struct parser_params *parser, ID id) { return (dyna_in_block() && dvar_defined_get(id)) || local_id(id); } /* emacsen -*- hack */ static long parser_encode_length(struct parser_params *parser, const char *name, long len) { long nlen; if (len > 5 && name[nlen = len - 5] == '-') { if (rb_memcicmp(name + nlen + 1, "unix", 4) == 0) return nlen; } if (len > 4 && name[nlen = len - 4] == '-') { if (rb_memcicmp(name + nlen + 1, "dos", 3) == 0) return nlen; if (rb_memcicmp(name + nlen + 1, "mac", 3) == 0 && !(len == 8 && rb_memcicmp(name, "utf8-mac", len) == 0)) /* exclude UTF8-MAC because the encoding named "UTF8" doesn't exist in Ruby */ return nlen; } return len; } static void parser_set_encode(struct parser_params *parser, const char *name) { int idx = rb_enc_find_index(name); rb_encoding *enc; VALUE excargs[3]; if (idx < 0) { excargs[1] = rb_sprintf("unknown encoding name: %s", name); error: excargs[0] = rb_eArgError; excargs[2] = rb_make_backtrace(); rb_ary_unshift(excargs[2], rb_sprintf("%s:%d", ruby_sourcefile, ruby_sourceline)); rb_exc_raise(rb_make_exception(3, excargs)); } enc = rb_enc_from_index(idx); if (!rb_enc_asciicompat(enc)) { excargs[1] = rb_sprintf("%s is not ASCII compatible", rb_enc_name(enc)); goto error; } parser->enc = enc; #ifndef RIPPER if (ruby_debug_lines) { VALUE lines = ruby_debug_lines; long i, n = RARRAY_LEN(lines); for (i = 0; i < n; ++i) { rb_enc_associate_index(RARRAY_AREF(lines, i), idx); } } #endif } static int comment_at_top(struct parser_params *parser) { const char *p = lex_pbeg, *pend = lex_p - 1; if (parser->line_count != (parser->has_shebang ? 2 : 1)) return 0; while (p < pend) { if (!ISSPACE(*p)) return 0; p++; } return 1; } #ifndef RIPPER typedef long (*rb_magic_comment_length_t)(struct parser_params *parser, const char *name, long len); typedef void (*rb_magic_comment_setter_t)(struct parser_params *parser, const char *name, const char *val); static void magic_comment_encoding(struct parser_params *parser, const char *name, const char *val) { if (!comment_at_top(parser)) { return; } parser_set_encode(parser, val); } static void parser_set_token_info(struct parser_params *parser, const char *name, const char *val) { int *p = &parser->parser_token_info_enabled; switch (*val) { case 't': case 'T': if (strcasecmp(val, "true") == 0) { *p = TRUE; return; } break; case 'f': case 'F': if (strcasecmp(val, "false") == 0) { *p = FALSE; return; } break; } rb_compile_warning(ruby_sourcefile, ruby_sourceline, "invalid value for %s: %s", name, val); } struct magic_comment { const char *name; rb_magic_comment_setter_t func; rb_magic_comment_length_t length; }; static const struct magic_comment magic_comments[] = { {"coding", magic_comment_encoding, parser_encode_length}, {"encoding", magic_comment_encoding, parser_encode_length}, {"warn_indent", parser_set_token_info}, }; #endif static const char * magic_comment_marker(const char *str, long len) { long i = 2; while (i < len) { switch (str[i]) { case '-': if (str[i-1] == '*' && str[i-2] == '-') { return str + i + 1; } i += 2; break; case '*': if (i + 1 >= len) return 0; if (str[i+1] != '-') { i += 4; } else if (str[i-1] != '-') { i += 2; } else { return str + i + 2; } break; default: i += 3; break; } } return 0; } static int parser_magic_comment(struct parser_params *parser, const char *str, long len) { VALUE name = 0, val = 0; const char *beg, *end, *vbeg, *vend; #define str_copy(_s, _p, _n) ((_s) \ ? (void)(rb_str_resize((_s), (_n)), \ MEMCPY(RSTRING_PTR(_s), (_p), char, (_n)), (_s)) \ : (void)((_s) = STR_NEW((_p), (_n)))) if (len <= 7) return FALSE; if (!(beg = magic_comment_marker(str, len))) return FALSE; if (!(end = magic_comment_marker(beg, str + len - beg))) return FALSE; str = beg; len = end - beg - 3; /* %r"([^\\s\'\":;]+)\\s*:\\s*(\"(?:\\\\.|[^\"])*\"|[^\"\\s;]+)[\\s;]*" */ while (len > 0) { #ifndef RIPPER const struct magic_comment *p = magic_comments; #endif char *s; int i; long n = 0; for (; len > 0 && *str; str++, --len) { switch (*str) { case '\'': case '"': case ':': case ';': continue; } if (!ISSPACE(*str)) break; } for (beg = str; len > 0; str++, --len) { switch (*str) { case '\'': case '"': case ':': case ';': break; default: if (ISSPACE(*str)) break; continue; } break; } for (end = str; len > 0 && ISSPACE(*str); str++, --len); if (!len) break; if (*str != ':') continue; do str++; while (--len > 0 && ISSPACE(*str)); if (!len) break; if (*str == '"') { for (vbeg = ++str; --len > 0 && *str != '"'; str++) { if (*str == '\\') { --len; ++str; } } vend = str; if (len) { --len; ++str; } } else { for (vbeg = str; len > 0 && *str != '"' && *str != ';' && !ISSPACE(*str); --len, str++); vend = str; } while (len > 0 && (*str == ';' || ISSPACE(*str))) --len, str++; n = end - beg; str_copy(name, beg, n); s = RSTRING_PTR(name); for (i = 0; i < n; ++i) { if (s[i] == '-') s[i] = '_'; } #ifndef RIPPER do { if (STRNCASECMP(p->name, s, n) == 0) { n = vend - vbeg; if (p->length) { n = (*p->length)(parser, vbeg, n); } str_copy(val, vbeg, n); (*p->func)(parser, s, RSTRING_PTR(val)); break; } } while (++p < magic_comments + numberof(magic_comments)); #else str_copy(val, vbeg, vend - vbeg); dispatch2(magic_comment, name, val); #endif } return TRUE; } static void set_file_encoding(struct parser_params *parser, const char *str, const char *send) { int sep = 0; const char *beg = str; VALUE s; for (;;) { if (send - str <= 6) return; switch (str[6]) { case 'C': case 'c': str += 6; continue; case 'O': case 'o': str += 5; continue; case 'D': case 'd': str += 4; continue; case 'I': case 'i': str += 3; continue; case 'N': case 'n': str += 2; continue; case 'G': case 'g': str += 1; continue; case '=': case ':': sep = 1; str += 6; break; default: str += 6; if (ISSPACE(*str)) break; continue; } if (STRNCASECMP(str-6, "coding", 6) == 0) break; } for (;;) { do { if (++str >= send) return; } while (ISSPACE(*str)); if (sep) break; if (*str != '=' && *str != ':') return; sep = 1; str++; } beg = str; while ((*str == '-' || *str == '_' || ISALNUM(*str)) && ++str < send); s = rb_str_new(beg, parser_encode_length(parser, beg, str - beg)); parser_set_encode(parser, RSTRING_PTR(s)); rb_str_resize(s, 0); } static void parser_prepare(struct parser_params *parser) { int c = nextc(); switch (c) { case '#': if (peek('!')) parser->has_shebang = 1; break; case 0xef: /* UTF-8 BOM marker */ if (lex_pend - lex_p >= 2 && (unsigned char)lex_p[0] == 0xbb && (unsigned char)lex_p[1] == 0xbf) { parser->enc = rb_utf8_encoding(); lex_p += 2; lex_pbeg = lex_p; return; } break; case EOF: return; } pushback(c); parser->enc = rb_enc_get(lex_lastline); } #define IS_ARG() IS_lex_state(EXPR_ARG_ANY) #define IS_END() IS_lex_state(EXPR_END_ANY) #define IS_BEG() IS_lex_state(EXPR_BEG_ANY) #define IS_SPCARG(c) (IS_ARG() && space_seen && !ISSPACE(c)) #define IS_LABEL_POSSIBLE() ((IS_lex_state(EXPR_BEG | EXPR_ENDFN) && !cmd_state) || IS_ARG()) #define IS_LABEL_SUFFIX(n) (peek_n(':',(n)) && !peek_n(':', (n)+1)) #define IS_AFTER_OPERATOR() IS_lex_state(EXPR_FNAME | EXPR_DOT) #ifndef RIPPER #define ambiguous_operator(op, syn) ( \ rb_warning0("`"op"' after local variable or literal is interpreted as binary operator"), \ rb_warning0("even though it seems like "syn"")) #else #define ambiguous_operator(op, syn) dispatch2(operator_ambiguous, ripper_intern(op), rb_str_new_cstr(syn)) #endif #define warn_balanced(op, syn) ((void) \ (!IS_lex_state_for(last_state, EXPR_CLASS|EXPR_DOT|EXPR_FNAME|EXPR_ENDFN|EXPR_ENDARG) && \ space_seen && !ISSPACE(c) && \ (ambiguous_operator(op, syn), 0))) static int parser_yylex(struct parser_params *parser) { register int c; int space_seen = 0; int cmd_state; enum lex_state_e last_state; rb_encoding *enc; int mb; #ifdef RIPPER int fallthru = FALSE; #endif if (lex_strterm) { int token; if (nd_type(lex_strterm) == NODE_HEREDOC) { token = here_document(lex_strterm); if (token == tSTRING_END) { lex_strterm = 0; lex_state = EXPR_END; } } else { token = parse_string(lex_strterm); if ((token == tSTRING_END && lex_strterm->nd_func) || token == tSTRING_SUFFIX || token == tREGEXP_END) { rb_gc_force_recycle((VALUE)lex_strterm); lex_strterm = 0; lex_state = EXPR_END; } } return token; } cmd_state = command_start; command_start = FALSE; retry: last_state = lex_state; switch (c = nextc()) { case '\0': /* NUL */ case '\004': /* ^D */ case '\032': /* ^Z */ case -1: /* end of script. */ return 0; /* white spaces */ case ' ': case '\t': case '\f': case '\r': case '\13': /* '\v' */ space_seen = 1; #ifdef RIPPER while ((c = nextc())) { switch (c) { case ' ': case '\t': case '\f': case '\r': case '\13': /* '\v' */ break; default: goto outofloop; } } outofloop: pushback(c); ripper_dispatch_scan_event(parser, tSP); #endif goto retry; case '#': /* it's a comment */ /* no magic_comment in shebang line */ if (!parser_magic_comment(parser, lex_p, lex_pend - lex_p)) { if (comment_at_top(parser)) { set_file_encoding(parser, lex_p, lex_pend); } } lex_p = lex_pend; #ifdef RIPPER ripper_dispatch_scan_event(parser, tCOMMENT); fallthru = TRUE; #endif /* fall through */ case '\n': if (IS_lex_state(EXPR_BEG | EXPR_VALUE | EXPR_CLASS | EXPR_FNAME | EXPR_DOT)) { #ifdef RIPPER if (!fallthru) { ripper_dispatch_scan_event(parser, tIGNORED_NL); } fallthru = FALSE; #endif goto retry; } while ((c = nextc())) { switch (c) { case ' ': case '\t': case '\f': case '\r': case '\13': /* '\v' */ space_seen = 1; break; case '.': { if ((c = nextc()) != '.') { pushback(c); pushback('.'); goto retry; } } default: --ruby_sourceline; lex_nextline = lex_lastline; case -1: /* EOF no decrement*/ lex_goto_eol(parser); #ifdef RIPPER if (c != -1) { parser->tokp = lex_p; } #endif goto normal_newline; } } normal_newline: command_start = TRUE; lex_state = EXPR_BEG; return '\n'; case '*': if ((c = nextc()) == '*') { if ((c = nextc()) == '=') { set_yylval_id(tPOW); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); if (IS_SPCARG(c)) { rb_warning0("`**' interpreted as argument prefix"); c = tDSTAR; } else if (IS_BEG()) { c = tDSTAR; } else { warn_balanced("**", "argument prefix"); c = tPOW; } } else { if (c == '=') { set_yylval_id('*'); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); if (IS_SPCARG(c)) { rb_warning0("`*' interpreted as argument prefix"); c = tSTAR; } else if (IS_BEG()) { c = tSTAR; } else { warn_balanced("*", "argument prefix"); c = '*'; } } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; return c; case '!': c = nextc(); if (IS_AFTER_OPERATOR()) { lex_state = EXPR_ARG; if (c == '@') { return '!'; } } else { lex_state = EXPR_BEG; } if (c == '=') { return tNEQ; } if (c == '~') { return tNMATCH; } pushback(c); return '!'; case '=': if (was_bol()) { /* skip embedded rd document */ if (strncmp(lex_p, "begin", 5) == 0 && ISSPACE(lex_p[5])) { #ifdef RIPPER int first_p = TRUE; lex_goto_eol(parser); ripper_dispatch_scan_event(parser, tEMBDOC_BEG); #endif for (;;) { lex_goto_eol(parser); #ifdef RIPPER if (!first_p) { ripper_dispatch_scan_event(parser, tEMBDOC); } first_p = FALSE; #endif c = nextc(); if (c == -1) { compile_error(PARSER_ARG "embedded document meets end of file"); return 0; } if (c != '=') continue; if (strncmp(lex_p, "end", 3) == 0 && (lex_p + 3 == lex_pend || ISSPACE(lex_p[3]))) { break; } } lex_goto_eol(parser); #ifdef RIPPER ripper_dispatch_scan_event(parser, tEMBDOC_END); #endif goto retry; } } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; if ((c = nextc()) == '=') { if ((c = nextc()) == '=') { return tEQQ; } pushback(c); return tEQ; } if (c == '~') { return tMATCH; } else if (c == '>') { return tASSOC; } pushback(c); return '='; case '<': last_state = lex_state; c = nextc(); if (c == '<' && !IS_lex_state(EXPR_DOT | EXPR_CLASS) && !IS_END() && (!IS_ARG() || space_seen)) { int token = heredoc_identifier(); if (token) return token; } if (IS_AFTER_OPERATOR()) { lex_state = EXPR_ARG; } else { if (IS_lex_state(EXPR_CLASS)) command_start = TRUE; lex_state = EXPR_BEG; } if (c == '=') { if ((c = nextc()) == '>') { return tCMP; } pushback(c); return tLEQ; } if (c == '<') { if ((c = nextc()) == '=') { set_yylval_id(tLSHFT); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); warn_balanced("<<", "here document"); return tLSHFT; } pushback(c); return '<'; case '>': lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; if ((c = nextc()) == '=') { return tGEQ; } if (c == '>') { if ((c = nextc()) == '=') { set_yylval_id(tRSHFT); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); return tRSHFT; } pushback(c); return '>'; case '"': lex_strterm = NEW_STRTERM(str_dquote, '"', 0); return tSTRING_BEG; case '`': if (IS_lex_state(EXPR_FNAME)) { lex_state = EXPR_ENDFN; return c; } if (IS_lex_state(EXPR_DOT)) { if (cmd_state) lex_state = EXPR_CMDARG; else lex_state = EXPR_ARG; return c; } lex_strterm = NEW_STRTERM(str_xquote, '`', 0); return tXSTRING_BEG; case '\'': lex_strterm = NEW_STRTERM(str_squote, '\'', 0); return tSTRING_BEG; case '?': if (IS_END()) { lex_state = EXPR_VALUE; return '?'; } c = nextc(); if (c == -1) { compile_error(PARSER_ARG "incomplete character syntax"); return 0; } if (rb_enc_isspace(c, current_enc)) { if (!IS_ARG()) { int c2 = 0; switch (c) { case ' ': c2 = 's'; break; case '\n': c2 = 'n'; break; case '\t': c2 = 't'; break; case '\v': c2 = 'v'; break; case '\r': c2 = 'r'; break; case '\f': c2 = 'f'; break; } if (c2) { rb_warnI("invalid character syntax; use ?\\%c", c2); } } ternary: pushback(c); lex_state = EXPR_VALUE; return '?'; } newtok(); enc = current_enc; if (!parser_isascii()) { if (tokadd_mbchar(c) == -1) return 0; } else if ((rb_enc_isalnum(c, current_enc) || c == '_') && lex_p < lex_pend && is_identchar(lex_p, lex_pend, current_enc)) { goto ternary; } else if (c == '\\') { if (peek('u')) { nextc(); c = parser_tokadd_utf8(parser, &enc, 0, 0, 0); if (0x80 <= c) { tokaddmbc(c, enc); } else { tokadd(c); } } else if (!lex_eol_p() && !(c = *lex_p, ISASCII(c))) { nextc(); if (tokadd_mbchar(c) == -1) return 0; } else { c = read_escape(0, &enc); tokadd(c); } } else { tokadd(c); } tokfix(); set_yylval_str(STR_NEW3(tok(), toklen(), enc, 0)); lex_state = EXPR_END; return tCHAR; case '&': if ((c = nextc()) == '&') { lex_state = EXPR_BEG; if ((c = nextc()) == '=') { set_yylval_id(tANDOP); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); return tANDOP; } else if (c == '=') { set_yylval_id('&'); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); if (IS_SPCARG(c)) { rb_warning0("`&' interpreted as argument prefix"); c = tAMPER; } else if (IS_BEG()) { c = tAMPER; } else { warn_balanced("&", "argument prefix"); c = '&'; } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; return c; case '|': if ((c = nextc()) == '|') { lex_state = EXPR_BEG; if ((c = nextc()) == '=') { set_yylval_id(tOROP); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); return tOROP; } if (c == '=') { set_yylval_id('|'); lex_state = EXPR_BEG; return tOP_ASGN; } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; pushback(c); return '|'; case '+': c = nextc(); if (IS_AFTER_OPERATOR()) { lex_state = EXPR_ARG; if (c == '@') { return tUPLUS; } pushback(c); return '+'; } if (c == '=') { set_yylval_id('+'); lex_state = EXPR_BEG; return tOP_ASGN; } if (IS_BEG() || (IS_SPCARG(c) && arg_ambiguous())) { lex_state = EXPR_BEG; pushback(c); if (c != -1 && ISDIGIT(c)) { c = '+'; goto start_num; } return tUPLUS; } lex_state = EXPR_BEG; pushback(c); warn_balanced("+", "unary operator"); return '+'; case '-': c = nextc(); if (IS_AFTER_OPERATOR()) { lex_state = EXPR_ARG; if (c == '@') { return tUMINUS; } pushback(c); return '-'; } if (c == '=') { set_yylval_id('-'); lex_state = EXPR_BEG; return tOP_ASGN; } if (c == '>') { lex_state = EXPR_ENDFN; return tLAMBDA; } if (IS_BEG() || (IS_SPCARG(c) && arg_ambiguous())) { lex_state = EXPR_BEG; pushback(c); if (c != -1 && ISDIGIT(c)) { return tUMINUS_NUM; } return tUMINUS; } lex_state = EXPR_BEG; pushback(c); warn_balanced("-", "unary operator"); return '-'; case '.': lex_state = EXPR_BEG; if ((c = nextc()) == '.') { if ((c = nextc()) == '.') { return tDOT3; } pushback(c); return tDOT2; } pushback(c); if (c != -1 && ISDIGIT(c)) { yyerror("no . floating literal anymore; put 0 before dot"); } lex_state = EXPR_DOT; return '.'; start_num: case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9': { int is_float, seen_point, seen_e, nondigit; int suffix; is_float = seen_point = seen_e = nondigit = 0; lex_state = EXPR_END; newtok(); if (c == '-' || c == '+') { tokadd(c); c = nextc(); } if (c == '0') { #define no_digits() do {yyerror("numeric literal without digits"); return 0;} while (0) int start = toklen(); c = nextc(); if (c == 'x' || c == 'X') { /* hexadecimal */ c = nextc(); if (c != -1 && ISXDIGIT(c)) { do { if (c == '_') { if (nondigit) break; nondigit = c; continue; } if (!ISXDIGIT(c)) break; nondigit = 0; tokadd(c); } while ((c = nextc()) != -1); } pushback(c); tokfix(); if (toklen() == start) { no_digits(); } else if (nondigit) goto trailing_uc; suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(rb_cstr_to_inum(tok(), 16, FALSE), suffix); } if (c == 'b' || c == 'B') { /* binary */ c = nextc(); if (c == '0' || c == '1') { do { if (c == '_') { if (nondigit) break; nondigit = c; continue; } if (c != '0' && c != '1') break; nondigit = 0; tokadd(c); } while ((c = nextc()) != -1); } pushback(c); tokfix(); if (toklen() == start) { no_digits(); } else if (nondigit) goto trailing_uc; suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(rb_cstr_to_inum(tok(), 2, FALSE), suffix); } if (c == 'd' || c == 'D') { /* decimal */ c = nextc(); if (c != -1 && ISDIGIT(c)) { do { if (c == '_') { if (nondigit) break; nondigit = c; continue; } if (!ISDIGIT(c)) break; nondigit = 0; tokadd(c); } while ((c = nextc()) != -1); } pushback(c); tokfix(); if (toklen() == start) { no_digits(); } else if (nondigit) goto trailing_uc; suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(rb_cstr_to_inum(tok(), 10, FALSE), suffix); } if (c == '_') { /* 0_0 */ goto octal_number; } if (c == 'o' || c == 'O') { /* prefixed octal */ c = nextc(); if (c == -1 || c == '_' || !ISDIGIT(c)) { no_digits(); } } if (c >= '0' && c <= '7') { /* octal */ octal_number: do { if (c == '_') { if (nondigit) break; nondigit = c; continue; } if (c < '0' || c > '9') break; if (c > '7') goto invalid_octal; nondigit = 0; tokadd(c); } while ((c = nextc()) != -1); if (toklen() > start) { pushback(c); tokfix(); if (nondigit) goto trailing_uc; suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(rb_cstr_to_inum(tok(), 8, FALSE), suffix); } if (nondigit) { pushback(c); goto trailing_uc; } } if (c > '7' && c <= '9') { invalid_octal: yyerror("Invalid octal digit"); } else if (c == '.' || c == 'e' || c == 'E') { tokadd('0'); } else { pushback(c); suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(INT2FIX(0), suffix); } } for (;;) { switch (c) { case '0': case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9': nondigit = 0; tokadd(c); break; case '.': if (nondigit) goto trailing_uc; if (seen_point || seen_e) { goto decode_num; } else { int c0 = nextc(); if (c0 == -1 || !ISDIGIT(c0)) { pushback(c0); goto decode_num; } c = c0; } seen_point = toklen(); tokadd('.'); tokadd(c); is_float++; nondigit = 0; break; case 'e': case 'E': if (nondigit) { pushback(c); c = nondigit; goto decode_num; } if (seen_e) { goto decode_num; } nondigit = c; c = nextc(); if (c != '-' && c != '+' && !ISDIGIT(c)) { pushback(c); nondigit = 0; goto decode_num; } tokadd(nondigit); seen_e++; is_float++; tokadd(c); nondigit = (c == '-' || c == '+') ? c : 0; break; case '_': /* `_' in number just ignored */ if (nondigit) goto decode_num; nondigit = c; break; default: goto decode_num; } c = nextc(); } decode_num: pushback(c); if (nondigit) { char tmp[30]; trailing_uc: snprintf(tmp, sizeof(tmp), "trailing `%c' in number", nondigit); yyerror(tmp); } tokfix(); if (is_float) { int type = tFLOAT; VALUE v; suffix = number_literal_suffix(seen_e ? NUM_SUFFIX_I : NUM_SUFFIX_ALL); if (suffix & NUM_SUFFIX_R) { char *point = &tok()[seen_point]; size_t fraclen = toklen()-seen_point-1; type = tRATIONAL; memmove(point, point+1, fraclen+1); v = rb_cstr_to_inum(tok(), 10, FALSE); v = rb_rational_new(v, rb_int_positive_pow(10, fraclen)); } else { double d = strtod(tok(), 0); if (errno == ERANGE) { rb_warningS("Float %s out of range", tok()); errno = 0; } v = DBL2NUM(d); } return set_number_literal(v, type, suffix); } suffix = number_literal_suffix(NUM_SUFFIX_ALL); return set_integer_literal(rb_cstr_to_inum(tok(), 10, FALSE), suffix); } case ')': case ']': paren_nest--; case '}': COND_LEXPOP(); CMDARG_LEXPOP(); if (c == ')') lex_state = EXPR_ENDFN; else lex_state = EXPR_ENDARG; if (c == '}') { if (!brace_nest--) c = tSTRING_DEND; } return c; case ':': c = nextc(); if (c == ':') { if (IS_BEG() || IS_lex_state(EXPR_CLASS) || IS_SPCARG(-1)) { lex_state = EXPR_BEG; return tCOLON3; } lex_state = EXPR_DOT; return tCOLON2; } if (IS_END() || ISSPACE(c)) { pushback(c); warn_balanced(":", "symbol literal"); lex_state = EXPR_BEG; return ':'; } switch (c) { case '\'': lex_strterm = NEW_STRTERM(str_ssym, c, 0); break; case '"': lex_strterm = NEW_STRTERM(str_dsym, c, 0); break; default: pushback(c); break; } lex_state = EXPR_FNAME; return tSYMBEG; case '/': if (IS_lex_state(EXPR_BEG_ANY)) { lex_strterm = NEW_STRTERM(str_regexp, '/', 0); return tREGEXP_BEG; } if ((c = nextc()) == '=') { set_yylval_id('/'); lex_state = EXPR_BEG; return tOP_ASGN; } pushback(c); if (IS_SPCARG(c)) { (void)arg_ambiguous(); lex_strterm = NEW_STRTERM(str_regexp, '/', 0); return tREGEXP_BEG; } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; warn_balanced("/", "regexp literal"); return '/'; case '^': if ((c = nextc()) == '=') { set_yylval_id('^'); lex_state = EXPR_BEG; return tOP_ASGN; } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; pushback(c); return '^'; case ';': lex_state = EXPR_BEG; command_start = TRUE; return ';'; case ',': lex_state = EXPR_BEG; return ','; case '~': if (IS_AFTER_OPERATOR()) { if ((c = nextc()) != '@') { pushback(c); } lex_state = EXPR_ARG; } else { lex_state = EXPR_BEG; } return '~'; case '(': if (IS_BEG()) { c = tLPAREN; } else if (IS_SPCARG(-1)) { c = tLPAREN_ARG; } paren_nest++; COND_PUSH(0); CMDARG_PUSH(0); lex_state = EXPR_BEG; return c; case '[': paren_nest++; if (IS_AFTER_OPERATOR()) { lex_state = EXPR_ARG; if ((c = nextc()) == ']') { if ((c = nextc()) == '=') { return tASET; } pushback(c); return tAREF; } pushback(c); return '['; } else if (IS_BEG()) { c = tLBRACK; } else if (IS_ARG() && space_seen) { c = tLBRACK; } lex_state = EXPR_BEG; COND_PUSH(0); CMDARG_PUSH(0); return c; case '{': ++brace_nest; if (lpar_beg && lpar_beg == paren_nest) { lex_state = EXPR_BEG; lpar_beg = 0; --paren_nest; COND_PUSH(0); CMDARG_PUSH(0); return tLAMBEG; } if (IS_ARG() || IS_lex_state(EXPR_END | EXPR_ENDFN)) c = '{'; /* block (primary) */ else if (IS_lex_state(EXPR_ENDARG)) c = tLBRACE_ARG; /* block (expr) */ else c = tLBRACE; /* hash */ COND_PUSH(0); CMDARG_PUSH(0); lex_state = EXPR_BEG; if (c != tLBRACE) command_start = TRUE; return c; case '\\': c = nextc(); if (c == '\n') { space_seen = 1; #ifdef RIPPER ripper_dispatch_scan_event(parser, tSP); #endif goto retry; /* skip \\n */ } pushback(c); return '\\'; case '%': if (IS_lex_state(EXPR_BEG_ANY)) { int term; int paren; c = nextc(); quotation: if (c == -1 || !ISALNUM(c)) { term = c; c = 'Q'; } else { term = nextc(); if (rb_enc_isalnum(term, current_enc) || !parser_isascii()) { yyerror("unknown type of %string"); return 0; } } if (c == -1 || term == -1) { compile_error(PARSER_ARG "unterminated quoted string meets end of file"); return 0; } paren = term; if (term == '(') term = ')'; else if (term == '[') term = ']'; else if (term == '{') term = '}'; else if (term == '<') term = '>'; else paren = 0; switch (c) { case 'Q': lex_strterm = NEW_STRTERM(str_dquote, term, paren); return tSTRING_BEG; case 'q': lex_strterm = NEW_STRTERM(str_squote, term, paren); return tSTRING_BEG; case 'W': lex_strterm = NEW_STRTERM(str_dword, term, paren); do {c = nextc();} while (ISSPACE(c)); pushback(c); return tWORDS_BEG; case 'w': lex_strterm = NEW_STRTERM(str_sword, term, paren); do {c = nextc();} while (ISSPACE(c)); pushback(c); return tQWORDS_BEG; case 'I': lex_strterm = NEW_STRTERM(str_dword, term, paren); do {c = nextc();} while (ISSPACE(c)); pushback(c); return tSYMBOLS_BEG; case 'i': lex_strterm = NEW_STRTERM(str_sword, term, paren); do {c = nextc();} while (ISSPACE(c)); pushback(c); return tQSYMBOLS_BEG; case 'x': lex_strterm = NEW_STRTERM(str_xquote, term, paren); return tXSTRING_BEG; case 'r': lex_strterm = NEW_STRTERM(str_regexp, term, paren); return tREGEXP_BEG; case 's': lex_strterm = NEW_STRTERM(str_ssym, term, paren); lex_state = EXPR_FNAME; return tSYMBEG; default: yyerror("unknown type of %string"); return 0; } } if ((c = nextc()) == '=') { set_yylval_id('%'); lex_state = EXPR_BEG; return tOP_ASGN; } if (IS_SPCARG(c)) { goto quotation; } lex_state = IS_AFTER_OPERATOR() ? EXPR_ARG : EXPR_BEG; pushback(c); warn_balanced("%%", "string literal"); return '%'; case '$': lex_state = EXPR_END; newtok(); c = nextc(); switch (c) { case '_': /* $_: last read line string */ c = nextc(); if (parser_is_identchar()) { tokadd('$'); tokadd('_'); break; } pushback(c); c = '_'; /* fall through */ case '~': /* $~: match-data */ case '*': /* $*: argv */ case '$': /* $$: pid */ case '?': /* $?: last status */ case '!': /* $!: error string */ case '@': /* $@: error position */ case '/': /* $/: input record separator */ case '\\': /* $\: output record separator */ case ';': /* $;: field separator */ case ',': /* $,: output field separator */ case '.': /* $.: last read line number */ case '=': /* $=: ignorecase */ case ':': /* $:: load path */ case '<': /* $<: reading filename */ case '>': /* $>: default output handle */ case '\"': /* $": already loaded files */ tokadd('$'); tokadd(c); goto gvar; case '-': tokadd('$'); tokadd(c); c = nextc(); if (parser_is_identchar()) { if (tokadd_mbchar(c) == -1) return 0; } else { pushback(c); pushback('-'); return '$'; } gvar: set_yylval_name(rb_intern3(tok(), tokidx, current_enc)); return tGVAR; case '&': /* $&: last match */ case '`': /* $`: string before last match */ case '\'': /* $': string after last match */ case '+': /* $+: string matches last paren. */ if (IS_lex_state_for(last_state, EXPR_FNAME)) { tokadd('$'); tokadd(c); goto gvar; } set_yylval_node(NEW_BACK_REF(c)); return tBACK_REF; case '1': case '2': case '3': case '4': case '5': case '6': case '7': case '8': case '9': tokadd('$'); do { tokadd(c); c = nextc(); } while (c != -1 && ISDIGIT(c)); pushback(c); if (IS_lex_state_for(last_state, EXPR_FNAME)) goto gvar; tokfix(); set_yylval_node(NEW_NTH_REF(atoi(tok()+1))); return tNTH_REF; default: if (!parser_is_identchar()) { pushback(c); compile_error(PARSER_ARG "`$%c' is not allowed as a global variable name", c); return 0; } case '0': tokadd('$'); } break; case '@': c = nextc(); newtok(); tokadd('@'); if (c == '@') { tokadd('@'); c = nextc(); } if (c != -1 && (ISDIGIT(c) || !parser_is_identchar())) { pushback(c); if (tokidx == 1) { compile_error(PARSER_ARG "`@%c' is not allowed as an instance variable name", c); } else { compile_error(PARSER_ARG "`@@%c' is not allowed as a class variable name", c); } return 0; } break; case '_': if (was_bol() && whole_match_p("__END__", 7, 0)) { ruby__end__seen = 1; parser->eofp = Qtrue; #ifndef RIPPER return -1; #else lex_goto_eol(parser); ripper_dispatch_scan_event(parser, k__END__); return 0; #endif } newtok(); break; default: if (!parser_is_identchar()) { rb_compile_error(PARSER_ARG "Invalid char `\\x%02X' in expression", c); goto retry; } newtok(); break; } mb = ENC_CODERANGE_7BIT; do { if (!ISASCII(c)) mb = ENC_CODERANGE_UNKNOWN; if (tokadd_mbchar(c) == -1) return 0; c = nextc(); } while (parser_is_identchar()); switch (tok()[0]) { case '@': case '$': pushback(c); break; default: if ((c == '!' || c == '?') && !peek('=')) { tokadd(c); } else { pushback(c); } } tokfix(); { int result = 0; last_state = lex_state; switch (tok()[0]) { case '$': lex_state = EXPR_END; result = tGVAR; break; case '@': lex_state = EXPR_END; if (tok()[1] == '@') result = tCVAR; else result = tIVAR; break; default: if (toklast() == '!' || toklast() == '?') { result = tFID; } else { if (IS_lex_state(EXPR_FNAME)) { if ((c = nextc()) == '=' && !peek('~') && !peek('>') && (!peek('=') || (peek_n('>', 1)))) { result = tIDENTIFIER; tokadd(c); tokfix(); } else { pushback(c); } } if (result == 0 && ISUPPER(tok()[0])) { result = tCONSTANT; } else { result = tIDENTIFIER; } } if (IS_LABEL_POSSIBLE()) { if (IS_LABEL_SUFFIX(0)) { lex_state = EXPR_BEG; nextc(); set_yylval_name(TOK_INTERN(!ENC_SINGLE(mb))); return tLABEL; } } if (mb == ENC_CODERANGE_7BIT && !IS_lex_state(EXPR_DOT)) { const struct kwtable *kw; /* See if it is a reserved word. */ kw = rb_reserved_word(tok(), toklen()); if (kw) { enum lex_state_e state = lex_state; lex_state = kw->state; if (IS_lex_state_for(state, EXPR_FNAME)) { set_yylval_name(rb_intern(kw->name)); return kw->id[0]; } if (IS_lex_state(EXPR_BEG)) { command_start = TRUE; } if (kw->id[0] == keyword_do) { if (lpar_beg && lpar_beg == paren_nest) { lpar_beg = 0; --paren_nest; return keyword_do_LAMBDA; } if (COND_P()) return keyword_do_cond; if (CMDARG_P() && !IS_lex_state_for(state, EXPR_CMDARG)) return keyword_do_block; if (IS_lex_state_for(state, (EXPR_BEG | EXPR_ENDARG))) return keyword_do_block; return keyword_do; } if (IS_lex_state_for(state, (EXPR_BEG | EXPR_VALUE))) return kw->id[0]; else { if (kw->id[0] != kw->id[1]) lex_state = EXPR_BEG; return kw->id[1]; } } } if (IS_lex_state(EXPR_BEG_ANY | EXPR_ARG_ANY | EXPR_DOT)) { if (cmd_state) { lex_state = EXPR_CMDARG; } else { lex_state = EXPR_ARG; } } else if (lex_state == EXPR_FNAME) { lex_state = EXPR_ENDFN; } else { lex_state = EXPR_END; } } { ID ident = TOK_INTERN(!ENC_SINGLE(mb)); set_yylval_name(ident); if (!IS_lex_state_for(last_state, EXPR_DOT|EXPR_FNAME) && is_local_id(ident) && lvar_defined(ident)) { lex_state = EXPR_END; } } return result; } } #if YYPURE static int yylex(void *lval, void *p) #else yylex(void *p) #endif { struct parser_params *parser = (struct parser_params*)p; int t; #if YYPURE parser->parser_yylval = lval; parser->parser_yylval->val = Qundef; #endif t = parser_yylex(parser); #ifdef RIPPER if (!NIL_P(parser->delayed)) { ripper_dispatch_delayed_token(parser, t); return t; } if (t != 0) ripper_dispatch_scan_event(parser, t); #endif return t; } #ifndef RIPPER static NODE* node_newnode(struct parser_params *parser, enum node_type type, VALUE a0, VALUE a1, VALUE a2) { NODE *n = (rb_node_newnode)(type, a0, a1, a2); nd_set_line(n, ruby_sourceline); return n; } static enum node_type nodetype(NODE *node) /* for debug */ { return (enum node_type)nd_type(node); } static int nodeline(NODE *node) { return nd_line(node); } static NODE* newline_node(NODE *node) { if (node) { node = remove_begin(node); node->flags |= NODE_FL_NEWLINE; } return node; } static void fixpos(NODE *node, NODE *orig) { if (!node) return; if (!orig) return; if (orig == (NODE*)1) return; nd_set_line(node, nd_line(orig)); } static void parser_warning(struct parser_params *parser, NODE *node, const char *mesg) { rb_compile_warning(ruby_sourcefile, nd_line(node), "%s", mesg); } #define parser_warning(node, mesg) parser_warning(parser, (node), (mesg)) static void parser_warn(struct parser_params *parser, NODE *node, const char *mesg) { rb_compile_warn(ruby_sourcefile, nd_line(node), "%s", mesg); } #define parser_warn(node, mesg) parser_warn(parser, (node), (mesg)) static NODE* block_append_gen(struct parser_params *parser, NODE *head, NODE *tail) { NODE *end, *h = head, *nd; if (tail == 0) return head; if (h == 0) return tail; switch (nd_type(h)) { case NODE_LIT: case NODE_STR: case NODE_SELF: case NODE_TRUE: case NODE_FALSE: case NODE_NIL: parser_warning(h, "unused literal ignored"); return tail; default: h = end = NEW_BLOCK(head); end->nd_end = end; fixpos(end, head); head = end; break; case NODE_BLOCK: end = h->nd_end; break; } nd = end->nd_head; switch (nd_type(nd)) { case NODE_RETURN: case NODE_BREAK: case NODE_NEXT: case NODE_REDO: case NODE_RETRY: if (RTEST(ruby_verbose)) { parser_warning(tail, "statement not reached"); } break; default: break; } if (nd_type(tail) != NODE_BLOCK) { tail = NEW_BLOCK(tail); tail->nd_end = tail; } end->nd_next = tail; h->nd_end = tail->nd_end; return head; } /* append item to the list */ static NODE* list_append_gen(struct parser_params *parser, NODE *list, NODE *item) { NODE *last; if (list == 0) return NEW_LIST(item); if (list->nd_next) { last = list->nd_next->nd_end; } else { last = list; } list->nd_alen += 1; last->nd_next = NEW_LIST(item); list->nd_next->nd_end = last->nd_next; return list; } /* concat two lists */ static NODE* list_concat_gen(struct parser_params *parser, NODE *head, NODE *tail) { NODE *last; if (head->nd_next) { last = head->nd_next->nd_end; } else { last = head; } head->nd_alen += tail->nd_alen; last->nd_next = tail; if (tail->nd_next) { head->nd_next->nd_end = tail->nd_next->nd_end; } else { head->nd_next->nd_end = tail; } return head; } static int literal_concat0(struct parser_params *parser, VALUE head, VALUE tail) { if (NIL_P(tail)) return 1; if (!rb_enc_compatible(head, tail)) { compile_error(PARSER_ARG "string literal encodings differ (%s / %s)", rb_enc_name(rb_enc_get(head)), rb_enc_name(rb_enc_get(tail))); rb_str_resize(head, 0); rb_str_resize(tail, 0); return 0; } rb_str_buf_append(head, tail); return 1; } /* concat two string literals */ static NODE * literal_concat_gen(struct parser_params *parser, NODE *head, NODE *tail) { enum node_type htype; NODE *headlast; VALUE lit; if (!head) return tail; if (!tail) return head; htype = nd_type(head); if (htype == NODE_EVSTR) { NODE *node = NEW_DSTR(Qnil); head = list_append(node, head); htype = NODE_DSTR; } switch (nd_type(tail)) { case NODE_STR: if (htype == NODE_DSTR && (headlast = head->nd_next->nd_end->nd_head) && nd_type(headlast) == NODE_STR) { htype = NODE_STR; lit = headlast->nd_lit; } else { lit = head->nd_lit; } if (htype == NODE_STR) { if (!literal_concat0(parser, lit, tail->nd_lit)) { error: rb_gc_force_recycle((VALUE)head); rb_gc_force_recycle((VALUE)tail); return 0; } rb_gc_force_recycle((VALUE)tail); } else { list_append(head, tail); } break; case NODE_DSTR: if (htype == NODE_STR) { if (!literal_concat0(parser, head->nd_lit, tail->nd_lit)) goto error; tail->nd_lit = head->nd_lit; rb_gc_force_recycle((VALUE)head); head = tail; } else if (NIL_P(tail->nd_lit)) { append: head->nd_alen += tail->nd_alen - 1; head->nd_next->nd_end->nd_next = tail->nd_next; head->nd_next->nd_end = tail->nd_next->nd_end; rb_gc_force_recycle((VALUE)tail); } else if (htype == NODE_DSTR && (headlast = head->nd_next->nd_end->nd_head) && nd_type(headlast) == NODE_STR) { lit = headlast->nd_lit; if (!literal_concat0(parser, lit, tail->nd_lit)) goto error; tail->nd_lit = Qnil; goto append; } else { nd_set_type(tail, NODE_ARRAY); tail->nd_head = NEW_STR(tail->nd_lit); list_concat(head, tail); } break; case NODE_EVSTR: if (htype == NODE_STR) { nd_set_type(head, NODE_DSTR); head->nd_alen = 1; } list_append(head, tail); break; } return head; } static NODE * evstr2dstr_gen(struct parser_params *parser, NODE *node) { if (nd_type(node) == NODE_EVSTR) { node = list_append(NEW_DSTR(Qnil), node); } return node; } static NODE * str_suffix_gen(struct parser_params *parser, NODE *node, long opt) { if (nd_type(node) == NODE_STR) { #if STR_OPTION_BINARY if (opt & STR_OPTION_BINARY) { rb_enc_associate_index(node->nd_lit, ENCINDEX_ASCII); } #endif #if STR_OPTION_FROZEN if (opt & STR_OPTION_FROZEN) { OBJ_FREEZE(node->nd_lit); node->nd_lit = rb_fstring(node->nd_lit); nd_set_type(node, NODE_LIT); } #endif } else { #if STR_OPTION_BINARY if (opt & STR_OPTION_BINARY) { VALUE ascii8bit = rb_enc_from_encoding(rb_ascii8bit_encoding()); NODE *args = NEW_LIST(NEW_LIT(ascii8bit)); node = NEW_CALL(node, rb_intern("force_encoding"), args); } #endif #if STR_OPTION_FROZEN if (opt & STR_OPTION_FROZEN) { node = NEW_CALL(node, rb_intern("freeze"), 0); } #endif } return node; } static NODE * new_evstr_gen(struct parser_params *parser, NODE *node) { NODE *head = node; if (node) { switch (nd_type(node)) { case NODE_STR: case NODE_DSTR: case NODE_EVSTR: return node; } } return NEW_EVSTR(head); } static NODE * call_bin_op_gen(struct parser_params *parser, NODE *recv, ID id, NODE *arg1) { value_expr(recv); value_expr(arg1); return NEW_CALL(recv, id, NEW_LIST(arg1)); } static NODE * call_uni_op_gen(struct parser_params *parser, NODE *recv, ID id) { value_expr(recv); return NEW_CALL(recv, id, 0); } static NODE* match_op_gen(struct parser_params *parser, NODE *node1, NODE *node2) { value_expr(node1); value_expr(node2); if (node1) { switch (nd_type(node1)) { case NODE_DREGX: case NODE_DREGX_ONCE: return NEW_MATCH2(node1, node2); case NODE_LIT: if (RB_TYPE_P(node1->nd_lit, T_REGEXP)) { return NEW_MATCH2(node1, node2); } } } if (node2) { switch (nd_type(node2)) { case NODE_DREGX: case NODE_DREGX_ONCE: return NEW_MATCH3(node2, node1); case NODE_LIT: if (RB_TYPE_P(node2->nd_lit, T_REGEXP)) { return NEW_MATCH3(node2, node1); } } } return NEW_CALL(node1, tMATCH, NEW_LIST(node2)); } static NODE* gettable_gen(struct parser_params *parser, ID id) { switch (id) { case keyword_self: return NEW_SELF(); case keyword_nil: return NEW_NIL(); case keyword_true: return NEW_TRUE(); case keyword_false: return NEW_FALSE(); case keyword__FILE__: return NEW_STR(rb_str_dup(ruby_sourcefile_string)); case keyword__LINE__: return NEW_LIT(INT2FIX(tokline)); case keyword__ENCODING__: return NEW_LIT(rb_enc_from_encoding(current_enc)); } switch (id_type(id)) { case ID_LOCAL: if (dyna_in_block() && dvar_defined(id)) return NEW_DVAR(id); if (local_id(id)) return NEW_LVAR(id); /* method call without arguments */ return NEW_VCALL(id); case ID_GLOBAL: return NEW_GVAR(id); case ID_INSTANCE: return NEW_IVAR(id); case ID_CONST: return NEW_CONST(id); case ID_CLASS: return NEW_CVAR(id); } compile_error(PARSER_ARG "identifier %s is not valid to get", rb_id2name(id)); return 0; } #else /* !RIPPER */ static int id_is_var_gen(struct parser_params *parser, ID id) { if (is_notop_id(id)) { switch (id & ID_SCOPE_MASK) { case ID_GLOBAL: case ID_INSTANCE: case ID_CONST: case ID_CLASS: return 1; case ID_LOCAL: if (dyna_in_block() && dvar_defined(id)) return 1; if (local_id(id)) return 1; /* method call without arguments */ return 0; } } compile_error(PARSER_ARG "identifier %s is not valid to get", rb_id2name(id)); return 0; } #endif /* !RIPPER */ #if PARSER_DEBUG static const char * lex_state_name(enum lex_state_e state) { static const char names[][12] = { "EXPR_BEG", "EXPR_END", "EXPR_ENDARG", "EXPR_ENDFN", "EXPR_ARG", "EXPR_CMDARG", "EXPR_MID", "EXPR_FNAME", "EXPR_DOT", "EXPR_CLASS", "EXPR_VALUE", }; if ((unsigned)state & ~(~0u << EXPR_MAX_STATE)) return names[ffs(state)]; return NULL; } #endif #ifdef RIPPER static VALUE assignable_gen(struct parser_params *parser, VALUE lhs) #else static NODE* assignable_gen(struct parser_params *parser, ID id, NODE *val) #endif { #ifdef RIPPER ID id = get_id(lhs); # define assignable_result(x) get_value(lhs) # define parser_yyerror(parser, x) dispatch1(assign_error, lhs) #else # define assignable_result(x) (x) #endif if (!id) return assignable_result(0); switch (id) { case keyword_self: yyerror("Can't change the value of self"); goto error; case keyword_nil: yyerror("Can't assign to nil"); goto error; case keyword_true: yyerror("Can't assign to true"); goto error; case keyword_false: yyerror("Can't assign to false"); goto error; case keyword__FILE__: yyerror("Can't assign to __FILE__"); goto error; case keyword__LINE__: yyerror("Can't assign to __LINE__"); goto error; case keyword__ENCODING__: yyerror("Can't assign to __ENCODING__"); goto error; } switch (id_type(id)) { case ID_LOCAL: if (dyna_in_block()) { if (dvar_curr(id)) { return assignable_result(NEW_DASGN_CURR(id, val)); } else if (dvar_defined(id)) { return assignable_result(NEW_DASGN(id, val)); } else if (local_id(id)) { return assignable_result(NEW_LASGN(id, val)); } else { dyna_var(id); return assignable_result(NEW_DASGN_CURR(id, val)); } } else { if (!local_id(id)) { local_var(id); } return assignable_result(NEW_LASGN(id, val)); } break; case ID_GLOBAL: return assignable_result(NEW_GASGN(id, val)); case ID_INSTANCE: return assignable_result(NEW_IASGN(id, val)); case ID_CONST: if (!in_def && !in_single) return assignable_result(NEW_CDECL(id, val, 0)); yyerror("dynamic constant assignment"); break; case ID_CLASS: return assignable_result(NEW_CVASGN(id, val)); default: compile_error(PARSER_ARG "identifier %s is not valid to set", rb_id2name(id)); } error: return assignable_result(0); #undef assignable_result #undef parser_yyerror } static int is_private_local_id(ID name) { VALUE s; if (name == idUScore) return 1; if (!is_local_id(name)) return 0; s = rb_id2str(name); if (!s) return 0; return RSTRING_PTR(s)[0] == '_'; } #define LVAR_USED ((ID)1 << (sizeof(ID) * CHAR_BIT - 1)) static ID shadowing_lvar_gen(struct parser_params *parser, ID name) { if (is_private_local_id(name)) return name; if (dyna_in_block()) { if (dvar_curr(name)) { yyerror("duplicated argument name"); } else if (dvar_defined_get(name) || local_id(name)) { rb_warningS("shadowing outer local variable - %s", rb_id2name(name)); vtable_add(lvtbl->vars, name); if (lvtbl->used) { vtable_add(lvtbl->used, (ID)ruby_sourceline | LVAR_USED); } } } else { if (local_id(name)) { yyerror("duplicated argument name"); } } return name; } static void new_bv_gen(struct parser_params *parser, ID name) { if (!name) return; if (!is_local_id(name)) { compile_error(PARSER_ARG "invalid local variable - %s", rb_id2name(name)); return; } shadowing_lvar(name); dyna_var(name); } #ifndef RIPPER static NODE * aryset_gen(struct parser_params *parser, NODE *recv, NODE *idx) { if (recv && nd_type(recv) == NODE_SELF) recv = (NODE *)1; return NEW_ATTRASGN(recv, tASET, idx); } static void block_dup_check_gen(struct parser_params *parser, NODE *node1, NODE *node2) { if (node2 && node1 && nd_type(node1) == NODE_BLOCK_PASS) { compile_error(PARSER_ARG "both block arg and actual block given"); } } static const char id_type_names[][9] = { "LOCAL", "INSTANCE", "", /* INSTANCE2 */ "GLOBAL", "ATTRSET", "CONST", "CLASS", "JUNK", }; ID rb_id_attrset(ID id) { if (!is_notop_id(id)) { rb_bug("rb_id_attrset: operator ID - %"PRIdVALUE, (VALUE)id); } else { int scope = (int)(id & ID_SCOPE_MASK); switch (scope) { case ID_LOCAL: case ID_INSTANCE: case ID_GLOBAL: case ID_CONST: case ID_CLASS: case ID_JUNK: break; default: rb_bug("rb_id_attrset: %s ID - %"PRIdVALUE, id_type_names[scope], (VALUE)id); } } id &= ~ID_SCOPE_MASK; id |= ID_ATTRSET; return id; } static NODE * attrset_gen(struct parser_params *parser, NODE *recv, ID id) { if (recv && nd_type(recv) == NODE_SELF) recv = (NODE *)1; return NEW_ATTRASGN(recv, rb_id_attrset(id), 0); } static void rb_backref_error_gen(struct parser_params *parser, NODE *node) { switch (nd_type(node)) { case NODE_NTH_REF: compile_error(PARSER_ARG "Can't set variable $%ld", node->nd_nth); break; case NODE_BACK_REF: compile_error(PARSER_ARG "Can't set variable $%c", (int)node->nd_nth); break; } } static NODE * arg_concat_gen(struct parser_params *parser, NODE *node1, NODE *node2) { if (!node2) return node1; switch (nd_type(node1)) { case NODE_BLOCK_PASS: if (node1->nd_head) node1->nd_head = arg_concat(node1->nd_head, node2); else node1->nd_head = NEW_LIST(node2); return node1; case NODE_ARGSPUSH: if (nd_type(node2) != NODE_ARRAY) break; node1->nd_body = list_concat(NEW_LIST(node1->nd_body), node2); nd_set_type(node1, NODE_ARGSCAT); return node1; case NODE_ARGSCAT: if (nd_type(node2) != NODE_ARRAY || nd_type(node1->nd_body) != NODE_ARRAY) break; node1->nd_body = list_concat(node1->nd_body, node2); return node1; } return NEW_ARGSCAT(node1, node2); } static NODE * arg_append_gen(struct parser_params *parser, NODE *node1, NODE *node2) { if (!node1) return NEW_LIST(node2); switch (nd_type(node1)) { case NODE_ARRAY: return list_append(node1, node2); case NODE_BLOCK_PASS: node1->nd_head = arg_append(node1->nd_head, node2); return node1; case NODE_ARGSPUSH: node1->nd_body = list_append(NEW_LIST(node1->nd_body), node2); nd_set_type(node1, NODE_ARGSCAT); return node1; } return NEW_ARGSPUSH(node1, node2); } static NODE * splat_array(NODE* node) { if (nd_type(node) == NODE_SPLAT) node = node->nd_head; if (nd_type(node) == NODE_ARRAY) return node; return 0; } static NODE * node_assign_gen(struct parser_params *parser, NODE *lhs, NODE *rhs) { if (!lhs) return 0; switch (nd_type(lhs)) { case NODE_GASGN: case NODE_IASGN: case NODE_IASGN2: case NODE_LASGN: case NODE_DASGN: case NODE_DASGN_CURR: case NODE_MASGN: case NODE_CDECL: case NODE_CVASGN: lhs->nd_value = rhs; break; case NODE_ATTRASGN: case NODE_CALL: lhs->nd_args = arg_append(lhs->nd_args, rhs); break; default: /* should not happen */ break; } return lhs; } static int value_expr_gen(struct parser_params *parser, NODE *node) { int cond = 0; if (!node) { rb_warning0("empty expression"); } while (node) { switch (nd_type(node)) { case NODE_RETURN: case NODE_BREAK: case NODE_NEXT: case NODE_REDO: case NODE_RETRY: if (!cond) yyerror("void value expression"); /* or "control never reach"? */ return FALSE; case NODE_BLOCK: while (node->nd_next) { node = node->nd_next; } node = node->nd_head; break; case NODE_BEGIN: node = node->nd_body; break; case NODE_IF: if (!node->nd_body) { node = node->nd_else; break; } else if (!node->nd_else) { node = node->nd_body; break; } if (!value_expr(node->nd_body)) return FALSE; node = node->nd_else; break; case NODE_AND: case NODE_OR: cond = 1; node = node->nd_2nd; break; default: return TRUE; } } return TRUE; } static void void_expr_gen(struct parser_params *parser, NODE *node) { const char *useless = 0; if (!RTEST(ruby_verbose)) return; if (!node) return; switch (nd_type(node)) { case NODE_CALL: switch (node->nd_mid) { case '+': case '-': case '*': case '/': case '%': case tPOW: case tUPLUS: case tUMINUS: case '|': case '^': case '&': case tCMP: case '>': case tGEQ: case '<': case tLEQ: case tEQ: case tNEQ: useless = rb_id2name(node->nd_mid); break; } break; case NODE_LVAR: case NODE_DVAR: case NODE_GVAR: case NODE_IVAR: case NODE_CVAR: case NODE_NTH_REF: case NODE_BACK_REF: useless = "a variable"; break; case NODE_CONST: useless = "a constant"; break; case NODE_LIT: case NODE_STR: case NODE_DSTR: case NODE_DREGX: case NODE_DREGX_ONCE: useless = "a literal"; break; case NODE_COLON2: case NODE_COLON3: useless = "::"; break; case NODE_DOT2: useless = ".."; break; case NODE_DOT3: useless = "..."; break; case NODE_SELF: useless = "self"; break; case NODE_NIL: useless = "nil"; break; case NODE_TRUE: useless = "true"; break; case NODE_FALSE: useless = "false"; break; case NODE_DEFINED: useless = "defined?"; break; } if (useless) { int line = ruby_sourceline; ruby_sourceline = nd_line(node); rb_warnS("possibly useless use of %s in void context", useless); ruby_sourceline = line; } } static void void_stmts_gen(struct parser_params *parser, NODE *node) { if (!RTEST(ruby_verbose)) return; if (!node) return; if (nd_type(node) != NODE_BLOCK) return; for (;;) { if (!node->nd_next) return; void_expr0(node->nd_head); node = node->nd_next; } } static NODE * remove_begin(NODE *node) { NODE **n = &node, *n1 = node; while (n1 && nd_type(n1) == NODE_BEGIN && n1->nd_body) { *n = n1 = n1->nd_body; } return node; } static NODE * remove_begin_all(NODE *node) { NODE **n = &node, *n1 = node; while (n1 && nd_type(n1) == NODE_BEGIN) { *n = n1 = n1->nd_body; } return node; } static void reduce_nodes_gen(struct parser_params *parser, NODE **body) { NODE *node = *body; if (!node) { *body = NEW_NIL(); return; } #define subnodes(n1, n2) \ ((!node->n1) ? (node->n2 ? (body = &node->n2, 1) : 0) : \ (!node->n2) ? (body = &node->n1, 1) : \ (reduce_nodes(&node->n1), body = &node->n2, 1)) while (node) { int newline = (int)(node->flags & NODE_FL_NEWLINE); switch (nd_type(node)) { end: case NODE_NIL: *body = 0; return; case NODE_RETURN: *body = node = node->nd_stts; if (newline && node) node->flags |= NODE_FL_NEWLINE; continue; case NODE_BEGIN: *body = node = node->nd_body; if (newline && node) node->flags |= NODE_FL_NEWLINE; continue; case NODE_BLOCK: body = &node->nd_end->nd_head; break; case NODE_IF: if (subnodes(nd_body, nd_else)) break; return; case NODE_CASE: body = &node->nd_body; break; case NODE_WHEN: if (!subnodes(nd_body, nd_next)) goto end; break; case NODE_ENSURE: if (!subnodes(nd_head, nd_resq)) goto end; break; case NODE_RESCUE: if (node->nd_else) { body = &node->nd_resq; break; } if (!subnodes(nd_head, nd_resq)) goto end; break; default: return; } node = *body; if (newline && node) node->flags |= NODE_FL_NEWLINE; } #undef subnodes } static int is_static_content(NODE *node) { if (!node) return 1; switch (nd_type(node)) { case NODE_HASH: if (!(node = node->nd_head)) break; case NODE_ARRAY: do { if (!is_static_content(node->nd_head)) return 0; } while ((node = node->nd_next) != 0); case NODE_LIT: case NODE_STR: case NODE_NIL: case NODE_TRUE: case NODE_FALSE: case NODE_ZARRAY: break; default: return 0; } return 1; } static int assign_in_cond(struct parser_params *parser, NODE *node) { switch (nd_type(node)) { case NODE_MASGN: yyerror("multiple assignment in conditional"); return 1; case NODE_LASGN: case NODE_DASGN: case NODE_DASGN_CURR: case NODE_GASGN: case NODE_IASGN: break; default: return 0; } if (!node->nd_value) return 1; if (is_static_content(node->nd_value)) { /* reports always */ parser_warn(node->nd_value, "found = in conditional, should be =="); } return 1; } static void warn_unless_e_option(struct parser_params *parser, NODE *node, const char *str) { if (!e_option_supplied(parser)) parser_warn(node, str); } static void warning_unless_e_option(struct parser_params *parser, NODE *node, const char *str) { if (!e_option_supplied(parser)) parser_warning(node, str); } static void fixup_nodes(NODE **rootnode) { NODE *node, *next, *head; for (node = *rootnode; node; node = next) { enum node_type type; VALUE val; next = node->nd_next; head = node->nd_head; rb_gc_force_recycle((VALUE)node); *rootnode = next; switch (type = nd_type(head)) { case NODE_DOT2: case NODE_DOT3: val = rb_range_new(head->nd_beg->nd_lit, head->nd_end->nd_lit, type == NODE_DOT3); rb_gc_force_recycle((VALUE)head->nd_beg); rb_gc_force_recycle((VALUE)head->nd_end); nd_set_type(head, NODE_LIT); head->nd_lit = val; break; default: break; } } } static NODE *cond0(struct parser_params*,NODE*); static NODE* range_op(struct parser_params *parser, NODE *node) { enum node_type type; if (node == 0) return 0; type = nd_type(node); value_expr(node); if (type == NODE_LIT && FIXNUM_P(node->nd_lit)) { warn_unless_e_option(parser, node, "integer literal in conditional range"); return NEW_CALL(node, tEQ, NEW_LIST(NEW_GVAR(rb_intern("$.")))); } return cond0(parser, node); } static int literal_node(NODE *node) { if (!node) return 1; /* same as NODE_NIL */ switch (nd_type(node)) { case NODE_LIT: case NODE_STR: case NODE_DSTR: case NODE_EVSTR: case NODE_DREGX: case NODE_DREGX_ONCE: case NODE_DSYM: return 2; case NODE_TRUE: case NODE_FALSE: case NODE_NIL: return 1; } return 0; } static NODE* cond0(struct parser_params *parser, NODE *node) { if (node == 0) return 0; assign_in_cond(parser, node); switch (nd_type(node)) { case NODE_DSTR: case NODE_EVSTR: case NODE_STR: rb_warn0("string literal in condition"); break; case NODE_DREGX: case NODE_DREGX_ONCE: warning_unless_e_option(parser, node, "regex literal in condition"); return NEW_MATCH2(node, NEW_GVAR(rb_intern("$_"))); case NODE_AND: case NODE_OR: node->nd_1st = cond0(parser, node->nd_1st); node->nd_2nd = cond0(parser, node->nd_2nd); break; case NODE_DOT2: case NODE_DOT3: node->nd_beg = range_op(parser, node->nd_beg); node->nd_end = range_op(parser, node->nd_end); if (nd_type(node) == NODE_DOT2) nd_set_type(node,NODE_FLIP2); else if (nd_type(node) == NODE_DOT3) nd_set_type(node, NODE_FLIP3); if (!e_option_supplied(parser)) { int b = literal_node(node->nd_beg); int e = literal_node(node->nd_end); if ((b == 1 && e == 1) || (b + e >= 2 && RTEST(ruby_verbose))) { parser_warn(node, "range literal in condition"); } } break; case NODE_DSYM: parser_warning(node, "literal in condition"); break; case NODE_LIT: if (RB_TYPE_P(node->nd_lit, T_REGEXP)) { warn_unless_e_option(parser, node, "regex literal in condition"); nd_set_type(node, NODE_MATCH); } else { parser_warning(node, "literal in condition"); } default: break; } return node; } static NODE* cond_gen(struct parser_params *parser, NODE *node) { if (node == 0) return 0; return cond0(parser, node); } static NODE* logop_gen(struct parser_params *parser, enum node_type type, NODE *left, NODE *right) { value_expr(left); if (left && (enum node_type)nd_type(left) == type) { NODE *node = left, *second; while ((second = node->nd_2nd) != 0 && (enum node_type)nd_type(second) == type) { node = second; } node->nd_2nd = NEW_NODE(type, second, right, 0); return left; } return NEW_NODE(type, left, right, 0); } static void no_blockarg(struct parser_params *parser, NODE *node) { if (node && nd_type(node) == NODE_BLOCK_PASS) { compile_error(PARSER_ARG "block argument should not be given"); } } static NODE * ret_args_gen(struct parser_params *parser, NODE *node) { if (node) { no_blockarg(parser, node); if (nd_type(node) == NODE_ARRAY) { if (node->nd_next == 0) { node = node->nd_head; } else { nd_set_type(node, NODE_VALUES); } } } return node; } static NODE * new_yield_gen(struct parser_params *parser, NODE *node) { if (node) no_blockarg(parser, node); return NEW_YIELD(node); } static NODE* negate_lit(NODE *node) { switch (TYPE(node->nd_lit)) { case T_FIXNUM: node->nd_lit = LONG2FIX(-FIX2LONG(node->nd_lit)); break; case T_BIGNUM: case T_RATIONAL: case T_COMPLEX: node->nd_lit = rb_funcall(node->nd_lit,tUMINUS,0,0); break; case T_FLOAT: #if USE_FLONUM if (FLONUM_P(node->nd_lit)) { node->nd_lit = DBL2NUM(-RFLOAT_VALUE(node->nd_lit)); } else { RFLOAT(node->nd_lit)->float_value = -RFLOAT_VALUE(node->nd_lit); } #else RFLOAT(node->nd_lit)->float_value = -RFLOAT_VALUE(node->nd_lit); #endif break; default: rb_bug("unknown literal type passed to negate_lit"); break; } return node; } static NODE * arg_blk_pass(NODE *node1, NODE *node2) { if (node2) { node2->nd_head = node1; return node2; } return node1; } static NODE* new_args_gen(struct parser_params *parser, NODE *m, NODE *o, ID r, NODE *p, NODE *tail) { int saved_line = ruby_sourceline; struct rb_args_info *args = tail->nd_ainfo; args->pre_args_num = m ? rb_long2int(m->nd_plen) : 0; args->pre_init = m ? m->nd_next : 0; args->post_args_num = p ? rb_long2int(p->nd_plen) : 0; args->post_init = p ? p->nd_next : 0; args->first_post_arg = p ? p->nd_pid : 0; args->rest_arg = r; args->opt_args = o; ruby_sourceline = saved_line; return tail; } static NODE* new_args_tail_gen(struct parser_params *parser, NODE *k, ID kr, ID b) { int saved_line = ruby_sourceline; struct rb_args_info *args; NODE *kw_rest_arg = 0; NODE *node; args = ALLOC(struct rb_args_info); MEMZERO(args, struct rb_args_info, 1); node = NEW_NODE(NODE_ARGS, 0, 0, args); args->block_arg = b; args->kw_args = k; if (k && !kr) kr = internal_id(); if (kr) { arg_var(kr); kw_rest_arg = NEW_DVAR(kr); } args->kw_rest_arg = kw_rest_arg; ruby_sourceline = saved_line; return node; } static NODE* dsym_node_gen(struct parser_params *parser, NODE *node) { VALUE lit; if (!node) { return NEW_LIT(ID2SYM(idNULL)); } switch (nd_type(node)) { case NODE_DSTR: nd_set_type(node, NODE_DSYM); break; case NODE_STR: lit = node->nd_lit; node->nd_lit = ID2SYM(rb_intern_str(lit)); nd_set_type(node, NODE_LIT); break; default: node = NEW_NODE(NODE_DSYM, Qnil, 1, NEW_LIST(node)); break; } return node; } #endif /* !RIPPER */ #ifndef RIPPER static NODE * new_op_assign_gen(struct parser_params *parser, NODE *lhs, ID op, NODE *rhs) { NODE *asgn; if (lhs) { ID vid = lhs->nd_vid; if (op == tOROP) { lhs->nd_value = rhs; asgn = NEW_OP_ASGN_OR(gettable(vid), lhs); if (is_asgn_or_id(vid)) { asgn->nd_aid = vid; } } else if (op == tANDOP) { lhs->nd_value = rhs; asgn = NEW_OP_ASGN_AND(gettable(vid), lhs); } else { asgn = lhs; asgn->nd_value = NEW_CALL(gettable(vid), op, NEW_LIST(rhs)); } } else { asgn = NEW_BEGIN(0); } return asgn; } static NODE * new_attr_op_assign_gen(struct parser_params *parser, NODE *lhs, ID attr, ID op, NODE *rhs) { NODE *asgn; if (op == tOROP) { op = 0; } else if (op == tANDOP) { op = 1; } asgn = NEW_OP_ASGN2(lhs, attr, op, rhs); fixpos(asgn, lhs); return asgn; } static NODE * new_const_op_assign_gen(struct parser_params *parser, NODE *lhs, ID op, NODE *rhs) { NODE *asgn; if (op == tOROP) { op = 0; } else if (op == tANDOP) { op = 1; } if (lhs) { asgn = NEW_OP_CDECL(lhs, op, rhs); } else { asgn = NEW_BEGIN(0); } fixpos(asgn, lhs); return asgn; } #else static VALUE new_op_assign_gen(struct parser_params *parser, VALUE lhs, VALUE op, VALUE rhs) { return dispatch3(opassign, lhs, op, rhs); } static VALUE new_attr_op_assign_gen(struct parser_params *parser, VALUE lhs, VALUE type, VALUE attr, VALUE op, VALUE rhs) { VALUE recv = dispatch3(field, lhs, type, attr); return dispatch3(opassign, recv, op, rhs); } #endif static void warn_unused_var(struct parser_params *parser, struct local_vars *local) { int i, cnt; ID *v, *u; if (!local->used) return; v = local->vars->tbl; u = local->used->tbl; cnt = local->used->pos; if (cnt != local->vars->pos) { rb_bug("local->used->pos != local->vars->pos"); } for (i = 0; i < cnt; ++i) { if (!v[i] || (u[i] & LVAR_USED)) continue; if (is_private_local_id(v[i])) continue; rb_warn4S(ruby_sourcefile, (int)u[i], "assigned but unused variable - %s", rb_id2name(v[i])); } } static void local_push_gen(struct parser_params *parser, int inherit_dvars) { struct local_vars *local; local = ALLOC(struct local_vars); local->prev = lvtbl; local->args = vtable_alloc(0); local->vars = vtable_alloc(inherit_dvars ? DVARS_INHERIT : DVARS_TOPSCOPE); local->used = !(inherit_dvars && (ifndef_ripper(compile_for_eval || e_option_supplied(parser))+0)) && RTEST(ruby_verbose) ? vtable_alloc(0) : 0; lvtbl = local; } static void local_pop_gen(struct parser_params *parser) { struct local_vars *local = lvtbl->prev; if (lvtbl->used) { warn_unused_var(parser, lvtbl); vtable_free(lvtbl->used); } vtable_free(lvtbl->args); vtable_free(lvtbl->vars); xfree(lvtbl); lvtbl = local; } #ifndef RIPPER static ID* vtable_tblcpy(ID *buf, const struct vtable *src) { int i, cnt = vtable_size(src); if (cnt > 0) { buf[0] = cnt; for (i = 0; i < cnt; i++) { buf[i] = src->tbl[i]; } return buf; } return 0; } static ID* local_tbl_gen(struct parser_params *parser) { int cnt = vtable_size(lvtbl->args) + vtable_size(lvtbl->vars); ID *buf; if (cnt <= 0) return 0; buf = ALLOC_N(ID, cnt + 1); vtable_tblcpy(buf+1, lvtbl->args); vtable_tblcpy(buf+vtable_size(lvtbl->args)+1, lvtbl->vars); buf[0] = cnt; return buf; } #endif static int arg_var_gen(struct parser_params *parser, ID id) { vtable_add(lvtbl->args, id); return vtable_size(lvtbl->args) - 1; } static int local_var_gen(struct parser_params *parser, ID id) { vtable_add(lvtbl->vars, id); if (lvtbl->used) { vtable_add(lvtbl->used, (ID)ruby_sourceline); } return vtable_size(lvtbl->vars) - 1; } static int local_id_gen(struct parser_params *parser, ID id) { struct vtable *vars, *args, *used; vars = lvtbl->vars; args = lvtbl->args; used = lvtbl->used; while (vars && POINTER_P(vars->prev)) { vars = vars->prev; args = args->prev; if (used) used = used->prev; } if (vars && vars->prev == DVARS_INHERIT) { return rb_local_defined(id); } else if (vtable_included(args, id)) { return 1; } else { int i = vtable_included(vars, id); if (i && used) used->tbl[i-1] |= LVAR_USED; return i != 0; } } static const struct vtable * dyna_push_gen(struct parser_params *parser) { lvtbl->args = vtable_alloc(lvtbl->args); lvtbl->vars = vtable_alloc(lvtbl->vars); if (lvtbl->used) { lvtbl->used = vtable_alloc(lvtbl->used); } return lvtbl->args; } static void dyna_pop_1(struct parser_params *parser) { struct vtable *tmp; if ((tmp = lvtbl->used) != 0) { warn_unused_var(parser, lvtbl); lvtbl->used = lvtbl->used->prev; vtable_free(tmp); } tmp = lvtbl->args; lvtbl->args = lvtbl->args->prev; vtable_free(tmp); tmp = lvtbl->vars; lvtbl->vars = lvtbl->vars->prev; vtable_free(tmp); } static void dyna_pop_gen(struct parser_params *parser, const struct vtable *lvargs) { while (lvtbl->args != lvargs) { dyna_pop_1(parser); if (!lvtbl->args) { struct local_vars *local = lvtbl->prev; xfree(lvtbl); lvtbl = local; } } dyna_pop_1(parser); } static int dyna_in_block_gen(struct parser_params *parser) { return POINTER_P(lvtbl->vars) && lvtbl->vars->prev != DVARS_TOPSCOPE; } static int dvar_defined_gen(struct parser_params *parser, ID id, int get) { struct vtable *vars, *args, *used; int i; args = lvtbl->args; vars = lvtbl->vars; used = lvtbl->used; while (POINTER_P(vars)) { if (vtable_included(args, id)) { return 1; } if ((i = vtable_included(vars, id)) != 0) { if (used) used->tbl[i-1] |= LVAR_USED; return 1; } args = args->prev; vars = vars->prev; if (get) used = 0; if (used) used = used->prev; } if (vars == DVARS_INHERIT) { return rb_dvar_defined(id); } return 0; } static int dvar_curr_gen(struct parser_params *parser, ID id) { return (vtable_included(lvtbl->args, id) || vtable_included(lvtbl->vars, id)); } #ifndef RIPPER static void reg_fragment_setenc_gen(struct parser_params* parser, VALUE str, int options) { int c = RE_OPTION_ENCODING_IDX(options); if (c) { int opt, idx; rb_char_to_option_kcode(c, &opt, &idx); if (idx != ENCODING_GET(str) && rb_enc_str_coderange(str) != ENC_CODERANGE_7BIT) { goto error; } ENCODING_SET(str, idx); } else if (RE_OPTION_ENCODING_NONE(options)) { if (!ENCODING_IS_ASCII8BIT(str) && rb_enc_str_coderange(str) != ENC_CODERANGE_7BIT) { c = 'n'; goto error; } rb_enc_associate(str, rb_ascii8bit_encoding()); } else if (current_enc == rb_usascii_encoding()) { if (rb_enc_str_coderange(str) != ENC_CODERANGE_7BIT) { /* raise in re.c */ rb_enc_associate(str, rb_usascii_encoding()); } else { rb_enc_associate(str, rb_ascii8bit_encoding()); } } return; error: compile_error(PARSER_ARG "regexp encoding option '%c' differs from source encoding '%s'", c, rb_enc_name(rb_enc_get(str))); } static int reg_fragment_check_gen(struct parser_params* parser, VALUE str, int options) { VALUE err; reg_fragment_setenc(str, options); err = rb_reg_check_preprocess(str); if (err != Qnil) { err = rb_obj_as_string(err); compile_error(PARSER_ARG "%"PRIsVALUE, err); return 0; } return 1; } typedef struct { struct parser_params* parser; rb_encoding *enc; NODE *succ_block; NODE *fail_block; int num; } reg_named_capture_assign_t; static int reg_named_capture_assign_iter(const OnigUChar *name, const OnigUChar *name_end, int back_num, int *back_refs, OnigRegex regex, void *arg0) { reg_named_capture_assign_t *arg = (reg_named_capture_assign_t*)arg0; struct parser_params* parser = arg->parser; rb_encoding *enc = arg->enc; long len = name_end - name; const char *s = (const char *)name; ID var; arg->num++; if (arg->succ_block == 0) { arg->succ_block = NEW_BEGIN(0); arg->fail_block = NEW_BEGIN(0); } if (!len || (*name != '_' && ISASCII(*name) && !rb_enc_islower(*name, enc)) || (len < MAX_WORD_LENGTH && rb_reserved_word(s, (int)len)) || !rb_enc_symname2_p(s, len, enc)) { return ST_CONTINUE; } var = rb_intern3(s, len, enc); if (dvar_defined(var) || local_id(var)) { rb_warningS("named capture conflicts a local variable - %s", rb_id2name(var)); } arg->succ_block = block_append(arg->succ_block, newline_node(node_assign(assignable(var,0), NEW_CALL( gettable(rb_intern("$~")), idAREF, NEW_LIST(NEW_LIT(ID2SYM(var)))) ))); arg->fail_block = block_append(arg->fail_block, newline_node(node_assign(assignable(var,0), NEW_LIT(Qnil)))); return ST_CONTINUE; } static NODE * reg_named_capture_assign_gen(struct parser_params* parser, VALUE regexp, NODE *match) { reg_named_capture_assign_t arg; arg.parser = parser; arg.enc = rb_enc_get(regexp); arg.succ_block = 0; arg.fail_block = 0; arg.num = 0; onig_foreach_name(RREGEXP(regexp)->ptr, reg_named_capture_assign_iter, (void*)&arg); if (arg.num == 0) return match; return block_append( newline_node(match), NEW_IF(gettable(rb_intern("$~")), block_append( newline_node(arg.succ_block), newline_node( NEW_CALL( gettable(rb_intern("$~")), rb_intern("begin"), NEW_LIST(NEW_LIT(INT2FIX(0)))))), block_append( newline_node(arg.fail_block), newline_node( NEW_LIT(Qnil))))); } static VALUE reg_compile_gen(struct parser_params* parser, VALUE str, int options) { VALUE re; VALUE err; reg_fragment_setenc(str, options); err = rb_errinfo(); re = rb_reg_compile(str, options & RE_OPTION_MASK, ruby_sourcefile, ruby_sourceline); if (NIL_P(re)) { ID mesg = rb_intern("mesg"); VALUE m = rb_attr_get(rb_errinfo(), mesg); rb_set_errinfo(err); if (!NIL_P(err)) { rb_str_append(rb_str_cat(rb_attr_get(err, mesg), "\n", 1), m); } else { compile_error(PARSER_ARG "%"PRIsVALUE, m); } return Qnil; } return re; } void rb_gc_mark_parser(void) { } NODE* rb_parser_append_print(VALUE vparser, NODE *node) { NODE *prelude = 0; NODE *scope = node; struct parser_params *parser; if (!node) return node; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); node = node->nd_body; if (nd_type(node) == NODE_PRELUDE) { prelude = node; node = node->nd_body; } node = block_append(node, NEW_FCALL(rb_intern("print"), NEW_ARRAY(NEW_GVAR(rb_intern("$_"))))); if (prelude) { prelude->nd_body = node; scope->nd_body = prelude; } else { scope->nd_body = node; } return scope; } NODE * rb_parser_while_loop(VALUE vparser, NODE *node, int chop, int split) { NODE *prelude = 0; NODE *scope = node; struct parser_params *parser; if (!node) return node; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); node = node->nd_body; if (nd_type(node) == NODE_PRELUDE) { prelude = node; node = node->nd_body; } if (split) { node = block_append(NEW_GASGN(rb_intern("$F"), NEW_CALL(NEW_GVAR(rb_intern("$_")), rb_intern("split"), 0)), node); } if (chop) { node = block_append(NEW_CALL(NEW_GVAR(rb_intern("$_")), rb_intern("chop!"), 0), node); } node = NEW_OPT_N(node); if (prelude) { prelude->nd_body = node; scope->nd_body = prelude; } else { scope->nd_body = node; } return scope; } static const struct { ID token; const char *name; } op_tbl[] = { {tDOT2, ".."}, {tDOT3, "..."}, {'+', "+(binary)"}, {'-', "-(binary)"}, {tPOW, "**"}, {tDSTAR, "**"}, {tUPLUS, "+@"}, {tUMINUS, "-@"}, {tCMP, "<=>"}, {tGEQ, ">="}, {tLEQ, "<="}, {tEQ, "=="}, {tEQQ, "==="}, {tNEQ, "!="}, {tMATCH, "=~"}, {tNMATCH, "!~"}, {tAREF, "[]"}, {tASET, "[]="}, {tLSHFT, "<<"}, {tRSHFT, ">>"}, {tCOLON2, "::"}, }; #define op_tbl_count numberof(op_tbl) #ifndef ENABLE_SELECTOR_NAMESPACE #define ENABLE_SELECTOR_NAMESPACE 0 #endif static struct symbols { ID last_id; st_table *sym_id; st_table *id_str; #if ENABLE_SELECTOR_NAMESPACE st_table *ivar2_id; st_table *id_ivar2; #endif VALUE op_sym[tLAST_OP_ID]; } global_symbols = {tLAST_TOKEN}; static const struct st_hash_type symhash = { rb_str_hash_cmp, rb_str_hash, }; #if ENABLE_SELECTOR_NAMESPACE struct ivar2_key { ID id; VALUE klass; }; static int ivar2_cmp(struct ivar2_key *key1, struct ivar2_key *key2) { if (key1->id == key2->id && key1->klass == key2->klass) { return 0; } return 1; } static int ivar2_hash(struct ivar2_key *key) { return (key->id << 8) ^ (key->klass >> 2); } static const struct st_hash_type ivar2_hash_type = { ivar2_cmp, ivar2_hash, }; #endif void Init_sym(void) { global_symbols.sym_id = st_init_table_with_size(&symhash, 1000); global_symbols.id_str = st_init_numtable_with_size(1000); #if ENABLE_SELECTOR_NAMESPACE global_symbols.ivar2_id = st_init_table_with_size(&ivar2_hash_type, 1000); global_symbols.id_ivar2 = st_init_numtable_with_size(1000); #endif (void)nodetype; (void)nodeline; #if PARSER_DEBUG (void)lex_state_name(-1); #endif Init_id(); } void rb_gc_mark_symbols(void) { rb_mark_tbl(global_symbols.id_str); rb_gc_mark_locations(global_symbols.op_sym, global_symbols.op_sym + numberof(global_symbols.op_sym)); } #endif /* !RIPPER */ static ID internal_id_gen(struct parser_params *parser) { ID id = (ID)vtable_size(lvtbl->args) + (ID)vtable_size(lvtbl->vars); id += ((tLAST_TOKEN - ID_INTERNAL) >> ID_SCOPE_SHIFT) + 1; return ID_INTERNAL | (id << ID_SCOPE_SHIFT); } #ifndef RIPPER static int is_special_global_name(const char *m, const char *e, rb_encoding *enc) { int mb = 0; if (m >= e) return 0; if (is_global_name_punct(*m)) { ++m; } else if (*m == '-') { if (++m >= e) return 0; if (is_identchar(m, e, enc)) { if (!ISASCII(*m)) mb = 1; m += rb_enc_mbclen(m, e, enc); } } else { if (!rb_enc_isdigit(*m, enc)) return 0; do { if (!ISASCII(*m)) mb = 1; ++m; } while (m < e && rb_enc_isdigit(*m, enc)); } return m == e ? mb + 1 : 0; } int rb_symname_p(const char *name) { return rb_enc_symname_p(name, rb_ascii8bit_encoding()); } int rb_enc_symname_p(const char *name, rb_encoding *enc) { return rb_enc_symname2_p(name, strlen(name), enc); } #define IDSET_ATTRSET_FOR_SYNTAX ((1U<') ++m; break; default: break; } break; case '>': switch (*++m) { case '>': case '=': ++m; break; } break; case '=': switch (*++m) { case '~': ++m; break; case '=': if (*++m == '=') ++m; break; default: return -1; } break; case '*': if (*++m == '*') ++m; break; case '+': case '-': if (*++m == '@') ++m; break; case '|': case '^': case '&': case '/': case '%': case '~': case '`': ++m; break; case '[': if (*++m != ']') return -1; if (*++m == '=') ++m; break; case '!': if (len == 1) return ID_JUNK; switch (*++m) { case '=': case '~': ++m; break; default: return -1; } break; default: type = rb_enc_isupper(*m, enc) ? ID_CONST : ID_LOCAL; id: if (m >= e || (*m != '_' && !rb_enc_isalpha(*m, enc) && ISASCII(*m))) return -1; while (m < e && is_identchar(m, e, enc)) m += rb_enc_mbclen(m, e, enc); switch (*m) { case '!': case '?': if (type == ID_GLOBAL || type == ID_CLASS || type == ID_INSTANCE) return -1; type = ID_JUNK; ++m; break; case '=': if (!(allowed_attrset & (1U << type))) return -1; type = ID_ATTRSET; ++m; break; } break; } return m == e ? type : -1; } int rb_enc_symname2_p(const char *name, long len, rb_encoding *enc) { return rb_enc_symname_type(name, len, enc, IDSET_ATTRSET_FOR_SYNTAX) != -1; } static int rb_str_symname_type(VALUE name, unsigned int allowed_attrset) { const char *ptr = StringValuePtr(name); long len = RSTRING_LEN(name); int type = rb_enc_symname_type(ptr, len, rb_enc_get(name), allowed_attrset); RB_GC_GUARD(name); return type; } static ID register_symid(ID id, const char *name, long len, rb_encoding *enc) { VALUE str = rb_enc_str_new(name, len, enc); return register_symid_str(id, str); } static ID register_symid_str(ID id, VALUE str) { OBJ_FREEZE(str); if (RUBY_DTRACE_SYMBOL_CREATE_ENABLED()) { RUBY_DTRACE_SYMBOL_CREATE(RSTRING_PTR(str), rb_sourcefile(), rb_sourceline()); } st_add_direct(global_symbols.sym_id, (st_data_t)str, id); st_add_direct(global_symbols.id_str, id, (st_data_t)str); return id; } static int sym_check_asciionly(VALUE str) { if (!rb_enc_asciicompat(rb_enc_get(str))) return FALSE; switch (rb_enc_str_coderange(str)) { case ENC_CODERANGE_BROKEN: rb_raise(rb_eEncodingError, "invalid encoding symbol"); case ENC_CODERANGE_7BIT: return TRUE; } return FALSE; } /* * _str_ itself will be registered at the global symbol table. _str_ * can be modified before the registration, since the encoding will be * set to ASCII-8BIT if it is a special global name. */ static ID intern_str(VALUE str); static VALUE setup_fake_str(struct RString *fake_str, const char *name, long len) { fake_str->basic.flags = T_STRING|RSTRING_NOEMBED; RBASIC_SET_CLASS((VALUE)fake_str, rb_cString); fake_str->as.heap.len = len; fake_str->as.heap.ptr = (char *)name; fake_str->as.heap.aux.capa = len; return (VALUE)fake_str; } ID rb_intern3(const char *name, long len, rb_encoding *enc) { st_data_t data; struct RString fake_str; VALUE str = setup_fake_str(&fake_str, name, len); rb_enc_associate(str, enc); OBJ_FREEZE(str); if (st_lookup(global_symbols.sym_id, str, &data)) return (ID)data; str = rb_enc_str_new(name, len, enc); /* make true string */ return intern_str(str); } static ID intern_str(VALUE str) { const char *name, *m, *e; long len, last; rb_encoding *enc, *symenc; unsigned char c; ID id; int mb; RSTRING_GETMEM(str, name, len); m = name; e = m + len; enc = rb_enc_get(str); symenc = enc; if (rb_cString && !rb_enc_asciicompat(enc)) { id = ID_JUNK; goto new_id; } last = len-1; id = 0; switch (*m) { case '$': id |= ID_GLOBAL; if ((mb = is_special_global_name(++m, e, enc)) != 0) { if (!--mb) symenc = rb_usascii_encoding(); goto new_id; } break; case '@': if (m[1] == '@') { m++; id |= ID_CLASS; } else { id |= ID_INSTANCE; } m++; break; default: c = m[0]; if (c != '_' && rb_enc_isascii(c, enc) && rb_enc_ispunct(c, enc)) { /* operators */ int i; if (len == 1) { id = c; goto id_register; } for (i = 0; i < op_tbl_count; i++) { if (*op_tbl[i].name == *m && strcmp(op_tbl[i].name, m) == 0) { id = op_tbl[i].token; goto id_register; } } } break; } if (name[last] == '=') { /* attribute assignment */ id = rb_intern3(name, last, enc); if (id > tLAST_OP_ID && !is_attrset_id(id)) { enc = rb_enc_get(rb_id2str(id)); id = rb_id_attrset(id); goto id_register; } id = ID_ATTRSET; } else if (id == 0) { if (rb_enc_isupper(m[0], enc)) { id = ID_CONST; } else { id = ID_LOCAL; } } if (!rb_enc_isdigit(*m, enc)) { while (m <= name + last && is_identchar(m, e, enc)) { if (ISASCII(*m)) { m++; } else { m += rb_enc_mbclen(m, e, enc); } } } if (id != ID_ATTRSET && m - name < len) id = ID_JUNK; if (sym_check_asciionly(str)) symenc = rb_usascii_encoding(); new_id: if (symenc != enc) rb_enc_associate(str, symenc); if (global_symbols.last_id >= ~(ID)0 >> (ID_SCOPE_SHIFT+RUBY_SPECIAL_SHIFT)) { if (len > 20) { rb_raise(rb_eRuntimeError, "symbol table overflow (symbol %.20s...)", name); } else { rb_raise(rb_eRuntimeError, "symbol table overflow (symbol %.*s)", (int)len, name); } } id |= ++global_symbols.last_id << ID_SCOPE_SHIFT; id_register: return register_symid_str(id, str); } ID rb_intern2(const char *name, long len) { return rb_intern3(name, len, rb_usascii_encoding()); } #undef rb_intern ID rb_intern(const char *name) { return rb_intern2(name, strlen(name)); } ID rb_intern_str(VALUE str) { st_data_t id; if (st_lookup(global_symbols.sym_id, str, &id)) return (ID)id; return intern_str(rb_str_dup(str)); } VALUE rb_id2str(ID id) { st_data_t data; if (id < tLAST_TOKEN) { int i = 0; if (id < INT_MAX && rb_ispunct((int)id)) { VALUE str = global_symbols.op_sym[i = (int)id]; if (!str) { char name[2]; name[0] = (char)id; name[1] = 0; str = rb_usascii_str_new(name, 1); OBJ_FREEZE(str); global_symbols.op_sym[i] = str; } return str; } for (i = 0; i < op_tbl_count; i++) { if (op_tbl[i].token == id) { VALUE str = global_symbols.op_sym[i]; if (!str) { str = rb_usascii_str_new2(op_tbl[i].name); OBJ_FREEZE(str); global_symbols.op_sym[i] = str; } return str; } } } if (st_lookup(global_symbols.id_str, id, &data)) { VALUE str = (VALUE)data; if (RBASIC(str)->klass == 0) RBASIC_SET_CLASS_RAW(str, rb_cString); return str; } if (is_attrset_id(id)) { ID id_stem = (id & ~ID_SCOPE_MASK); VALUE str; do { if (!!(str = rb_id2str(id_stem | ID_LOCAL))) break; if (!!(str = rb_id2str(id_stem | ID_CONST))) break; if (!!(str = rb_id2str(id_stem | ID_INSTANCE))) break; if (!!(str = rb_id2str(id_stem | ID_GLOBAL))) break; if (!!(str = rb_id2str(id_stem | ID_CLASS))) break; if (!!(str = rb_id2str(id_stem | ID_JUNK))) break; return 0; } while (0); str = rb_str_dup(str); rb_str_cat(str, "=", 1); register_symid_str(id, str); if (st_lookup(global_symbols.id_str, id, &data)) { VALUE str = (VALUE)data; if (RBASIC(str)->klass == 0) RBASIC_SET_CLASS_RAW(str, rb_cString); return str; } } return 0; } const char * rb_id2name(ID id) { VALUE str = rb_id2str(id); if (!str) return 0; return RSTRING_PTR(str); } static int symbols_i(VALUE sym, ID value, VALUE ary) { rb_ary_push(ary, ID2SYM(value)); return ST_CONTINUE; } /* * call-seq: * Symbol.all_symbols => array * * Returns an array of all the symbols currently in Ruby's symbol * table. * * Symbol.all_symbols.size #=> 903 * Symbol.all_symbols[1,20] #=> [:floor, :ARGV, :Binding, :symlink, * :chown, :EOFError, :$;, :String, * :LOCK_SH, :"setuid?", :$<, * :default_proc, :compact, :extend, * :Tms, :getwd, :$=, :ThreadGroup, * :wait2, :$>] */ VALUE rb_sym_all_symbols(void) { VALUE ary = rb_ary_new2(global_symbols.sym_id->num_entries); st_foreach(global_symbols.sym_id, symbols_i, ary); return ary; } int rb_is_const_id(ID id) { return is_const_id(id); } int rb_is_class_id(ID id) { return is_class_id(id); } int rb_is_global_id(ID id) { return is_global_id(id); } int rb_is_instance_id(ID id) { return is_instance_id(id); } int rb_is_attrset_id(ID id) { return is_attrset_id(id); } int rb_is_local_id(ID id) { return is_local_id(id); } int rb_is_junk_id(ID id) { return is_junk_id(id); } /** * Returns ID for the given name if it is interned already, or 0. * * \param namep the pointer to the name object * \return the ID for *namep * \pre the object referred by \p namep must be a Symbol or * a String, or possible to convert with to_str method. * \post the object referred by \p namep is a Symbol or a * String if non-zero value is returned, or is a String * if 0 is returned. */ ID rb_check_id(volatile VALUE *namep) { st_data_t id; VALUE tmp; VALUE name = *namep; if (SYMBOL_P(name)) { return SYM2ID(name); } else if (!RB_TYPE_P(name, T_STRING)) { tmp = rb_check_string_type(name); if (NIL_P(tmp)) { tmp = rb_inspect(name); rb_raise(rb_eTypeError, "%s is not a symbol", RSTRING_PTR(tmp)); } name = tmp; *namep = name; } sym_check_asciionly(name); if (st_lookup(global_symbols.sym_id, (st_data_t)name, &id)) return (ID)id; if (rb_is_attrset_name(name)) { struct RString fake_str; /* make local name by chopping '=' */ const VALUE localname = setup_fake_str(&fake_str, RSTRING_PTR(name), RSTRING_LEN(name) - 1); rb_enc_copy(localname, name); OBJ_FREEZE(localname); if (st_lookup(global_symbols.sym_id, (st_data_t)localname, &id)) { return rb_id_attrset((ID)id); } RB_GC_GUARD(name); } return (ID)0; } ID rb_check_id_cstr(const char *ptr, long len, rb_encoding *enc) { st_data_t id; struct RString fake_str; const VALUE name = setup_fake_str(&fake_str, ptr, len); rb_enc_associate(name, enc); sym_check_asciionly(name); if (st_lookup(global_symbols.sym_id, (st_data_t)name, &id)) return (ID)id; if (rb_is_attrset_name(name)) { fake_str.as.heap.len = len - 1; if (st_lookup(global_symbols.sym_id, (st_data_t)name, &id)) { return rb_id_attrset((ID)id); } } return (ID)0; } int rb_is_const_name(VALUE name) { return rb_str_symname_type(name, 0) == ID_CONST; } int rb_is_class_name(VALUE name) { return rb_str_symname_type(name, 0) == ID_CLASS; } int rb_is_global_name(VALUE name) { return rb_str_symname_type(name, 0) == ID_GLOBAL; } int rb_is_instance_name(VALUE name) { return rb_str_symname_type(name, 0) == ID_INSTANCE; } int rb_is_attrset_name(VALUE name) { return rb_str_symname_type(name, IDSET_ATTRSET_FOR_INTERN) == ID_ATTRSET; } int rb_is_local_name(VALUE name) { return rb_str_symname_type(name, 0) == ID_LOCAL; } int rb_is_method_name(VALUE name) { switch (rb_str_symname_type(name, 0)) { case ID_LOCAL: case ID_ATTRSET: case ID_JUNK: return TRUE; } return FALSE; } int rb_is_junk_name(VALUE name) { return rb_str_symname_type(name, IDSET_ATTRSET_FOR_SYNTAX) == -1; } #endif /* !RIPPER */ static void parser_initialize(struct parser_params *parser) { parser->eofp = Qfalse; parser->parser_lex_strterm = 0; parser->parser_cond_stack = 0; parser->parser_cmdarg_stack = 0; parser->parser_class_nest = 0; parser->parser_paren_nest = 0; parser->parser_lpar_beg = 0; parser->parser_brace_nest = 0; parser->parser_in_single = 0; parser->parser_in_def = 0; parser->parser_in_defined = 0; parser->parser_compile_for_eval = 0; parser->parser_cur_mid = 0; parser->parser_tokenbuf = NULL; parser->parser_tokidx = 0; parser->parser_toksiz = 0; parser->parser_heredoc_end = 0; parser->parser_command_start = TRUE; parser->parser_deferred_nodes = 0; parser->parser_lex_pbeg = 0; parser->parser_lex_p = 0; parser->parser_lex_pend = 0; parser->parser_lvtbl = 0; parser->parser_ruby__end__seen = 0; parser->parser_ruby_sourcefile = 0; parser->parser_ruby_sourcefile_string = Qnil; #ifndef RIPPER parser->is_ripper = 0; parser->parser_eval_tree_begin = 0; parser->parser_eval_tree = 0; #else parser->is_ripper = 1; parser->delayed = Qnil; parser->result = Qnil; parser->parsing_thread = Qnil; parser->toplevel_p = TRUE; #endif #ifdef YYMALLOC parser->heap = NULL; #endif parser->enc = rb_utf8_encoding(); } #ifdef RIPPER #define parser_mark ripper_parser_mark #define parser_free ripper_parser_free #endif static void parser_mark(void *ptr) { struct parser_params *p = (struct parser_params*)ptr; rb_gc_mark((VALUE)p->parser_lex_strterm); rb_gc_mark((VALUE)p->parser_deferred_nodes); rb_gc_mark(p->parser_lex_input); rb_gc_mark(p->parser_lex_lastline); rb_gc_mark(p->parser_lex_nextline); rb_gc_mark(p->parser_ruby_sourcefile_string); #ifndef RIPPER rb_gc_mark((VALUE)p->parser_eval_tree_begin) ; rb_gc_mark((VALUE)p->parser_eval_tree) ; rb_gc_mark(p->debug_lines); #else rb_gc_mark(p->delayed); rb_gc_mark(p->value); rb_gc_mark(p->result); rb_gc_mark(p->parsing_thread); #endif #ifdef YYMALLOC rb_gc_mark((VALUE)p->heap); #endif } static void parser_free(void *ptr) { struct parser_params *p = (struct parser_params*)ptr; struct local_vars *local, *prev; if (p->parser_tokenbuf) { xfree(p->parser_tokenbuf); } for (local = p->parser_lvtbl; local; local = prev) { if (local->vars) xfree(local->vars); prev = local->prev; xfree(local); } xfree(p); } static size_t parser_memsize(const void *ptr) { struct parser_params *p = (struct parser_params*)ptr; struct local_vars *local; size_t size = sizeof(*p); if (!ptr) return 0; size += p->parser_toksiz; for (local = p->parser_lvtbl; local; local = local->prev) { size += sizeof(*local); if (local->vars) size += local->vars->capa * sizeof(ID); } return size; } static #ifndef RIPPER const #endif rb_data_type_t parser_data_type = { "parser", { parser_mark, parser_free, parser_memsize, }, }; #ifndef RIPPER #undef rb_reserved_word const struct kwtable * rb_reserved_word(const char *str, unsigned int len) { return reserved_word(str, len); } static struct parser_params * parser_new(void) { struct parser_params *p; p = ALLOC_N(struct parser_params, 1); MEMZERO(p, struct parser_params, 1); parser_initialize(p); return p; } VALUE rb_parser_new(void) { struct parser_params *p = parser_new(); return TypedData_Wrap_Struct(0, &parser_data_type, p); } /* * call-seq: * ripper#end_seen? -> Boolean * * Return true if parsed source ended by +\_\_END\_\_+. */ VALUE rb_parser_end_seen_p(VALUE vparser) { struct parser_params *parser; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); return ruby__end__seen ? Qtrue : Qfalse; } /* * call-seq: * ripper#encoding -> encoding * * Return encoding of the source. */ VALUE rb_parser_encoding(VALUE vparser) { struct parser_params *parser; TypedData_Get_Struct(vparser, struct parser_params, &parser_data_type, parser); return rb_enc_from_encoding(current_enc); } /* * call-seq: * ripper.yydebug -> true or false * * Get yydebug. */ VALUE rb_parser_get_yydebug(VALUE self) { struct parser_params *parser; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); return yydebug ? Qtrue : Qfalse; } /* * call-seq: * ripper.yydebug = flag * * Set yydebug. */ VALUE rb_parser_set_yydebug(VALUE self, VALUE flag) { struct parser_params *parser; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); yydebug = RTEST(flag); return flag; } #ifdef YYMALLOC #define HEAPCNT(n, size) ((n) * (size) / sizeof(YYSTYPE)) #define NEWHEAP() rb_node_newnode(NODE_ALLOCA, 0, (VALUE)parser->heap, 0) #define ADD2HEAP(n, c, p) ((parser->heap = (n))->u1.node = (p), \ (n)->u3.cnt = (c), (p)) void * rb_parser_malloc(struct parser_params *parser, size_t size) { size_t cnt = HEAPCNT(1, size); NODE *n = NEWHEAP(); void *ptr = xmalloc(size); return ADD2HEAP(n, cnt, ptr); } void * rb_parser_calloc(struct parser_params *parser, size_t nelem, size_t size) { size_t cnt = HEAPCNT(nelem, size); NODE *n = NEWHEAP(); void *ptr = xcalloc(nelem, size); return ADD2HEAP(n, cnt, ptr); } void * rb_parser_realloc(struct parser_params *parser, void *ptr, size_t size) { NODE *n; size_t cnt = HEAPCNT(1, size); if (ptr && (n = parser->heap) != NULL) { do { if (n->u1.node == ptr) { n->u1.node = ptr = xrealloc(ptr, size); if (n->u3.cnt) n->u3.cnt = cnt; return ptr; } } while ((n = n->u2.node) != NULL); } n = NEWHEAP(); ptr = xrealloc(ptr, size); return ADD2HEAP(n, cnt, ptr); } void rb_parser_free(struct parser_params *parser, void *ptr) { NODE **prev = &parser->heap, *n; while ((n = *prev) != NULL) { if (n->u1.node == ptr) { *prev = n->u2.node; rb_gc_force_recycle((VALUE)n); break; } prev = &n->u2.node; } xfree(ptr); } #endif #endif #ifdef RIPPER #ifdef RIPPER_DEBUG extern int rb_is_pointer_to_heap(VALUE); /* :nodoc: */ static VALUE ripper_validate_object(VALUE self, VALUE x) { if (x == Qfalse) return x; if (x == Qtrue) return x; if (x == Qnil) return x; if (x == Qundef) rb_raise(rb_eArgError, "Qundef given"); if (FIXNUM_P(x)) return x; if (SYMBOL_P(x)) return x; if (!rb_is_pointer_to_heap(x)) rb_raise(rb_eArgError, "invalid pointer: %p", x); switch (BUILTIN_TYPE(x)) { case T_STRING: case T_OBJECT: case T_ARRAY: case T_BIGNUM: case T_FLOAT: case T_COMPLEX: case T_RATIONAL: return x; case T_NODE: if (nd_type(x) != NODE_LASGN) { rb_raise(rb_eArgError, "NODE given: %p", x); } return ((NODE *)x)->nd_rval; default: rb_raise(rb_eArgError, "wrong type of ruby object: %p (%s)", x, rb_obj_classname(x)); } return x; } #endif #define validate(x) ((x) = get_value(x)) static VALUE ripper_dispatch0(struct parser_params *parser, ID mid) { return rb_funcall(parser->value, mid, 0); } static VALUE ripper_dispatch1(struct parser_params *parser, ID mid, VALUE a) { validate(a); return rb_funcall(parser->value, mid, 1, a); } static VALUE ripper_dispatch2(struct parser_params *parser, ID mid, VALUE a, VALUE b) { validate(a); validate(b); return rb_funcall(parser->value, mid, 2, a, b); } static VALUE ripper_dispatch3(struct parser_params *parser, ID mid, VALUE a, VALUE b, VALUE c) { validate(a); validate(b); validate(c); return rb_funcall(parser->value, mid, 3, a, b, c); } static VALUE ripper_dispatch4(struct parser_params *parser, ID mid, VALUE a, VALUE b, VALUE c, VALUE d) { validate(a); validate(b); validate(c); validate(d); return rb_funcall(parser->value, mid, 4, a, b, c, d); } static VALUE ripper_dispatch5(struct parser_params *parser, ID mid, VALUE a, VALUE b, VALUE c, VALUE d, VALUE e) { validate(a); validate(b); validate(c); validate(d); validate(e); return rb_funcall(parser->value, mid, 5, a, b, c, d, e); } static VALUE ripper_dispatch7(struct parser_params *parser, ID mid, VALUE a, VALUE b, VALUE c, VALUE d, VALUE e, VALUE f, VALUE g) { validate(a); validate(b); validate(c); validate(d); validate(e); validate(f); validate(g); return rb_funcall(parser->value, mid, 7, a, b, c, d, e, f, g); } static const struct kw_assoc { ID id; const char *name; } keyword_to_name[] = { {keyword_class, "class"}, {keyword_module, "module"}, {keyword_def, "def"}, {keyword_undef, "undef"}, {keyword_begin, "begin"}, {keyword_rescue, "rescue"}, {keyword_ensure, "ensure"}, {keyword_end, "end"}, {keyword_if, "if"}, {keyword_unless, "unless"}, {keyword_then, "then"}, {keyword_elsif, "elsif"}, {keyword_else, "else"}, {keyword_case, "case"}, {keyword_when, "when"}, {keyword_while, "while"}, {keyword_until, "until"}, {keyword_for, "for"}, {keyword_break, "break"}, {keyword_next, "next"}, {keyword_redo, "redo"}, {keyword_retry, "retry"}, {keyword_in, "in"}, {keyword_do, "do"}, {keyword_do_cond, "do"}, {keyword_do_block, "do"}, {keyword_return, "return"}, {keyword_yield, "yield"}, {keyword_super, "super"}, {keyword_self, "self"}, {keyword_nil, "nil"}, {keyword_true, "true"}, {keyword_false, "false"}, {keyword_and, "and"}, {keyword_or, "or"}, {keyword_not, "not"}, {modifier_if, "if"}, {modifier_unless, "unless"}, {modifier_while, "while"}, {modifier_until, "until"}, {modifier_rescue, "rescue"}, {keyword_alias, "alias"}, {keyword_defined, "defined?"}, {keyword_BEGIN, "BEGIN"}, {keyword_END, "END"}, {keyword__LINE__, "__LINE__"}, {keyword__FILE__, "__FILE__"}, {keyword__ENCODING__, "__ENCODING__"}, {0, NULL} }; static const char* keyword_id_to_str(ID id) { const struct kw_assoc *a; for (a = keyword_to_name; a->id; a++) { if (a->id == id) return a->name; } return NULL; } #undef ripper_id2sym static VALUE ripper_id2sym(ID id) { const char *name; char buf[8]; if (id <= 256) { buf[0] = (char)id; buf[1] = '\0'; return ID2SYM(rb_intern2(buf, 1)); } if ((name = keyword_id_to_str(id))) { return ID2SYM(rb_intern(name)); } switch (id) { case tOROP: name = "||"; break; case tANDOP: name = "&&"; break; default: name = rb_id2name(id); if (!name) { rb_bug("cannot convert ID to string: %ld", (unsigned long)id); } return ID2SYM(id); } return ID2SYM(rb_intern(name)); } static ID ripper_get_id(VALUE v) { NODE *nd; if (!RB_TYPE_P(v, T_NODE)) return 0; nd = (NODE *)v; if (nd_type(nd) != NODE_LASGN) return 0; return nd->nd_vid; } static VALUE ripper_get_value(VALUE v) { NODE *nd; if (v == Qundef) return Qnil; if (!RB_TYPE_P(v, T_NODE)) return v; nd = (NODE *)v; if (nd_type(nd) != NODE_LASGN) return Qnil; return nd->nd_rval; } static void ripper_compile_error(struct parser_params *parser, const char *fmt, ...) { VALUE str; va_list args; va_start(args, fmt); str = rb_vsprintf(fmt, args); va_end(args); rb_funcall(parser->value, rb_intern("compile_error"), 1, str); } static void ripper_warn0(struct parser_params *parser, const char *fmt) { rb_funcall(parser->value, rb_intern("warn"), 1, STR_NEW2(fmt)); } static void ripper_warnI(struct parser_params *parser, const char *fmt, int a) { rb_funcall(parser->value, rb_intern("warn"), 2, STR_NEW2(fmt), INT2NUM(a)); } static void ripper_warnS(struct parser_params *parser, const char *fmt, const char *str) { rb_funcall(parser->value, rb_intern("warn"), 2, STR_NEW2(fmt), STR_NEW2(str)); } static void ripper_warning0(struct parser_params *parser, const char *fmt) { rb_funcall(parser->value, rb_intern("warning"), 1, STR_NEW2(fmt)); } static void ripper_warningS(struct parser_params *parser, const char *fmt, const char *str) { rb_funcall(parser->value, rb_intern("warning"), 2, STR_NEW2(fmt), STR_NEW2(str)); } static VALUE ripper_lex_get_generic(struct parser_params *parser, VALUE src) { return rb_io_gets(src); } static VALUE ripper_s_allocate(VALUE klass) { struct parser_params *p; VALUE self; p = ALLOC_N(struct parser_params, 1); MEMZERO(p, struct parser_params, 1); self = TypedData_Wrap_Struct(klass, &parser_data_type, p); p->value = self; return self; } #define ripper_initialized_p(r) ((r)->parser_lex_input != 0) /* * call-seq: * Ripper.new(src, filename="(ripper)", lineno=1) -> ripper * * Create a new Ripper object. * _src_ must be a String, an IO, or an Object which has #gets method. * * This method does not starts parsing. * See also Ripper#parse and Ripper.parse. */ static VALUE ripper_initialize(int argc, VALUE *argv, VALUE self) { struct parser_params *parser; VALUE src, fname, lineno; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); rb_scan_args(argc, argv, "12", &src, &fname, &lineno); if (RB_TYPE_P(src, T_FILE)) { parser->parser_lex_gets = ripper_lex_get_generic; } else { StringValue(src); parser->parser_lex_gets = lex_get_str; } parser->parser_lex_input = src; parser->eofp = Qfalse; if (NIL_P(fname)) { fname = STR_NEW2("(ripper)"); } else { StringValue(fname); } parser_initialize(parser); parser->parser_ruby_sourcefile_string = fname; parser->parser_ruby_sourcefile = RSTRING_PTR(fname); parser->parser_ruby_sourceline = NIL_P(lineno) ? 0 : NUM2INT(lineno) - 1; return Qnil; } struct ripper_args { struct parser_params *parser; int argc; VALUE *argv; }; static VALUE ripper_parse0(VALUE parser_v) { struct parser_params *parser; TypedData_Get_Struct(parser_v, struct parser_params, &parser_data_type, parser); parser_prepare(parser); ripper_yyparse((void*)parser); return parser->result; } static VALUE ripper_ensure(VALUE parser_v) { struct parser_params *parser; TypedData_Get_Struct(parser_v, struct parser_params, &parser_data_type, parser); parser->parsing_thread = Qnil; return Qnil; } /* * call-seq: * ripper#parse * * Start parsing and returns the value of the root action. */ static VALUE ripper_parse(VALUE self) { struct parser_params *parser; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); if (!ripper_initialized_p(parser)) { rb_raise(rb_eArgError, "method called for uninitialized object"); } if (!NIL_P(parser->parsing_thread)) { if (parser->parsing_thread == rb_thread_current()) rb_raise(rb_eArgError, "Ripper#parse is not reentrant"); else rb_raise(rb_eArgError, "Ripper#parse is not multithread-safe"); } parser->parsing_thread = rb_thread_current(); rb_ensure(ripper_parse0, self, ripper_ensure, self); return parser->result; } /* * call-seq: * ripper#column -> Integer * * Return column number of current parsing line. * This number starts from 0. */ static VALUE ripper_column(VALUE self) { struct parser_params *parser; long col; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); if (!ripper_initialized_p(parser)) { rb_raise(rb_eArgError, "method called for uninitialized object"); } if (NIL_P(parser->parsing_thread)) return Qnil; col = parser->tokp - parser->parser_lex_pbeg; return LONG2NUM(col); } /* * call-seq: * ripper#filename -> String * * Return current parsing filename. */ static VALUE ripper_filename(VALUE self) { struct parser_params *parser; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); if (!ripper_initialized_p(parser)) { rb_raise(rb_eArgError, "method called for uninitialized object"); } return parser->parser_ruby_sourcefile_string; } /* * call-seq: * ripper#lineno -> Integer * * Return line number of current parsing line. * This number starts from 1. */ static VALUE ripper_lineno(VALUE self) { struct parser_params *parser; TypedData_Get_Struct(self, struct parser_params, &parser_data_type, parser); if (!ripper_initialized_p(parser)) { rb_raise(rb_eArgError, "method called for uninitialized object"); } if (NIL_P(parser->parsing_thread)) return Qnil; return INT2NUM(parser->parser_ruby_sourceline); } #ifdef RIPPER_DEBUG /* :nodoc: */ static VALUE ripper_assert_Qundef(VALUE self, VALUE obj, VALUE msg) { StringValue(msg); if (obj == Qundef) { rb_raise(rb_eArgError, "%"PRIsVALUE, msg); } return Qnil; } /* :nodoc: */ static VALUE ripper_value(VALUE self, VALUE obj) { return ULONG2NUM(obj); } #endif void Init_ripper(void) { parser_data_type.parent = RTYPEDDATA_TYPE(rb_parser_new()); ripper_init_eventids1(); ripper_init_eventids2(); /* ensure existing in symbol table */ (void)rb_intern("||"); (void)rb_intern("&&"); InitVM(ripper); } void InitVM_ripper(void) { VALUE Ripper; Ripper = rb_define_class("Ripper", rb_cObject); /* version of Ripper */ rb_define_const(Ripper, "Version", rb_usascii_str_new2(RIPPER_VERSION)); rb_define_alloc_func(Ripper, ripper_s_allocate); rb_define_method(Ripper, "initialize", ripper_initialize, -1); rb_define_method(Ripper, "parse", ripper_parse, 0); rb_define_method(Ripper, "column", ripper_column, 0); rb_define_method(Ripper, "filename", ripper_filename, 0); rb_define_method(Ripper, "lineno", ripper_lineno, 0); rb_define_method(Ripper, "end_seen?", rb_parser_end_seen_p, 0); rb_define_method(Ripper, "encoding", rb_parser_encoding, 0); rb_define_method(Ripper, "yydebug", rb_parser_get_yydebug, 0); rb_define_method(Ripper, "yydebug=", rb_parser_set_yydebug, 1); #ifdef RIPPER_DEBUG rb_define_method(rb_mKernel, "assert_Qundef", ripper_assert_Qundef, 2); rb_define_method(rb_mKernel, "rawVALUE", ripper_value, 1); rb_define_method(rb_mKernel, "validate_object", ripper_validate_object, 1); #endif ripper_init_eventids1_table(Ripper); ripper_init_eventids2_table(Ripper); # if 0 /* Hack to let RDoc document SCRIPT_LINES__ */ /* * When a Hash is assigned to +SCRIPT_LINES__+ the contents of files loaded * after the assignment will be added as an Array of lines with the file * name as the key. */ rb_define_global_const("SCRIPT_LINES__", Qnil); #endif } #endif /* RIPPER */