diff options
Diffstat (limited to 'array.c')
| -rw-r--r-- | array.c | 4035 |
1 files changed, 2475 insertions, 1560 deletions
@@ -3,29 +3,36 @@ array.c - $Author$ - $Date$ created at: Fri Aug 6 09:46:12 JST 1993 - Copyright (C) 1993-2003 Yukihiro Matsumoto + Copyright (C) 1993-2007 Yukihiro Matsumoto Copyright (C) 2000 Network Applied Communication Laboratory, Inc. Copyright (C) 2000 Information-technology Promotion Agency, Japan **********************************************************************/ -#include "ruby.h" -#include "util.h" -#include "st.h" +#include "ruby/ruby.h" +#include "ruby/util.h" +#include "ruby/st.h" +#include "ruby/encoding.h" +#include "internal.h" + +#ifndef ARRAY_DEBUG +# define NDEBUG +#endif +#include <assert.h> + +#define numberof(array) (int)(sizeof(array) / sizeof((array)[0])) VALUE rb_cArray; + static ID id_cmp; #define ARY_DEFAULT_SIZE 16 -#define ARY_MAX_SIZE (LONG_MAX / sizeof(VALUE)) +#define ARY_MAX_SIZE (LONG_MAX / (int)sizeof(VALUE)) void -rb_mem_clear(mem, size) - register VALUE *mem; - register long size; +rb_mem_clear(register VALUE *mem, register long size) { while (size--) { *mem++ = Qnil; @@ -33,134 +40,311 @@ rb_mem_clear(mem, size) } static inline void -memfill(mem, size, val) - register VALUE *mem; - register long size; - register VALUE val; +memfill(register VALUE *mem, register long size, register VALUE val) { while (size--) { *mem++ = val; } } -#define ARY_TMPLOCK FL_USER1 +# define ARY_SHARED_P(ary) \ + (assert(!FL_TEST((ary), ELTS_SHARED) || !FL_TEST((ary), RARRAY_EMBED_FLAG)), \ + FL_TEST((ary),ELTS_SHARED)!=0) +# define ARY_EMBED_P(ary) \ + (assert(!FL_TEST((ary), ELTS_SHARED) || !FL_TEST((ary), RARRAY_EMBED_FLAG)), \ + FL_TEST((ary), RARRAY_EMBED_FLAG)!=0) + +#define ARY_HEAP_PTR(a) (assert(!ARY_EMBED_P(a)), RARRAY(a)->as.heap.ptr) +#define ARY_HEAP_LEN(a) (assert(!ARY_EMBED_P(a)), RARRAY(a)->as.heap.len) +#define ARY_EMBED_PTR(a) (assert(ARY_EMBED_P(a)), RARRAY(a)->as.ary) +#define ARY_EMBED_LEN(a) \ + (assert(ARY_EMBED_P(a)), \ + (long)((RBASIC(a)->flags >> RARRAY_EMBED_LEN_SHIFT) & \ + (RARRAY_EMBED_LEN_MASK >> RARRAY_EMBED_LEN_SHIFT))) + +#define ARY_OWNS_HEAP_P(a) (!FL_TEST((a), ELTS_SHARED|RARRAY_EMBED_FLAG)) +#define FL_SET_EMBED(a) do { \ + assert(!ARY_SHARED_P(a)); \ + assert(!OBJ_FROZEN(a)); \ + FL_SET((a), RARRAY_EMBED_FLAG); \ +} while (0) +#define FL_UNSET_EMBED(ary) FL_UNSET((ary), RARRAY_EMBED_FLAG|RARRAY_EMBED_LEN_MASK) +#define FL_SET_SHARED(ary) do { \ + assert(!ARY_EMBED_P(ary)); \ + FL_SET((ary), ELTS_SHARED); \ +} while (0) +#define FL_UNSET_SHARED(ary) FL_UNSET((ary), ELTS_SHARED) + +#define ARY_SET_PTR(ary, p) do { \ + assert(!ARY_EMBED_P(ary)); \ + assert(!OBJ_FROZEN(ary)); \ + RARRAY(ary)->as.heap.ptr = (p); \ +} while (0) +#define ARY_SET_EMBED_LEN(ary, n) do { \ + long tmp_n = (n); \ + assert(ARY_EMBED_P(ary)); \ + assert(!OBJ_FROZEN(ary)); \ + RBASIC(ary)->flags &= ~RARRAY_EMBED_LEN_MASK; \ + RBASIC(ary)->flags |= (tmp_n) << RARRAY_EMBED_LEN_SHIFT; \ +} while (0) +#define ARY_SET_HEAP_LEN(ary, n) do { \ + assert(!ARY_EMBED_P(ary)); \ + RARRAY(ary)->as.heap.len = (n); \ +} while (0) +#define ARY_SET_LEN(ary, n) do { \ + if (ARY_EMBED_P(ary)) { \ + ARY_SET_EMBED_LEN((ary), (n)); \ + } \ + else { \ + ARY_SET_HEAP_LEN((ary), (n)); \ + } \ + assert(RARRAY_LEN(ary) == (n)); \ +} while (0) +#define ARY_INCREASE_PTR(ary, n) do { \ + assert(!ARY_EMBED_P(ary)); \ + assert(!OBJ_FROZEN(ary)); \ + RARRAY(ary)->as.heap.ptr += (n); \ +} while (0) +#define ARY_INCREASE_LEN(ary, n) do { \ + assert(!OBJ_FROZEN(ary)); \ + if (ARY_EMBED_P(ary)) { \ + ARY_SET_EMBED_LEN((ary), RARRAY_LEN(ary)+(n)); \ + } \ + else { \ + RARRAY(ary)->as.heap.len += (n); \ + } \ +} while (0) + +#define ARY_CAPA(ary) (ARY_EMBED_P(ary) ? RARRAY_EMBED_LEN_MAX : \ + ARY_SHARED_ROOT_P(ary) ? RARRAY_LEN(ary) : RARRAY(ary)->as.heap.aux.capa) +#define ARY_SET_CAPA(ary, n) do { \ + assert(!ARY_EMBED_P(ary)); \ + assert(!ARY_SHARED_P(ary)); \ + assert(!OBJ_FROZEN(ary)); \ + RARRAY(ary)->as.heap.aux.capa = (n); \ +} while (0) + +#define ARY_SHARED(ary) (assert(ARY_SHARED_P(ary)), RARRAY(ary)->as.heap.aux.shared) +#define ARY_SET_SHARED(ary, value) do { \ + assert(!ARY_EMBED_P(ary)); \ + assert(ARY_SHARED_P(ary)); \ + assert(ARY_SHARED_ROOT_P(value)); \ + RARRAY(ary)->as.heap.aux.shared = (value); \ +} while (0) +#define RARRAY_SHARED_ROOT_FLAG FL_USER5 +#define ARY_SHARED_ROOT_P(ary) (FL_TEST((ary), RARRAY_SHARED_ROOT_FLAG)) +#define ARY_SHARED_NUM(ary) \ + (assert(ARY_SHARED_ROOT_P(ary)), RARRAY(ary)->as.heap.aux.capa) +#define ARY_SET_SHARED_NUM(ary, value) do { \ + assert(ARY_SHARED_ROOT_P(ary)); \ + RARRAY(ary)->as.heap.aux.capa = (value); \ +} while (0) +#define FL_SET_SHARED_ROOT(ary) do { \ + assert(!ARY_EMBED_P(ary)); \ + FL_SET((ary), RARRAY_SHARED_ROOT_FLAG); \ +} while (0) + +static void +ary_resize_capa(VALUE ary, long capacity) +{ + assert(RARRAY_LEN(ary) <= capacity); + assert(!OBJ_FROZEN(ary)); + assert(!ARY_SHARED_P(ary)); + if (capacity > RARRAY_EMBED_LEN_MAX) { + if (ARY_EMBED_P(ary)) { + long len = ARY_EMBED_LEN(ary); + VALUE *ptr = ALLOC_N(VALUE, (capacity)); + MEMCPY(ptr, ARY_EMBED_PTR(ary), VALUE, len); + FL_UNSET_EMBED(ary); + ARY_SET_PTR(ary, ptr); + ARY_SET_HEAP_LEN(ary, len); + } + else { + REALLOC_N(RARRAY(ary)->as.heap.ptr, VALUE, (capacity)); + } + ARY_SET_CAPA(ary, (capacity)); + } + else { + if (!ARY_EMBED_P(ary)) { + long len = RARRAY_LEN(ary); + VALUE *ptr = RARRAY_PTR(ary); + if (len > capacity) len = capacity; + MEMCPY(RARRAY(ary)->as.ary, ptr, VALUE, len); + FL_SET_EMBED(ary); + ARY_SET_LEN(ary, len); + xfree(ptr); + } + } +} + +static void +ary_double_capa(VALUE ary, long min) +{ + long new_capa = ARY_CAPA(ary) / 2; + + if (new_capa < ARY_DEFAULT_SIZE) { + new_capa = ARY_DEFAULT_SIZE; + } + if (new_capa >= ARY_MAX_SIZE - min) { + new_capa = (ARY_MAX_SIZE - min) / 2; + } + new_capa += min; + ary_resize_capa(ary, new_capa); +} + +static void +rb_ary_decrement_share(VALUE shared) +{ + if (shared) { + long num = ARY_SHARED_NUM(shared) - 1; + if (num == 0) { + rb_ary_free(shared); + rb_gc_force_recycle(shared); + } + else if (num > 0) { + ARY_SET_SHARED_NUM(shared, num); + } + } +} + +static void +rb_ary_unshare(VALUE ary) +{ + VALUE shared = RARRAY(ary)->as.heap.aux.shared; + rb_ary_decrement_share(shared); + FL_UNSET_SHARED(ary); +} static inline void -rb_ary_modify_check(ary) - VALUE ary; +rb_ary_unshare_safe(VALUE ary) { - if (OBJ_FROZEN(ary)) rb_error_frozen("array"); - if (FL_TEST(ary, ARY_TMPLOCK)) - rb_raise(rb_eRuntimeError, "can't modify array during iteration"); - if (!OBJ_TAINTED(ary) && rb_safe_level() >= 4) - rb_raise(rb_eSecurityError, "Insecure: can't modify array"); + if (ARY_SHARED_P(ary) && !ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + } +} + +static VALUE +rb_ary_increment_share(VALUE shared) +{ + long num = ARY_SHARED_NUM(shared); + if (num >= 0) { + ARY_SET_SHARED_NUM(shared, num + 1); + } + return shared; } static void -rb_ary_modify(ary) - VALUE ary; +rb_ary_set_shared(VALUE ary, VALUE shared) { - VALUE *ptr; + rb_ary_increment_share(shared); + FL_SET_SHARED(ary); + ARY_SET_SHARED(ary, shared); +} +static inline void +rb_ary_modify_check(VALUE ary) +{ + rb_check_frozen(ary); + if (!OBJ_UNTRUSTED(ary) && rb_safe_level() >= 4) + rb_raise(rb_eSecurityError, "Insecure: can't modify array"); +} + +void +rb_ary_modify(VALUE ary) +{ rb_ary_modify_check(ary); - if (FL_TEST(ary, ELTS_SHARED)) { - ptr = ALLOC_N(VALUE, RARRAY(ary)->len); - FL_UNSET(ary, ELTS_SHARED); - RARRAY(ary)->aux.capa = RARRAY(ary)->len; - MEMCPY(ptr, RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); - RARRAY(ary)->ptr = ptr; + if (ARY_SHARED_P(ary)) { + long len = RARRAY_LEN(ary); + if (len <= RARRAY_EMBED_LEN_MAX) { + VALUE *ptr = ARY_HEAP_PTR(ary); + VALUE shared = ARY_SHARED(ary); + FL_UNSET_SHARED(ary); + FL_SET_EMBED(ary); + MEMCPY(ARY_EMBED_PTR(ary), ptr, VALUE, len); + rb_ary_decrement_share(shared); + ARY_SET_EMBED_LEN(ary, len); + } + else { + VALUE *ptr = ALLOC_N(VALUE, len); + MEMCPY(ptr, RARRAY_PTR(ary), VALUE, len); + rb_ary_unshare(ary); + ARY_SET_CAPA(ary, len); + ARY_SET_PTR(ary, ptr); + } } } VALUE -rb_ary_freeze(ary) - VALUE ary; +rb_ary_freeze(VALUE ary) { return rb_obj_freeze(ary); } /* * call-seq: - * array.frozen? -> true or false + * ary.frozen? -> true or false * * Return <code>true</code> if this array is frozen (or temporarily frozen * while being sorted). */ static VALUE -rb_ary_frozen_p(ary) - VALUE ary; +rb_ary_frozen_p(VALUE ary) { if (OBJ_FROZEN(ary)) return Qtrue; - if (FL_TEST(ary, ARY_TMPLOCK)) return Qtrue; return Qfalse; } -static VALUE ary_alloc _((VALUE)); static VALUE -ary_alloc(klass) - VALUE klass; +ary_alloc(VALUE klass) { NEWOBJ(ary, struct RArray); OBJSETUP(ary, klass, T_ARRAY); - - ary->len = 0; - ary->ptr = 0; - ary->aux.capa = 0; + FL_SET_EMBED((VALUE)ary); + ARY_SET_EMBED_LEN((VALUE)ary, 0); return (VALUE)ary; } static VALUE -ary_new(klass, len) - VALUE klass; - long len; +ary_new(VALUE klass, long capa) { - VALUE ary = ary_alloc(klass); + VALUE ary; - if (len < 0) { + if (capa < 0) { rb_raise(rb_eArgError, "negative array size (or size too big)"); } - if (len > ARY_MAX_SIZE) { + if (capa > ARY_MAX_SIZE) { rb_raise(rb_eArgError, "array size too big"); } - if (len == 0) len++; - RARRAY(ary)->ptr = ALLOC_N(VALUE, len); - RARRAY(ary)->aux.capa = len; + ary = ary_alloc(klass); + if (capa > RARRAY_EMBED_LEN_MAX) { + FL_UNSET_EMBED(ary); + ARY_SET_PTR(ary, ALLOC_N(VALUE, capa)); + ARY_SET_CAPA(ary, capa); + ARY_SET_HEAP_LEN(ary, 0); + } return ary; } VALUE -rb_ary_new2(len) - long len; +rb_ary_new2(long capa) { - return ary_new(rb_cArray, len); + return ary_new(rb_cArray, capa); } VALUE -rb_ary_new() +rb_ary_new(void) { - return rb_ary_new2(ARY_DEFAULT_SIZE); + return rb_ary_new2(RARRAY_EMBED_LEN_MAX); } -#ifdef HAVE_STDARG_PROTOTYPES #include <stdarg.h> -#define va_init_list(a,b) va_start(a,b) -#else -#include <varargs.h> -#define va_init_list(a,b) va_start(a) -#endif VALUE -#ifdef HAVE_STDARG_PROTOTYPES rb_ary_new3(long n, ...) -#else -rb_ary_new3(n, va_alist) - long n; - va_dcl -#endif { va_list ar; VALUE ary; @@ -168,84 +352,152 @@ rb_ary_new3(n, va_alist) ary = rb_ary_new2(n); - va_init_list(ar, n); + va_start(ar, n); for (i=0; i<n; i++) { - RARRAY(ary)->ptr[i] = va_arg(ar, VALUE); + RARRAY_PTR(ary)[i] = va_arg(ar, VALUE); } va_end(ar); - RARRAY(ary)->len = n; + ARY_SET_LEN(ary, n); return ary; } VALUE -rb_ary_new4(n, elts) - long n; - const VALUE *elts; +rb_ary_new4(long n, const VALUE *elts) { VALUE ary; ary = rb_ary_new2(n); if (n > 0 && elts) { - MEMCPY(RARRAY(ary)->ptr, elts, VALUE, n); + MEMCPY(RARRAY_PTR(ary), elts, VALUE, n); + ARY_SET_LEN(ary, n); } - /* This assignment to len will be moved to the above "if" block in Ruby 1.9 */ - RARRAY(ary)->len = n; - return ary; } +VALUE +rb_ary_tmp_new(long capa) +{ + return ary_new(0, capa); +} + +void +rb_ary_free(VALUE ary) +{ + if (ARY_OWNS_HEAP_P(ary)) { + xfree(ARY_HEAP_PTR(ary)); + } +} + +RUBY_FUNC_EXPORTED size_t +rb_ary_memsize(VALUE ary) +{ + if (ARY_OWNS_HEAP_P(ary)) { + return RARRAY(ary)->as.heap.aux.capa * sizeof(VALUE); + } + else { + return 0; + } +} + +static inline void +ary_discard(VALUE ary) +{ + rb_ary_free(ary); + RBASIC(ary)->flags |= RARRAY_EMBED_FLAG; + RBASIC(ary)->flags &= ~RARRAY_EMBED_LEN_MASK; +} + static VALUE -ary_make_shared(ary) - VALUE ary; +ary_make_shared(VALUE ary) { - if (!FL_TEST(ary, ELTS_SHARED)) { + assert(!ARY_EMBED_P(ary)); + if (ARY_SHARED_P(ary)) { + return ARY_SHARED(ary); + } + else if (ARY_SHARED_ROOT_P(ary)) { + return ary; + } + else if (OBJ_FROZEN(ary)) { + ary_resize_capa(ary, ARY_HEAP_LEN(ary)); + FL_SET_SHARED_ROOT(ary); + ARY_SET_SHARED_NUM(ary, 1); + return ary; + } + else { NEWOBJ(shared, struct RArray); - OBJSETUP(shared, rb_cArray, T_ARRAY); - - shared->len = RARRAY(ary)->len; - shared->ptr = RARRAY(ary)->ptr; - shared->aux.capa = RARRAY(ary)->aux.capa; - RARRAY(ary)->aux.shared = (VALUE)shared; - FL_SET(ary, ELTS_SHARED); + OBJSETUP(shared, 0, T_ARRAY); + FL_UNSET_EMBED(shared); + + ARY_SET_LEN((VALUE)shared, RARRAY_LEN(ary)); + ARY_SET_PTR((VALUE)shared, RARRAY_PTR(ary)); + FL_SET_SHARED_ROOT(shared); + ARY_SET_SHARED_NUM((VALUE)shared, 1); + FL_SET_SHARED(ary); + ARY_SET_SHARED(ary, (VALUE)shared); OBJ_FREEZE(shared); return (VALUE)shared; } +} + + +static VALUE +ary_make_substitution(VALUE ary) +{ + if (RARRAY_LEN(ary) <= RARRAY_EMBED_LEN_MAX) { + VALUE subst = rb_ary_new2(RARRAY_LEN(ary)); + MEMCPY(ARY_EMBED_PTR(subst), RARRAY_PTR(ary), VALUE, RARRAY_LEN(ary)); + ARY_SET_EMBED_LEN(subst, RARRAY_LEN(ary)); + return subst; + } else { - return RARRAY(ary)->aux.shared; + return rb_ary_increment_share(ary_make_shared(ary)); } } VALUE -rb_assoc_new(car, cdr) - VALUE car, cdr; +rb_assoc_new(VALUE car, VALUE cdr) { - VALUE ary; - - ary = rb_ary_new2(2); - RARRAY(ary)->ptr[0] = car; - RARRAY(ary)->ptr[1] = cdr; - RARRAY(ary)->len = 2; - - return ary; + return rb_ary_new3(2, car, cdr); } static VALUE -to_ary(ary) - VALUE ary; +to_ary(VALUE ary) { return rb_convert_type(ary, T_ARRAY, "Array", "to_ary"); } VALUE -rb_check_array_type(ary) - VALUE ary; +rb_check_array_type(VALUE ary) { return rb_check_convert_type(ary, T_ARRAY, "Array", "to_ary"); } -static VALUE rb_ary_replace _((VALUE, VALUE)); +/* + * call-seq: + * Array.try_convert(obj) -> array or nil + * + * Tries to convert +obj+ into an array, using +to_ary+ method. Returns the + * converted array or +nil+ if +obj+ cannot be converted for any reason. + * This method can be used to check if an argument is an array. + * + * Array.try_convert([1]) #=> [1] + * Array.try_convert("1") #=> nil + * + * if tmp = Array.try_convert(arg) + * # the argument is an array + * elsif tmp = String.try_convert(arg) + * # the argument is a string + * end + * + */ + +static VALUE +rb_ary_s_try_convert(VALUE dummy, VALUE ary) +{ + return rb_check_array_type(ary); +} /* * call-seq: @@ -253,56 +505,76 @@ static VALUE rb_ary_replace _((VALUE, VALUE)); * Array.new(array) * Array.new(size) {|index| block } * - * Returns a new array. In the first form, the new array is - * empty. In the second it is created with _size_ copies of _obj_ - * (that is, _size_ references to the same - * _obj_). The third form creates a copy of the array - * passed as a parameter (the array is generated by calling - * to_ary on the parameter). In the last form, an array - * of the given size is created. Each element in this array is - * calculated by passing the element's index to the given block and - * storing the return value. - * - * Array.new - * Array.new(2) - * Array.new(5, "A") - * - * # only one copy of the object is created + * Returns a new array. + * + * In the first form, if no arguments are sent, the new array will be empty. + * When a +size+ and an optional +obj+ are sent, an array is created with + * +size+ copies of +obj+. Take notice that all elements will reference the + * same object +obj+. + * + * The second form creates a copy of the array passed as a parameter (the + * array is generated by calling to_ary on the parameter). + * + * first_array = ["Matz", "Guido"] + * + * second_array = Array.new(first_array) #=> ["Matz", "Guido"] + * + * first_array.equal? second_array #=> false + * + * In the last form, an array of the given size is created. Each element in + * this array is created by passing the element's index to the given block + * and storing the return value. + * + * Array.new(3){ |index| index ** 2 } + * # => [0, 1, 4] + * + * == Common gotchas + * + * When sending the second parameter, the same object will be used as the + * value for all the array elements: + * * a = Array.new(2, Hash.new) + * # => [{}, {}] + * * a[0]['cat'] = 'feline' - * a + * a # => [{"cat"=>"feline"}, {"cat"=>"feline"}] + * * a[1]['cat'] = 'Felix' - * a - * - * # here multiple copies are created + * a # => [{"cat"=>"Felix"}, {"cat"=>"Felix"}] + * + * Since all the Array elements store the same hash, changes to one of them + * will affect them all. + * + * If multiple copies are what you want, you should use the block + * version which uses the result of that block each time an element + * of the array needs to be initialized: + * * a = Array.new(2) { Hash.new } * a[0]['cat'] = 'feline' - * a - * - * squares = Array.new(5) {|i| i*i} - * squares - * - * copy = Array.new(squares) + * a # => [{"cat"=>"feline"}, {}] + * */ static VALUE -rb_ary_initialize(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_initialize(int argc, VALUE *argv, VALUE ary) { long len; VALUE size, val; rb_ary_modify(ary); - if (rb_scan_args(argc, argv, "02", &size, &val) == 0) { - RARRAY(ary)->len = 0; + if (argc == 0) { + if (ARY_OWNS_HEAP_P(ary) && RARRAY_PTR(ary)) { + xfree(RARRAY_PTR(ary)); + } + rb_ary_unshare_safe(ary); + FL_SET_EMBED(ary); + ARY_SET_EMBED_LEN(ary, 0); if (rb_block_given_p()) { rb_warning("given block not used"); } return ary; } - + rb_scan_args(argc, argv, "02", &size, &val); if (argc == 1 && !FIXNUM_P(size)) { val = rb_check_array_type(size); if (!NIL_P(val)) { @@ -318,10 +590,8 @@ rb_ary_initialize(argc, argv, ary) if (len > ARY_MAX_SIZE) { rb_raise(rb_eArgError, "array size too big"); } - if (len > RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, len); - RARRAY(ary)->aux.capa = len; - } + rb_ary_modify(ary); + ary_resize_capa(ary, len); if (rb_block_given_p()) { long i; @@ -330,20 +600,19 @@ rb_ary_initialize(argc, argv, ary) } for (i=0; i<len; i++) { rb_ary_store(ary, i, rb_yield(LONG2NUM(i))); - RARRAY(ary)->len = i + 1; + ARY_SET_LEN(ary, i + 1); } } else { - memfill(RARRAY(ary)->ptr, len, val); - RARRAY(ary)->len = len; + memfill(RARRAY_PTR(ary), len, val); + ARY_SET_LEN(ary, len); } - return ary; } -/* -* Returns a new array populated with the given objects. +/* +* Returns a new array populated with the given objects. * * Array.[]( 1, 'a', /^A/ ) * Array[ 1, 'a', /^A/ ] @@ -351,33 +620,25 @@ rb_ary_initialize(argc, argv, ary) */ static VALUE -rb_ary_s_create(argc, argv, klass) - int argc; - VALUE *argv; - VALUE klass; +rb_ary_s_create(int argc, VALUE *argv, VALUE klass) { - VALUE ary = ary_alloc(klass); - - if (argc > 0) { - RARRAY(ary)->ptr = ALLOC_N(VALUE, argc); - MEMCPY(RARRAY(ary)->ptr, argv, VALUE, argc); + VALUE ary = ary_new(klass, argc); + if (argc > 0 && argv) { + MEMCPY(RARRAY_PTR(ary), argv, VALUE, argc); + ARY_SET_LEN(ary, argc); } - RARRAY(ary)->len = RARRAY(ary)->aux.capa = argc; return ary; } void -rb_ary_store(ary, idx, val) - VALUE ary; - long idx; - VALUE val; +rb_ary_store(VALUE ary, long idx, VALUE val) { if (idx < 0) { - idx += RARRAY(ary)->len; + idx += RARRAY_LEN(ary); if (idx < 0) { - rb_raise(rb_eIndexError, "index %ld out of array", - idx - RARRAY(ary)->len); + rb_raise(rb_eIndexError, "index %ld too small for array; minimum: %ld", + idx - RARRAY_LEN(ary), -RARRAY_LEN(ary)); } } else if (idx >= ARY_MAX_SIZE) { @@ -385,78 +646,87 @@ rb_ary_store(ary, idx, val) } rb_ary_modify(ary); - if (idx >= RARRAY(ary)->aux.capa) { - long new_capa = RARRAY(ary)->aux.capa / 2; - - if (new_capa < ARY_DEFAULT_SIZE) { - new_capa = ARY_DEFAULT_SIZE; - } - if (new_capa >= ARY_MAX_SIZE - idx) { - new_capa = (ARY_MAX_SIZE - idx) / 2; - } - new_capa += idx; - REALLOC_N(RARRAY(ary)->ptr, VALUE, new_capa); - RARRAY(ary)->aux.capa = new_capa; + if (idx >= ARY_CAPA(ary)) { + ary_double_capa(ary, idx); } - if (idx > RARRAY(ary)->len) { - rb_mem_clear(RARRAY(ary)->ptr + RARRAY(ary)->len, - idx-RARRAY(ary)->len + 1); + if (idx > RARRAY_LEN(ary)) { + rb_mem_clear(RARRAY_PTR(ary) + RARRAY_LEN(ary), + idx-RARRAY_LEN(ary) + 1); } - if (idx >= RARRAY(ary)->len) { - RARRAY(ary)->len = idx + 1; + if (idx >= RARRAY_LEN(ary)) { + ARY_SET_LEN(ary, idx + 1); } - RARRAY(ary)->ptr[idx] = val; + RARRAY_PTR(ary)[idx] = val; } static VALUE -ary_shared_array(klass, ary) - VALUE klass; - VALUE ary; +ary_make_partial(VALUE ary, VALUE klass, long offset, long len) { - VALUE val = ary_alloc(klass); + assert(offset >= 0); + assert(len >= 0); + assert(offset+len <= RARRAY_LEN(ary)); + + if (len <= RARRAY_EMBED_LEN_MAX) { + VALUE result = ary_alloc(klass); + MEMCPY(ARY_EMBED_PTR(result), RARRAY_PTR(ary) + offset, VALUE, len); + ARY_SET_EMBED_LEN(result, len); + return result; + } + else { + VALUE shared, result = ary_alloc(klass); + FL_UNSET_EMBED(result); + + shared = ary_make_shared(ary); + ARY_SET_PTR(result, RARRAY_PTR(ary)); + ARY_SET_LEN(result, RARRAY_LEN(ary)); + rb_ary_set_shared(result, shared); - ary_make_shared(ary); - RARRAY(val)->ptr = RARRAY(ary)->ptr; - RARRAY(val)->len = RARRAY(ary)->len; - RARRAY(val)->aux.shared = RARRAY(ary)->aux.shared; - FL_SET(val, ELTS_SHARED); - return val; + ARY_INCREASE_PTR(result, offset); + ARY_SET_LEN(result, len); + return result; + } } static VALUE -ary_shared_first(argc, argv, ary, last) - int argc; - VALUE *argv; - VALUE ary; - int last; +ary_make_shared_copy(VALUE ary) +{ + return ary_make_partial(ary, rb_obj_class(ary), 0, RARRAY_LEN(ary)); +} + +enum ary_take_pos_flags +{ + ARY_TAKE_FIRST = 0, + ARY_TAKE_LAST = 1 +}; + +static VALUE +ary_take_first_or_last(int argc, VALUE *argv, VALUE ary, enum ary_take_pos_flags last) { - VALUE nv, result; + VALUE nv; long n; long offset = 0; rb_scan_args(argc, argv, "1", &nv); n = NUM2LONG(nv); - if (n > RARRAY(ary)->len) { - n = RARRAY(ary)->len; + if (n > RARRAY_LEN(ary)) { + n = RARRAY_LEN(ary); } else if (n < 0) { rb_raise(rb_eArgError, "negative array size"); } if (last) { - offset = RARRAY(ary)->len - n; + offset = RARRAY_LEN(ary) - n; } - result = ary_shared_array(rb_cArray, ary); - RARRAY(result)->ptr += offset; - RARRAY(result)->len = n; - - return result; + return ary_make_partial(ary, rb_cArray, offset, n); } +static VALUE rb_ary_push_1(VALUE ary, VALUE item); + /* * call-seq: - * array << obj -> array - * + * ary << obj -> ary + * * Append---Pushes the given object on to the end of this array. This * expression returns the array itself, so several appends * may be chained together. @@ -467,65 +737,76 @@ ary_shared_first(argc, argv, ary, last) */ VALUE -rb_ary_push(ary, item) - VALUE ary; - VALUE item; +rb_ary_push(VALUE ary, VALUE item) { - rb_ary_store(ary, RARRAY(ary)->len, item); + rb_ary_modify(ary); + return rb_ary_push_1(ary, item); +} + +static VALUE +rb_ary_push_1(VALUE ary, VALUE item) +{ + long idx = RARRAY_LEN(ary); + + if (idx >= ARY_CAPA(ary)) { + ary_double_capa(ary, idx); + } + RARRAY_PTR(ary)[idx] = item; + ARY_SET_LEN(ary, idx + 1); return ary; } -/* +/* * call-seq: - * array.push(obj, ... ) -> array - * + * ary.push(obj, ... ) -> ary + * * Append---Pushes the given object(s) on to the end of this array. This * expression returns the array itself, so several appends * may be chained together. * * a = [ "a", "b", "c" ] - * a.push("d", "e", "f") + * a.push("d", "e", "f") * #=> ["a", "b", "c", "d", "e", "f"] */ static VALUE -rb_ary_push_m(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_push_m(int argc, VALUE *argv, VALUE ary) { + rb_ary_modify(ary); while (argc--) { - rb_ary_push(ary, *argv++); + rb_ary_push_1(ary, *argv++); } return ary; } VALUE -rb_ary_pop(ary) - VALUE ary; +rb_ary_pop(VALUE ary) { + long n; rb_ary_modify_check(ary); - if (RARRAY(ary)->len == 0) return Qnil; - if (!FL_TEST(ary, ELTS_SHARED) && - RARRAY(ary)->len * 3 < RARRAY(ary)->aux.capa && - RARRAY(ary)->aux.capa > ARY_DEFAULT_SIZE) { - RARRAY(ary)->aux.capa = RARRAY(ary)->len * 2; - REALLOC_N(RARRAY(ary)->ptr, VALUE, RARRAY(ary)->aux.capa); + if (RARRAY_LEN(ary) == 0) return Qnil; + if (ARY_OWNS_HEAP_P(ary) && + RARRAY_LEN(ary) * 3 < ARY_CAPA(ary) && + ARY_CAPA(ary) > ARY_DEFAULT_SIZE) + { + ary_resize_capa(ary, RARRAY_LEN(ary) * 2); } - return RARRAY(ary)->ptr[--RARRAY(ary)->len]; + n = RARRAY_LEN(ary)-1; + ARY_SET_LEN(ary, n); + return RARRAY_PTR(ary)[n]; } /* * call-seq: - * array.pop -> obj or nil - * array.pop(n) -> array - * - * Removes the last element from <i>self</i> and returns it, or + * ary.pop -> obj or nil + * ary.pop(n) -> new_ary + * + * Removes the last element from +self+ and returns it, or * <code>nil</code> if the array is empty. * * If a number _n_ is given, returns an array of the last n elements * (or less) just like <code>array.slice!(-n, n)</code> does. - * + * * a = [ "a", "b", "c", "d" ] * a.pop #=> "d" * a.pop(2) #=> ["b", "c"] @@ -533,10 +814,7 @@ rb_ary_pop(ary) */ static VALUE -rb_ary_pop_m(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_pop_m(int argc, VALUE *argv, VALUE ary) { VALUE result; @@ -545,47 +823,51 @@ rb_ary_pop_m(argc, argv, ary) } rb_ary_modify_check(ary); - result = ary_shared_first(argc, argv, ary, Qtrue); - RARRAY(ary)->len -= RARRAY(result)->len; + result = ary_take_first_or_last(argc, argv, ary, ARY_TAKE_LAST); + ARY_INCREASE_LEN(ary, -RARRAY_LEN(result)); return result; } VALUE -rb_ary_shift(ary) - VALUE ary; +rb_ary_shift(VALUE ary) { VALUE top; rb_ary_modify_check(ary); - if (RARRAY(ary)->len == 0) return Qnil; - top = RARRAY(ary)->ptr[0]; - if (!FL_TEST(ary, ELTS_SHARED)) { - if (RARRAY(ary)->len < ARY_DEFAULT_SIZE) { - MEMMOVE(RARRAY(ary)->ptr, RARRAY(ary)->ptr+1, VALUE, RARRAY(ary)->len-1); - RARRAY(ary)->len--; - return top; - } - RARRAY(ary)->ptr[0] = Qnil; + if (RARRAY_LEN(ary) == 0) return Qnil; + top = RARRAY_PTR(ary)[0]; + if (!ARY_SHARED_P(ary)) { + if (RARRAY_LEN(ary) < ARY_DEFAULT_SIZE) { + MEMMOVE(RARRAY_PTR(ary), RARRAY_PTR(ary)+1, VALUE, RARRAY_LEN(ary)-1); + ARY_INCREASE_LEN(ary, -1); + return top; + } + assert(!ARY_EMBED_P(ary)); /* ARY_EMBED_LEN_MAX < ARY_DEFAULT_SIZE */ + + RARRAY_PTR(ary)[0] = Qnil; ary_make_shared(ary); } - RARRAY(ary)->ptr++; /* shift ptr */ - RARRAY(ary)->len--; + else if (ARY_SHARED_NUM(ARY_SHARED(ary)) == 1) { + RARRAY_PTR(ary)[0] = Qnil; + } + ARY_INCREASE_PTR(ary, 1); /* shift ptr */ + ARY_INCREASE_LEN(ary, -1); return top; } /* * call-seq: - * array.shift -> obj or nil - * array.shift(n) -> array - * - * Returns the first element of <i>self</i> and removes it (shifting all + * ary.shift -> obj or nil + * ary.shift(n) -> new_ary + * + * Returns the first element of +self+ and removes it (shifting all * other elements down by one). Returns <code>nil</code> if the array * is empty. * * If a number _n_ is given, returns an array of the first n elements * (or less) just like <code>array.slice!(0, n)</code> does. - * + * * args = [ "-m", "-q", "filename" ] * args.shift #=> "-m" * args #=> ["-q", "filename"] @@ -596,10 +878,7 @@ rb_ary_shift(ary) */ static VALUE -rb_ary_shift_m(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_shift_m(int argc, VALUE *argv, VALUE ary) { VALUE result; long n; @@ -609,144 +888,111 @@ rb_ary_shift_m(argc, argv, ary) } rb_ary_modify_check(ary); - result = ary_shared_first(argc, argv, ary, Qfalse); - n = RARRAY(result)->len; - if (FL_TEST(ary, ELTS_SHARED)) { - RARRAY(ary)->ptr += n; - RARRAY(ary)->len -= n; + result = ary_take_first_or_last(argc, argv, ary, ARY_TAKE_FIRST); + n = RARRAY_LEN(result); + if (ARY_SHARED_P(ary)) { + if (ARY_SHARED_NUM(ARY_SHARED(ary)) == 1) { + rb_mem_clear(RARRAY_PTR(ary), n); } + ARY_INCREASE_PTR(ary, n); + } else { - MEMMOVE(RARRAY(ary)->ptr, RARRAY(ary)->ptr+n, VALUE, RARRAY(ary)->len-n); - RARRAY(ary)->len -= n; + MEMMOVE(RARRAY_PTR(ary), RARRAY_PTR(ary)+n, VALUE, RARRAY_LEN(ary)-n); } + ARY_INCREASE_LEN(ary, -n); return result; } -VALUE -rb_ary_unshift(ary, item) - VALUE ary, item; -{ - rb_ary_modify(ary); - if (RARRAY(ary)->len == RARRAY(ary)->aux.capa) { - long capa_inc = RARRAY(ary)->aux.capa / 2; - if (capa_inc < ARY_DEFAULT_SIZE) { - capa_inc = ARY_DEFAULT_SIZE; - } - RARRAY(ary)->aux.capa += capa_inc; - REALLOC_N(RARRAY(ary)->ptr, VALUE, RARRAY(ary)->aux.capa); - } - - /* sliding items */ - MEMMOVE(RARRAY(ary)->ptr + 1, RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); - - RARRAY(ary)->len++; - RARRAY(ary)->ptr[0] = item; - - return ary; -} - /* * call-seq: - * array.unshift(obj, ...) -> array - * - * Prepends objects to the front of <i>array</i>. - * other elements up one. - * + * ary.unshift(obj, ...) -> ary + * + * Prepends objects to the front of +self+, + * moving other elements upwards. + * * a = [ "b", "c", "d" ] * a.unshift("a") #=> ["a", "b", "c", "d"] * a.unshift(1, 2) #=> [ 1, 2, "a", "b", "c", "d"] */ static VALUE -rb_ary_unshift_m(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_unshift_m(int argc, VALUE *argv, VALUE ary) { - long len = RARRAY(ary)->len; + long len; + rb_ary_modify(ary); if (argc == 0) return ary; - - /* make rooms by setting the last item */ - rb_ary_store(ary, len + argc - 1, Qnil); + if (ARY_CAPA(ary) <= (len = RARRAY_LEN(ary)) + argc) { + ary_double_capa(ary, len + argc); + } /* sliding items */ - MEMMOVE(RARRAY(ary)->ptr + argc, RARRAY(ary)->ptr, VALUE, len); - MEMCPY(RARRAY(ary)->ptr, argv, VALUE, argc); - + MEMMOVE(RARRAY_PTR(ary) + argc, RARRAY_PTR(ary), VALUE, len); + MEMCPY(RARRAY_PTR(ary), argv, VALUE, argc); + ARY_INCREASE_LEN(ary, argc); + return ary; } +VALUE +rb_ary_unshift(VALUE ary, VALUE item) +{ + return rb_ary_unshift_m(1,&item,ary); +} + /* faster version - use this if you don't need to treat negative offset */ static inline VALUE -rb_ary_elt(ary, offset) - VALUE ary; - long offset; +rb_ary_elt(VALUE ary, long offset) { - if (RARRAY(ary)->len == 0) return Qnil; - if (offset < 0 || RARRAY(ary)->len <= offset) { + if (RARRAY_LEN(ary) == 0) return Qnil; + if (offset < 0 || RARRAY_LEN(ary) <= offset) { return Qnil; } - return RARRAY(ary)->ptr[offset]; + return RARRAY_PTR(ary)[offset]; } VALUE -rb_ary_entry(ary, offset) - VALUE ary; - long offset; +rb_ary_entry(VALUE ary, long offset) { if (offset < 0) { - offset += RARRAY(ary)->len; + offset += RARRAY_LEN(ary); } return rb_ary_elt(ary, offset); } -static VALUE -rb_ary_subseq(ary, beg, len) - VALUE ary; - long beg, len; +VALUE +rb_ary_subseq(VALUE ary, long beg, long len) { - VALUE klass, ary2, shared; - VALUE *ptr; + VALUE klass; - if (beg > RARRAY(ary)->len) return Qnil; + if (beg > RARRAY_LEN(ary)) return Qnil; if (beg < 0 || len < 0) return Qnil; - if (RARRAY(ary)->len < len || RARRAY(ary)->len < beg + len) { - len = RARRAY(ary)->len - beg; - if (len < 0) - len = 0; + if (RARRAY_LEN(ary) < len || RARRAY_LEN(ary) < beg + len) { + len = RARRAY_LEN(ary) - beg; } klass = rb_obj_class(ary); if (len == 0) return ary_new(klass, 0); - shared = ary_make_shared(ary); - ptr = RARRAY(ary)->ptr; - ary2 = ary_alloc(klass); - RARRAY(ary2)->ptr = ptr + beg; - RARRAY(ary2)->len = len; - RARRAY(ary2)->aux.shared = shared; - FL_SET(ary2, ELTS_SHARED); - - return ary2; + return ary_make_partial(ary, klass, beg, len); } -/* +/* * call-seq: - * array[index] -> obj or nil - * array[start, length] -> an_array or nil - * array[range] -> an_array or nil - * array.slice(index) -> obj or nil - * array.slice(start, length) -> an_array or nil - * array.slice(range) -> an_array or nil + * ary[index] -> obj or nil + * ary[start, length] -> new_ary or nil + * ary[range] -> new_ary or nil + * ary.slice(index) -> obj or nil + * ary.slice(start, length) -> new_ary or nil + * ary.slice(range) -> new_ary or nil * * Element Reference---Returns the element at _index_, * or returns a subarray starting at _start_ and * continuing for _length_ elements, or returns a subarray * specified by _range_. * Negative indices count backward from the end of the - * array (-1 is the last element). Returns nil if the index + * array (-1 is the last element). Returns +nil+ if the index * (or starting index) are out of range. * * a = [ "a", "b", "c", "d", "e" ] @@ -765,22 +1011,16 @@ rb_ary_subseq(ary, beg, len) */ VALUE -rb_ary_aref(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_aref(int argc, VALUE *argv, VALUE ary) { VALUE arg; long beg, len; if (argc == 2) { - if (SYMBOL_P(argv[0])) { - rb_raise(rb_eTypeError, "Symbol as array index"); - } beg = NUM2LONG(argv[0]); len = NUM2LONG(argv[1]); if (beg < 0) { - beg += RARRAY(ary)->len; + beg += RARRAY_LEN(ary); } return rb_ary_subseq(ary, beg, len); } @@ -792,11 +1032,8 @@ rb_ary_aref(argc, argv, ary) if (FIXNUM_P(arg)) { return rb_ary_entry(ary, FIX2LONG(arg)); } - if (SYMBOL_P(arg)) { - rb_raise(rb_eTypeError, "Symbol as array index"); - } /* check if idx is Range */ - switch (rb_range_beg_len(arg, &beg, &len, RARRAY(ary)->len, 0)) { + switch (rb_range_beg_len(arg, &beg, &len, RARRAY_LEN(ary), 0)) { case Qfalse: break; case Qnil: @@ -807,15 +1044,13 @@ rb_ary_aref(argc, argv, ary) return rb_ary_entry(ary, NUM2LONG(arg)); } -/* +/* * call-seq: - * array.at(index) -> obj or nil + * ary.at(index) -> obj or nil * * Returns the element at _index_. A - * negative index counts from the end of _self_. Returns +nil+ + * negative index counts from the end of +self+. Returns +nil+ * if the index is out of range. See also <code>Array#[]</code>. - * (<code>Array#at</code> is slightly faster than <code>Array#[]</code>, - * as it does not accept ranges and so on.) * * a = [ "a", "b", "c", "d", "e" ] * a.at(0) #=> "a" @@ -823,81 +1058,75 @@ rb_ary_aref(argc, argv, ary) */ static VALUE -rb_ary_at(ary, pos) - VALUE ary, pos; +rb_ary_at(VALUE ary, VALUE pos) { return rb_ary_entry(ary, NUM2LONG(pos)); } /* * call-seq: - * array.first -> obj or nil - * array.first(n) -> an_array + * ary.first -> obj or nil + * ary.first(n) -> new_ary * * Returns the first element, or the first +n+ elements, of the array. * If the array is empty, the first form returns <code>nil</code>, and the * second form returns an empty array. * * a = [ "q", "r", "s", "t" ] - * a.first #=> "q" - * a.first(1) #=> ["q"] - * a.first(3) #=> ["q", "r", "s"] + * a.first #=> "q" + * a.first(2) #=> ["q", "r"] */ static VALUE -rb_ary_first(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_first(int argc, VALUE *argv, VALUE ary) { if (argc == 0) { - if (RARRAY(ary)->len == 0) return Qnil; - return RARRAY(ary)->ptr[0]; + if (RARRAY_LEN(ary) == 0) return Qnil; + return RARRAY_PTR(ary)[0]; } else { - return ary_shared_first(argc, argv, ary, Qfalse); + return ary_take_first_or_last(argc, argv, ary, ARY_TAKE_FIRST); } } /* * call-seq: - * array.last -> obj or nil - * array.last(n) -> an_array - * - * Returns the last element(s) of <i>self</i>. If the array is empty, + * ary.last -> obj or nil + * ary.last(n) -> new_ary + * + * Returns the last element(s) of +self+. If the array is empty, * the first form returns <code>nil</code>. - * - * [ "w", "x", "y", "z" ].last #=> "z" + * + * a = [ "w", "x", "y", "z" ] + * a.last #=> "z" + * a.last(2) #=> ["y", "z"] */ -static VALUE -rb_ary_last(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +VALUE +rb_ary_last(int argc, VALUE *argv, VALUE ary) { if (argc == 0) { - if (RARRAY(ary)->len == 0) return Qnil; - return RARRAY(ary)->ptr[RARRAY(ary)->len-1]; + if (RARRAY_LEN(ary) == 0) return Qnil; + return RARRAY_PTR(ary)[RARRAY_LEN(ary)-1]; } else { - return ary_shared_first(argc, argv, ary, Qtrue); + return ary_take_first_or_last(argc, argv, ary, ARY_TAKE_LAST); } } /* * call-seq: - * array.fetch(index) -> obj - * array.fetch(index, default ) -> obj - * array.fetch(index) {|index| block } -> obj - * + * ary.fetch(index) -> obj + * ary.fetch(index, default ) -> obj + * ary.fetch(index) {|index| block } -> obj + * * Tries to return the element at position <i>index</i>. If the index * lies outside the array, the first form throws an * <code>IndexError</code> exception, the second form returns * <i>default</i>, and the third form returns the value of invoking * the block, passing in the index. Negative values of <i>index</i> * count from the end of the array. - * + * * a = [ 11, 22, 33, 44 ] * a.fetch(1) #=> 22 * a.fetch(-1) #=> 44 @@ -906,10 +1135,7 @@ rb_ary_last(argc, argv, ary) */ static VALUE -rb_ary_fetch(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_fetch(int argc, VALUE *argv, VALUE ary) { VALUE pos, ifnone; long block_given; @@ -923,28 +1149,33 @@ rb_ary_fetch(argc, argv, ary) idx = NUM2LONG(pos); if (idx < 0) { - idx += RARRAY(ary)->len; + idx += RARRAY_LEN(ary); } - if (idx < 0 || RARRAY(ary)->len <= idx) { + if (idx < 0 || RARRAY_LEN(ary) <= idx) { if (block_given) return rb_yield(pos); if (argc == 1) { - rb_raise(rb_eIndexError, "index %ld out of array", idx); + rb_raise(rb_eIndexError, "index %ld outside of array bounds: %ld...%ld", + idx - (idx < 0 ? RARRAY_LEN(ary) : 0), -RARRAY_LEN(ary), RARRAY_LEN(ary)); } return ifnone; } - return RARRAY(ary)->ptr[idx]; + return RARRAY_PTR(ary)[idx]; } /* * call-seq: - * array.index(obj) -> int or nil - * array.index {|item| block} -> int or nil - * - * Returns the index of the first object in <i>self</i> such that is + * ary.index(obj) -> int or nil + * ary.index {|item| block} -> int or nil + * ary.index -> an_enumerator + * + * Returns the index of the first object in +self+ such that the object is * <code>==</code> to <i>obj</i>. If a block is given instead of an - * argument, returns first object for which <em>block</em> is true. + * argument, returns index of first object for which <em>block</em> is true. * Returns <code>nil</code> if no match is found. - * + * See also <code>Array#rindex</code>. + * + * If neither block nor argument is given, an enumerator is returned instead. + * * a = [ "a", "b", "c" ] * a.index("b") #=> 1 * a.index("z") #=> nil @@ -954,26 +1185,25 @@ rb_ary_fetch(argc, argv, ary) */ static VALUE -rb_ary_index(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_index(int argc, VALUE *argv, VALUE ary) { VALUE val; long i; - if (argc == 0) { + if (argc == 0) { RETURN_ENUMERATOR(ary, 0, 0); - for (i=0; i<RARRAY(ary)->len; i++) { - if (RTEST(rb_yield(RARRAY(ary)->ptr[i]))) { + for (i=0; i<RARRAY_LEN(ary); i++) { + if (RTEST(rb_yield(RARRAY_PTR(ary)[i]))) { return LONG2NUM(i); } } return Qnil; } - rb_scan_args(argc, argv, "01", &val); - for (i=0; i<RARRAY(ary)->len; i++) { - if (rb_equal(RARRAY(ary)->ptr[i], val)) + rb_scan_args(argc, argv, "1", &val); + if (rb_block_given_p()) + rb_warn("given block not used"); + for (i=0; i<RARRAY_LEN(ary); i++) { + if (rb_equal(RARRAY_PTR(ary)[i], val)) return LONG2NUM(i); } return Qnil; @@ -981,162 +1211,173 @@ rb_ary_index(argc, argv, ary) /* * call-seq: - * array.rindex(obj) -> int or nil - * - * Returns the index of the last object in <i>array</i> + * ary.rindex(obj) -> int or nil + * ary.rindex {|item| block} -> int or nil + * ary.rindex -> an_enumerator + * + * Returns the index of the last object in +self+ * <code>==</code> to <i>obj</i>. If a block is given instead of an - * argument, returns first object for which <em>block</em> is - * true. Returns <code>nil</code> if no match is found. - * + * argument, returns index of first object for which <em>block</em> is + * true, starting from the last object. + * Returns <code>nil</code> if no match is found. + * See also <code>Array#index</code>. + * + * If neither block nor argument is given, an enumerator is returned instead. + * * a = [ "a", "b", "b", "b", "c" ] - * a.rindex("b") #=> 3 - * a.rindex("z") #=> nil - * a.rindex{|x|x=="b"} #=> 3 + * a.rindex("b") #=> 3 + * a.rindex("z") #=> nil + * a.rindex { |x| x == "b" } #=> 3 */ static VALUE -rb_ary_rindex(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_rindex(int argc, VALUE *argv, VALUE ary) { VALUE val; - long i = RARRAY(ary)->len; + long i = RARRAY_LEN(ary); if (argc == 0) { RETURN_ENUMERATOR(ary, 0, 0); while (i--) { - if (RTEST(rb_yield(RARRAY(ary)->ptr[i]))) + if (RTEST(rb_yield(RARRAY_PTR(ary)[i]))) return LONG2NUM(i); - if (i > RARRAY(ary)->len) { - i = RARRAY(ary)->len; + if (i > RARRAY_LEN(ary)) { + i = RARRAY_LEN(ary); } } return Qnil; } - rb_scan_args(argc, argv, "01", &val); + rb_scan_args(argc, argv, "1", &val); + if (rb_block_given_p()) + rb_warn("given block not used"); while (i--) { - if (rb_equal(RARRAY(ary)->ptr[i], val)) + if (rb_equal(RARRAY_PTR(ary)[i], val)) return LONG2NUM(i); - if (i > RARRAY(ary)->len) { - i = RARRAY(ary)->len; + if (i > RARRAY_LEN(ary)) { + i = RARRAY_LEN(ary); } } return Qnil; } -/* - * call-seq: - * array.indexes( i1, i2, ... iN ) -> an_array - * array.indices( i1, i2, ... iN ) -> an_array - * - * Deprecated; use <code>Array#values_at</code>. - */ - -static VALUE -rb_ary_indexes(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; -{ - VALUE new_ary; - long i; - - rb_warn("Array#%s is deprecated; use Array#values_at", rb_id2name(rb_frame_last_func())); - new_ary = rb_ary_new2(argc); - for (i=0; i<argc; i++) { - rb_ary_push(new_ary, rb_ary_aref(1, argv+i, ary)); - } - - return new_ary; -} - VALUE -rb_ary_to_ary(obj) - VALUE obj; +rb_ary_to_ary(VALUE obj) { - if (TYPE(obj) == T_ARRAY) { - return obj; - } - if (rb_respond_to(obj, rb_intern("to_ary"))) { - return rb_convert_type(obj, T_ARRAY, "Array", "to_ary"); - } + VALUE tmp = rb_check_array_type(obj); + + if (!NIL_P(tmp)) return tmp; return rb_ary_new3(1, obj); } static void -rb_ary_splice(ary, beg, len, rpl) - VALUE ary; - long beg, len; - VALUE rpl; +rb_ary_splice(VALUE ary, long beg, long len, VALUE rpl) { long rlen; if (len < 0) rb_raise(rb_eIndexError, "negative length (%ld)", len); if (beg < 0) { - beg += RARRAY(ary)->len; + beg += RARRAY_LEN(ary); if (beg < 0) { - beg -= RARRAY(ary)->len; - rb_raise(rb_eIndexError, "index %ld out of array", beg); + rb_raise(rb_eIndexError, "index %ld too small for array; minimum: %ld", + beg - RARRAY_LEN(ary), -RARRAY_LEN(ary)); } } - if (RARRAY(ary)->len < len || RARRAY(ary)->len < beg + len) { - len = RARRAY(ary)->len - beg; + if (RARRAY_LEN(ary) < len || RARRAY_LEN(ary) < beg + len) { + len = RARRAY_LEN(ary) - beg; } - if (NIL_P(rpl)) { + if (rpl == Qundef) { rlen = 0; } else { rpl = rb_ary_to_ary(rpl); - rlen = RARRAY(rpl)->len; + rlen = RARRAY_LEN(rpl); } rb_ary_modify(ary); - - if (beg >= RARRAY(ary)->len) { + if (beg >= RARRAY_LEN(ary)) { if (beg > ARY_MAX_SIZE - rlen) { rb_raise(rb_eIndexError, "index %ld too big", beg); } len = beg + rlen; - if (len >= RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, len); - RARRAY(ary)->aux.capa = len; + if (len >= ARY_CAPA(ary)) { + ary_double_capa(ary, len); } - rb_mem_clear(RARRAY(ary)->ptr + RARRAY(ary)->len, beg - RARRAY(ary)->len); + rb_mem_clear(RARRAY_PTR(ary) + RARRAY_LEN(ary), beg - RARRAY_LEN(ary)); if (rlen > 0) { - MEMCPY(RARRAY(ary)->ptr + beg, RARRAY(rpl)->ptr, VALUE, rlen); + MEMCPY(RARRAY_PTR(ary) + beg, RARRAY_PTR(rpl), VALUE, rlen); } - RARRAY(ary)->len = len; + ARY_SET_LEN(ary, len); } else { long alen; - if (beg + len > RARRAY(ary)->len) { - len = RARRAY(ary)->len - beg; - } - - alen = RARRAY(ary)->len + rlen - len; - if (alen >= RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, alen); - RARRAY(ary)->aux.capa = alen; + alen = RARRAY_LEN(ary) + rlen - len; + if (alen >= ARY_CAPA(ary)) { + ary_double_capa(ary, alen); } if (len != rlen) { - MEMMOVE(RARRAY(ary)->ptr + beg + rlen, RARRAY(ary)->ptr + beg + len, - VALUE, RARRAY(ary)->len - (beg + len)); - RARRAY(ary)->len = alen; + MEMMOVE(RARRAY_PTR(ary) + beg + rlen, RARRAY_PTR(ary) + beg + len, + VALUE, RARRAY_LEN(ary) - (beg + len)); + ARY_SET_LEN(ary, alen); } if (rlen > 0) { - MEMMOVE(RARRAY(ary)->ptr + beg, RARRAY(rpl)->ptr, VALUE, rlen); + MEMMOVE(RARRAY_PTR(ary) + beg, RARRAY_PTR(rpl), VALUE, rlen); } } } -/* +/*! + * expands or shrinks \a ary to \a len elements. + * expanded region will be filled with Qnil. + * \param ary an array + * \param len new size + * \return \a ary + * \post the size of \a ary is \a len. + */ +VALUE +rb_ary_resize(VALUE ary, long len) +{ + long olen; + + rb_ary_modify(ary); + olen = RARRAY_LEN(ary); + if (len == olen) return ary; + if (len > ARY_MAX_SIZE) { + rb_raise(rb_eIndexError, "index %ld too big", len); + } + if (len > olen) { + if (len >= ARY_CAPA(ary)) { + ary_double_capa(ary, len); + } + rb_mem_clear(RARRAY_PTR(ary) + olen, len - olen); + ARY_SET_LEN(ary, len); + } + else if (ARY_EMBED_P(ary)) { + ARY_SET_EMBED_LEN(ary, len); + } + else if (len <= RARRAY_EMBED_LEN_MAX) { + VALUE tmp[RARRAY_EMBED_LEN_MAX]; + MEMCPY(tmp, ARY_HEAP_PTR(ary), VALUE, len); + ary_discard(ary); + MEMCPY(ARY_EMBED_PTR(ary), tmp, VALUE, len); + ARY_SET_EMBED_LEN(ary, len); + } + else { + if (olen > len + ARY_DEFAULT_SIZE) { + REALLOC_N(RARRAY(ary)->as.heap.ptr, VALUE, len); + ARY_SET_CAPA(ary, len); + } + ARY_SET_HEAP_LEN(ary, len); + } + return ary; +} + +/* * call-seq: - * array[index] = obj -> obj - * array[start, length] = obj or an_array or nil -> obj or an_array or nil - * array[range] = obj or an_array or nil -> obj or an_array or nil + * ary[index] = obj -> obj + * ary[start, length] = obj or other_ary or nil -> obj or other_ary or nil + * ary[range] = obj or other_ary or nil -> obj or other_ary or nil * * Element Assignment---Sets the element at _index_, * or replaces a subarray starting at _start_ and @@ -1145,11 +1386,10 @@ rb_ary_splice(ary, beg, len, rpl) * the current capacity of the array, the array grows * automatically. A negative indices will count backward * from the end of the array. Inserts elements if _length_ is - * zero. If +nil+ is used in the second and third form, - * deletes elements from _self_. An +IndexError+ is raised if a - * negative index points past the beginning of the array. See also + * zero. An +IndexError+ is raised if a negative index points + * past the beginning of the array. See also * <code>Array#push</code>, and <code>Array#unshift</code>. - * + * * a = Array.new * a[4] = "4"; #=> [nil, nil, nil, nil, "4"] * a[0, 3] = [ 'a', 'b', 'c' ] #=> ["a", "b", "c", nil, "4"] @@ -1157,38 +1397,31 @@ rb_ary_splice(ary, beg, len, rpl) * a[0, 2] = "?" #=> ["?", 2, nil, "4"] * a[0..2] = "A" #=> ["A", "4"] * a[-1] = "Z" #=> ["A", "Z"] - * a[1..-1] = nil #=> ["A"] + * a[1..-1] = nil #=> ["A", nil] + * a[1..-1] = [] #=> ["A"] */ static VALUE -rb_ary_aset(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_aset(int argc, VALUE *argv, VALUE ary) { long offset, beg, len; if (argc == 3) { - if (SYMBOL_P(argv[0])) { - rb_raise(rb_eTypeError, "Symbol as array index"); - } - if (SYMBOL_P(argv[1])) { - rb_raise(rb_eTypeError, "Symbol as subarray length"); - } - rb_ary_splice(ary, NUM2LONG(argv[0]), NUM2LONG(argv[1]), argv[2]); + rb_ary_modify_check(ary); + beg = NUM2LONG(argv[0]); + len = NUM2LONG(argv[1]); + rb_ary_splice(ary, beg, len, argv[2]); return argv[2]; } if (argc != 2) { rb_raise(rb_eArgError, "wrong number of arguments (%d for 2)", argc); } + rb_ary_modify_check(ary); if (FIXNUM_P(argv[0])) { offset = FIX2LONG(argv[0]); goto fixnum; } - if (SYMBOL_P(argv[0])) { - rb_raise(rb_eTypeError, "Symbol as array index"); - } - if (rb_range_beg_len(argv[0], &beg, &len, RARRAY(ary)->len, 1)) { + if (rb_range_beg_len(argv[0], &beg, &len, RARRAY_LEN(ary), 1)) { /* check if idx is Range */ rb_ary_splice(ary, beg, len, argv[1]); return argv[1]; @@ -1202,31 +1435,29 @@ fixnum: /* * call-seq: - * array.insert(index, obj...) -> array - * + * ary.insert(index, obj...) -> ary + * * Inserts the given values before the element with the given index * (which may be negative). - * + * * a = %w{ a b c d } * a.insert(2, 99) #=> ["a", "b", 99, "c", "d"] * a.insert(-2, 1, 2, 3) #=> ["a", "b", 99, "c", 1, 2, 3, "d"] */ static VALUE -rb_ary_insert(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_insert(int argc, VALUE *argv, VALUE ary) { long pos; - if (argc == 1) return ary; if (argc < 1) { rb_raise(rb_eArgError, "wrong number of arguments (at least 1)"); } + rb_ary_modify_check(ary); + if (argc == 1) return ary; pos = NUM2LONG(argv[0]); if (pos == -1) { - pos = RARRAY(ary)->len; + pos = RARRAY_LEN(ary); } if (pos < 0) { pos++; @@ -1237,55 +1468,61 @@ rb_ary_insert(argc, argv, ary) /* * call-seq: - * array.each {|item| block } -> array - * - * Calls <i>block</i> once for each element in <i>self</i>, passing that + * ary.each {|item| block } -> ary + * ary.each -> an_enumerator + * + * Calls <i>block</i> once for each element in +self+, passing that * element as a parameter. - * + * + * If no block is given, an enumerator is returned instead. + * * a = [ "a", "b", "c" ] * a.each {|x| print x, " -- " } - * + * * produces: - * + * * a -- b -- c -- */ VALUE -rb_ary_each(ary) - VALUE ary; +rb_ary_each(VALUE array) { long i; + volatile VALUE ary = array; RETURN_ENUMERATOR(ary, 0, 0); - for (i=0; i<RARRAY(ary)->len; i++) { - rb_yield(RARRAY(ary)->ptr[i]); + for (i=0; i<RARRAY_LEN(ary); i++) { + rb_yield(RARRAY_PTR(ary)[i]); } return ary; } /* * call-seq: - * array.each_index {|index| block } -> array - * + * ary.each_index {|index| block } -> ary + * ary.each_index -> an_enumerator + * * Same as <code>Array#each</code>, but passes the index of the element * instead of the element itself. - * + * + * If no block is given, an enumerator is returned instead. + * + * * a = [ "a", "b", "c" ] * a.each_index {|x| print x, " -- " } - * + * * produces: - * + * * 0 -- 1 -- 2 -- */ static VALUE -rb_ary_each_index(ary) - VALUE ary; +rb_ary_each_index(VALUE ary) { long i; - RETURN_ENUMERATOR(ary, 0, 0); - for (i=0; i<RARRAY(ary)->len; i++) { + + for (i=0; i<RARRAY_LEN(ary); i++) { rb_yield(LONG2NUM(i)); } return ary; @@ -1293,31 +1530,31 @@ rb_ary_each_index(ary) /* * call-seq: - * array.reverse_each {|item| block } - * - * Same as <code>Array#each</code>, but traverses <i>self</i> in reverse + * ary.reverse_each {|item| block } -> ary + * ary.reverse_each -> an_enumerator + * + * Same as <code>Array#each</code>, but traverses +self+ in reverse * order. - * + * * a = [ "a", "b", "c" ] * a.reverse_each {|x| print x, " " } - * + * * produces: - * + * * c b a */ static VALUE -rb_ary_reverse_each(ary) - VALUE ary; +rb_ary_reverse_each(VALUE ary) { long len; RETURN_ENUMERATOR(ary, 0, 0); - len = RARRAY(ary)->len; + len = RARRAY_LEN(ary); while (len--) { - rb_yield(RARRAY(ary)->ptr[len]); - if (RARRAY(ary)->len < len) { - len = RARRAY(ary)->len; + rb_yield(RARRAY_PTR(ary)[len]); + if (RARRAY_LEN(ary) < len) { + len = RARRAY_LEN(ary); } } return ary; @@ -1325,286 +1562,268 @@ rb_ary_reverse_each(ary) /* * call-seq: - * array.length -> int - * - * Returns the number of elements in <i>self</i>. May be zero. - * + * ary.length -> int + * + * Returns the number of elements in +self+. May be zero. + * * [ 1, 2, 3, 4, 5 ].length #=> 5 */ static VALUE -rb_ary_length(ary) - VALUE ary; +rb_ary_length(VALUE ary) { - return LONG2NUM(RARRAY(ary)->len); + long len = RARRAY_LEN(ary); + return LONG2NUM(len); } /* * call-seq: - * array.empty? -> true or false - * - * Returns <code>true</code> if <i>self</i> array contains no elements. - * + * ary.empty? -> true or false + * + * Returns <code>true</code> if +self+ contains no elements. + * * [].empty? #=> true */ static VALUE -rb_ary_empty_p(ary) - VALUE ary; +rb_ary_empty_p(VALUE ary) { - if (RARRAY(ary)->len == 0) + if (RARRAY_LEN(ary) == 0) return Qtrue; return Qfalse; } VALUE -rb_ary_dup(ary) - VALUE ary; +rb_ary_dup(VALUE ary) { - VALUE dup = rb_ary_new2(RARRAY(ary)->len); - - DUPSETUP(dup, ary); - MEMCPY(RARRAY(dup)->ptr, RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); - RARRAY(dup)->len = RARRAY(ary)->len; + VALUE dup = rb_ary_new2(RARRAY_LEN(ary)); + MEMCPY(RARRAY_PTR(dup), RARRAY_PTR(ary), VALUE, RARRAY_LEN(ary)); + ARY_SET_LEN(dup, RARRAY_LEN(ary)); return dup; } +VALUE +rb_ary_resurrect(VALUE ary) +{ + return rb_ary_new4(RARRAY_LEN(ary), RARRAY_PTR(ary)); +} + extern VALUE rb_output_fs; +static void ary_join_1(VALUE obj, VALUE ary, VALUE sep, long i, VALUE result, int *first); + static VALUE -inspect_join(ary, arg) - VALUE ary; - VALUE *arg; +recursive_join(VALUE obj, VALUE argp, int recur) { - return rb_ary_join(arg[0], arg[1]); + VALUE *arg = (VALUE *)argp; + VALUE ary = arg[0]; + VALUE sep = arg[1]; + VALUE result = arg[2]; + int *first = (int *)arg[3]; + + if (recur) { + rb_raise(rb_eArgError, "recursive array join"); + } + else { + ary_join_1(obj, ary, sep, 0, result, first); + } + return Qnil; } -VALUE -rb_ary_join(ary, sep) - VALUE ary, sep; +static void +ary_join_0(VALUE ary, VALUE sep, long max, VALUE result) { - long len = 1, i; - int taint = Qfalse; - VALUE result, tmp; - - if (RARRAY(ary)->len == 0) return rb_str_new(0, 0); - if (OBJ_TAINTED(ary) || OBJ_TAINTED(sep)) taint = Qtrue; + long i; + VALUE val; - for (i=0; i<RARRAY(ary)->len; i++) { - tmp = rb_check_string_type(RARRAY(ary)->ptr[i]); - len += NIL_P(tmp) ? 10 : RSTRING(tmp)->len; - } - if (!NIL_P(sep)) { - StringValue(sep); - len += RSTRING(sep)->len * (RARRAY(ary)->len - 1); + if (max > 0) rb_enc_copy(result, RARRAY_PTR(ary)[0]); + for (i=0; i<max; i++) { + val = RARRAY_PTR(ary)[i]; + if (i > 0 && !NIL_P(sep)) + rb_str_buf_append(result, sep); + rb_str_buf_append(result, val); + if (OBJ_TAINTED(val)) OBJ_TAINT(result); + if (OBJ_UNTRUSTED(val)) OBJ_TAINT(result); } - result = rb_str_buf_new(len); - for (i=0; i<RARRAY(ary)->len; i++) { - tmp = RARRAY(ary)->ptr[i]; - switch (TYPE(tmp)) { +} + +static void +ary_join_1(VALUE obj, VALUE ary, VALUE sep, long i, VALUE result, int *first) +{ + VALUE val, tmp; + + for (; i<RARRAY_LEN(ary); i++) { + if (i > 0 && !NIL_P(sep)) + rb_str_buf_append(result, sep); + + val = RARRAY_PTR(ary)[i]; + switch (TYPE(val)) { case T_STRING: + str_join: + rb_str_buf_append(result, val); + *first = FALSE; break; case T_ARRAY: - if (tmp == ary || rb_inspecting_p(tmp)) { - tmp = rb_str_new2("[...]"); + obj = val; + ary_join: + if (val == ary) { + rb_raise(rb_eArgError, "recursive array join"); } else { - VALUE args[2]; + VALUE args[4]; - args[0] = tmp; + args[0] = val; args[1] = sep; - tmp = rb_protect_inspect(inspect_join, ary, (VALUE)args); + args[2] = result; + args[3] = (VALUE)first; + rb_exec_recursive(recursive_join, obj, (VALUE)args); } break; default: - tmp = rb_obj_as_string(tmp); + tmp = rb_check_string_type(val); + if (!NIL_P(tmp)) { + val = tmp; + goto str_join; + } + tmp = rb_check_convert_type(val, T_ARRAY, "Array", "to_ary"); + if (!NIL_P(tmp)) { + obj = val; + val = tmp; + goto ary_join; + } + val = rb_obj_as_string(val); + if (*first) { + rb_enc_copy(result, val); + *first = FALSE; + } + goto str_join; } - if (i > 0 && !NIL_P(sep)) - rb_str_buf_append(result, sep); - rb_str_buf_append(result, tmp); - if (OBJ_TAINTED(tmp)) taint = Qtrue; } +} + +VALUE +rb_ary_join(VALUE ary, VALUE sep) +{ + long len = 1, i; + int taint = FALSE; + int untrust = FALSE; + VALUE val, tmp, result; + if (RARRAY_LEN(ary) == 0) return rb_usascii_str_new(0, 0); + if (OBJ_TAINTED(ary) || OBJ_TAINTED(sep)) taint = TRUE; + if (OBJ_UNTRUSTED(ary) || OBJ_UNTRUSTED(sep)) untrust = TRUE; + + if (!NIL_P(sep)) { + StringValue(sep); + len += RSTRING_LEN(sep) * (RARRAY_LEN(ary) - 1); + } + for (i=0; i<RARRAY_LEN(ary); i++) { + val = RARRAY_PTR(ary)[i]; + tmp = rb_check_string_type(val); + + if (NIL_P(tmp) || tmp != val) { + int first; + result = rb_str_buf_new(len + (RARRAY_LEN(ary)-i)*10); + rb_enc_associate(result, rb_usascii_encoding()); + if (taint) OBJ_TAINT(result); + if (untrust) OBJ_UNTRUST(result); + ary_join_0(ary, sep, i, result); + first = i == 0; + ary_join_1(ary, ary, sep, i, result, &first); + return result; + } + + len += RSTRING_LEN(tmp); + } + + result = rb_str_buf_new(len); if (taint) OBJ_TAINT(result); + if (untrust) OBJ_UNTRUST(result); + ary_join_0(ary, sep, RARRAY_LEN(ary), result); + return result; } /* * call-seq: - * array.join(sep=$,) -> str - * + * ary.join(sep=$,) -> str + * * Returns a string created by converting each element of the array to * a string, separated by <i>sep</i>. - * + * * [ "a", "b", "c" ].join #=> "abc" * [ "a", "b", "c" ].join("-") #=> "a-b-c" */ static VALUE -rb_ary_join_m(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_join_m(int argc, VALUE *argv, VALUE ary) { VALUE sep; rb_scan_args(argc, argv, "01", &sep); if (NIL_P(sep)) sep = rb_output_fs; - - return rb_ary_join(ary, sep); -} - -/* - * call-seq: - * array.to_s -> string - * - * Returns _self_<code>.join</code>. - * - * [ "a", "e", "i", "o" ].to_s #=> "aeio" - * - */ - -VALUE -rb_ary_to_s(ary) - VALUE ary; -{ - if (RARRAY(ary)->len == 0) return rb_str_new(0, 0); - - return rb_ary_join(ary, rb_output_fs); -} - -static ID inspect_key; - -struct inspect_arg { - VALUE (*func)(); - VALUE arg1, arg2; -}; - -static VALUE -inspect_call(arg) - struct inspect_arg *arg; -{ - return (*arg->func)(arg->arg1, arg->arg2); -} - -static VALUE -get_inspect_tbl(create) - int create; -{ - VALUE inspect_tbl = rb_thread_local_aref(rb_thread_current(), inspect_key); - - if (NIL_P(inspect_tbl)) { - if (create) { - tbl_init: - inspect_tbl = rb_ary_new(); - rb_thread_local_aset(rb_thread_current(), inspect_key, inspect_tbl); - } - } - else if (TYPE(inspect_tbl) != T_ARRAY) { - rb_warn("invalid inspect_tbl value"); - if (create) goto tbl_init; - rb_thread_local_aset(rb_thread_current(), inspect_key, Qnil); - return Qnil; - } - return inspect_tbl; -} - -static VALUE -inspect_ensure(obj) - VALUE obj; -{ - VALUE inspect_tbl; - - inspect_tbl = get_inspect_tbl(Qfalse); - if (!NIL_P(inspect_tbl)) { - rb_ary_pop(inspect_tbl); - } - return 0; -} - -VALUE -rb_protect_inspect(func, obj, arg) - VALUE (*func)(ANYARGS); - VALUE obj, arg; -{ - struct inspect_arg iarg; - VALUE inspect_tbl; - VALUE id; - - inspect_tbl = get_inspect_tbl(Qtrue); - id = rb_obj_id(obj); - if (rb_ary_includes(inspect_tbl, id)) { - return (*func)(obj, arg); - } - rb_ary_push(inspect_tbl, id); - iarg.func = func; - iarg.arg1 = obj; - iarg.arg2 = arg; - - return rb_ensure(inspect_call, (VALUE)&iarg, inspect_ensure, obj); -} -VALUE -rb_inspecting_p(obj) - VALUE obj; -{ - VALUE inspect_tbl; - - inspect_tbl = get_inspect_tbl(Qfalse); - if (NIL_P(inspect_tbl)) return Qfalse; - return rb_ary_includes(inspect_tbl, rb_obj_id(obj)); + return rb_ary_join(ary, sep); } static VALUE -inspect_ary(ary) - VALUE ary; +inspect_ary(VALUE ary, VALUE dummy, int recur) { int tainted = OBJ_TAINTED(ary); + int untrust = OBJ_UNTRUSTED(ary); long i; VALUE s, str; + if (recur) return rb_usascii_str_new_cstr("[...]"); str = rb_str_buf_new2("["); - for (i=0; i<RARRAY(ary)->len; i++) { - s = rb_inspect(RARRAY(ary)->ptr[i]); - if (OBJ_TAINTED(s)) tainted = Qtrue; + for (i=0; i<RARRAY_LEN(ary); i++) { + s = rb_inspect(RARRAY_PTR(ary)[i]); + if (OBJ_TAINTED(s)) tainted = TRUE; + if (OBJ_UNTRUSTED(s)) untrust = TRUE; if (i > 0) rb_str_buf_cat2(str, ", "); + else rb_enc_copy(str, s); rb_str_buf_append(str, s); } rb_str_buf_cat2(str, "]"); if (tainted) OBJ_TAINT(str); + if (untrust) OBJ_UNTRUST(str); return str; } /* * call-seq: - * array.inspect -> string + * ary.to_s -> string + * ary.inspect -> string * - * Create a printable version of <i>array</i>. + * Creates a string representation of +self+. */ static VALUE -rb_ary_inspect(ary) - VALUE ary; +rb_ary_inspect(VALUE ary) { - if (RARRAY(ary)->len == 0) return rb_str_new2("[]"); - if (rb_inspecting_p(ary)) return rb_str_new2("[...]"); - return rb_protect_inspect(inspect_ary, ary, 0); + if (RARRAY_LEN(ary) == 0) return rb_usascii_str_new2("[]"); + return rb_exec_recursive(inspect_ary, ary, 0); +} + +VALUE +rb_ary_to_s(VALUE ary) +{ + return rb_ary_inspect(ary); } /* * call-seq: - * array.to_a -> array - * - * Returns _self_. If called on a subclass of Array, converts + * ary.to_a -> ary + * + * Returns +self+. If called on a subclass of Array, converts * the receiver to an Array object. */ static VALUE -rb_ary_to_a(ary) - VALUE ary; +rb_ary_to_a(VALUE ary) { if (rb_obj_class(ary) != rb_cArray) { - VALUE dup = rb_ary_new2(RARRAY(ary)->len); + VALUE dup = rb_ary_new2(RARRAY_LEN(ary)); rb_ary_replace(dup, ary); return dup; } @@ -1613,268 +1832,448 @@ rb_ary_to_a(ary) /* * call-seq: - * array.to_ary -> array - * - * Returns _self_. + * ary.to_ary -> ary + * + * Returns +self+. */ static VALUE -rb_ary_to_ary_m(ary) - VALUE ary; +rb_ary_to_ary_m(VALUE ary) { return ary; } +static void +ary_reverse(p1, p2) + VALUE *p1, *p2; +{ + while (p1 < p2) { + VALUE tmp = *p1; + *p1++ = *p2; + *p2-- = tmp; + } +} + VALUE -rb_ary_reverse(ary) - VALUE ary; +rb_ary_reverse(VALUE ary) { VALUE *p1, *p2; - VALUE tmp; rb_ary_modify(ary); - if (RARRAY(ary)->len > 1) { - p1 = RARRAY(ary)->ptr; - p2 = p1 + RARRAY(ary)->len - 1; /* points last item */ - - while (p1 < p2) { - tmp = *p1; - *p1++ = *p2; - *p2-- = tmp; - } + if (RARRAY_LEN(ary) > 1) { + p1 = RARRAY_PTR(ary); + p2 = p1 + RARRAY_LEN(ary) - 1; /* points last item */ + ary_reverse(p1, p2); } return ary; } /* * call-seq: - * array.reverse! -> array - * - * Reverses _self_ in place. - * + * ary.reverse! -> ary + * + * Reverses +self+ in place. + * * a = [ "a", "b", "c" ] * a.reverse! #=> ["c", "b", "a"] * a #=> ["c", "b", "a"] */ static VALUE -rb_ary_reverse_bang(ary) - VALUE ary; +rb_ary_reverse_bang(VALUE ary) { return rb_ary_reverse(ary); } /* * call-seq: - * array.reverse -> an_array - * - * Returns a new array containing <i>self</i>'s elements in reverse order. - * + * ary.reverse -> new_ary + * + * Returns a new array containing +self+'s elements in reverse order. + * * [ "a", "b", "c" ].reverse #=> ["c", "b", "a"] * [ 1 ].reverse #=> [1] */ static VALUE -rb_ary_reverse_m(ary) - VALUE ary; +rb_ary_reverse_m(VALUE ary) +{ + long len = RARRAY_LEN(ary); + VALUE dup = rb_ary_new2(len); + + if (len > 0) { + VALUE *p1 = RARRAY_PTR(ary); + VALUE *p2 = RARRAY_PTR(dup) + len - 1; + do *p2-- = *p1++; while (--len > 0); + } + ARY_SET_LEN(dup, RARRAY_LEN(ary)); + return dup; +} + +static inline long +rotate_count(long cnt, long len) +{ + return (cnt < 0) ? (len - (~cnt % len) - 1) : (cnt % len); +} + +VALUE +rb_ary_rotate(VALUE ary, long cnt) +{ + rb_ary_modify(ary); + + if (cnt != 0) { + VALUE *ptr = RARRAY_PTR(ary); + long len = RARRAY_LEN(ary); + + if (len > 0 && (cnt = rotate_count(cnt, len)) > 0) { + --len; + if (cnt < len) ary_reverse(ptr + cnt, ptr + len); + if (--cnt > 0) ary_reverse(ptr, ptr + cnt); + if (len > 0) ary_reverse(ptr, ptr + len); + return ary; + } + } + + return Qnil; +} + +/* + * call-seq: + * ary.rotate!(cnt=1) -> ary + * + * Rotates +self+ in place so that the element at +cnt+ comes first, + * and returns +self+. If +cnt+ is negative then it rotates in + * the opposite direction. + * + * a = [ "a", "b", "c", "d" ] + * a.rotate! #=> ["b", "c", "d", "a"] + * a #=> ["b", "c", "d", "a"] + * a.rotate!(2) #=> ["d", "a", "b", "c"] + * a.rotate!(-3) #=> ["a", "b", "c", "d"] + */ + +static VALUE +rb_ary_rotate_bang(int argc, VALUE *argv, VALUE ary) +{ + long n = 1; + + switch (argc) { + case 1: n = NUM2LONG(argv[0]); + case 0: break; + default: rb_scan_args(argc, argv, "01", NULL); + } + rb_ary_rotate(ary, n); + return ary; +} + +/* + * call-seq: + * ary.rotate(cnt=1) -> new_ary + * + * Returns new array by rotating +self+ so that the element at + * +cnt+ in +self+ is the first element of the new array. If +cnt+ + * is negative then it rotates in the opposite direction. + * + * a = [ "a", "b", "c", "d" ] + * a.rotate #=> ["b", "c", "d", "a"] + * a #=> ["a", "b", "c", "d"] + * a.rotate(2) #=> ["c", "d", "a", "b"] + * a.rotate(-3) #=> ["b", "c", "d", "a"] + */ + +static VALUE +rb_ary_rotate_m(int argc, VALUE *argv, VALUE ary) { - return rb_ary_reverse(rb_ary_dup(ary)); + VALUE rotated, *ptr, *ptr2; + long len, cnt = 1; + + switch (argc) { + case 1: cnt = NUM2LONG(argv[0]); + case 0: break; + default: rb_scan_args(argc, argv, "01", NULL); + } + + len = RARRAY_LEN(ary); + rotated = rb_ary_new2(len); + if (len > 0) { + cnt = rotate_count(cnt, len); + ptr = RARRAY_PTR(ary); + ptr2 = RARRAY_PTR(rotated); + len -= cnt; + MEMCPY(ptr2, ptr + cnt, VALUE, len); + MEMCPY(ptr2 + len, ptr, VALUE, cnt); + } + ARY_SET_LEN(rotated, RARRAY_LEN(ary)); + return rotated; } struct ary_sort_data { VALUE ary; - VALUE *ptr; - long len; + int opt_methods; + int opt_inited; }; -static void -ary_sort_check(data) - struct ary_sort_data *data; +enum { + sort_opt_Fixnum, + sort_opt_String, + sort_optimizable_count +}; + +#define STRING_P(s) (TYPE(s) == T_STRING && CLASS_OF(s) == rb_cString) + +#define SORT_OPTIMIZABLE_BIT(type) (1U << TOKEN_PASTE(sort_opt_,type)) +#define SORT_OPTIMIZABLE(data, type) \ + (((data)->opt_inited & SORT_OPTIMIZABLE_BIT(type)) ? \ + ((data)->opt_methods & SORT_OPTIMIZABLE_BIT(type)) : \ + (((data)->opt_inited |= SORT_OPTIMIZABLE_BIT(type)), \ + rb_method_basic_definition_p(TOKEN_PASTE(rb_c,type), id_cmp) && \ + ((data)->opt_methods |= SORT_OPTIMIZABLE_BIT(type)))) + +static VALUE +sort_reentered(VALUE ary) { - if (RARRAY(data->ary)->ptr != data->ptr || RARRAY(data->ary)->len != data->len) { - rb_raise(rb_eArgError, "array modified during sort"); + if (RBASIC(ary)->klass) { + rb_raise(rb_eRuntimeError, "sort reentered"); } + return Qnil; } static int -sort_1(a, b, data) - VALUE *a, *b; - struct ary_sort_data *data; +sort_1(const void *ap, const void *bp, void *dummy) { - VALUE retval = rb_yield_values(2, *a, *b); + struct ary_sort_data *data = dummy; + VALUE retval = sort_reentered(data->ary); + VALUE a = *(const VALUE *)ap, b = *(const VALUE *)bp; int n; - n = rb_cmpint(retval, *a, *b); - ary_sort_check(data); + retval = rb_yield_values(2, a, b); + n = rb_cmpint(retval, a, b); + sort_reentered(data->ary); return n; } static int -sort_2(ap, bp, data) - VALUE *ap, *bp; - struct ary_sort_data *data; +sort_2(const void *ap, const void *bp, void *dummy) { - VALUE retval; - VALUE a = *ap, b = *bp; + struct ary_sort_data *data = dummy; + VALUE retval = sort_reentered(data->ary); + VALUE a = *(const VALUE *)ap, b = *(const VALUE *)bp; int n; - if (FIXNUM_P(a) && FIXNUM_P(b)) { + if (FIXNUM_P(a) && FIXNUM_P(b) && SORT_OPTIMIZABLE(data, Fixnum)) { if ((long)a > (long)b) return 1; if ((long)a < (long)b) return -1; return 0; } - if (TYPE(a) == T_STRING) { - if (TYPE(b) == T_STRING) return rb_str_cmp(a, b); + if (STRING_P(a) && STRING_P(b) && SORT_OPTIMIZABLE(data, String)) { + return rb_str_cmp(a, b); } retval = rb_funcall(a, id_cmp, 1, b); n = rb_cmpint(retval, a, b); - ary_sort_check(data); + sort_reentered(data->ary); return n; } -static VALUE -sort_internal(ary) - VALUE ary; -{ - struct ary_sort_data data; - - data.ary = ary; - data.ptr = RARRAY(ary)->ptr; data.len = RARRAY(ary)->len; - qsort(RARRAY(ary)->ptr, RARRAY(ary)->len, sizeof(VALUE), - rb_block_given_p()?sort_1:sort_2, &data); - return ary; -} - -static VALUE -sort_unlock(ary) - VALUE ary; -{ - FL_UNSET(ary, ARY_TMPLOCK); - return ary; -} - /* * call-seq: - * array.sort! -> array - * array.sort! {| a,b | block } -> array - * - * Sorts _self_. Comparisons for + * ary.sort! -> ary + * ary.sort! {| a,b | block } -> ary + * + * Sorts +self+. Comparisons for * the sort will be done using the <code><=></code> operator or using * an optional code block. The block implements a comparison between * <i>a</i> and <i>b</i>, returning -1, 0, or +1. See also * <code>Enumerable#sort_by</code>. - * + * * a = [ "d", "a", "e", "c", "b" ] - * a.sort #=> ["a", "b", "c", "d", "e"] - * a.sort {|x,y| y <=> x } #=> ["e", "d", "c", "b", "a"] + * a.sort! #=> ["a", "b", "c", "d", "e"] + * a.sort! {|x,y| y <=> x } #=> ["e", "d", "c", "b", "a"] */ VALUE -rb_ary_sort_bang(ary) - VALUE ary; +rb_ary_sort_bang(VALUE ary) { rb_ary_modify(ary); - if (RARRAY(ary)->len > 1) { - FL_SET(ary, ARY_TMPLOCK); /* prohibit modification during sort */ - rb_ensure(sort_internal, ary, sort_unlock, ary); + assert(!ARY_SHARED_P(ary)); + if (RARRAY_LEN(ary) > 1) { + VALUE tmp = ary_make_substitution(ary); /* only ary refers tmp */ + struct ary_sort_data data; + + RBASIC(tmp)->klass = 0; + data.ary = tmp; + data.opt_methods = 0; + data.opt_inited = 0; + ruby_qsort(RARRAY_PTR(tmp), RARRAY_LEN(tmp), sizeof(VALUE), + rb_block_given_p()?sort_1:sort_2, &data); + + if (ARY_EMBED_P(tmp)) { + assert(ARY_EMBED_P(tmp)); + if (ARY_SHARED_P(ary)) { /* ary might be destructively operated in the given block */ + rb_ary_unshare(ary); + } + FL_SET_EMBED(ary); + MEMCPY(RARRAY_PTR(ary), ARY_EMBED_PTR(tmp), VALUE, ARY_EMBED_LEN(tmp)); + ARY_SET_LEN(ary, ARY_EMBED_LEN(tmp)); + } + else { + assert(!ARY_EMBED_P(tmp)); + if (ARY_HEAP_PTR(ary) == ARY_HEAP_PTR(tmp)) { + assert(!ARY_EMBED_P(ary)); + FL_UNSET_SHARED(ary); + ARY_SET_CAPA(ary, ARY_CAPA(tmp)); + } + else { + assert(!ARY_SHARED_P(tmp)); + if (ARY_EMBED_P(ary)) { + FL_UNSET_EMBED(ary); + } + else if (ARY_SHARED_P(ary)) { + /* ary might be destructively operated in the given block */ + rb_ary_unshare(ary); + } + else { + xfree(ARY_HEAP_PTR(ary)); + } + ARY_SET_PTR(ary, RARRAY_PTR(tmp)); + ARY_SET_HEAP_LEN(ary, RARRAY_LEN(tmp)); + ARY_SET_CAPA(ary, ARY_CAPA(tmp)); + } + /* tmp was lost ownership for the ptr */ + FL_UNSET(tmp, FL_FREEZE); + FL_SET_EMBED(tmp); + ARY_SET_EMBED_LEN(tmp, 0); + FL_SET(tmp, FL_FREEZE); + } + /* tmp will be GC'ed. */ + RBASIC(tmp)->klass = rb_cArray; } return ary; } /* * call-seq: - * array.sort -> an_array - * array.sort {| a,b | block } -> an_array - * - * Returns a new array created by sorting <i>self</i>. Comparisons for + * ary.sort -> new_ary + * ary.sort {| a,b | block } -> new_ary + * + * Returns a new array created by sorting +self+. Comparisons for * the sort will be done using the <code><=></code> operator or using * an optional code block. The block implements a comparison between * <i>a</i> and <i>b</i>, returning -1, 0, or +1. See also * <code>Enumerable#sort_by</code>. - * + * * a = [ "d", "a", "e", "c", "b" ] * a.sort #=> ["a", "b", "c", "d", "e"] * a.sort {|x,y| y <=> x } #=> ["e", "d", "c", "b", "a"] */ VALUE -rb_ary_sort(ary) - VALUE ary; +rb_ary_sort(VALUE ary) { ary = rb_ary_dup(ary); rb_ary_sort_bang(ary); return ary; } + +static VALUE +sort_by_i(VALUE i) +{ + return rb_yield(i); +} + +/* + * call-seq: + * ary.sort_by! {| obj | block } -> ary + * ary.sort_by! -> an_enumerator + * + * Sorts +self+ in place using a set of keys generated by mapping the + * values in +self+ through the given block. + * + * If no block is given, an enumerator is returned instead. + * + */ + +static VALUE +rb_ary_sort_by_bang(VALUE ary) +{ + VALUE sorted; + + RETURN_ENUMERATOR(ary, 0, 0); + rb_ary_modify(ary); + sorted = rb_block_call(ary, rb_intern("sort_by"), 0, 0, sort_by_i, 0); + rb_ary_replace(ary, sorted); + return ary; +} + + /* * call-seq: - * array.collect {|item| block } -> an_array - * array.map {|item| block } -> an_array - * - * Invokes <i>block</i> once for each element of <i>self</i>. Creates a + * ary.collect {|item| block } -> new_ary + * ary.map {|item| block } -> new_ary + * ary.collect -> an_enumerator + * ary.map -> an_enumerator + * + * Invokes <i>block</i> once for each element of +self+. Creates a * new array containing the values returned by the block. * See also <code>Enumerable#collect</code>. - * + * + * If no block is given, an enumerator is returned instead. + * * a = [ "a", "b", "c", "d" ] * a.collect {|x| x + "!" } #=> ["a!", "b!", "c!", "d!"] * a #=> ["a", "b", "c", "d"] */ static VALUE -rb_ary_collect(ary) - VALUE ary; +rb_ary_collect(VALUE ary) { long i; VALUE collect; - if (!rb_block_given_p()) { - return rb_ary_new4(RARRAY(ary)->len, RARRAY(ary)->ptr); - } - - collect = rb_ary_new2(RARRAY(ary)->len); - for (i = 0; i < RARRAY(ary)->len; i++) { - rb_ary_push(collect, rb_yield(RARRAY(ary)->ptr[i])); + RETURN_ENUMERATOR(ary, 0, 0); + collect = rb_ary_new2(RARRAY_LEN(ary)); + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_ary_push(collect, rb_yield(RARRAY_PTR(ary)[i])); } return collect; } -/* + +/* * call-seq: - * array.collect! {|item| block } -> array - * array.map! {|item| block } -> array + * ary.collect! {|item| block } -> ary + * ary.map! {|item| block } -> ary + * ary.collect -> an_enumerator + * ary.map -> an_enumerator * - * Invokes the block once for each element of _self_, replacing the + * Invokes the block once for each element of +self+, replacing the * element with the value returned by _block_. * See also <code>Enumerable#collect</code>. - * + * + * If no block is given, an enumerator is returned instead. + * * a = [ "a", "b", "c", "d" ] * a.collect! {|x| x + "!" } * a #=> [ "a!", "b!", "c!", "d!" ] */ static VALUE -rb_ary_collect_bang(ary) - VALUE ary; +rb_ary_collect_bang(VALUE ary) { long i; RETURN_ENUMERATOR(ary, 0, 0); rb_ary_modify(ary); - for (i = 0; i < RARRAY(ary)->len; i++) { - rb_ary_store(ary, i, rb_yield(RARRAY(ary)->ptr[i])); + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_ary_store(ary, i, rb_yield(RARRAY_PTR(ary)[i])); } return ary; } VALUE -rb_values_at(obj, olen, argc, argv, func) - VALUE obj; - long olen; - int argc; - VALUE *argv; - VALUE (*func) _((VALUE,long)); +rb_get_values_at(VALUE obj, long olen, int argc, VALUE *argv, VALUE (*func) (VALUE, long)) { VALUE result = rb_ary_new2(argc); long beg, len, i, j; @@ -1901,15 +2300,15 @@ rb_values_at(obj, olen, argc, argv, func) return result; } -/* +/* * call-seq: - * array.values_at(selector,... ) -> an_array + * ary.values_at(selector,... ) -> new_ary * * Returns an array containing the elements in - * _self_ corresponding to the given selector(s). The selectors - * may be either integer indices or ranges. + * +self+ corresponding to the given selector(s). The selectors + * may be either integer indices or ranges. * See also <code>Array#select</code>. - * + * * a = %w{ a b c d e f } * a.values_at(1, 3, 5) * a.values_at(1, 3, 5, 7) @@ -1918,37 +2317,37 @@ rb_values_at(obj, olen, argc, argv, func) */ static VALUE -rb_ary_values_at(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_values_at(int argc, VALUE *argv, VALUE ary) { - return rb_values_at(ary, RARRAY(ary)->len, argc, argv, rb_ary_entry); + return rb_get_values_at(ary, RARRAY_LEN(ary), argc, argv, rb_ary_entry); } + /* * call-seq: - * array.select {|item| block } -> an_array - * - * Invokes the block passing in successive elements from <i>array</i>, + * ary.select {|item| block } -> new_ary + * ary.select -> an_enumerator + * + * Invokes the block passing in successive elements from +self+, * returning an array containing those elements for which the block * returns a true value (equivalent to <code>Enumerable#select</code>). - * + * + * If no block is given, an enumerator is returned instead. + * * a = %w{ a b c d e f } * a.select {|v| v =~ /[aeiou]/} #=> ["a", "e"] */ static VALUE -rb_ary_select(ary) - VALUE ary; +rb_ary_select(VALUE ary) { VALUE result; long i; RETURN_ENUMERATOR(ary, 0, 0); - result = rb_ary_new2(RARRAY(ary)->len); - for (i = 0; i < RARRAY(ary)->len; i++) { - if (RTEST(rb_yield(RARRAY(ary)->ptr[i]))) { + result = rb_ary_new2(RARRAY_LEN(ary)); + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (RTEST(rb_yield(RARRAY_PTR(ary)[i]))) { rb_ary_push(result, rb_ary_elt(ary, i)); } } @@ -1957,14 +2356,76 @@ rb_ary_select(ary) /* * call-seq: - * array.delete(obj) -> obj or nil - * array.delete(obj) { block } -> obj or nil - * - * Deletes items from <i>self</i> that are equal to <i>obj</i>. If + * ary.select! {|item| block } -> ary or nil + * ary.select! -> an_enumerator + * + * Invokes the block passing in successive elements from + * +self+, deleting elements for which the block returns a + * false value. It returns +self+ if changes were made, + * otherwise it returns <code>nil</code>. + * See also <code>Array#keep_if</code> + * + * If no block is given, an enumerator is returned instead. + * + */ + +static VALUE +rb_ary_select_bang(VALUE ary) +{ + long i1, i2; + + RETURN_ENUMERATOR(ary, 0, 0); + rb_ary_modify(ary); + for (i1 = i2 = 0; i1 < RARRAY_LEN(ary); i1++) { + VALUE v = RARRAY_PTR(ary)[i1]; + if (!RTEST(rb_yield(v))) continue; + if (i1 != i2) { + rb_ary_store(ary, i2, v); + } + i2++; + } + + if (RARRAY_LEN(ary) == i2) return Qnil; + if (i2 < RARRAY_LEN(ary)) + ARY_SET_LEN(ary, i2); + return ary; +} + +/* + * call-seq: + * ary.keep_if {|item| block } -> ary + * ary.keep_if -> an_enumerator + * + * Deletes every element of +self+ for which <i>block</i> evaluates + * to false. + * See also <code>Array#select!</code> + * + * If no block is given, an enumerator is returned instead. + * + * a = %w{ a b c d e f } + * a.keep_if {|v| v =~ /[aeiou]/} #=> ["a", "e"] + */ + +static VALUE +rb_ary_keep_if(VALUE ary) +{ + RETURN_ENUMERATOR(ary, 0, 0); + rb_ary_select_bang(ary); + return ary; +} + +/* + * call-seq: + * ary.delete(obj) -> obj or nil + * ary.delete(obj) { block } -> obj or nil + * + * Deletes items from +self+ that are equal to <i>obj</i>. + * If any items are found, returns <i>obj</i>. If * the item is not found, returns <code>nil</code>. If the optional * code block is given, returns the result of <i>block</i> if the item - * is not found. - * + * is not found. (To remove <code>nil</code> elements and + * get an informative return value, use #compact!) + * * a = [ "a", "b", "b", "b", "c" ] * a.delete("b") #=> "b" * a #=> ["a", "c"] @@ -1973,22 +2434,24 @@ rb_ary_select(ary) */ VALUE -rb_ary_delete(ary, item) - VALUE ary; - VALUE item; +rb_ary_delete(VALUE ary, VALUE item) { + VALUE v = item; long i1, i2; - for (i1 = i2 = 0; i1 < RARRAY(ary)->len; i1++) { - VALUE e = RARRAY(ary)->ptr[i1]; + for (i1 = i2 = 0; i1 < RARRAY_LEN(ary); i1++) { + VALUE e = RARRAY_PTR(ary)[i1]; - if (rb_equal(e, item)) continue; + if (rb_equal(e, item)) { + v = e; + continue; + } if (i1 != i2) { rb_ary_store(ary, i2, e); } i2++; } - if (RARRAY(ary)->len == i2) { + if (RARRAY_LEN(ary) == i2) { if (rb_block_given_p()) { return rb_yield(item); } @@ -1996,24 +2459,21 @@ rb_ary_delete(ary, item) } rb_ary_modify(ary); - if (RARRAY(ary)->len > i2) { - RARRAY(ary)->len = i2; - if (i2 * 2 < RARRAY(ary)->aux.capa && - RARRAY(ary)->aux.capa > ARY_DEFAULT_SIZE) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, i2 * 2); - RARRAY(ary)->aux.capa = i2 * 2; + if (RARRAY_LEN(ary) > i2) { + ARY_SET_LEN(ary, i2); + if (i2 * 2 < ARY_CAPA(ary) && + ARY_CAPA(ary) > ARY_DEFAULT_SIZE) { + ary_resize_capa(ary, i2*2); } } - return item; + return v; } VALUE -rb_ary_delete_at(ary, pos) - VALUE ary; - long pos; +rb_ary_delete_at(VALUE ary, long pos) { - long i, len = RARRAY(ary)->len; + long len = RARRAY_LEN(ary); VALUE del; if (pos >= len) return Qnil; @@ -2023,23 +2483,22 @@ rb_ary_delete_at(ary, pos) } rb_ary_modify(ary); - del = RARRAY(ary)->ptr[pos]; - for (i = pos + 1; i < len; i++, pos++) { - RARRAY(ary)->ptr[pos] = RARRAY(ary)->ptr[i]; - } - RARRAY(ary)->len = pos; + del = RARRAY_PTR(ary)[pos]; + MEMMOVE(RARRAY_PTR(ary)+pos, RARRAY_PTR(ary)+pos+1, VALUE, + RARRAY_LEN(ary)-pos-1); + ARY_INCREASE_LEN(ary, -1); return del; } /* * call-seq: - * array.delete_at(index) -> obj or nil - * + * ary.delete_at(index) -> obj or nil + * * Deletes the element at the specified index, returning that element, * or <code>nil</code> if the index is out of range. See also * <code>Array#slice!</code>. - * + * * a = %w( ant bat cat dog ) * a.delete_at(2) #=> "cat" * a #=> ["ant", "bat", "dog"] @@ -2047,28 +2506,21 @@ rb_ary_delete_at(ary, pos) */ static VALUE -rb_ary_delete_at_m(ary, pos) - VALUE ary, pos; +rb_ary_delete_at_m(VALUE ary, VALUE pos) { return rb_ary_delete_at(ary, NUM2LONG(pos)); } /* * call-seq: - * array.slice!(index) -> obj or nil - * array.slice!(start, length) -> sub_array or nil - * array.slice!(range) -> sub_array or nil - * + * ary.slice!(index) -> obj or nil + * ary.slice!(start, length) -> new_ary or nil + * ary.slice!(range) -> new_ary or nil + * * Deletes the element(s) given by an index (optionally with a length) - * or by a range. Returns the deleted object, subarray, or - * <code>nil</code> if the index is out of range. Equivalent to: - * - * def slice!(*args) - * result = self[*args] - * self[*args] = nil - * result - * end - * + * or by a range. Returns the deleted object (or objects), or + * <code>nil</code> if the index is out of range. + * * a = [ "a", "b", "c" ] * a.slice!(1) #=> "b" * a #=> ["a", "c"] @@ -2079,18 +2531,15 @@ rb_ary_delete_at_m(ary, pos) */ static VALUE -rb_ary_slice_bang(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_slice_bang(int argc, VALUE *argv, VALUE ary) { VALUE arg1, arg2; long pos, len, orig_len; rb_ary_modify_check(ary); - if (rb_scan_args(argc, argv, "11", &arg1, &arg2) == 2) { - pos = NUM2LONG(arg1); - len = NUM2LONG(arg2); + if (argc == 2) { + pos = NUM2LONG(argv[0]); + len = NUM2LONG(argv[1]); delete_pos_len: if (len < 0) return Qnil; orig_len = RARRAY_LEN(ary); @@ -2105,10 +2554,16 @@ rb_ary_slice_bang(argc, argv, ary) if (len == 0) return rb_ary_new2(0); arg2 = rb_ary_new4(len, RARRAY_PTR(ary)+pos); RBASIC(arg2)->klass = rb_obj_class(ary); - rb_ary_splice(ary, pos, len, Qnil); /* Qundef in 1.9 */ + rb_ary_splice(ary, pos, len, Qundef); return arg2; } + if (argc != 1) { + /* error report */ + rb_scan_args(argc, argv, "11", NULL, NULL); + } + arg1 = argv[0]; + if (!FIXNUM_P(arg1)) { switch (rb_range_beg_len(arg1, &pos, &len, RARRAY_LEN(ary), 0)) { case Qtrue: @@ -2126,170 +2581,219 @@ rb_ary_slice_bang(argc, argv, ary) return rb_ary_delete_at(ary, NUM2LONG(arg1)); } +static VALUE +ary_reject(VALUE orig, VALUE result) +{ + long i; + + for (i = 0; i < RARRAY_LEN(orig); i++) { + VALUE v = RARRAY_PTR(orig)[i]; + if (!RTEST(rb_yield(v))) { + rb_ary_push_1(result, v); + } + } + return result; +} + +static VALUE +ary_reject_bang(VALUE ary) +{ + long i; + VALUE result = Qnil; + + rb_ary_modify_check(ary); + for (i = 0; i < RARRAY_LEN(ary); ) { + VALUE v = RARRAY_PTR(ary)[i]; + if (RTEST(rb_yield(v))) { + rb_ary_delete_at(ary, i); + result = ary; + } + else { + i++; + } + } + return result; +} + /* * call-seq: - * array.reject! {|item| block } -> array or nil - * + * ary.reject! {|item| block } -> ary or nil + * ary.reject! -> an_enumerator + * * Equivalent to <code>Array#delete_if</code>, deleting elements from - * _self_ for which the block evaluates to true, but returns - * <code>nil</code> if no changes were made. Also see - * <code>Enumerable#reject</code>. + * +self+ for which the block evaluates to true, but returns + * <code>nil</code> if no changes were made. + * The array is changed instantly every time the block is called and + * not after the iteration is over. + * See also <code>Enumerable#reject</code> and <code>Array#delete_if</code>. + * + * If no block is given, an enumerator is returned instead. + * */ static VALUE -rb_ary_reject_bang(ary) - VALUE ary; +rb_ary_reject_bang(VALUE ary) { - long i1, i2; - RETURN_ENUMERATOR(ary, 0, 0); - rb_ary_modify(ary); - for (i1 = i2 = 0; i1 < RARRAY(ary)->len; i1++) { - VALUE v = RARRAY(ary)->ptr[i1]; - if (RTEST(rb_yield(v))) continue; - if (i1 != i2) { - rb_ary_store(ary, i2, v); - } - i2++; - } - if (RARRAY(ary)->len == i2) return Qnil; - if (i2 < RARRAY(ary)->len) - RARRAY(ary)->len = i2; - - return ary; + return ary_reject_bang(ary); } /* * call-seq: - * array.reject {|item| block } -> an_array - * - * Returns a new array containing the items in _self_ + * ary.reject {|item| block } -> new_ary + * ary.reject -> an_enumerator + * + * Returns a new array containing the items in +self+ * for which the block is not true. + * See also <code>Array#delete_if</code> + * + * If no block is given, an enumerator is returned instead. + * */ static VALUE -rb_ary_reject(ary) - VALUE ary; +rb_ary_reject(VALUE ary) { + VALUE rejected_ary; + RETURN_ENUMERATOR(ary, 0, 0); - ary = rb_ary_dup(ary); - rb_ary_reject_bang(ary); - return ary; + rejected_ary = rb_ary_new(); + ary_reject(ary, rejected_ary); + return rejected_ary; } /* * call-seq: - * array.delete_if {|item| block } -> array - * - * Deletes every element of <i>self</i> for which <i>block</i> evaluates - * to <code>true</code>. - * + * ary.delete_if {|item| block } -> ary + * ary.delete_if -> an_enumerator + * + * Deletes every element of +self+ for which <i>block</i> evaluates + * to true. + * The array is changed instantly every time the block is called and + * not after the iteration is over. + * See also <code>Array#reject!</code> + * + * If no block is given, an enumerator is returned instead. + * * a = [ "a", "b", "c" ] * a.delete_if {|x| x >= "b" } #=> ["a"] */ static VALUE -rb_ary_delete_if(ary) - VALUE ary; +rb_ary_delete_if(VALUE ary) { RETURN_ENUMERATOR(ary, 0, 0); - rb_ary_reject_bang(ary); + ary_reject_bang(ary); return ary; } +static VALUE +take_i(VALUE val, VALUE *args, int argc, VALUE *argv) +{ + if (args[1]-- == 0) rb_iter_break(); + if (argc > 1) val = rb_ary_new4(argc, argv); + rb_ary_push(args[0], val); + return Qnil; +} + +static VALUE +take_items(VALUE obj, long n) +{ + VALUE result = rb_check_array_type(obj); + VALUE args[2]; + + if (!NIL_P(result)) return rb_ary_subseq(result, 0, n); + result = rb_ary_new2(n); + args[0] = result; args[1] = (VALUE)n; + rb_block_call(obj, rb_intern("each"), 0, 0, take_i, (VALUE)args); + return result; +} + + /* * call-seq: - * array.zip(arg, ...) -> an_array - * array.zip(arg, ...) {| arr | block } -> nil - * + * ary.zip(arg, ...) -> new_ary + * ary.zip(arg, ...) {| arr | block } -> nil + * * Converts any arguments to arrays, then merges elements of - * <i>self</i> with corresponding elements from each argument. This + * +self+ with corresponding elements from each argument. This * generates a sequence of <code>self.size</code> <em>n</em>-element * arrays, where <em>n</em> is one more that the count of arguments. If * the size of any argument is less than <code>enumObj.size</code>, - * <code>nil</code> values are supplied. If a block given, it is + * <code>nil</code> values are supplied. If a block is given, it is * invoked for each output array, otherwise an array of arrays is * returned. - * + * * a = [ 4, 5, 6 ] * b = [ 7, 8, 9 ] - * * [1,2,3].zip(a, b) #=> [[1, 4, 7], [2, 5, 8], [3, 6, 9]] * [1,2].zip(a,b) #=> [[1, 4, 7], [2, 5, 8]] * a.zip([1,2],[8]) #=> [[4,1,8], [5,2,nil], [6,nil,nil]] */ static VALUE -rb_ary_zip(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_zip(int argc, VALUE *argv, VALUE ary) { int i, j; long len; - VALUE result; + VALUE result = Qnil; + len = RARRAY_LEN(ary); for (i=0; i<argc; i++) { - argv[i] = to_ary(argv[i]); + argv[i] = take_items(argv[i], len); } - if (rb_block_given_p()) { - for (i=0; i<RARRAY(ary)->len; i++) { - VALUE tmp = rb_ary_new2(argc+1); - - rb_ary_push(tmp, rb_ary_elt(ary, i)); - for (j=0; j<argc; j++) { - rb_ary_push(tmp, rb_ary_elt(argv[j], i)); - } - rb_yield(tmp); - } - return Qnil; + if (!rb_block_given_p()) { + result = rb_ary_new2(len); } - len = RARRAY(ary)->len; - result = rb_ary_new2(len); - for (i=0; i<len; i++) { + + for (i=0; i<RARRAY_LEN(ary); i++) { VALUE tmp = rb_ary_new2(argc+1); rb_ary_push(tmp, rb_ary_elt(ary, i)); for (j=0; j<argc; j++) { rb_ary_push(tmp, rb_ary_elt(argv[j], i)); } - rb_ary_push(result, tmp); + if (NIL_P(result)) { + rb_yield(tmp); + } + else { + rb_ary_push(result, tmp); + } } return result; } /* * call-seq: - * array.transpose -> an_array - * - * Assumes that <i>self</i> is an array of arrays and transposes the + * ary.transpose -> new_ary + * + * Assumes that +self+ is an array of arrays and transposes the * rows and columns. - * + * * a = [[1,2], [3,4], [5,6]] * a.transpose #=> [[1, 3, 5], [2, 4, 6]] */ static VALUE -rb_ary_transpose(ary) - VALUE ary; +rb_ary_transpose(VALUE ary) { long elen = -1, alen, i, j; VALUE tmp, result = 0; - alen = RARRAY(ary)->len; + alen = RARRAY_LEN(ary); if (alen == 0) return rb_ary_dup(ary); for (i=0; i<alen; i++) { tmp = to_ary(rb_ary_elt(ary, i)); if (elen < 0) { /* first element */ - elen = RARRAY(tmp)->len; + elen = RARRAY_LEN(tmp); result = rb_ary_new2(elen); for (j=0; j<elen; j++) { rb_ary_store(result, j, rb_ary_new2(alen)); } } - else if (elen != RARRAY(tmp)->len) { - rb_raise(rb_eIndexError, "element size differs (%d should be %d)", - RARRAY(tmp)->len, elen); + else if (elen != RARRAY_LEN(tmp)) { + rb_raise(rb_eIndexError, "element size differs (%ld should be %ld)", + RARRAY_LEN(tmp), elen); } for (j=0; j<elen; j++) { rb_ary_store(rb_ary_elt(result, j), i, rb_ary_elt(tmp, j)); @@ -2300,75 +2804,102 @@ rb_ary_transpose(ary) /* * call-seq: - * array.replace(other_array) -> array - * - * Replaces the contents of <i>self</i> with the contents of - * <i>other_array</i>, truncating or expanding if necessary. - * + * ary.replace(other_ary) -> ary + * + * Replaces the contents of +self+ with the contents of + * <i>other_ary</i>, truncating or expanding if necessary. + * * a = [ "a", "b", "c", "d", "e" ] * a.replace([ "x", "y", "z" ]) #=> ["x", "y", "z"] * a #=> ["x", "y", "z"] */ -static VALUE -rb_ary_replace(copy, orig) - VALUE copy, orig; +VALUE +rb_ary_replace(VALUE copy, VALUE orig) { - VALUE shared; - - rb_ary_modify(copy); + rb_ary_modify_check(copy); orig = to_ary(orig); if (copy == orig) return copy; - shared = ary_make_shared(orig); - if (RARRAY(copy)->ptr && !FL_TEST(copy, ELTS_SHARED)) - free(RARRAY(copy)->ptr); - RARRAY(copy)->ptr = RARRAY(orig)->ptr; - RARRAY(copy)->len = RARRAY(orig)->len; - RARRAY(copy)->aux.shared = shared; - FL_SET(copy, ELTS_SHARED); + if (RARRAY_LEN(orig) <= RARRAY_EMBED_LEN_MAX) { + VALUE *ptr; + VALUE shared = 0; + + if (ARY_OWNS_HEAP_P(copy)) { + xfree(RARRAY_PTR(copy)); + } + else if (ARY_SHARED_P(copy)) { + shared = ARY_SHARED(copy); + FL_UNSET_SHARED(copy); + } + FL_SET_EMBED(copy); + ptr = RARRAY_PTR(orig); + MEMCPY(RARRAY_PTR(copy), ptr, VALUE, RARRAY_LEN(orig)); + if (shared) { + rb_ary_decrement_share(shared); + } + ARY_SET_LEN(copy, RARRAY_LEN(orig)); + } + else { + VALUE shared = ary_make_shared(orig); + if (ARY_OWNS_HEAP_P(copy)) { + xfree(RARRAY_PTR(copy)); + } + else { + rb_ary_unshare_safe(copy); + } + FL_UNSET_EMBED(copy); + ARY_SET_PTR(copy, RARRAY_PTR(orig)); + ARY_SET_LEN(copy, RARRAY_LEN(orig)); + rb_ary_set_shared(copy, shared); + } return copy; } -/* +/* * call-seq: - * array.clear -> array + * ary.clear -> ary * - * Removes all elements from _self_. + * Removes all elements from +self+. * * a = [ "a", "b", "c", "d", "e" ] * a.clear #=> [ ] */ VALUE -rb_ary_clear(ary) - VALUE ary; +rb_ary_clear(VALUE ary) { - rb_ary_modify(ary); - RARRAY(ary)->len = 0; - if (ARY_DEFAULT_SIZE * 2 < RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, ARY_DEFAULT_SIZE * 2); - RARRAY(ary)->aux.capa = ARY_DEFAULT_SIZE * 2; + rb_ary_modify_check(ary); + ARY_SET_LEN(ary, 0); + if (ARY_SHARED_P(ary)) { + if (!ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + FL_SET_EMBED(ary); + } + } + else if (ARY_DEFAULT_SIZE * 2 < ARY_CAPA(ary)) { + ary_resize_capa(ary, ARY_DEFAULT_SIZE * 2); } return ary; } /* * call-seq: - * array.fill(obj) -> array - * array.fill(obj, start [, length]) -> array - * array.fill(obj, range ) -> array - * array.fill {|index| block } -> array - * array.fill(start [, length] ) {|index| block } -> array - * array.fill(range) {|index| block } -> array - * - * The first three forms set the selected elements of <i>self</i> (which + * ary.fill(obj) -> ary + * ary.fill(obj, start [, length]) -> ary + * ary.fill(obj, range ) -> ary + * ary.fill {|index| block } -> ary + * ary.fill(start [, length] ) {|index| block } -> ary + * ary.fill(range) {|index| block } -> ary + * + * The first three forms set the selected elements of +self+ (which * may be the entire array) to <i>obj</i>. A <i>start</i> of * <code>nil</code> is equivalent to zero. A <i>length</i> of * <code>nil</code> is equivalent to <i>self.length</i>. The last three * forms fill the array with the value of the block. The block is * passed the absolute index of each element to be filled. - * + * Negative values of <i>start</i> count from the end of the array. + * * a = [ "a", "b", "c", "d" ] * a.fill("x") #=> ["x", "x", "x", "x"] * a.fill("z", 2, 2) #=> ["x", "x", "z", "z"] @@ -2378,18 +2909,15 @@ rb_ary_clear(ary) */ static VALUE -rb_ary_fill(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_fill(int argc, VALUE *argv, VALUE ary) { VALUE item, arg1, arg2; long beg = 0, end = 0, len = 0; VALUE *p, *pend; - int block_p = Qfalse; + int block_p = FALSE; if (rb_block_given_p()) { - block_p = Qtrue; + block_p = TRUE; rb_scan_args(argc, argv, "02", &arg1, &arg2); argc += 1; /* hackish */ } @@ -2399,20 +2927,20 @@ rb_ary_fill(argc, argv, ary) switch (argc) { case 1: beg = 0; - len = RARRAY(ary)->len; + len = RARRAY_LEN(ary); break; case 2: - if (rb_range_beg_len(arg1, &beg, &len, RARRAY(ary)->len, 1)) { + if (rb_range_beg_len(arg1, &beg, &len, RARRAY_LEN(ary), 1)) { break; } /* fall through */ case 3: beg = NIL_P(arg1) ? 0 : NUM2LONG(arg1); if (beg < 0) { - beg = RARRAY(ary)->len + beg; + beg = RARRAY_LEN(ary) + beg; if (beg < 0) beg = 0; } - len = NIL_P(arg2) ? RARRAY(ary)->len - beg : NUM2LONG(arg2); + len = NIL_P(arg2) ? RARRAY_LEN(ary) - beg : NUM2LONG(arg2); break; } rb_ary_modify(ary); @@ -2423,13 +2951,12 @@ rb_ary_fill(argc, argv, ary) rb_raise(rb_eArgError, "argument too big"); } end = beg + len; - if (end > RARRAY(ary)->len) { - if (end >= RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, end); - RARRAY(ary)->aux.capa = end; + if (RARRAY_LEN(ary) < end) { + if (end >= ARY_CAPA(ary)) { + ary_resize_capa(ary, end); } - rb_mem_clear(RARRAY(ary)->ptr + RARRAY(ary)->len, end - RARRAY(ary)->len); - RARRAY(ary)->len = end; + rb_mem_clear(RARRAY_PTR(ary) + RARRAY_LEN(ary), end - RARRAY_LEN(ary)); + ARY_SET_LEN(ary, end); } if (block_p) { @@ -2438,12 +2965,12 @@ rb_ary_fill(argc, argv, ary) for (i=beg; i<end; i++) { v = rb_yield(LONG2NUM(i)); - if (i>=RARRAY(ary)->len) break; - RARRAY(ary)->ptr[i] = v; + if (i>=RARRAY_LEN(ary)) break; + RARRAY_PTR(ary)[i] = v; } } else { - p = RARRAY(ary)->ptr + beg; + p = RARRAY_PTR(ary) + beg; pend = p + len; while (p < pend) { *p++ = item; @@ -2452,62 +2979,61 @@ rb_ary_fill(argc, argv, ary) return ary; } -/* +/* * call-seq: - * array + other_array -> an_array + * ary + other_ary -> new_ary * * Concatenation---Returns a new array built by concatenating the * two arrays together to produce a third array. - * + * * [ 1, 2, 3 ] + [ 4, 5 ] #=> [ 1, 2, 3, 4, 5 ] */ VALUE -rb_ary_plus(x, y) - VALUE x, y; +rb_ary_plus(VALUE x, VALUE y) { VALUE z; long len; y = to_ary(y); - len = RARRAY(x)->len + RARRAY(y)->len; + len = RARRAY_LEN(x) + RARRAY_LEN(y); z = rb_ary_new2(len); - MEMCPY(RARRAY(z)->ptr, RARRAY(x)->ptr, VALUE, RARRAY(x)->len); - MEMCPY(RARRAY(z)->ptr + RARRAY(x)->len, RARRAY(y)->ptr, VALUE, RARRAY(y)->len); - RARRAY(z)->len = len; + MEMCPY(RARRAY_PTR(z), RARRAY_PTR(x), VALUE, RARRAY_LEN(x)); + MEMCPY(RARRAY_PTR(z) + RARRAY_LEN(x), RARRAY_PTR(y), VALUE, RARRAY_LEN(y)); + ARY_SET_LEN(z, len); return z; } -/* +/* * call-seq: - * array.concat(other_array) -> array + * ary.concat(other_ary) -> ary + * + * Appends the elements of <i>other_ary</i> to +self+. * - * Appends the elements in other_array to _self_. - * * [ "a", "b" ].concat( ["c", "d"] ) #=> [ "a", "b", "c", "d" ] */ VALUE -rb_ary_concat(x, y) - VALUE x, y; +rb_ary_concat(VALUE x, VALUE y) { + rb_ary_modify_check(x); y = to_ary(y); - if (RARRAY(y)->len > 0) { - rb_ary_splice(x, RARRAY(x)->len, 0, y); + if (RARRAY_LEN(y) > 0) { + rb_ary_splice(x, RARRAY_LEN(x), 0, y); } return x; } -/* +/* * call-seq: - * array * int -> an_array - * array * str -> a_string + * ary * int -> new_ary + * ary * str -> new_string * * Repetition---With a String argument, equivalent to * self.join(str). Otherwise, returns a new array - * built by concatenating the _int_ copies of _self_. + * built by concatenating the _int_ copies of +self+. * * * [ 1, 2, 3 ] * 3 #=> [ 1, 2, 3, 1, 2, 3, 1, 2, 3 ] @@ -2516,11 +3042,10 @@ rb_ary_concat(x, y) */ static VALUE -rb_ary_times(ary, times) - VALUE ary, times; +rb_ary_times(VALUE ary, VALUE times) { - VALUE ary2, tmp; - long i, len; + VALUE ary2, tmp, *ptr, *ptr2; + long t, len; tmp = rb_check_string_type(times); if (!NIL_P(tmp)) { @@ -2528,29 +3053,43 @@ rb_ary_times(ary, times) } len = NUM2LONG(times); - if (len == 0) return ary_new(rb_obj_class(ary), 0); + if (len == 0) { + ary2 = ary_new(rb_obj_class(ary), 0); + goto out; + } if (len < 0) { rb_raise(rb_eArgError, "negative argument"); } - if (ARY_MAX_SIZE/len < RARRAY(ary)->len) { + if (ARY_MAX_SIZE/len < RARRAY_LEN(ary)) { rb_raise(rb_eArgError, "argument too big"); } - len *= RARRAY(ary)->len; + len *= RARRAY_LEN(ary); ary2 = ary_new(rb_obj_class(ary), len); - RARRAY(ary2)->len = len; - - for (i=0; i<len; i+=RARRAY(ary)->len) { - MEMCPY(RARRAY(ary2)->ptr+i, RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); + ARY_SET_LEN(ary2, len); + + ptr = RARRAY_PTR(ary); + ptr2 = RARRAY_PTR(ary2); + t = RARRAY_LEN(ary); + if (0 < t) { + MEMCPY(ptr2, ptr, VALUE, t); + while (t <= len/2) { + MEMCPY(ptr2+t, ptr2, VALUE, t); + t *= 2; + } + if (t < len) { + MEMCPY(ptr2+t, ptr2, VALUE, len-t); + } } + out: OBJ_INFECT(ary2, ary); return ary2; } -/* +/* * call-seq: - * array.assoc(obj) -> an_array or nil + * ary.assoc(obj) -> new_ary or nil * * Searches through an array whose elements are also arrays * comparing _obj_ with the first element of each contained array @@ -2569,16 +3108,15 @@ rb_ary_times(ary, times) */ VALUE -rb_ary_assoc(ary, key) - VALUE ary, key; +rb_ary_assoc(VALUE ary, VALUE key) { long i; VALUE v; - for (i = 0; i < RARRAY(ary)->len; ++i) { - v = rb_check_array_type(RARRAY(ary)->ptr[i]); - if (!NIL_P(v) && RARRAY(v)->len > 0 && - rb_equal(RARRAY(v)->ptr[0], key)) + for (i = 0; i < RARRAY_LEN(ary); ++i) { + v = rb_check_array_type(RARRAY_PTR(ary)[i]); + if (!NIL_P(v) && RARRAY_LEN(v) > 0 && + rb_equal(RARRAY_PTR(v)[0], key)) return v; } return Qnil; @@ -2586,54 +3124,50 @@ rb_ary_assoc(ary, key) /* * call-seq: - * array.rassoc(key) -> an_array or nil - * + * ary.rassoc(obj) -> new_ary or nil + * * Searches through the array whose elements are also arrays. Compares - * <em>key</em> with the second element of each contained array using + * _obj_ with the second element of each contained array using * <code>==</code>. Returns the first contained array that matches. See * also <code>Array#assoc</code>. - * + * * a = [ [ 1, "one"], [2, "two"], [3, "three"], ["ii", "two"] ] * a.rassoc("two") #=> [2, "two"] * a.rassoc("four") #=> nil */ VALUE -rb_ary_rassoc(ary, value) - VALUE ary, value; +rb_ary_rassoc(VALUE ary, VALUE value) { long i; VALUE v; - for (i = 0; i < RARRAY(ary)->len; ++i) { - v = RARRAY(ary)->ptr[i]; + for (i = 0; i < RARRAY_LEN(ary); ++i) { + v = RARRAY_PTR(ary)[i]; if (TYPE(v) == T_ARRAY && - RARRAY(v)->len > 1 && - rb_equal(RARRAY(v)->ptr[1], value)) + RARRAY_LEN(v) > 1 && + rb_equal(RARRAY_PTR(v)[1], value)) return v; } return Qnil; } -static VALUE recursive_equal _((VALUE, VALUE, int)); static VALUE -recursive_equal(ary1, ary2, recur) - VALUE ary1, ary2; - int recur; +recursive_equal(VALUE ary1, VALUE ary2, int recur) { long i; - if (recur) return Qfalse; - for (i=0; i<RARRAY(ary1)->len; i++) { + if (recur) return Qtrue; /* Subtle! */ + for (i=0; i<RARRAY_LEN(ary1); i++) { if (!rb_equal(rb_ary_elt(ary1, i), rb_ary_elt(ary2, i))) return Qfalse; } return Qtrue; } -/* +/* * call-seq: - * array == other_array -> bool + * ary == other_ary -> bool * * Equality---Two arrays are equal if they contain the same number * of elements and if each element is equal to (according to @@ -2646,8 +3180,7 @@ recursive_equal(ary1, ary2, recur) */ static VALUE -rb_ary_equal(ary1, ary2) - VALUE ary1, ary2; +rb_ary_equal(VALUE ary1, VALUE ary2) { if (ary1 == ary2) return Qtrue; if (TYPE(ary2) != T_ARRAY) { @@ -2656,20 +3189,17 @@ rb_ary_equal(ary1, ary2) } return rb_equal(ary2, ary1); } - if (RARRAY(ary1)->len != RARRAY(ary2)->len) return Qfalse; - return rb_exec_recursive(recursive_equal, ary1, ary2); + if (RARRAY_LEN(ary1) != RARRAY_LEN(ary2)) return Qfalse; + return rb_exec_recursive_paired(recursive_equal, ary1, ary2, ary2); } -static VALUE recursive_eql _((VALUE, VALUE, int)); static VALUE -recursive_eql(ary1, ary2, recur) - VALUE ary1, ary2; - int recur; +recursive_eql(VALUE ary1, VALUE ary2, int recur) { long i; - if (recur) return Qfalse; - for (i=0; i<RARRAY(ary1)->len; i++) { + if (recur) return Qtrue; /* Subtle! */ + for (i=0; i<RARRAY_LEN(ary1); i++) { if (!rb_eql(rb_ary_elt(ary1, i), rb_ary_elt(ary2, i))) return Qfalse; } @@ -2678,82 +3208,76 @@ recursive_eql(ary1, ary2, recur) /* * call-seq: - * array.eql?(other) -> true or false + * ary.eql?(other) -> true or false * - * Returns <code>true</code> if _array_ and _other_ are the same object, + * Returns <code>true</code> if +self+ and _other_ are the same object, * or are both arrays with the same content. */ static VALUE -rb_ary_eql(ary1, ary2) - VALUE ary1, ary2; +rb_ary_eql(VALUE ary1, VALUE ary2) { if (ary1 == ary2) return Qtrue; if (TYPE(ary2) != T_ARRAY) return Qfalse; - if (RARRAY(ary1)->len != RARRAY(ary2)->len) return Qfalse; - return rb_exec_recursive(recursive_eql, ary1, ary2); + if (RARRAY_LEN(ary1) != RARRAY_LEN(ary2)) return Qfalse; + return rb_exec_recursive_paired(recursive_eql, ary1, ary2, ary2); } -static VALUE recursive_hash _((VALUE, VALUE, int)); static VALUE -recursive_hash(ary, dummy, recur) - VALUE ary; - VALUE dummy; - int recur; +recursive_hash(VALUE ary, VALUE dummy, int recur) { - long i, h; + long i; + st_index_t h; VALUE n; + h = rb_hash_start(RARRAY_LEN(ary)); if (recur) { - return LONG2FIX(0); + h = rb_hash_uint(h, NUM2LONG(rb_hash(rb_cArray))); } - - h = RARRAY(ary)->len; - for (i=0; i<RARRAY(ary)->len; i++) { - h = (h << 1) | (h<0 ? 1 : 0); - n = rb_hash(RARRAY(ary)->ptr[i]); - h ^= NUM2LONG(n); + else { + for (i=0; i<RARRAY_LEN(ary); i++) { + n = rb_hash(RARRAY_PTR(ary)[i]); + h = rb_hash_uint(h, NUM2LONG(n)); + } } + h = rb_hash_end(h); return LONG2FIX(h); } /* * call-seq: - * array.hash -> fixnum + * ary.hash -> fixnum * * Compute a hash-code for this array. Two arrays with the same content * will have the same hash code (and will compare using <code>eql?</code>). */ static VALUE -rb_ary_hash(ary) - VALUE ary; +rb_ary_hash(VALUE ary) { - return rb_exec_recursive(recursive_hash, ary, 0); + return rb_exec_recursive_outer(recursive_hash, ary, 0); } /* * call-seq: - * array.include?(obj) -> true or false - * + * ary.include?(obj) -> true or false + * * Returns <code>true</code> if the given object is present in - * <i>self</i> (that is, if any object <code>==</code> <i>anObject</i>), + * +self+ (that is, if any object <code>==</code> <i>anObject</i>), * <code>false</code> otherwise. - * + * * a = [ "a", "b", "c" ] * a.include?("b") #=> true * a.include?("z") #=> false */ VALUE -rb_ary_includes(ary, item) - VALUE ary; - VALUE item; +rb_ary_includes(VALUE ary, VALUE item) { long i; - - for (i=0; i<RARRAY(ary)->len; i++) { - if (rb_equal(RARRAY(ary)->ptr[i], item)) { + + for (i=0; i<RARRAY_LEN(ary); i++) { + if (rb_equal(RARRAY_PTR(ary)[i], item)) { return Qtrue; } } @@ -2761,19 +3285,15 @@ rb_ary_includes(ary, item) } -static VALUE recursive_cmp _((VALUE, VALUE, int)); static VALUE -recursive_cmp(ary1, ary2, recur) - VALUE ary1; - VALUE ary2; - int recur; +recursive_cmp(VALUE ary1, VALUE ary2, int recur) { long i, len; - if (recur) return Qnil; - len = RARRAY(ary1)->len; - if (len > RARRAY(ary2)->len) { - len = RARRAY(ary2)->len; + if (recur) return Qundef; /* Subtle! */ + len = RARRAY_LEN(ary1); + if (len > RARRAY_LEN(ary2)) { + len = RARRAY_LEN(ary2); } for (i=0; i<len; i++) { VALUE v = rb_funcall(rb_ary_elt(ary1, i), id_cmp, 1, rb_ary_elt(ary2, i)); @@ -2784,13 +3304,13 @@ recursive_cmp(ary1, ary2, recur) return Qundef; } -/* +/* * call-seq: - * array <=> other_array -> -1, 0, +1 + * ary <=> other_ary -> -1, 0, +1 or nil * * Comparison---Returns an integer (-1, 0, * or +1) if this array is less than, equal to, or greater than - * other_array. Each object in each array is compared + * <i>other_ary</i>. Each object in each array is compared * (using <=>). If any value isn't * equal, then that inequality is the return value. If all the * values found are equal, then the return is based on a @@ -2798,80 +3318,120 @@ recursive_cmp(ary1, ary2, recur) * ``equal'' according to <code>Array#<=></code> if and only if they have * the same length and the value of each element is equal to the * value of the corresponding element in the other array. - * + * * [ "a", "a", "c" ] <=> [ "a", "b", "c" ] #=> -1 * [ 1, 2, 3, 4, 5, 6 ] <=> [ 1, 2 ] #=> +1 * */ VALUE -rb_ary_cmp(ary1, ary2) - VALUE ary1, ary2; +rb_ary_cmp(VALUE ary1, VALUE ary2) { long len; VALUE v; - ary2 = to_ary(ary2); + ary2 = rb_check_array_type(ary2); + if (NIL_P(ary2)) return Qnil; if (ary1 == ary2) return INT2FIX(0); - v = rb_exec_recursive(recursive_cmp, ary1, ary2); + v = rb_exec_recursive_paired(recursive_cmp, ary1, ary2, ary2); if (v != Qundef) return v; - len = RARRAY(ary1)->len - RARRAY(ary2)->len; + len = RARRAY_LEN(ary1) - RARRAY_LEN(ary2); if (len == 0) return INT2FIX(0); if (len > 0) return INT2FIX(1); return INT2FIX(-1); } static VALUE -ary_make_hash(ary1, ary2) - VALUE ary1, ary2; +ary_add_hash(VALUE hash, VALUE ary) { - VALUE hash = rb_hash_new(); long i; - for (i=0; i<RARRAY(ary1)->len; i++) { - rb_hash_aset(hash, RARRAY(ary1)->ptr[i], Qtrue); + for (i=0; i<RARRAY_LEN(ary); i++) { + rb_hash_aset(hash, RARRAY_PTR(ary)[i], Qtrue); } - if (ary2) { - for (i=0; i<RARRAY(ary2)->len; i++) { - rb_hash_aset(hash, RARRAY(ary2)->ptr[i], Qtrue); + return hash; +} + +static inline VALUE +ary_tmp_hash_new(void) +{ + VALUE hash = rb_hash_new(); + + RBASIC(hash)->klass = 0; + return hash; +} + +static VALUE +ary_make_hash(VALUE ary) +{ + VALUE hash = ary_tmp_hash_new(); + return ary_add_hash(hash, ary); +} + +static VALUE +ary_add_hash_by(VALUE hash, VALUE ary) +{ + long i; + + for (i = 0; i < RARRAY_LEN(ary); ++i) { + VALUE v = rb_ary_elt(ary, i), k = rb_yield(v); + if (rb_hash_lookup2(hash, k, Qundef) == Qundef) { + rb_hash_aset(hash, k, v); } } return hash; } -/* +static VALUE +ary_make_hash_by(VALUE ary) +{ + VALUE hash = ary_tmp_hash_new(); + return ary_add_hash_by(hash, ary); +} + +static inline void +ary_recycle_hash(VALUE hash) +{ + if (RHASH(hash)->ntbl) { + st_table *tbl = RHASH(hash)->ntbl; + RHASH(hash)->ntbl = 0; + st_free_table(tbl); + } +} + +/* * call-seq: - * array - other_array -> an_array + * ary - other_ary -> new_ary * * Array Difference---Returns a new array that is a copy of * the original array, removing any items that also appear in - * other_array. (If you need set-like behavior, see the + * <i>other_ary</i>. (If you need set-like behavior, see the * library class Set.) * * [ 1, 1, 2, 2, 3, 3, 4, 5 ] - [ 1, 2, 4 ] #=> [ 3, 3, 5 ] */ static VALUE -rb_ary_diff(ary1, ary2) - VALUE ary1, ary2; +rb_ary_diff(VALUE ary1, VALUE ary2) { VALUE ary3; volatile VALUE hash; long i; - hash = ary_make_hash(to_ary(ary2), 0); + hash = ary_make_hash(to_ary(ary2)); ary3 = rb_ary_new(); - for (i=0; i<RARRAY(ary1)->len; i++) { - if (st_lookup(RHASH(hash)->tbl, RARRAY(ary1)->ptr[i], 0)) continue; + for (i=0; i<RARRAY_LEN(ary1); i++) { + if (st_lookup(RHASH_TBL(hash), RARRAY_PTR(ary1)[i], 0)) continue; rb_ary_push(ary3, rb_ary_elt(ary1, i)); } + ary_recycle_hash(hash); return ary3; } -/* +/* * call-seq: - * array & other_array + * ary & other_ary -> new_ary * * Set Intersection---Returns a new array * containing elements common to the two arrays, with no duplicates. @@ -2881,168 +3441,231 @@ rb_ary_diff(ary1, ary2) static VALUE -rb_ary_and(ary1, ary2) - VALUE ary1, ary2; +rb_ary_and(VALUE ary1, VALUE ary2) { - VALUE hash, ary3, v, vv; + VALUE hash, ary3, v; + st_data_t vv; long i; ary2 = to_ary(ary2); - ary3 = rb_ary_new2(RARRAY(ary1)->len < RARRAY(ary2)->len ? - RARRAY(ary1)->len : RARRAY(ary2)->len); - hash = ary_make_hash(ary2, 0); + ary3 = rb_ary_new2(RARRAY_LEN(ary1) < RARRAY_LEN(ary2) ? + RARRAY_LEN(ary1) : RARRAY_LEN(ary2)); + hash = ary_make_hash(ary2); + + if (RHASH_EMPTY_P(hash)) + return ary3; - for (i=0; i<RARRAY(ary1)->len; i++) { - v = vv = rb_ary_elt(ary1, i); - if (st_delete(RHASH(hash)->tbl, (st_data_t*)&vv, 0)) { + for (i=0; i<RARRAY_LEN(ary1); i++) { + vv = (st_data_t)(v = rb_ary_elt(ary1, i)); + if (st_delete(RHASH_TBL(hash), &vv, 0)) { rb_ary_push(ary3, v); } } + ary_recycle_hash(hash); return ary3; } -/* +/* * call-seq: - * array | other_array -> an_array + * ary | other_ary -> new_ary * * Set Union---Returns a new array by joining this array with - * other_array, removing duplicates. + * <i>other_ary</i>, removing duplicates. * * [ "a", "b", "c" ] | [ "c", "d", "a" ] * #=> [ "a", "b", "c", "d" ] */ static VALUE -rb_ary_or(ary1, ary2) - VALUE ary1, ary2; +rb_ary_or(VALUE ary1, VALUE ary2) { - VALUE hash, ary3; - VALUE v, vv; + VALUE hash, ary3, v; + st_data_t vv; long i; ary2 = to_ary(ary2); - ary3 = rb_ary_new2(RARRAY(ary1)->len+RARRAY(ary2)->len); - hash = ary_make_hash(ary1, ary2); + ary3 = rb_ary_new2(RARRAY_LEN(ary1)+RARRAY_LEN(ary2)); + hash = ary_add_hash(ary_make_hash(ary1), ary2); - for (i=0; i<RARRAY(ary1)->len; i++) { - v = vv = rb_ary_elt(ary1, i); - if (st_delete(RHASH(hash)->tbl, (st_data_t*)&vv, 0)) { + for (i=0; i<RARRAY_LEN(ary1); i++) { + vv = (st_data_t)(v = rb_ary_elt(ary1, i)); + if (st_delete(RHASH_TBL(hash), &vv, 0)) { rb_ary_push(ary3, v); } } - for (i=0; i<RARRAY(ary2)->len; i++) { - v = vv = rb_ary_elt(ary2, i); - if (st_delete(RHASH(hash)->tbl, (st_data_t*)&vv, 0)) { + for (i=0; i<RARRAY_LEN(ary2); i++) { + vv = (st_data_t)(v = rb_ary_elt(ary2, i)); + if (st_delete(RHASH_TBL(hash), &vv, 0)) { rb_ary_push(ary3, v); } } + ary_recycle_hash(hash); return ary3; } +static int +push_value(st_data_t key, st_data_t val, st_data_t ary) +{ + rb_ary_push((VALUE)ary, (VALUE)val); + return ST_CONTINUE; +} + /* * call-seq: - * array.uniq! -> array or nil - * - * Removes duplicate elements from _self_. + * ary.uniq! -> ary or nil + * ary.uniq! { |item| ... } -> ary or nil + * + * Removes duplicate elements from +self+. If a block is given, + * it will use the return value of the block for comparison. * Returns <code>nil</code> if no changes are made (that is, no * duplicates are found). - * + * * a = [ "a", "a", "b", "b", "c" ] - * a.uniq! #=> ["a", "b", "c"] + * a.uniq! # => ["a", "b", "c"] + * * b = [ "a", "b", "c" ] - * b.uniq! #=> nil + * b.uniq! # => nil + * + * c = [["student","sam"], ["student","george"], ["teacher","matz"]] + * c.uniq! { |s| s.first } # => [["student", "sam"], ["teacher", "matz"]] + * */ static VALUE -rb_ary_uniq_bang(ary) - VALUE ary; +rb_ary_uniq_bang(VALUE ary) { - VALUE hash, v, vv; + VALUE hash, v; long i, j; - hash = ary_make_hash(ary, 0); - - if (RARRAY(ary)->len == RHASH(hash)->tbl->num_entries) { - return Qnil; + rb_ary_modify_check(ary); + if (RARRAY_LEN(ary) <= 1) + return Qnil; + if (rb_block_given_p()) { + hash = ary_make_hash_by(ary); + if (RARRAY_LEN(ary) == (i = RHASH_SIZE(hash))) { + return Qnil; + } + ARY_SET_LEN(ary, 0); + if (ARY_SHARED_P(ary) && !ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + FL_SET_EMBED(ary); + } + ary_resize_capa(ary, i); + st_foreach(RHASH_TBL(hash), push_value, ary); } - for (i=j=0; i<RARRAY(ary)->len; i++) { - v = vv = rb_ary_elt(ary, i); - if (st_delete(RHASH(hash)->tbl, (st_data_t*)&vv, 0)) { - rb_ary_store(ary, j++, v); + else { + hash = ary_make_hash(ary); + if (RARRAY_LEN(ary) == (long)RHASH_SIZE(hash)) { + return Qnil; } + for (i=j=0; i<RARRAY_LEN(ary); i++) { + st_data_t vv = (st_data_t)(v = rb_ary_elt(ary, i)); + if (st_delete(RHASH_TBL(hash), &vv, 0)) { + rb_ary_store(ary, j++, v); + } + } + ARY_SET_LEN(ary, j); } - RARRAY(ary)->len = j; + ary_recycle_hash(hash); return ary; } /* * call-seq: - * array.uniq -> an_array - * - * Returns a new array by removing duplicate values in <i>self</i>. - * + * ary.uniq -> new_ary + * ary.uniq { |item| ... } -> new_ary + * + * Returns a new array by removing duplicate values in +self+. If a block + * is given, it will use the return value of the block for comparison. + * * a = [ "a", "a", "b", "b", "c" ] - * a.uniq #=> ["a", "b", "c"] + * a.uniq # => ["a", "b", "c"] + * + * b = [["student","sam"], ["student","george"], ["teacher","matz"]] + * b.uniq { |s| s.first } # => [["student", "sam"], ["teacher", "matz"]] + * */ static VALUE -rb_ary_uniq(ary) - VALUE ary; +rb_ary_uniq(VALUE ary) { - ary = rb_ary_dup(ary); - rb_ary_uniq_bang(ary); - return ary; + VALUE hash, uniq, v; + long i; + + if (RARRAY_LEN(ary) <= 1) + return rb_ary_dup(ary); + if (rb_block_given_p()) { + hash = ary_make_hash_by(ary); + uniq = ary_new(rb_obj_class(ary), RHASH_SIZE(hash)); + st_foreach(RHASH_TBL(hash), push_value, uniq); + } + else { + hash = ary_make_hash(ary); + uniq = ary_new(rb_obj_class(ary), RHASH_SIZE(hash)); + for (i=0; i<RARRAY_LEN(ary); i++) { + st_data_t vv = (st_data_t)(v = rb_ary_elt(ary, i)); + if (st_delete(RHASH_TBL(hash), &vv, 0)) { + rb_ary_push(uniq, v); + } + } + } + ary_recycle_hash(hash); + + return uniq; } -/* +/* * call-seq: - * array.compact! -> array or nil + * ary.compact! -> ary or nil * - * Removes +nil+ elements from array. - * Returns +nil+ if no changes were made. + * Removes +nil+ elements from the array. + * Returns +nil+ if no changes were made, otherwise returns + * <i>ary</i>. * * [ "a", nil, "b", nil, "c" ].compact! #=> [ "a", "b", "c" ] * [ "a", "b", "c" ].compact! #=> nil */ static VALUE -rb_ary_compact_bang(ary) - VALUE ary; +rb_ary_compact_bang(VALUE ary) { VALUE *p, *t, *end; + long n; rb_ary_modify(ary); - p = t = RARRAY(ary)->ptr; - end = p + RARRAY(ary)->len; - + p = t = RARRAY_PTR(ary); + end = p + RARRAY_LEN(ary); + while (t < end) { if (NIL_P(*t)) t++; else *p++ = *t++; } - if (RARRAY(ary)->len == (p - RARRAY(ary)->ptr)) { + n = p - RARRAY_PTR(ary); + if (RARRAY_LEN(ary) == n) { return Qnil; } - RARRAY(ary)->len = RARRAY(ary)->aux.capa = (p - RARRAY(ary)->ptr); - REALLOC_N(RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); + ARY_SET_LEN(ary, n); + if (n * 2 < ARY_CAPA(ary) && ARY_DEFAULT_SIZE * 2 < ARY_CAPA(ary)) { + ary_resize_capa(ary, n * 2); + } return ary; } /* * call-seq: - * array.compact -> an_array + * ary.compact -> new_ary * - * Returns a copy of _self_ with all +nil+ elements removed. + * Returns a copy of +self+ with all +nil+ elements removed. * * [ "a", nil, "b", nil, "c", nil ].compact * #=> [ "a", "b", "c" ] */ static VALUE -rb_ary_compact(ary) - VALUE ary; +rb_ary_compact(VALUE ary) { ary = rb_ary_dup(ary); rb_ary_compact_bang(ary); @@ -3051,72 +3674,46 @@ rb_ary_compact(ary) /* * call-seq: - * array.nitems -> int - * - * Returns the number of non-<code>nil</code> elements in _self_. + * ary.count -> int + * ary.count(obj) -> int + * ary.count { |item| block } -> int * - * May be zero. - * - * [ 1, nil, 3, nil, 5 ].nitems #=> 3 - */ - -static VALUE -rb_ary_nitems(ary) - VALUE ary; -{ - long n = 0; - VALUE *p, *pend; - - for (p = RARRAY(ary)->ptr, pend = p + RARRAY(ary)->len; p < pend; p++) { - if (!NIL_P(*p)) n++; - } - return LONG2NUM(n); -} - -/* - * call-seq: - * array.count -> int - * array.count(obj) -> int - * array.count { |item| block } -> int - * * Returns the number of elements. If an argument is given, counts * the number of elements which equals to <i>obj</i>. If a block is * given, counts the number of elements yielding a true value. * * ary = [1, 2, 4, 2] - * ary.count # => 4 - * ary.count(2) # => 2 - * ary.count{|x|x%2==0} # => 3 + * ary.count #=> 4 + * ary.count(2) #=> 2 + * ary.count{|x|x%2==0} #=> 3 * */ static VALUE -rb_ary_count(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_count(int argc, VALUE *argv, VALUE ary) { - long n = 0; + long i, n = 0; if (argc == 0) { - VALUE *p, *pend; + VALUE v; if (!rb_block_given_p()) return LONG2NUM(RARRAY_LEN(ary)); - for (p = RARRAY_PTR(ary), pend = p + RARRAY_LEN(ary); p < pend; p++) { - if (RTEST(rb_yield(*p))) n++; + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_PTR(ary)[i]; + if (RTEST(rb_yield(v))) n++; } } else { - VALUE obj, *p, *pend; + VALUE obj; rb_scan_args(argc, argv, "1", &obj); if (rb_block_given_p()) { rb_warn("given block not used"); } - for (p = RARRAY_PTR(ary), pend = p + RARRAY_LEN(ary); p < pend; p++) { - if (rb_equal(*p, obj)) n++; + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (rb_equal(RARRAY_PTR(ary)[i], obj)) n++; } } @@ -3124,10 +3721,7 @@ rb_ary_count(argc, argv, ary) } static VALUE -flatten(ary, level, modified) - VALUE ary; - int level; - int *modified; +flatten(VALUE ary, int level, int *modified) { long i = 0; VALUE stack, result, tmp, elt; @@ -3141,13 +3735,13 @@ flatten(ary, level, modified) *modified = 0; while (1) { - while (i < RARRAY(ary)->len) { - elt = RARRAY(ary)->ptr[i++]; + while (i < RARRAY_LEN(ary)) { + elt = RARRAY_PTR(ary)[i++]; tmp = rb_check_array_type(elt); if (RBASIC(result)->klass) { rb_raise(rb_eRuntimeError, "flatten reentered"); } - if (NIL_P(tmp) || (level >= 0 && RARRAY(stack)->len / 2 >= level)) { + if (NIL_P(tmp) || (level >= 0 && RARRAY_LEN(stack) / 2 >= level)) { rb_ary_push(result, elt); } else { @@ -3164,7 +3758,7 @@ flatten(ary, level, modified) i = 0; } } - if (RARRAY(stack)->len == 0) { + if (RARRAY_LEN(stack) == 0) { break; } id = (st_data_t)ary; @@ -3182,14 +3776,14 @@ flatten(ary, level, modified) /* * call-seq: - * array.flatten! -> array or nil - * array.flatten!(level) -> array or nil - * - * Flattens _self_ in place. + * ary.flatten! -> ary or nil + * ary.flatten!(level) -> array or nil + * + * Flattens +self+ in place. * Returns <code>nil</code> if no modifications were made (i.e., - * <i>array</i> contains no subarrays.) If the optional <i>level</i> + * <i>ary</i> contains no subarrays.) If the optional <i>level</i> * argument determines the level of recursion to flatten. - * + * * a = [ 1, 2, [3, [4, 5] ] ] * a.flatten! #=> [1, 2, 3, 4, 5] * a.flatten! #=> nil @@ -3199,35 +3793,38 @@ flatten(ary, level, modified) */ static VALUE -rb_ary_flatten_bang(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_flatten_bang(int argc, VALUE *argv, VALUE ary) { int mod = 0, level = -1; VALUE result, lv; rb_scan_args(argc, argv, "01", &lv); + rb_ary_modify_check(ary); if (!NIL_P(lv)) level = NUM2INT(lv); - if (level == 0) return ary; + if (level == 0) return Qnil; result = flatten(ary, level, &mod); - if (mod == 0) return Qnil; + if (mod == 0) { + ary_discard(result); + return Qnil; + } + if (!(mod = ARY_EMBED_P(result))) rb_obj_freeze(result); rb_ary_replace(ary, result); + if (mod) ARY_SET_EMBED_LEN(result, 0); return ary; } /* * call-seq: - * array.flatten -> an_array - * array.flatten(level) -> an_array - * + * ary.flatten -> new_ary + * ary.flatten(level) -> new_ary + * * Returns a new array that is a one-dimensional flattening of this * array (recursively). That is, for every element that is an array, * extract its elements into the new array. If the optional * <i>level</i> argument determines the level of recursion to flatten. - * + * * s = [ 1, 2, 3 ] #=> [1, 2, 3] * t = [ 4, 5, 6, [7, 8] ] #=> [4, 5, 6, [7, 8]] * a = [ s, t, 9, 10 ] #=> [[1, 2, 3], [4, 5, 6, [7, 8]], 9, 10] @@ -3237,44 +3834,62 @@ rb_ary_flatten_bang(argc, argv, ary) */ static VALUE -rb_ary_flatten(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_flatten(int argc, VALUE *argv, VALUE ary) { int mod = 0, level = -1; VALUE result, lv; rb_scan_args(argc, argv, "01", &lv); if (!NIL_P(lv)) level = NUM2INT(lv); - if (level == 0) return ary; + if (level == 0) return ary_make_shared_copy(ary); result = flatten(ary, level, &mod); - if (OBJ_TAINTED(ary)) OBJ_TAINT(result); + OBJ_INFECT(result, ary); return result; } +#define OPTHASH_GIVEN_P(opts) \ + (argc > 0 && !NIL_P((opts) = rb_check_hash_type(argv[argc-1])) && (--argc, 1)) +static VALUE sym_random; + +#define RAND_UPTO(max) (long)(rb_random_real(randgen)*(max)) + /* * call-seq: - * array.shuffle! -> array or nil - * - * Shuffles elements in _self_ in place. + * ary.shuffle! -> ary + * ary.shuffle!(random: rng) -> ary + * + * Shuffles elements in +self+ in place. + * If +rng+ is given, it will be used as the random number generator. */ - static VALUE -rb_ary_shuffle_bang(ary) - VALUE ary; +rb_ary_shuffle_bang(int argc, VALUE *argv, VALUE ary) { - long i = RARRAY(ary)->len; + VALUE *ptr, opts, *snap_ptr, randgen = rb_cRandom; + long i, snap_len; + if (OPTHASH_GIVEN_P(opts)) { + randgen = rb_hash_lookup2(opts, sym_random, randgen); + } + if (argc > 0) { + rb_raise(rb_eArgError, "wrong number of arguments (%d for 0)", argc); + } rb_ary_modify(ary); + i = RARRAY_LEN(ary); + ptr = RARRAY_PTR(ary); + snap_len = i; + snap_ptr = ptr; while (i) { - long j = rb_genrand_real()*i; - VALUE tmp = RARRAY(ary)->ptr[--i]; - RARRAY(ary)->ptr[i] = RARRAY(ary)->ptr[j]; - RARRAY(ary)->ptr[j] = tmp; + long j = RAND_UPTO(i); + VALUE tmp; + if (snap_len != RARRAY_LEN(ary) || snap_ptr != RARRAY_PTR(ary)) { + rb_raise(rb_eRuntimeError, "modified during shuffle"); + } + tmp = ptr[--i]; + ptr[i] = ptr[j]; + ptr[j] = tmp; } return ary; } @@ -3282,65 +3897,164 @@ rb_ary_shuffle_bang(ary) /* * call-seq: - * array.shuffle -> an_array - * + * ary.shuffle -> new_ary + * ary.shuffle(random: rng) -> new_ary + * * Returns a new array with elements of this array shuffled. - * + * * a = [ 1, 2, 3 ] #=> [1, 2, 3] * a.shuffle #=> [2, 3, 1] + * + * If +rng+ is given, it will be used as the random number generator. + * + * a.shuffle(random: Random.new(1)) #=> [1, 3, 2] */ static VALUE -rb_ary_shuffle(VALUE ary) +rb_ary_shuffle(int argc, VALUE *argv, VALUE ary) { ary = rb_ary_dup(ary); - rb_ary_shuffle_bang(ary); + rb_ary_shuffle_bang(argc, argv, ary); return ary; } /* * call-seq: - * array.choice -> obj - * - * Choose a random element from an array. + * ary.sample -> obj + * ary.sample(random: rng) -> obj + * ary.sample(n) -> new_ary + * ary.sample(n, random: rng) -> new_ary + * + * Choose a random element or +n+ random elements from the array. The elements + * are chosen by using random and unique indices into the array in order to + * ensure that an element doesn't repeat itself unless the array already + * contained duplicate elements. If the array is empty the first form returns + * <code>nil</code> and the second form returns an empty array. + * + * If +rng+ is given, it will be used as the random number generator. */ static VALUE -rb_ary_choice(ary) - VALUE ary; +rb_ary_sample(int argc, VALUE *argv, VALUE ary) { - long i, j; + VALUE nv, result, *ptr; + VALUE opts, randgen = rb_cRandom; + long n, len, i, j, k, idx[10]; + double rnds[numberof(idx)]; + + if (OPTHASH_GIVEN_P(opts)) { + randgen = rb_hash_lookup2(opts, sym_random, randgen); + } + ptr = RARRAY_PTR(ary); + len = RARRAY_LEN(ary); + if (argc == 0) { + if (len == 0) return Qnil; + if (len == 1) { + i = 0; + } + else { + double x = rb_random_real(randgen); + if ((len = RARRAY_LEN(ary)) == 0) return Qnil; + i = (long)(x * len); + } + return RARRAY_PTR(ary)[i]; + } + rb_scan_args(argc, argv, "1", &nv); + n = NUM2LONG(nv); + if (n < 0) rb_raise(rb_eArgError, "negative sample number"); + if (n > len) n = len; + if (n <= numberof(idx)) { + for (i = 0; i < n; ++i) { + rnds[i] = rb_random_real(randgen); + } + } + len = RARRAY_LEN(ary); + ptr = RARRAY_PTR(ary); + if (n > len) n = len; + switch (n) { + case 0: + return rb_ary_new2(0); + case 1: + i = (long)(rnds[0] * len); + return rb_ary_new4(1, &ptr[i]); + case 2: + i = (long)(rnds[0] * len); + j = (long)(rnds[1] * (len-1)); + if (j >= i) j++; + return rb_ary_new3(2, ptr[i], ptr[j]); + case 3: + i = (long)(rnds[0] * len); + j = (long)(rnds[1] * (len-1)); + k = (long)(rnds[2] * (len-2)); + { + long l = j, g = i; + if (j >= i) l = i, g = ++j; + if (k >= l && (++k >= g)) ++k; + } + return rb_ary_new3(3, ptr[i], ptr[j], ptr[k]); + } + if (n <= numberof(idx)) { + VALUE *ptr_result; + long sorted[numberof(idx)]; + sorted[0] = idx[0] = (long)(rnds[0] * len); + for (i=1; i<n; i++) { + k = (long)(rnds[i] * --len); + for (j = 0; j < i; ++j) { + if (k < sorted[j]) break; + ++k; + } + memmove(&sorted[j+1], &sorted[j], sizeof(sorted[0])*(i-j)); + sorted[j] = idx[i] = k; + } + result = rb_ary_new2(n); + ptr_result = RARRAY_PTR(result); + for (i=0; i<n; i++) { + ptr_result[i] = ptr[idx[i]]; + } + } + else { + VALUE *ptr_result; + result = rb_ary_new4(len, ptr); + RBASIC(result)->klass = 0; + ptr_result = RARRAY_PTR(result); + RB_GC_GUARD(ary); + for (i=0; i<n; i++) { + j = RAND_UPTO(len-i) + i; + nv = ptr_result[j]; + ptr_result[j] = ptr_result[i]; + ptr_result[i] = nv; + } + RBASIC(result)->klass = rb_cArray; + } + ARY_SET_LEN(result, n); - i = RARRAY(ary)->len; - if (i == 0) return Qnil; - j = rb_genrand_real()*i; - return RARRAY(ary)->ptr[j]; + return result; } /* * call-seq: - * ary.cycle {|obj| block } - * ary.cycle(n) {|obj| block } - * + * ary.cycle(n=nil) {|obj| block } -> nil + * ary.cycle(n=nil) -> an_enumerator + * * Calls <i>block</i> for each element repeatedly _n_ times or - * forever if none or nil is given. If a non-positive number is - * given or the array is empty, does nothing. Returns nil if the + * forever if none or +nil+ is given. If a non-positive number is + * given or the array is empty, does nothing. Returns +nil+ if the * loop has finished without getting interrupted. - * + * + * If no block is given, an enumerator is returned instead. + * + * * a = ["a", "b", "c"] * a.cycle {|x| puts x } # print, a, b, c, a, b, c,.. forever. * a.cycle(2) {|x| puts x } # print, a, b, c, a, b, c. - * + * */ static VALUE -rb_ary_cycle(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_cycle(int argc, VALUE *argv, VALUE ary) { long n, i; VALUE nv = Qnil; @@ -3356,22 +4070,25 @@ rb_ary_cycle(argc, argv, ary) if (n <= 0) return Qnil; } - while (RARRAY(ary)->len > 0 && (n < 0 || 0 < n--)) { - for (i=0; i<RARRAY(ary)->len; i++) { - rb_yield(RARRAY(ary)->ptr[i]); + while (RARRAY_LEN(ary) > 0 && (n < 0 || 0 < n--)) { + for (i=0; i<RARRAY_LEN(ary); i++) { + rb_yield(RARRAY_PTR(ary)[i]); } } return Qnil; } #define tmpbuf(n, size) rb_str_tmp_new((n)*(size)) +#define tmpbuf_discard(s) (rb_str_resize((s), 0L), RBASIC(s)->klass = rb_cString) +#define tmpary(n) rb_ary_tmp_new(n) +#define tmpary_discard(a) (ary_discard(a), RBASIC(a)->klass = rb_cArray) /* * Recursively compute permutations of r elements of the set [0..n-1]. * When we have a complete permutation of array indexes, copy the values - * at those indexes into a new array and yield that array. + * at those indexes into a new array and yield that array. * - * n: the size of the set + * n: the size of the set * r: the number of elements in each permutation * p: the array (of size r) that we're filling in * index: what index we're filling in now @@ -3379,10 +4096,7 @@ rb_ary_cycle(argc, argv, ary) * values: the Ruby array that holds the actual values to permute */ static void -permute0(n, r, p, index, used, values) - long n, r, *p, index; - int *used; - VALUE values; +permute0(long n, long r, long *p, long index, char *used, VALUE values) { long i,j; for (i = 0; i < n; i++) { @@ -3391,7 +4105,7 @@ permute0(n, r, p, index, used, values) if (index < r-1) { /* if not done yet */ used[i] = 1; /* mark index used */ permute0(n, r, p, index+1, /* recurse */ - used, values); + used, values); used[i] = 0; /* index unused */ } else { @@ -3399,12 +4113,15 @@ permute0(n, r, p, index, used, values) /* Build a ruby array of the corresponding values */ /* And yield it to the associated block */ VALUE result = rb_ary_new2(r); - VALUE *result_array = RARRAY(result)->ptr; - const VALUE *values_array = RARRAY(values)->ptr; + VALUE *result_array = RARRAY_PTR(result); + const VALUE *values_array = RARRAY_PTR(values); for (j = 0; j < r; j++) result_array[j] = values_array[p[j]]; - RARRAY(result)->len = r; + ARY_SET_LEN(result, r); rb_yield(result); + if (RBASIC(values)->klass) { + rb_raise(rb_eRuntimeError, "permute reentered"); + } } } } @@ -3412,19 +4129,19 @@ permute0(n, r, p, index, used, values) /* * call-seq: - * ary.permutation { |p| block } -> array - * ary.permutation -> enumerator - * ary.permutation(n) { |p| block } -> array - * ary.permutation(n) -> enumerator - * + * ary.permutation { |p| block } -> ary + * ary.permutation -> an_enumerator + * ary.permutation(n) { |p| block } -> ary + * ary.permutation(n) -> an_enumerator + * * When invoked with a block, yield all permutations of length <i>n</i> * of the elements of <i>ary</i>, then return the array itself. * If <i>n</i> is not specified, yield all permutations of all elements. - * The implementation makes no guarantees about the order in which + * The implementation makes no guarantees about the order in which * the permutations are yielded. * - * When invoked without a block, return an enumerator object instead. - * + * If no block is given, an enumerator is returned instead. + * * Examples: * * a = [1, 2, 3] @@ -3437,79 +4154,57 @@ permute0(n, r, p, index, used, values) */ static VALUE -rb_ary_permutation(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_permutation(int argc, VALUE *argv, VALUE ary) { VALUE num; long r, n, i; - n = RARRAY(ary)->len; /* Array length */ + n = RARRAY_LEN(ary); /* Array length */ RETURN_ENUMERATOR(ary, argc, argv); /* Return enumerator if no block */ rb_scan_args(argc, argv, "01", &num); r = NIL_P(num) ? n : NUM2LONG(num); /* Permutation size from argument */ - if (r < 0 || n < r) { + if (r < 0 || n < r) { /* no permutations: yield nothing */ } else if (r == 0) { /* exactly one permutation: the zero-length array */ rb_yield(rb_ary_new2(0)); } else if (r == 1) { /* this is a special, easy case */ - for (i = 0; i < RARRAY(ary)->len; i++) { - rb_yield(rb_ary_new3(1, RARRAY(ary)->ptr[i])); + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_yield(rb_ary_new3(1, RARRAY_PTR(ary)[i])); } } else { /* this is the general case */ volatile VALUE t0 = tmpbuf(n,sizeof(long)); - long *p = (long*)RSTRING(t0)->ptr; - volatile VALUE t1 = tmpbuf(n,sizeof(int)); - int *used = (int*)RSTRING(t1)->ptr; - VALUE ary0 = ary_make_shared(ary); /* private defensive copy of ary */ + long *p = (long*)RSTRING_PTR(t0); + volatile VALUE t1 = tmpbuf(n,sizeof(char)); + char *used = (char*)RSTRING_PTR(t1); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC(ary0)->klass = 0; - for (i = 0; i < n; i++) used[i] = 0; /* initialize array */ + MEMZERO(used, char, n); /* initialize array */ permute0(n, r, p, 0, used, ary0); /* compute and yield permutations */ - RB_GC_GUARD(t0); - RB_GC_GUARD(t1); + tmpbuf_discard(t0); + tmpbuf_discard(t1); + RBASIC(ary0)->klass = rb_cArray; } return ary; } -static long -combi_len(n, k) - long n, k; -{ - long i, val = 1; - - if (k*2 > n) k = n-k; - if (k == 0) return 1; - if (k < 0) return 0; - val = 1; - for (i=1; i <= k; i++,n--) { - long m = val; - val *= n; - if (val < m) { - rb_raise(rb_eRangeError, "too big for combination"); - } - val /= i; - } - return val; -} - /* * call-seq: * ary.combination(n) { |c| block } -> ary - * ary.combination(n) -> enumerator - * - * When invoked with a block, yields all combinations of length <i>n</i> + * ary.combination(n) -> an_enumerator + * + * When invoked with a block, yields all combinations of length <i>n</i> * of elements from <i>ary</i> and then returns <i>ary</i> itself. - * The implementation makes no guarantees about the order in which + * The implementation makes no guarantees about the order in which * the combinations are yielded. * - * When invoked without a block, returns an enumerator object instead. - * + * If no block is given, an enumerator is returned instead. + * * Examples: * * a = [1, 2, 3, 4] @@ -3519,19 +4214,17 @@ combi_len(n, k) * a.combination(4).to_a #=> [[1,2,3,4]] * a.combination(0).to_a #=> [[]] # one combination of length 0 * a.combination(5).to_a #=> [] # no combinations of length 5 - * + * */ static VALUE -rb_ary_combination(ary, num) - VALUE ary; - VALUE num; +rb_ary_combination(VALUE ary, VALUE num) { long n, i, len; n = NUM2LONG(num); RETURN_ENUMERATOR(ary, 1, &num); - len = RARRAY(ary)->len; + len = RARRAY_LEN(ary); if (n < 0 || len < n) { /* yield nothing */ } @@ -3540,62 +4233,248 @@ rb_ary_combination(ary, num) } else if (n == 1) { for (i = 0; i < len; i++) { - rb_yield(rb_ary_new3(1, RARRAY(ary)->ptr[i])); + rb_yield(rb_ary_new3(1, RARRAY_PTR(ary)[i])); } } else { volatile VALUE t0 = tmpbuf(n+1, sizeof(long)); - long *stack = (long*)RSTRING(t0)->ptr; - long nlen = combi_len(len, n); - volatile VALUE cc = rb_ary_new2(n); - VALUE *chosen = RARRAY(cc)->ptr; + long *stack = (long*)RSTRING_PTR(t0); + volatile VALUE cc = tmpary(n); + VALUE *chosen = RARRAY_PTR(cc); long lev = 0; - RBASIC(cc)->klass = 0; MEMZERO(stack, long, n); stack[0] = -1; - for (i = 0; i < nlen; i++) { - chosen[lev] = RARRAY(ary)->ptr[stack[lev+1]]; + for (;;) { + chosen[lev] = RARRAY_PTR(ary)[stack[lev+1]]; for (lev++; lev < n; lev++) { - chosen[lev] = RARRAY(ary)->ptr[stack[lev+1] = stack[lev]+1]; + chosen[lev] = RARRAY_PTR(ary)[stack[lev+1] = stack[lev]+1]; } rb_yield(rb_ary_new4(n, chosen)); + if (RBASIC(t0)->klass) { + rb_raise(rb_eRuntimeError, "combination reentered"); + } do { + if (lev == 0) goto done; stack[lev--]++; - } while (lev && (stack[lev+1]+n == len+lev+1)); + } while (stack[lev+1]+n == len+lev+1); + } + done: + tmpbuf_discard(t0); + tmpary_discard(cc); + } + return ary; +} + +/* + * Recursively compute repeated permutations of r elements of the set + * [0..n-1]. + * When we have a complete repeated permutation of array indexes, copy the + * values at those indexes into a new array and yield that array. + * + * n: the size of the set + * r: the number of elements in each permutation + * p: the array (of size r) that we're filling in + * index: what index we're filling in now + * values: the Ruby array that holds the actual values to permute + */ +static void +rpermute0(long n, long r, long *p, long index, VALUE values) +{ + long i, j; + for (i = 0; i < n; i++) { + p[index] = i; + if (index < r-1) { /* if not done yet */ + rpermute0(n, r, p, index+1, values); /* recurse */ + } + else { + /* We have a complete permutation of array indexes */ + /* Build a ruby array of the corresponding values */ + /* And yield it to the associated block */ + VALUE result = rb_ary_new2(r); + VALUE *result_array = RARRAY_PTR(result); + const VALUE *values_array = RARRAY_PTR(values); + + for (j = 0; j < r; j++) result_array[j] = values_array[p[j]]; + ARY_SET_LEN(result, r); + rb_yield(result); + if (RBASIC(values)->klass) { + rb_raise(rb_eRuntimeError, "repeated permute reentered"); + } + } + } +} + +/* + * call-seq: + * ary.repeated_permutation(n) { |p| block } -> ary + * ary.repeated_permutation(n) -> an_enumerator + * + * When invoked with a block, yield all repeated permutations of length + * <i>n</i> of the elements of <i>ary</i>, then return the array itself. + * The implementation makes no guarantees about the order in which + * the repeated permutations are yielded. + * + * If no block is given, an enumerator is returned instead. + * + * Examples: + * + * a = [1, 2] + * a.repeated_permutation(1).to_a #=> [[1], [2]] + * a.repeated_permutation(2).to_a #=> [[1,1],[1,2],[2,1],[2,2]] + * a.repeated_permutation(3).to_a #=> [[1,1,1],[1,1,2],[1,2,1],[1,2,2], + * # [2,1,1],[2,1,2],[2,2,1],[2,2,2]] + * a.repeated_permutation(0).to_a #=> [[]] # one permutation of length 0 + */ + +static VALUE +rb_ary_repeated_permutation(VALUE ary, VALUE num) +{ + long r, n, i; + + n = RARRAY_LEN(ary); /* Array length */ + RETURN_ENUMERATOR(ary, 1, &num); /* Return enumerator if no block */ + r = NUM2LONG(num); /* Permutation size from argument */ + + if (r < 0) { + /* no permutations: yield nothing */ + } + else if (r == 0) { /* exactly one permutation: the zero-length array */ + rb_yield(rb_ary_new2(0)); + } + else if (r == 1) { /* this is a special, easy case */ + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_yield(rb_ary_new3(1, RARRAY_PTR(ary)[i])); } } + else { /* this is the general case */ + volatile VALUE t0 = tmpbuf(r, sizeof(long)); + long *p = (long*)RSTRING_PTR(t0); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC(ary0)->klass = 0; + + rpermute0(n, r, p, 0, ary0); /* compute and yield repeated permutations */ + tmpbuf_discard(t0); + RBASIC(ary0)->klass = rb_cArray; + } return ary; } +static void +rcombinate0(long n, long r, long *p, long index, long rest, VALUE values) +{ + long j; + if (rest > 0) { + for (; index < n; ++index) { + p[r-rest] = index; + rcombinate0(n, r, p, index, rest-1, values); + } + } + else { + VALUE result = rb_ary_new2(r); + VALUE *result_array = RARRAY_PTR(result); + const VALUE *values_array = RARRAY_PTR(values); + + for (j = 0; j < r; ++j) result_array[j] = values_array[p[j]]; + ARY_SET_LEN(result, r); + rb_yield(result); + if (RBASIC(values)->klass) { + rb_raise(rb_eRuntimeError, "repeated combination reentered"); + } + } +} + /* * call-seq: - * ary.product(other_ary, ...) - * + * ary.repeated_combination(n) { |c| block } -> ary + * ary.repeated_combination(n) -> an_enumerator + * + * When invoked with a block, yields all repeated combinations of + * length <i>n</i> of elements from <i>ary</i> and then returns + * <i>ary</i> itself. + * The implementation makes no guarantees about the order in which + * the repeated combinations are yielded. + * + * If no block is given, an enumerator is returned instead. + * + * Examples: + * + * a = [1, 2, 3] + * a.repeated_combination(1).to_a #=> [[1], [2], [3]] + * a.repeated_combination(2).to_a #=> [[1,1],[1,2],[1,3],[2,2],[2,3],[3,3]] + * a.repeated_combination(3).to_a #=> [[1,1,1],[1,1,2],[1,1,3],[1,2,2],[1,2,3], + * # [1,3,3],[2,2,2],[2,2,3],[2,3,3],[3,3,3]] + * a.repeated_combination(4).to_a #=> [[1,1,1,1],[1,1,1,2],[1,1,1,3],[1,1,2,2],[1,1,2,3], + * # [1,1,3,3],[1,2,2,2],[1,2,2,3],[1,2,3,3],[1,3,3,3], + * # [2,2,2,2],[2,2,2,3],[2,2,3,3],[2,3,3,3],[3,3,3,3]] + * a.repeated_combination(0).to_a #=> [[]] # one combination of length 0 + * + */ + +static VALUE +rb_ary_repeated_combination(VALUE ary, VALUE num) +{ + long n, i, len; + + n = NUM2LONG(num); /* Combination size from argument */ + RETURN_ENUMERATOR(ary, 1, &num); /* Return enumerator if no block */ + len = RARRAY_LEN(ary); + if (n < 0) { + /* yield nothing */ + } + else if (n == 0) { + rb_yield(rb_ary_new2(0)); + } + else if (n == 1) { + for (i = 0; i < len; i++) { + rb_yield(rb_ary_new3(1, RARRAY_PTR(ary)[i])); + } + } + else if (len == 0) { + /* yield nothing */ + } + else { + volatile VALUE t0 = tmpbuf(n, sizeof(long)); + long *p = (long*)RSTRING_PTR(t0); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC(ary0)->klass = 0; + + rcombinate0(len, n, p, 0, n, ary0); /* compute and yield repeated combinations */ + tmpbuf_discard(t0); + RBASIC(ary0)->klass = rb_cArray; + } + return ary; +} + +/* + * call-seq: + * ary.product(other_ary, ...) -> new_ary + * ary.product(other_ary, ...) { |p| block } -> ary + * * Returns an array of all combinations of elements from all arrays. * The length of the returned array is the product of the length - * of ary and the argument arrays - * - * [1,2,3].product([4,5]) # => [[1,4],[1,5],[2,4],[2,5],[3,4],[3,5]] - * [1,2].product([1,2]) # => [[1,1],[1,2],[2,1],[2,2]] - * [1,2].product([3,4],[5,6]) # => [[1,3,5],[1,3,6],[1,4,5],[1,4,6], + * of +self+ and the argument arrays. + * If given a block, <i>product</i> will yield all combinations + * and return +self+ instead. + * + * + * [1,2,3].product([4,5]) #=> [[1,4],[1,5],[2,4],[2,5],[3,4],[3,5]] + * [1,2].product([1,2]) #=> [[1,1],[1,2],[2,1],[2,2]] + * [1,2].product([3,4],[5,6]) #=> [[1,3,5],[1,3,6],[1,4,5],[1,4,6], * # [2,3,5],[2,3,6],[2,4,5],[2,4,6]] - * [1,2].product() # => [[1],[2]] - * [1,2].product([]) # => [] + * [1,2].product() #=> [[1],[2]] + * [1,2].product([]) #=> [] */ static VALUE -rb_ary_product(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_product(int argc, VALUE *argv, VALUE ary) { int n = argc+1; /* How many arrays we're operating on */ - volatile VALUE t0 = tmpbuf(n, sizeof(VALUE)); + volatile VALUE t0 = tmpary(n); volatile VALUE t1 = tmpbuf(n, sizeof(int)); - VALUE *arrays = (VALUE*)RSTRING(t0)->ptr; /* The arrays we're computing the product of */ - int *counters = (int*)RSTRING(t1)->ptr; /* The current position in each one */ - VALUE result; /* The array we'll be returning */ + VALUE *arrays = RARRAY_PTR(t0); /* The arrays we're computing the product of */ + int *counters = (int*)RSTRING_PTR(t1); /* The current position in each one */ + VALUE result = Qnil; /* The array we'll be returning, when no block given */ long i,j; long resultlen = 1; @@ -3603,25 +4482,38 @@ rb_ary_product(argc, argv, ary) RBASIC(t1)->klass = 0; /* initialize the arrays of arrays */ + ARY_SET_LEN(t0, n); arrays[0] = ary; + for (i = 1; i < n; i++) arrays[i] = Qnil; for (i = 1; i < n; i++) arrays[i] = to_ary(argv[i-1]); - + /* initialize the counters for the arrays */ for (i = 0; i < n; i++) counters[i] = 0; - /* Compute the length of the result array; return [] if any is empty */ - for (i = 0; i < n; i++) { - long k = RARRAY(arrays[i])->len, l = resultlen; - if (k == 0) return rb_ary_new2(0); - resultlen *= k; - if (resultlen < k || resultlen < l || resultlen / k != l) { - rb_raise(rb_eRangeError, "too big to product"); + /* Otherwise, allocate and fill in an array of results */ + if (rb_block_given_p()) { + /* Make defensive copies of arrays; exit if any is empty */ + for (i = 0; i < n; i++) { + if (RARRAY_LEN(arrays[i]) == 0) goto done; + arrays[i] = ary_make_shared_copy(arrays[i]); } } - - /* Otherwise, allocate and fill in an array of results */ - result = rb_ary_new2(resultlen); - for (i = 0; i < resultlen; i++) { + else { + /* Compute the length of the result array; return [] if any is empty */ + for (i = 0; i < n; i++) { + long k = RARRAY_LEN(arrays[i]), l = resultlen; + if (k == 0) { + result = rb_ary_new2(0); + goto done; + } + resultlen *= k; + if (resultlen < k || resultlen < l || resultlen / k != l) { + rb_raise(rb_eRangeError, "too big to product"); + } + } + result = rb_ary_new2(resultlen); + } + for (;;) { int m; /* fill in one subarray */ VALUE subarray = rb_ary_new2(n); @@ -3630,7 +4522,19 @@ rb_ary_product(argc, argv, ary) } /* put it on the result array */ - rb_ary_push(result, subarray); + if(NIL_P(result)) { + FL_SET(t0, FL_USER5); + rb_yield(subarray); + if (! FL_TEST(t0, FL_USER5)) { + rb_raise(rb_eRuntimeError, "product reentered"); + } + else { + FL_UNSET(t0, FL_USER5); + } + } + else { + rb_ary_push(result, subarray); + } /* * Increment the last counter. If it overflows, reset to 0 @@ -3638,81 +4542,82 @@ rb_ary_product(argc, argv, ary) */ m = n-1; counters[m]++; - while (m > 0 && counters[m] == RARRAY(arrays[m])->len) { + while (counters[m] == RARRAY_LEN(arrays[m])) { counters[m] = 0; - m--; + /* If the first counter overflows, we are done */ + if (--m < 0) goto done; counters[m]++; } } +done: + tmpary_discard(t0); + tmpbuf_discard(t1); - return result; + return NIL_P(result) ? ary : result; } /* * call-seq: - * ary.take(n) => array - * + * ary.take(n) -> new_ary + * * Returns first n elements from <i>ary</i>. - * + * * a = [1, 2, 3, 4, 5, 0] - * a.take(3) # => [1, 2, 3] - * + * a.take(3) #=> [1, 2, 3] + * */ static VALUE -rb_ary_take(obj, n) - VALUE obj; - VALUE n; +rb_ary_take(VALUE obj, VALUE n) { long len = NUM2LONG(n); if (len < 0) { rb_raise(rb_eArgError, "attempt to take negative size"); } - return rb_ary_subseq(obj, 0, len); } /* * call-seq: - * ary.take_while {|arr| block } => array - * - * Passes elements to the block until the block returns nil or false, + * ary.take_while {|arr| block } -> new_ary + * ary.take_while -> an_enumerator + * + * Passes elements to the block until the block returns +nil+ or +false+, * then stops iterating and returns an array of all prior elements. - * + * + * If no block is given, an enumerator is returned instead. + * * a = [1, 2, 3, 4, 5, 0] - * a.take_while {|i| i < 3 } # => [1, 2] - * + * a.take_while {|i| i < 3 } #=> [1, 2] + * */ static VALUE -rb_ary_take_while(ary) - VALUE ary; +rb_ary_take_while(VALUE ary) { long i; RETURN_ENUMERATOR(ary, 0, 0); - for (i = 0; i < RARRAY(ary)->len; i++) { - if (!RTEST(rb_yield(RARRAY(ary)->ptr[i]))) break; + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (!RTEST(rb_yield(RARRAY_PTR(ary)[i]))) break; } return rb_ary_take(ary, LONG2FIX(i)); } /* * call-seq: - * ary.drop(n) => array - * - * Drops first n elements from <i>ary</i>, and returns rest elements - * in an array. - * + * ary.drop(n) -> new_ary + * + * Drops first n elements from +ary+ and returns the rest of + * the elements in an array. + * * a = [1, 2, 3, 4, 5, 0] - * a.drop(3) # => [4, 5, 0] - * + * a.drop(3) #=> [4, 5, 0] + * */ static VALUE -rb_ary_drop(ary, n) - VALUE ary; - VALUE n; +rb_ary_drop(VALUE ary, VALUE n) { VALUE result; long pos = NUM2LONG(n); @@ -3720,59 +4625,65 @@ rb_ary_drop(ary, n) rb_raise(rb_eArgError, "attempt to drop negative size"); } - result = rb_ary_subseq(ary, pos, RARRAY(ary)->len); + result = rb_ary_subseq(ary, pos, RARRAY_LEN(ary)); if (result == Qnil) result = rb_ary_new(); return result; } /* * call-seq: - * ary.drop_while {|arr| block } => array - * + * ary.drop_while {|arr| block } -> new_ary + * ary.drop_while -> an_enumerator + * * Drops elements up to, but not including, the first element for - * which the block returns nil or false and returns an array + * which the block returns +nil+ or +false+ and returns an array * containing the remaining elements. - * + * + * If no block is given, an enumerator is returned instead. + * * a = [1, 2, 3, 4, 5, 0] - * a.drop_while {|i| i < 3 } # => [3, 4, 5, 0] - * + * a.drop_while {|i| i < 3 } #=> [3, 4, 5, 0] + * */ static VALUE -rb_ary_drop_while(ary) - VALUE ary; +rb_ary_drop_while(VALUE ary) { long i; RETURN_ENUMERATOR(ary, 0, 0); - for (i = 0; i < RARRAY(ary)->len; i++) { - if (!RTEST(rb_yield(RARRAY(ary)->ptr[i]))) break; + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (!RTEST(rb_yield(RARRAY_PTR(ary)[i]))) break; } return rb_ary_drop(ary, LONG2FIX(i)); } -/* Arrays are ordered, integer-indexed collections of any object. - * Array indexing starts at 0, as in C or Java. A negative index is - * assumed to be relative to the end of the array---that is, an index of -1 - * indicates the last element of the array, -2 is the next to last - * element in the array, and so on. +/* Arrays are ordered, integer-indexed collections of any object. + * Array indexing starts at 0, as in C or Java. A negative index is + * assumed to be relative to the end of the array---that is, an index of -1 + * indicates the last element of the array, -2 is the next to last + * element in the array, and so on. */ void -Init_Array() +Init_Array(void) { +#undef rb_intern +#define rb_intern(str) rb_intern_const(str) + rb_cArray = rb_define_class("Array", rb_cObject); rb_include_module(rb_cArray, rb_mEnumerable); rb_define_alloc_func(rb_cArray, ary_alloc); rb_define_singleton_method(rb_cArray, "[]", rb_ary_s_create, -1); + rb_define_singleton_method(rb_cArray, "try_convert", rb_ary_s_try_convert, 1); rb_define_method(rb_cArray, "initialize", rb_ary_initialize, -1); rb_define_method(rb_cArray, "initialize_copy", rb_ary_replace, 1); - rb_define_method(rb_cArray, "to_s", rb_ary_to_s, 0); rb_define_method(rb_cArray, "inspect", rb_ary_inspect, 0); + rb_define_alias(rb_cArray, "to_s", "inspect"); rb_define_method(rb_cArray, "to_a", rb_ary_to_a, 0); rb_define_method(rb_cArray, "to_ary", rb_ary_to_ary_m, 0); rb_define_method(rb_cArray, "frozen?", rb_ary_frozen_p, 0); @@ -3803,18 +4714,21 @@ Init_Array() rb_define_method(rb_cArray, "find_index", rb_ary_index, -1); rb_define_method(rb_cArray, "index", rb_ary_index, -1); rb_define_method(rb_cArray, "rindex", rb_ary_rindex, -1); - rb_define_method(rb_cArray, "indexes", rb_ary_indexes, -1); - rb_define_method(rb_cArray, "indices", rb_ary_indexes, -1); rb_define_method(rb_cArray, "join", rb_ary_join_m, -1); rb_define_method(rb_cArray, "reverse", rb_ary_reverse_m, 0); rb_define_method(rb_cArray, "reverse!", rb_ary_reverse_bang, 0); + rb_define_method(rb_cArray, "rotate", rb_ary_rotate_m, -1); + rb_define_method(rb_cArray, "rotate!", rb_ary_rotate_bang, -1); rb_define_method(rb_cArray, "sort", rb_ary_sort, 0); rb_define_method(rb_cArray, "sort!", rb_ary_sort_bang, 0); + rb_define_method(rb_cArray, "sort_by!", rb_ary_sort_by_bang, 0); rb_define_method(rb_cArray, "collect", rb_ary_collect, 0); rb_define_method(rb_cArray, "collect!", rb_ary_collect_bang, 0); rb_define_method(rb_cArray, "map", rb_ary_collect, 0); rb_define_method(rb_cArray, "map!", rb_ary_collect_bang, 0); rb_define_method(rb_cArray, "select", rb_ary_select, 0); + rb_define_method(rb_cArray, "select!", rb_ary_select_bang, 0); + rb_define_method(rb_cArray, "keep_if", rb_ary_keep_if, 0); rb_define_method(rb_cArray, "values_at", rb_ary_values_at, -1); rb_define_method(rb_cArray, "delete", rb_ary_delete, 1); rb_define_method(rb_cArray, "delete_at", rb_ary_delete_at_m, 1); @@ -3848,14 +4762,15 @@ Init_Array() rb_define_method(rb_cArray, "compact!", rb_ary_compact_bang, 0); rb_define_method(rb_cArray, "flatten", rb_ary_flatten, -1); rb_define_method(rb_cArray, "flatten!", rb_ary_flatten_bang, -1); - rb_define_method(rb_cArray, "nitems", rb_ary_nitems, 0); rb_define_method(rb_cArray, "count", rb_ary_count, -1); - rb_define_method(rb_cArray, "shuffle!", rb_ary_shuffle_bang, 0); - rb_define_method(rb_cArray, "shuffle", rb_ary_shuffle, 0); - rb_define_method(rb_cArray, "choice", rb_ary_choice, 0); + rb_define_method(rb_cArray, "shuffle!", rb_ary_shuffle_bang, -1); + rb_define_method(rb_cArray, "shuffle", rb_ary_shuffle, -1); + rb_define_method(rb_cArray, "sample", rb_ary_sample, -1); rb_define_method(rb_cArray, "cycle", rb_ary_cycle, -1); rb_define_method(rb_cArray, "permutation", rb_ary_permutation, -1); rb_define_method(rb_cArray, "combination", rb_ary_combination, 1); + rb_define_method(rb_cArray, "repeated_permutation", rb_ary_repeated_permutation, 1); + rb_define_method(rb_cArray, "repeated_combination", rb_ary_repeated_combination, 1); rb_define_method(rb_cArray, "product", rb_ary_product, -1); rb_define_method(rb_cArray, "take", rb_ary_take, 1); @@ -3864,5 +4779,5 @@ Init_Array() rb_define_method(rb_cArray, "drop_while", rb_ary_drop_while, 0); id_cmp = rb_intern("<=>"); - inspect_key = rb_intern("__inspect_key__"); + sym_random = ID2SYM(rb_intern("random")); } |