diff options
Diffstat (limited to 'array.c')
| -rw-r--r-- | array.c | 7130 |
1 files changed, 5531 insertions, 1599 deletions
@@ -3,164 +3,730 @@ 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/encoding.h" +#include "ruby/util.h" +#include "ruby/st.h" +#include "probes.h" +#include "id.h" +#include "debug_counter.h" +#include "transient_heap.h" +#include "internal.h" + +#if !ARRAY_DEBUG +# define NDEBUG +#endif +#include "ruby_assert.h" VALUE rb_cArray; -static ID id_cmp; + +/* for OPTIMIZED_CMP: */ +#define id_cmp idCmp #define ARY_DEFAULT_SIZE 16 -#define ARY_MAX_SIZE (LONG_MAX / sizeof(VALUE)) +#define ARY_MAX_SIZE (LONG_MAX / (int)sizeof(VALUE)) +#define SMALL_ARRAY_LEN 16 + +static int +should_be_T_ARRAY(VALUE ary) +{ + return RB_TYPE_P(ary, T_ARRAY); +} + +static int +should_not_be_shared_and_embedded(VALUE ary) +{ + return !FL_TEST((ary), ELTS_SHARED) || !FL_TEST((ary), RARRAY_EMBED_FLAG); +} + +#define ARY_SHARED_P(ary) \ + (assert(should_be_T_ARRAY((VALUE)(ary))), \ + assert(should_not_be_shared_and_embedded((VALUE)ary)), \ + FL_TEST_RAW((ary),ELTS_SHARED)!=0) + +#define ARY_EMBED_P(ary) \ + (assert(should_be_T_ARRAY((VALUE)(ary))), \ + assert(should_not_be_shared_and_embedded((VALUE)ary)), \ + FL_TEST_RAW((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_HEAP_CAPA(a) (assert(!ARY_EMBED_P(a)), assert(!ARY_SHARED_ROOT_P(a)), \ + RARRAY(a)->as.heap.aux.capa) + +#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_HEAP_SIZE(a) (assert(!ARY_EMBED_P(a)), assert(ARY_OWNS_HEAP_P(a)), ARY_CAPA(a) * sizeof(VALUE)) + +#define ARY_OWNS_HEAP_P(a) (assert(should_be_T_ARRAY((VALUE)(a))), \ + !FL_TEST_RAW((a), ELTS_SHARED|RARRAY_EMBED_FLAG)) + +#define FL_SET_EMBED(a) do { \ + assert(!ARY_SHARED_P(a)); \ + FL_SET((a), RARRAY_EMBED_FLAG); \ + RARY_TRANSIENT_UNSET(a); \ + ary_verify(a); \ +} 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) : ARY_HEAP_CAPA(ary)) +#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_ROOT(ary) (assert(ARY_SHARED_P(ary)), RARRAY(ary)->as.heap.aux.shared_root) +#define ARY_SET_SHARED(ary, value) do { \ + const VALUE _ary_ = (ary); \ + const VALUE _value_ = (value); \ + assert(!ARY_EMBED_P(_ary_)); \ + assert(ARY_SHARED_P(_ary_)); \ + assert(ARY_SHARED_ROOT_P(_value_)); \ + RB_OBJ_WRITE(_ary_, &RARRAY(_ary_)->as.heap.aux.shared_root, _value_); \ +} while (0) +#define RARRAY_SHARED_ROOT_FLAG FL_USER5 +#define ARY_SHARED_ROOT_P(ary) (assert(should_be_T_ARRAY((VALUE)(ary))), \ + FL_TEST_RAW((ary), RARRAY_SHARED_ROOT_FLAG)) +#define ARY_SHARED_ROOT_REFCNT(ary) \ + (assert(ARY_SHARED_ROOT_P(ary)), RARRAY(ary)->as.heap.aux.capa) +#define ARY_SHARED_ROOT_OCCUPIED(ary) (ARY_SHARED_ROOT_REFCNT(ary) == 1) +#define ARY_SET_SHARED_ROOT_REFCNT(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)); \ + assert(!RARRAY_TRANSIENT_P(ary)); \ + FL_SET((ary), RARRAY_SHARED_ROOT_FLAG); \ +} while (0) + +static inline void +ARY_SET(VALUE a, long i, VALUE v) +{ + assert(!ARY_SHARED_P(a)); + assert(!OBJ_FROZEN(a)); + + RARRAY_ASET(a, i, v); +} +#undef RARRAY_ASET + + +#if ARRAY_DEBUG +#define ary_verify(ary) ary_verify_(ary, __FILE__, __LINE__) + +static VALUE +ary_verify_(VALUE ary, const char *file, int line) +{ + assert(RB_TYPE_P(ary, T_ARRAY)); + + if (FL_TEST(ary, ELTS_SHARED)) { + VALUE root = RARRAY(ary)->as.heap.aux.shared_root; + const VALUE *ptr = ARY_HEAP_PTR(ary); + const VALUE *root_ptr = RARRAY_CONST_PTR_TRANSIENT(root); + long len = ARY_HEAP_LEN(ary), root_len = RARRAY_LEN(root); + assert(FL_TEST(root, RARRAY_SHARED_ROOT_FLAG)); + assert(root_ptr <= ptr && ptr + len <= root_ptr + root_len); + ary_verify(root); + } + else if (ARY_EMBED_P(ary)) { + assert(!RARRAY_TRANSIENT_P(ary)); + assert(!ARY_SHARED_P(ary)); + assert(RARRAY_LEN(ary) <= RARRAY_EMBED_LEN_MAX); + } + else { +#if 1 + const VALUE *ptr = RARRAY_CONST_PTR_TRANSIENT(ary); + long i, len = RARRAY_LEN(ary); + volatile VALUE v; + if (len > 1) len = 1; /* check only HEAD */ + for (i=0; i<len; i++) { + v = ptr[i]; /* access check */ + } + v = v; +#endif + } + + if (RARRAY_TRANSIENT_P(ary)) { + assert(rb_transient_heap_managed_ptr_p(RARRAY_CONST_PTR_TRANSIENT(ary))); + } + + rb_transient_heap_verify(); + + return ary; +} void -rb_mem_clear(mem, size) - register VALUE *mem; - register long size; +rb_ary_verify(VALUE ary) +{ + ary_verify(ary); +} +#else +#define ary_verify(ary) ((void)0) +#endif + +VALUE * +rb_ary_ptr_use_start(VALUE ary) +{ +#if ARRAY_DEBUG + FL_SET_RAW(ary, RARRAY_PTR_IN_USE_FLAG); +#endif + return (VALUE *)RARRAY_CONST_PTR_TRANSIENT(ary); +} + +void +rb_ary_ptr_use_end(VALUE ary) +{ +#if ARRAY_DEBUG + FL_UNSET_RAW(ary, RARRAY_PTR_IN_USE_FLAG); +#endif +} + +void +rb_mem_clear(VALUE *mem, long size) { while (size--) { *mem++ = Qnil; } } +static void +ary_mem_clear(VALUE ary, long beg, long size) +{ + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + rb_mem_clear(ptr + beg, 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 +static void +ary_memfill(VALUE ary, long beg, long size, VALUE val) +{ + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + memfill(ptr + beg, size, val); + RB_OBJ_WRITTEN(ary, Qundef, val); + }); +} + +static void +ary_memcpy0(VALUE ary, long beg, long argc, const VALUE *argv, VALUE buff_owner_ary) +{ + assert(!ARY_SHARED_P(buff_owner_ary)); + + if (argc > (int)(128/sizeof(VALUE)) /* is magic number (cache line size) */) { + rb_gc_writebarrier_remember(buff_owner_ary); + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMCPY(ptr+beg, argv, VALUE, argc); + }); + } + else { + int i; + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + for (i=0; i<argc; i++) { + RB_OBJ_WRITE(buff_owner_ary, &ptr[i+beg], argv[i]); + } + }); + } +} +static void +ary_memcpy(VALUE ary, long beg, long argc, const VALUE *argv) +{ + ary_memcpy0(ary, beg, argc, argv, ary); +} + +static VALUE * +ary_heap_alloc(VALUE ary, size_t capa) +{ + VALUE *ptr = rb_transient_heap_alloc(ary, sizeof(VALUE) * capa); + + if (ptr != NULL) { + RARY_TRANSIENT_SET(ary); + } + else { + RARY_TRANSIENT_UNSET(ary); + ptr = ALLOC_N(VALUE, capa); + } + + return ptr; +} + +static void +ary_heap_free_ptr(VALUE ary, const VALUE *ptr, long size) +{ + if (RARRAY_TRANSIENT_P(ary)) { + /* ignore it */ + } + else { + ruby_sized_xfree((void *)ptr, size); + } +} + +static void +ary_heap_free(VALUE ary) +{ + if (RARRAY_TRANSIENT_P(ary)) { + RARY_TRANSIENT_UNSET(ary); + } + else { + ary_heap_free_ptr(ary, ARY_HEAP_PTR(ary), ARY_HEAP_SIZE(ary)); + } +} + +static void +ary_heap_realloc(VALUE ary, size_t new_capa) +{ + size_t old_capa = ARY_HEAP_CAPA(ary); + + if (RARRAY_TRANSIENT_P(ary)) { + if (new_capa <= old_capa) { + /* do nothing */ + } + else { + VALUE *new_ptr = rb_transient_heap_alloc(ary, sizeof(VALUE) * new_capa); + + if (new_ptr == NULL) { + new_ptr = ALLOC_N(VALUE, new_capa); + RARY_TRANSIENT_UNSET(ary); + } + + MEMCPY(new_ptr, ARY_HEAP_PTR(ary), VALUE, old_capa); + ARY_SET_PTR(ary, new_ptr); + } + } + else { + SIZED_REALLOC_N(RARRAY(ary)->as.heap.ptr, VALUE, new_capa, old_capa); + } + ary_verify(ary); +} + +#if USE_TRANSIENT_HEAP static inline void -rb_ary_modify_check(ary) - VALUE ary; +rb_ary_transient_heap_evacuate_(VALUE ary, int transient, int promote) +{ + if (transient) { + VALUE *new_ptr; + const VALUE *old_ptr = ARY_HEAP_PTR(ary); + long capa = ARY_HEAP_CAPA(ary); + long len = ARY_HEAP_LEN(ary); + + if (ARY_SHARED_ROOT_P(ary)) { + capa = len; + } + + assert(ARY_OWNS_HEAP_P(ary)); + assert(RARRAY_TRANSIENT_P(ary)); + assert(!ARY_PTR_USING_P(ary)); + + if (promote) { + new_ptr = ALLOC_N(VALUE, capa); + RARY_TRANSIENT_UNSET(ary); + } + else { + new_ptr = ary_heap_alloc(ary, capa); + } + + MEMCPY(new_ptr, old_ptr, VALUE, capa); + /* do not use ARY_SET_PTR() because they assert !frozen */ + RARRAY(ary)->as.heap.ptr = new_ptr; + } + + ary_verify(ary); +} + +void +rb_ary_transient_heap_evacuate(VALUE ary, int promote) { - 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"); + rb_ary_transient_heap_evacuate_(ary, RARRAY_TRANSIENT_P(ary), promote); } +void +rb_ary_detransient(VALUE ary) +{ + assert(RARRAY_TRANSIENT_P(ary)); + rb_ary_transient_heap_evacuate_(ary, TRUE, TRUE); +} +#else +void +rb_ary_detransient(VALUE ary) +{ + /* do nothing */ +} +#endif + static void -rb_ary_modify(ary) - VALUE ary; +ary_resize_capa(VALUE ary, long capacity) { - VALUE *ptr; + 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 = ary_heap_alloc(ary, 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 { + ary_heap_realloc(ary, capacity); + } + ARY_SET_CAPA(ary, capacity); + } + else { + if (!ARY_EMBED_P(ary)) { + long len = ARY_HEAP_LEN(ary); + long old_capa = ARY_HEAP_CAPA(ary); + const VALUE *ptr = ARY_HEAP_PTR(ary); + + if (len > capacity) len = capacity; + MEMCPY((VALUE *)RARRAY(ary)->as.ary, ptr, VALUE, len); + ary_heap_free_ptr(ary, ptr, old_capa); + + FL_SET_EMBED(ary); + ARY_SET_LEN(ary, len); + } + } + + ary_verify(ary); +} + +static inline void +ary_shrink_capa(VALUE ary) +{ + long capacity = ARY_HEAP_LEN(ary); + long old_capa = ARY_HEAP_CAPA(ary); + assert(!ARY_SHARED_P(ary)); + assert(old_capa >= capacity); + if (old_capa > capacity) ary_heap_realloc(ary, capacity); + + ary_verify(ary); +} + +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); + ary_verify(ary); +} + +static void +rb_ary_decrement_share(VALUE shared_root) +{ + if (shared_root) { + long num = ARY_SHARED_ROOT_REFCNT(shared_root) - 1; + if (num == 0) { + rb_ary_free(shared_root); + rb_gc_force_recycle(shared_root); + } + else if (num > 0) { + ARY_SET_SHARED_ROOT_REFCNT(shared_root, num); + } + } +} + +static void +rb_ary_unshare(VALUE ary) +{ + VALUE shared_root = RARRAY(ary)->as.heap.aux.shared_root; + rb_ary_decrement_share(shared_root); + FL_UNSET_SHARED(ary); +} + +static inline void +rb_ary_unshare_safe(VALUE ary) +{ + if (ARY_SHARED_P(ary) && !ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + } +} + +static VALUE +rb_ary_increment_share(VALUE shared_root) +{ + long num = ARY_SHARED_ROOT_REFCNT(shared_root); + if (num >= 0) { + ARY_SET_SHARED_ROOT_REFCNT(shared_root, num + 1); + } + return shared_root; +} + +static void +rb_ary_set_shared(VALUE ary, VALUE shared_root) +{ + rb_ary_increment_share(shared_root); + FL_SET_SHARED(ary); + RB_DEBUG_COUNTER_INC(obj_ary_shared_create); + ARY_SET_SHARED(ary, shared_root); +} + +static inline void +rb_ary_modify_check(VALUE ary) +{ + rb_check_frozen(ary); + ary_verify(ary); +} + +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 shared_len, len = RARRAY_LEN(ary); + VALUE shared_root = ARY_SHARED_ROOT(ary); + + ary_verify(shared_root); + + if (len <= RARRAY_EMBED_LEN_MAX) { + const VALUE *ptr = ARY_HEAP_PTR(ary); + FL_UNSET_SHARED(ary); + FL_SET_EMBED(ary); + MEMCPY((VALUE *)ARY_EMBED_PTR(ary), ptr, VALUE, len); + rb_ary_decrement_share(shared_root); + ARY_SET_EMBED_LEN(ary, len); + } + else if (ARY_SHARED_ROOT_OCCUPIED(shared_root) && len > ((shared_len = RARRAY_LEN(shared_root))>>1)) { + long shift = RARRAY_CONST_PTR_TRANSIENT(ary) - RARRAY_CONST_PTR_TRANSIENT(shared_root); + FL_UNSET_SHARED(ary); + ARY_SET_PTR(ary, RARRAY_CONST_PTR_TRANSIENT(shared_root)); + ARY_SET_CAPA(ary, shared_len); + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr, ptr+shift, VALUE, len); + }); + FL_SET_EMBED(shared_root); + rb_ary_decrement_share(shared_root); + } + else { + VALUE *ptr = ary_heap_alloc(ary, len); + MEMCPY(ptr, ARY_HEAP_PTR(ary), VALUE, len); + rb_ary_unshare(ary); + ARY_SET_CAPA(ary, len); + ARY_SET_PTR(ary, ptr); + } + + rb_gc_writebarrier_remember(ary); } + ary_verify(ary); } -VALUE -rb_ary_freeze(ary) - VALUE ary; +static VALUE +ary_ensure_room_for_push(VALUE ary, long add_len) { - return rb_obj_freeze(ary); + long old_len = RARRAY_LEN(ary); + long new_len = old_len + add_len; + long capa; + + if (old_len > ARY_MAX_SIZE - add_len) { + rb_raise(rb_eIndexError, "index %ld too big", new_len); + } + if (ARY_SHARED_P(ary)) { + if (new_len > RARRAY_EMBED_LEN_MAX) { + VALUE shared_root = ARY_SHARED_ROOT(ary); + if (ARY_SHARED_ROOT_OCCUPIED(shared_root)) { + if (ARY_HEAP_PTR(ary) - RARRAY_CONST_PTR_TRANSIENT(shared_root) + new_len <= RARRAY_LEN(shared_root)) { + rb_ary_modify_check(ary); + + ary_verify(ary); + ary_verify(shared_root); + return shared_root; + } + else { + /* if array is shared, then it is likely it participate in push/shift pattern */ + rb_ary_modify(ary); + capa = ARY_CAPA(ary); + if (new_len > capa - (capa >> 6)) { + ary_double_capa(ary, new_len); + } + ary_verify(ary); + return ary; + } + } + } + ary_verify(ary); + rb_ary_modify(ary); + } + else { + rb_ary_modify_check(ary); + } + capa = ARY_CAPA(ary); + if (new_len > capa) { + ary_double_capa(ary, new_len); + } + + ary_verify(ary); + return ary; } /* * call-seq: - * array.frozen? -> true or false + * ary.freeze -> ary + * + * Calls Object#freeze on +ary+ to prevent any further + * modification. A RuntimeError will be raised if a modification + * attempt is made. * - * Return <code>true</code> if this array is frozen (or temporarily frozen - * while being sorted). */ -static VALUE -rb_ary_frozen_p(ary) - VALUE ary; +VALUE +rb_ary_freeze(VALUE ary) +{ + return rb_obj_freeze(ary); +} + +/* This can be used to take a snapshot of an array (with + e.g. rb_ary_replace) and check later whether the array has been + modified from the snapshot. The snapshot is cheap, though if + something does modify the array it will pay the cost of copying + it. If Array#pop or Array#shift has been called, the array will + be still shared with the snapshot, but the array length will + differ. */ +VALUE +rb_ary_shared_with_p(VALUE ary1, VALUE ary2) { - if (OBJ_FROZEN(ary)) return Qtrue; - if (FL_TEST(ary, ARY_TMPLOCK)) return Qtrue; + if (!ARY_EMBED_P(ary1) && ARY_SHARED_P(ary1) && + !ARY_EMBED_P(ary2) && ARY_SHARED_P(ary2) && + RARRAY(ary1)->as.heap.aux.shared_root == RARRAY(ary2)->as.heap.aux.shared_root && + RARRAY(ary1)->as.heap.len == RARRAY(ary2)->as.heap.len) { + 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; - + NEWOBJ_OF(ary, struct RArray, klass, T_ARRAY | RARRAY_EMBED_FLAG | (RGENGC_WB_PROTECTED_ARRAY ? FL_WB_PROTECTED : 0)); + /* Created array is: + * 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; +empty_ary_alloc(VALUE klass) { - VALUE ary = ary_alloc(klass); + RUBY_DTRACE_CREATE_HOOK(ARRAY, 0); + return ary_alloc(klass); +} - if (len < 0) { +static VALUE +ary_new(VALUE klass, long capa) +{ + VALUE ary,*ptr; + + 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; + + RUBY_DTRACE_CREATE_HOOK(ARRAY, capa); + + ary = ary_alloc(klass); + if (capa > RARRAY_EMBED_LEN_MAX) { + ptr = ary_heap_alloc(ary, capa); + FL_UNSET_EMBED(ary); + ARY_SET_PTR(ary, ptr); + ARY_SET_CAPA(ary, capa); + ARY_SET_HEAP_LEN(ary, 0); + } return ary; } VALUE -rb_ary_new2(len) - long len; +rb_ary_new_capa(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 +(rb_ary_new_from_args)(long n, ...) { va_list ar; VALUE ary; @@ -168,120 +734,290 @@ 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); + ARY_SET(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; +MJIT_FUNC_EXPORTED VALUE +rb_ary_tmp_new_from_values(VALUE klass, long n, const VALUE *elts) { VALUE ary; - ary = rb_ary_new2(n); + ary = ary_new(klass, n); if (n > 0 && elts) { - MEMCPY(RARRAY(ary)->ptr, elts, VALUE, n); + ary_memcpy(ary, 0, n, elts); + 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_assoc_new(car, cdr) - VALUE car, cdr; +rb_ary_new_from_values(long n, const VALUE *elts) { - VALUE ary; + return rb_ary_tmp_new_from_values(rb_cArray, n, elts); +} - ary = rb_ary_new2(2); - RARRAY(ary)->ptr[0] = car; - RARRAY(ary)->ptr[1] = cdr; - RARRAY(ary)->len = 2; +VALUE +rb_ary_tmp_new(long capa) +{ + VALUE ary = ary_new(0, capa); + rb_ary_transient_heap_evacuate(ary, TRUE); + return ary; +} +VALUE +rb_ary_tmp_new_fill(long capa) +{ + VALUE ary = ary_new(0, capa); + ary_memfill(ary, 0, capa, Qnil); + ARY_SET_LEN(ary, capa); + rb_ary_transient_heap_evacuate(ary, TRUE); return ary; } +void +rb_ary_free(VALUE ary) +{ + if (ARY_OWNS_HEAP_P(ary)) { + if (USE_DEBUG_COUNTER && + !ARY_SHARED_ROOT_P(ary) && + ARY_HEAP_CAPA(ary) > RARRAY_LEN(ary)) { + RB_DEBUG_COUNTER_INC(obj_ary_extracapa); + } + + if (RARRAY_TRANSIENT_P(ary)) { + RB_DEBUG_COUNTER_INC(obj_ary_transient); + } + else { + RB_DEBUG_COUNTER_INC(obj_ary_ptr); + ary_heap_free(ary); + } + } + else { + RB_DEBUG_COUNTER_INC(obj_ary_embed); + } + + if (ARY_SHARED_P(ary)) { + RB_DEBUG_COUNTER_INC(obj_ary_shared); + } + if (ARY_SHARED_ROOT_P(ary) && ARY_SHARED_ROOT_OCCUPIED(ary)) { + RB_DEBUG_COUNTER_INC(obj_ary_shared_root_occupied); + } +} + +RUBY_FUNC_EXPORTED size_t +rb_ary_memsize(VALUE ary) +{ + if (ARY_OWNS_HEAP_P(ary)) { + return ARY_CAPA(ary) * 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 | RARRAY_TRANSIENT_FLAG); +} + static VALUE -to_ary(ary) - VALUE ary; +ary_make_shared(VALUE ary) { - return rb_convert_type(ary, T_ARRAY, "Array", "to_ary"); + assert(!ARY_EMBED_P(ary)); + ary_verify(ary); + + if (ARY_SHARED_P(ary)) { + return ARY_SHARED_ROOT(ary); + } + else if (ARY_SHARED_ROOT_P(ary)) { + return ary; + } + else if (OBJ_FROZEN(ary)) { + rb_ary_transient_heap_evacuate(ary, TRUE); + ary_shrink_capa(ary); + FL_SET_SHARED_ROOT(ary); + ARY_SET_SHARED_ROOT_REFCNT(ary, 1); + return ary; + } + else { + long capa = ARY_CAPA(ary), len = RARRAY_LEN(ary); + const VALUE *ptr; + NEWOBJ_OF(shared, struct RArray, 0, T_ARRAY | (RGENGC_WB_PROTECTED_ARRAY ? FL_WB_PROTECTED : 0)); + + rb_ary_transient_heap_evacuate(ary, TRUE); + ptr = ARY_HEAP_PTR(ary); + + FL_UNSET_EMBED(shared); + ARY_SET_LEN((VALUE)shared, capa); + ARY_SET_PTR((VALUE)shared, ptr); + ary_mem_clear((VALUE)shared, len, capa - len); + FL_SET_SHARED_ROOT(shared); + ARY_SET_SHARED_ROOT_REFCNT((VALUE)shared, 1); + FL_SET_SHARED(ary); + RB_DEBUG_COUNTER_INC(obj_ary_shared_create); + ARY_SET_SHARED(ary, (VALUE)shared); + OBJ_FREEZE(shared); + + ary_verify((VALUE)shared); + ary_verify(ary); + + return (VALUE)shared; + } +} + +static VALUE +ary_make_substitution(VALUE ary) +{ + long len = RARRAY_LEN(ary); + + if (len <= RARRAY_EMBED_LEN_MAX) { + VALUE subst = rb_ary_new2(len); + ary_memcpy(subst, 0, len, RARRAY_CONST_PTR_TRANSIENT(ary)); + ARY_SET_EMBED_LEN(subst, len); + return subst; + } + else { + return rb_ary_increment_share(ary_make_shared(ary)); + } } VALUE -rb_check_array_type(ary) - VALUE ary; +rb_assoc_new(VALUE car, VALUE cdr) { - return rb_check_convert_type(ary, T_ARRAY, "Array", "to_ary"); + return rb_ary_new3(2, car, cdr); } -static VALUE rb_ary_replace _((VALUE, VALUE)); +VALUE +rb_to_array_type(VALUE ary) +{ + return rb_convert_type_with_id(ary, T_ARRAY, "Array", idTo_ary); +} +#define to_ary rb_to_array_type + +VALUE +rb_check_array_type(VALUE ary) +{ + return rb_check_convert_type_with_id(ary, T_ARRAY, "Array", idTo_ary); +} + +MJIT_FUNC_EXPORTED VALUE +rb_check_to_array(VALUE ary) +{ + return rb_check_convert_type_with_id(ary, T_ARRAY, "Array", idTo_a); +} /* * call-seq: - * Array.new(size=0, obj=nil) + * Array.try_convert(obj) -> array or nil + * + * Tries to convert +obj+ into an array, using the +to_ary+ method. Returns + * the converted array or +nil+ if +obj+ cannot be converted. + * 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: + * Array.new(size=0, default=nil) * 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 +default+ are sent, an array is created with + * +size+ copies of +default+. Take notice that all elements will reference the + * same object +default+. + * + * 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 = Array.new(2) { Hash.new } + * 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) && ARY_HEAP_PTR(ary) != NULL) { + ary_heap_free(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)) { @@ -291,16 +1027,16 @@ rb_ary_initialize(argc, argv, ary) } len = NUM2LONG(size); + /* NUM2LONG() may call size.to_int, ary can be frozen, modified, etc */ if (len < 0) { rb_raise(rb_eArgError, "negative array size"); } 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; - } + /* recheck after argument conversion */ + rb_ary_modify(ary); + ary_resize_capa(ary, len); if (rb_block_given_p()) { long i; @@ -309,54 +1045,46 @@ 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; + ary_memfill(ary, 0, len, val); + ARY_SET_LEN(ary, len); } - return ary; } - -/* -* Returns a new array populated with the given objects. -* -* Array.[]( 1, 'a', /^A/ ) -* Array[ 1, 'a', /^A/ ] -* [ 1, 'a', /^A/ ] -*/ +/* + * Returns a new array populated with the given objects. + * + * Array.[]( 1, 'a', /^A/) # => [1, "a", /^A/] + * Array[ 1, 'a', /^A/ ] # => [1, "a", /^A/] + * [ 1, 'a', /^A/ ] # => [1, "a", /^A/] + */ 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) { + ary_memcpy(ary, 0, argc, argv); + 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) { + long len = RARRAY_LEN(ary); + if (idx < 0) { - idx += RARRAY(ary)->len; + idx += len; 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 - len, -len); } } else if (idx >= ARY_MAX_SIZE) { @@ -364,287 +1092,467 @@ 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 (idx >= ARY_CAPA(ary)) { + ary_double_capa(ary, idx); + } + if (idx > len) { + ary_mem_clear(ary, len, idx - len + 1); + } - 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 >= len) { + ARY_SET_LEN(ary, idx + 1); } - if (idx > RARRAY(ary)->len) { - rb_mem_clear(RARRAY(ary)->ptr + RARRAY(ary)->len, - idx-RARRAY(ary)->len + 1); + ARY_SET(ary, idx, val); +} + +static VALUE +ary_make_partial(VALUE ary, VALUE klass, long offset, long len) +{ + assert(offset >= 0); + assert(len >= 0); + assert(offset+len <= RARRAY_LEN(ary)); + + if (len <= RARRAY_EMBED_LEN_MAX) { + VALUE result = ary_alloc(klass); + ary_memcpy(result, 0, len, RARRAY_CONST_PTR_TRANSIENT(ary) + offset); + 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_CONST_PTR_TRANSIENT(ary)); + ARY_SET_LEN(result, RARRAY_LEN(ary)); + rb_ary_set_shared(result, shared); - if (idx >= RARRAY(ary)->len) { - RARRAY(ary)->len = idx + 1; + ARY_INCREASE_PTR(result, offset); + ARY_SET_LEN(result, len); + + ary_verify(shared); + ary_verify(result); + return result; } - RARRAY(ary)->ptr[idx] = val; +} + +static VALUE +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, const VALUE *argv, VALUE ary, enum ary_take_pos_flags last) +{ + long n; + long len; + long offset = 0; + + argc = rb_check_arity(argc, 0, 1); + /* the case optional argument is omitted should be handled in + * callers of this function. if another arity case is added, + * this arity check needs to rewrite. */ + RUBY_ASSERT_ALWAYS(argc == 1); + + n = NUM2LONG(argv[0]); + len = RARRAY_LEN(ary); + if (n > len) { + n = len; + } + else if (n < 0) { + rb_raise(rb_eArgError, "negative array size"); + } + if (last) { + offset = len - n; + } + return ary_make_partial(ary, rb_cArray, offset, n); } /* * 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. * - * [ 1, 2 ] << "c" << "d" << [ 3, 4 ] + * a = [ 1, 2 ] + * a << "c" << "d" << [ 3, 4 ] + * #=> [ 1, 2, "c", "d", [ 3, 4 ] ] + * a * #=> [ 1, 2, "c", "d", [ 3, 4 ] ] * */ VALUE -rb_ary_push(ary, item) - VALUE ary; - VALUE item; +rb_ary_push(VALUE ary, VALUE item) +{ + long idx = RARRAY_LEN((ary_verify(ary), ary)); + VALUE target_ary = ary_ensure_room_for_push(ary, 1); + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + RB_OBJ_WRITE(target_ary, &ptr[idx], item); + }); + ARY_SET_LEN(ary, idx + 1); + ary_verify(ary); + return ary; +} + +VALUE +rb_ary_cat(VALUE ary, const VALUE *argv, long len) { - rb_ary_store(ary, RARRAY(ary)->len, item); + long oldlen = RARRAY_LEN(ary); + VALUE target_ary = ary_ensure_room_for_push(ary, len); + ary_memcpy0(ary, oldlen, len, argv, target_ary); + ARY_SET_LEN(ary, oldlen + len); return ary; } -/* +/* * call-seq: - * array.push(obj, ... ) -> array - * - * Append---Pushes the given object(s) on to the end of this array. This + * ary.push(obj, ...) -> ary + * ary.append(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. + * may be chained together. See also Array#pop for the opposite + * effect. * * a = [ "a", "b", "c" ] - * a.push("d", "e", "f") + * a.push("d", "e", "f") * #=> ["a", "b", "c", "d", "e", "f"] + * [1, 2, 3].push(4).push(5) + * #=> [1, 2, 3, 4, 5] */ 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) { - while (argc--) { - rb_ary_push(ary, *argv++); - } - return ary; + return rb_ary_cat(ary, argv, argc); +} + +VALUE +rb_ary_pop(VALUE ary) +{ + long n; + rb_ary_modify_check(ary); + n = RARRAY_LEN(ary); + if (n == 0) return Qnil; + if (ARY_OWNS_HEAP_P(ary) && + n * 3 < ARY_CAPA(ary) && + ARY_CAPA(ary) > ARY_DEFAULT_SIZE) + { + ary_resize_capa(ary, n * 2); + } + --n; + ARY_SET_LEN(ary, n); + ary_verify(ary); + return RARRAY_AREF(ary, n); } /* * call-seq: - * array.pop -> obj or nil - * - * Removes the last element from <i>self</i> and returns it, or - * <code>nil</code> if the array is empty. - * - * a = [ "a", "m", "z" ] - * a.pop #=> "z" - * a #=> ["a", "m"] + * ary.pop -> obj or nil + * ary.pop(n) -> new_ary + * + * Removes the last element from +self+ and returns it, or + * +nil+ 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. See also + * Array#push for the opposite effect. + * + * a = [ "a", "b", "c", "d" ] + * a.pop #=> "d" + * a.pop(2) #=> ["b", "c"] + * a #=> ["a"] */ -VALUE -rb_ary_pop(ary) - VALUE ary; +static VALUE +rb_ary_pop_m(int argc, VALUE *argv, VALUE ary) { - rb_ary_modify_check(ary); - if (RARRAY(ary)->len == 0) return Qnil; - if (!FL_TEST(ary, ELTS_SHARED) && - RARRAY(ary)->len * 2 < 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); + VALUE result; + + if (argc == 0) { + return rb_ary_pop(ary); } - return RARRAY(ary)->ptr[--RARRAY(ary)->len]; + + rb_ary_modify_check(ary); + result = ary_take_first_or_last(argc, argv, ary, ARY_TAKE_LAST); + ARY_INCREASE_LEN(ary, -RARRAY_LEN(result)); + ary_verify(ary); + return result; } -static VALUE -ary_make_shared(ary) - VALUE ary; +VALUE +rb_ary_shift(VALUE ary) { - if (!FL_TEST(ary, ELTS_SHARED)) { - NEWOBJ(shared, struct RArray); - OBJSETUP(shared, rb_cArray, T_ARRAY); + VALUE top; + long len = RARRAY_LEN(ary); - 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); - OBJ_FREEZE(shared); - return (VALUE)shared; + rb_ary_modify_check(ary); + if (len == 0) return Qnil; + top = RARRAY_AREF(ary, 0); + if (!ARY_SHARED_P(ary)) { + if (len < ARY_DEFAULT_SIZE) { + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr, ptr+1, VALUE, len-1); + }); /* WB: no new reference */ + ARY_INCREASE_LEN(ary, -1); + ary_verify(ary); + return top; + } + assert(!ARY_EMBED_P(ary)); /* ARY_EMBED_LEN_MAX < ARY_DEFAULT_SIZE */ + + ARY_SET(ary, 0, Qnil); + ary_make_shared(ary); } - else { - return RARRAY(ary)->aux.shared; + else if (ARY_SHARED_ROOT_OCCUPIED(ARY_SHARED_ROOT(ary))) { + RARRAY_PTR_USE_TRANSIENT(ary, ptr, ptr[0] = Qnil); } + ARY_INCREASE_PTR(ary, 1); /* shift ptr */ + ARY_INCREASE_LEN(ary, -1); + + ary_verify(ary); + + return top; } /* * call-seq: - * array.shift -> obj or nil - * - * Returns the first element of <i>self</i> and removes it (shifting all - * other elements down by one). Returns <code>nil</code> if the array + * ary.shift -> obj or nil + * ary.shift(n) -> new_ary + * + * Removes the first element of +self+ and returns it (shifting all + * other elements down by one). Returns +nil+ 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. With +ary+ + * containing only the remainder elements, not including what was shifted to + * +new_ary+. See also Array#unshift for the opposite effect. + * + * args = [ "-m", "-q", "filename" ] + * args.shift #=> "-m" + * args #=> ["-q", "filename"] + * * args = [ "-m", "-q", "filename" ] - * args.shift #=> "-m" - * args #=> ["-q", "filename"] + * args.shift(2) #=> ["-m", "-q"] + * args #=> ["filename"] */ -VALUE -rb_ary_shift(ary) - VALUE ary; +static VALUE +rb_ary_shift_m(int argc, VALUE *argv, VALUE ary) { - VALUE top; + VALUE result; + long n; + + if (argc == 0) { + return rb_ary_shift(ary); + } + + rb_ary_modify_check(ary); + result = ary_take_first_or_last(argc, argv, ary, ARY_TAKE_FIRST); + n = RARRAY_LEN(result); + rb_ary_behead(ary,n); + + return result; +} + +MJIT_FUNC_EXPORTED VALUE +rb_ary_behead(VALUE ary, long n) +{ + if (n<=0) return ary; rb_ary_modify_check(ary); - if (RARRAY(ary)->len == 0) return Qnil; - top = RARRAY(ary)->ptr[0]; - if (RARRAY_LEN(ary) < ARY_DEFAULT_SIZE && !FL_TEST(ary, ELTS_SHARED)) { - MEMMOVE(RARRAY_PTR(ary), RARRAY_PTR(ary)+1, VALUE, RARRAY_LEN(ary)-1); + if (ARY_SHARED_P(ary)) { + if (ARY_SHARED_ROOT_OCCUPIED(ARY_SHARED_ROOT(ary))) { + setup_occupied_shared: + ary_mem_clear(ary, 0, n); + } + ARY_INCREASE_PTR(ary, n); } else { - if (!FL_TEST(ary, ELTS_SHARED)) { - RARRAY(ary)->ptr[0] = Qnil; + if (RARRAY_LEN(ary) < ARY_DEFAULT_SIZE) { + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr, ptr+n, VALUE, RARRAY_LEN(ary)-n); + }); /* WB: no new reference */ + } + else { + ary_make_shared(ary); + goto setup_occupied_shared; } - ary_make_shared(ary); - RARRAY(ary)->ptr++; /* shift ptr */ } - RARRAY(ary)->len--; + ARY_INCREASE_LEN(ary, -n); - return top; + ary_verify(ary); + return ary; } -VALUE -rb_ary_unshift(ary, item) - VALUE ary, item; +static VALUE +ary_ensure_room_for_unshift(VALUE ary, int argc) { - 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; + long len = RARRAY_LEN(ary); + long new_len = len + argc; + long capa; + const VALUE *head, *sharedp; + + if (len > ARY_MAX_SIZE - argc) { + rb_raise(rb_eIndexError, "index %ld too big", new_len); + } + + if (ARY_SHARED_P(ary)) { + VALUE shared_root = ARY_SHARED_ROOT(ary); + capa = RARRAY_LEN(shared_root); + if (ARY_SHARED_ROOT_OCCUPIED(shared_root) && capa > new_len) { + rb_ary_modify_check(ary); + head = RARRAY_CONST_PTR_TRANSIENT(ary); + sharedp = RARRAY_CONST_PTR_TRANSIENT(shared_root); + goto makeroom_if_need; } - 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); + rb_ary_modify(ary); + capa = ARY_CAPA(ary); + if (capa - (capa >> 6) <= new_len) { + ary_double_capa(ary, new_len); + } - RARRAY(ary)->len++; - RARRAY(ary)->ptr[0] = item; + /* use shared array for big "queues" */ + if (new_len > ARY_DEFAULT_SIZE * 4) { + ary_verify(ary); - return ary; + /* make a room for unshifted items */ + capa = ARY_CAPA(ary); + ary_make_shared(ary); + + head = sharedp = RARRAY_CONST_PTR_TRANSIENT(ary); + goto makeroom; + makeroom_if_need: + if (head - sharedp < argc) { + long room; + makeroom: + room = capa - new_len; + room -= room >> 4; + MEMMOVE((VALUE *)sharedp + argc + room, head, VALUE, len); + head = sharedp + argc + room; + } + ARY_SET_PTR(ary, head - argc); + assert(ARY_SHARED_ROOT_OCCUPIED(ARY_SHARED_ROOT(ary))); + + ary_verify(ary); + return ARY_SHARED_ROOT(ary); + } + else { + /* sliding items */ + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr + argc, ptr, VALUE, len); + }); + + ary_verify(ary); + 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 + * ary.prepend(obj, ...) -> ary + * + * Prepends objects to the front of +self+, moving other elements upwards. + * See also Array#shift for the opposite effect. + * * 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; - - if (argc == 0) return ary; + long len = RARRAY_LEN(ary); + VALUE target_ary; - /* make rooms by setting the last item */ - rb_ary_store(ary, len + argc - 1, Qnil); + if (argc == 0) { + rb_ary_modify_check(ary); + return ary; + } - /* sliding items */ - MEMMOVE(RARRAY(ary)->ptr + argc, RARRAY(ary)->ptr, VALUE, len); - MEMCPY(RARRAY(ary)->ptr, argv, VALUE, argc); - + target_ary = ary_ensure_room_for_unshift(ary, argc); + ary_memcpy0(ary, 0, argc, argv, target_ary); + ARY_SET_LEN(ary, len + 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) { + long len = RARRAY_LEN(ary); + if (len == 0) return Qnil; + if (offset < 0 || len <= offset) { return Qnil; } - return RARRAY(ary)->ptr[offset]; + return RARRAY_AREF(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; - } - return rb_ary_elt(ary, offset); + return rb_ary_entry_internal(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; + long alen = RARRAY_LEN(ary); - if (beg > RARRAY(ary)->len) return Qnil; + if (beg > alen) 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 (alen < len || alen < beg + len) { + len = alen - 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); } -/* +static VALUE rb_ary_aref2(VALUE ary, VALUE b, VALUE e); + +/* * 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 the +start+ index and continuing for +length+ + * elements, or returns a subarray specified by +range+ of indices. + * + * Negative indices count backward from the end of the array (-1 is the last + * element). For +start+ and +range+ cases the starting index is just before + * an element. Additionally, an empty array is returned when the starting + * index for an element range is at the end of the array. * - * 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 - * (or starting index) are out of range. + * Returns +nil+ if the index (or starting index) are out of range. * * a = [ "a", "b", "c", "d", "e" ] * a[2] + a[0] + a[1] #=> "cab" @@ -656,44 +1564,44 @@ rb_ary_subseq(ary, beg, len) * a[-3, 3] #=> [ "c", "d", "e" ] * # special cases * a[5] #=> nil + * a[6, 1] #=> nil * a[5, 1] #=> [] * a[5..10] #=> [] * */ VALUE -rb_ary_aref(argc, argv, ary) - int argc; - VALUE *argv; - VALUE ary; +rb_ary_aref(int argc, const VALUE *argv, VALUE ary) { - VALUE arg; - long beg, len; - + rb_check_arity(argc, 1, 2); 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; - } - return rb_ary_subseq(ary, beg, len); + return rb_ary_aref2(ary, argv[0], argv[1]); } - if (argc != 1) { - rb_scan_args(argc, argv, "11", 0, 0); + return rb_ary_aref1(ary, argv[0]); +} + +VALUE +rb_ary_aref2(VALUE ary, VALUE b, VALUE e) +{ + long beg = NUM2LONG(b); + long len = NUM2LONG(e); + if (beg < 0) { + beg += RARRAY_LEN(ary); } - arg = argv[0]; + return rb_ary_subseq(ary, beg, len); +} + +MJIT_FUNC_EXPORTED VALUE +rb_ary_aref1(VALUE ary, VALUE arg) +{ + long beg, len; + /* special case - speeding up */ 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: @@ -704,129 +1612,106 @@ 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+ - * 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.) + * Returns the element at +index+. A negative index counts from the end of + * +self+. Returns +nil+ if the index is out of range. See also + * Array#[]. * * a = [ "a", "b", "c", "d", "e" ] * a.at(0) #=> "a" * a.at(-1) #=> "e" */ -static VALUE -rb_ary_at(ary, pos) - VALUE ary, pos; +VALUE +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. + * If the array is empty, the first form returns +nil+, and the + * second form returns an empty array. See also Array#last for + * the opposite effect. * * 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_AREF(ary, 0); } else { - VALUE nv, result; - long n, i; - - rb_scan_args(argc, argv, "01", &nv); - n = NUM2LONG(nv); - if (n > RARRAY(ary)->len) n = RARRAY(ary)->len; - result = rb_ary_new2(n); - for (i=0; i<n; i++) { - rb_ary_push(result, RARRAY(ary)->ptr[i]); - } - return result; + 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, - * the first form returns <code>nil</code>. - * - * [ "w", "x", "y", "z" ].last #=> "z" + * 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 +nil+. + * + * See also Array#first for the opposite effect. + * + * 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, const VALUE *argv, VALUE ary) { if (argc == 0) { - if (RARRAY(ary)->len == 0) return Qnil; - return RARRAY(ary)->ptr[RARRAY(ary)->len-1]; + long len = RARRAY_LEN(ary); + if (len == 0) return Qnil; + return RARRAY_AREF(ary, len-1); } else { - VALUE nv, result; - long n, i; - - rb_scan_args(argc, argv, "01", &nv); - n = NUM2LONG(nv); - if (n > RARRAY(ary)->len) n = RARRAY(ary)->len; - result = rb_ary_new2(n); - for (i=RARRAY(ary)->len-n; n--; i++) { - rb_ary_push(result, RARRAY(ary)->ptr[i]); - } - return result; + 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 - * - * 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. - * + * ary.fetch(index) -> obj + * ary.fetch(index, default) -> obj + * ary.fetch(index) {|index| block} -> obj + * + * Tries to return the element at position +index+, but throws an IndexError + * exception if the referenced +index+ lies outside of the array bounds. This + * error can be prevented by supplying a second argument, which will act as a + * +default+ value. + * + * Alternatively, if a block is given it will only be executed when an + * invalid +index+ is referenced. + * + * Negative values of +index+ count from the end of the array. + * * a = [ 11, 22, 33, 44 ] * a.fetch(1) #=> 22 * a.fetch(-1) #=> 44 * a.fetch(4, 'cat') #=> "cat" - * a.fetch(4) { |i| i*i } #=> 16 + * a.fetch(100) {|i| puts "#{i} is out of bounds"} + * #=> "100 is out of bounds" */ 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; @@ -840,201 +1725,293 @@ 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_AREF(ary, idx); } /* * call-seq: - * array.index(obj) -> int or nil - * - * Returns the index of the first object in <i>self</i> such that is - * <code>==</code> to <i>obj</i>. Returns <code>nil</code> if - * no match is found. - * + * ary.find_index(obj) -> int or nil + * ary.find_index {|item| block} -> int or nil + * ary.find_index -> Enumerator + * ary.index(obj) -> int or nil + * ary.index {|item| block} -> int or nil + * ary.index -> Enumerator + * + * Returns the _index_ of the first object in +ary+ such that the object is + * <code>==</code> to +obj+. + * + * If a block is given instead of an argument, returns the _index_ of the + * first object for which the block returns +true+. Returns +nil+ if no + * match is found. + * + * See also Array#rindex. + * + * An Enumerator is returned if neither a block nor argument is given. + * * a = [ "a", "b", "c" ] - * a.index("b") #=> 1 - * a.index("z") #=> nil + * a.index("b") #=> 1 + * a.index("z") #=> nil + * a.index {|x| x == "b"} #=> 1 */ static VALUE -rb_ary_index(ary, val) - VALUE ary; - VALUE val; +rb_ary_index(int argc, VALUE *argv, VALUE ary) { + VALUE val; long i; - for (i=0; i<RARRAY(ary)->len; i++) { - if (rb_equal(RARRAY(ary)->ptr[i], val)) + if (argc == 0) { + RETURN_ENUMERATOR(ary, 0, 0); + for (i=0; i<RARRAY_LEN(ary); i++) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) { + return LONG2NUM(i); + } + } + return Qnil; + } + rb_check_arity(argc, 0, 1); + val = argv[0]; + if (rb_block_given_p()) + rb_warn("given block not used"); + for (i=0; i<RARRAY_LEN(ary); i++) { + VALUE e = RARRAY_AREF(ary, i); + if (rb_equal(e, val)) { return LONG2NUM(i); + } } return Qnil; } /* * call-seq: - * array.rindex(obj) -> int or nil - * - * Returns the index of the last object in <i>array</i> - * <code>==</code> to <i>obj</i>. Returns <code>nil</code> if - * no match is found. - * + * ary.rindex(obj) -> int or nil + * ary.rindex {|item| block} -> int or nil + * ary.rindex -> Enumerator + * + * Returns the _index_ of the last object in +self+ <code>==</code> to +obj+. + * + * If a block is given instead of an argument, returns the _index_ of the + * first object for which the block returns +true+, starting from the last + * object. + * + * Returns +nil+ if no match is found. + * + * See also Array#index. + * + * 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("b") #=> 3 + * a.rindex("z") #=> nil + * a.rindex {|x| x == "b"} #=> 3 */ static VALUE -rb_ary_rindex(ary, val) - VALUE ary; - VALUE val; +rb_ary_rindex(int argc, VALUE *argv, VALUE ary) { - long i = RARRAY(ary)->len; + VALUE val; + long i = RARRAY_LEN(ary), len; - while (i--) { - if (i > RARRAY(ary)->len) { - i = RARRAY(ary)->len; - continue; + if (argc == 0) { + RETURN_ENUMERATOR(ary, 0, 0); + while (i--) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) + return LONG2NUM(i); + if (i > (len = RARRAY_LEN(ary))) { + i = len; + } } - if (rb_equal(RARRAY(ary)->ptr[i], val)) + return Qnil; + } + rb_check_arity(argc, 0, 1); + val = argv[0]; + if (rb_block_given_p()) + rb_warn("given block not used"); + while (i--) { + VALUE e = RARRAY_AREF(ary, i); + if (rb_equal(e, val)) { return LONG2NUM(i); + } + if (i > RARRAY_LEN(ary)) { + break; + } } 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, const VALUE *rptr, long rlen) { - long rlen; + long olen; + long rofs; if (len < 0) rb_raise(rb_eIndexError, "negative length (%ld)", len); + olen = RARRAY_LEN(ary); if (beg < 0) { - beg += RARRAY(ary)->len; + beg += olen; 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 - olen, -olen); } } - if (RARRAY(ary)->len < len || RARRAY(ary)->len < beg + len) { - len = RARRAY(ary)->len - beg; + if (olen < len || olen < beg + len) { + len = olen - beg; } - if (NIL_P(rpl)) { - rlen = 0; - } - else { - rpl = rb_ary_to_ary(rpl); - rlen = RARRAY(rpl)->len; + { + const VALUE *optr = RARRAY_CONST_PTR_TRANSIENT(ary); + rofs = (rptr >= optr && rptr < optr + olen) ? rptr - optr : -1; } - rb_ary_modify(ary); - if (beg >= RARRAY(ary)->len) { + if (beg >= olen) { + VALUE target_ary; if (beg > ARY_MAX_SIZE - rlen) { rb_raise(rb_eIndexError, "index %ld too big", beg); } + target_ary = ary_ensure_room_for_push(ary, rlen-len); /* len is 0 or negative */ len = beg + rlen; - if (len >= RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, len); - RARRAY(ary)->aux.capa = len; - } - rb_mem_clear(RARRAY(ary)->ptr + RARRAY(ary)->len, beg - RARRAY(ary)->len); + ary_mem_clear(ary, olen, beg - olen); if (rlen > 0) { - MEMCPY(RARRAY(ary)->ptr + beg, RARRAY(rpl)->ptr, VALUE, rlen); + if (rofs != -1) rptr = RARRAY_CONST_PTR_TRANSIENT(ary) + rofs; + ary_memcpy0(ary, beg, rlen, rptr, target_ary); } - RARRAY(ary)->len = len; + ARY_SET_LEN(ary, len); } else { long alen; - if (beg + len > RARRAY(ary)->len) { - len = RARRAY(ary)->len - beg; + if (olen - len > ARY_MAX_SIZE - rlen) { + rb_raise(rb_eIndexError, "index %ld too big", olen + rlen - len); } - - alen = RARRAY(ary)->len + rlen - len; - if (alen >= RARRAY(ary)->aux.capa) { - REALLOC_N(RARRAY(ary)->ptr, VALUE, alen); - RARRAY(ary)->aux.capa = alen; + rb_ary_modify(ary); + alen = olen + 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; + RARRAY_PTR_USE_TRANSIENT(ary, ptr, + MEMMOVE(ptr + beg + rlen, ptr + beg + len, + VALUE, olen - (beg + len))); + ARY_SET_LEN(ary, alen); } if (rlen > 0) { - MEMMOVE(RARRAY(ary)->ptr + beg, RARRAY(rpl)->ptr, VALUE, rlen); + if (rofs != -1) rptr = RARRAY_CONST_PTR_TRANSIENT(ary) + rofs; + /* give up wb-protected ary */ + RB_OBJ_WB_UNPROTECT_FOR(ARRAY, ary); + + /* do not use RARRAY_PTR() because it can causes GC. + * ary can contain T_NONE object because it is not cleared. + */ + RARRAY_PTR_USE_TRANSIENT(ary, ptr, + MEMMOVE(ptr + beg, rptr, VALUE, rlen)); + } + } +} + +void +rb_ary_set_len(VALUE ary, long len) +{ + long capa; + + rb_ary_modify_check(ary); + if (ARY_SHARED_P(ary)) { + rb_raise(rb_eRuntimeError, "can't set length of shared "); + } + if (len > (capa = (long)ARY_CAPA(ary))) { + rb_bug("probable buffer overflow: %ld for %ld", len, capa); + } + ARY_SET_LEN(ary, len); +} + +/*! + * 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); + } + ary_mem_clear(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((VALUE *)ARY_EMBED_PTR(ary), tmp, VALUE, len); /* WB: no new reference */ + ARY_SET_EMBED_LEN(ary, len); + } + else { + if (olen > len + ARY_DEFAULT_SIZE) { + ary_heap_realloc(ary, len); + ARY_SET_CAPA(ary, len); } + ARY_SET_HEAP_LEN(ary, len); } + ary_verify(ary); + 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 + * from the +start+ index for +length+ elements, or replaces a subarray + * specified by the +range+ of indices. + * + * If indices are greater than the current capacity of the array, the array + * grows automatically. Elements are inserted into the array at +start+ if + * +length+ is zero. + * + * Negative indices will count backward from the end of the array. For + * +start+ and +range+ cases the starting index is just before an element. + * + * An IndexError is raised if a negative index points past the beginning of + * the array. + * + * See also Array#push, and Array#unshift. * - * Element Assignment---Sets the element at _index_, - * or replaces a subarray starting at _start_ and - * continuing for _length_ elements, or replaces a subarray - * specified by _range_. If indices are greater than - * 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 - * <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"] @@ -1042,41 +2019,37 @@ 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"] + * a[0, 0] = [ 1, 2 ] #=> [1, 2, "A"] + * a[3, 0] = "B" #=> [1, 2, "A", "B"] */ 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; + VALUE rpl; 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]); - return argv[2]; - } - if (argc != 2) { - rb_raise(rb_eArgError, "wrong number of arguments (%d for 2)", argc); + rb_ary_modify_check(ary); + beg = NUM2LONG(argv[0]); + len = NUM2LONG(argv[1]); + goto range; } + rb_check_arity(argc, 2, 2); + 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]; + range: + rpl = rb_ary_to_ary(argv[argc-1]); + rb_ary_splice(ary, beg, len, RARRAY_CONST_PTR_TRANSIENT(rpl), RARRAY_LEN(rpl)); + RB_GC_GUARD(rpl); + return argv[argc-1]; } offset = NUM2LONG(argv[0]); @@ -1087,88 +2060,108 @@ fixnum: /* * call-seq: - * array.insert(index, obj...) -> array - * - * Inserts the given values before the element with the given index - * (which may be negative). - * + * ary.insert(index, obj...) -> ary + * + * Inserts the given values before the element with the given +index+. + * + * Negative indices count backwards from the end of the array, where +-1+ is + * the last element. If a negative index is used, the given values will be + * inserted after that element, so using an index of +-1+ will insert the + * values at the end of the array. + * * 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_check_arity(argc, 1, UNLIMITED_ARGUMENTS); + rb_ary_modify_check(ary); pos = NUM2LONG(argv[0]); + if (argc == 1) return ary; if (pos == -1) { - pos = RARRAY(ary)->len; + pos = RARRAY_LEN(ary); } - if (pos < 0) { + else if (pos < 0) { + long minpos = -RARRAY_LEN(ary) - 1; + if (pos < minpos) { + rb_raise(rb_eIndexError, "index %ld too small for array; minimum: %ld", + pos, minpos); + } pos++; } - rb_ary_splice(ary, pos, 0, rb_ary_new4(argc - 1, argv + 1)); + rb_ary_splice(ary, pos, 0, argv + 1, argc - 1); return ary; } +static VALUE +rb_ary_length(VALUE ary); + +static VALUE +ary_enum_length(VALUE ary, VALUE args, VALUE eobj) +{ + return rb_ary_length(ary); +} + /* * call-seq: - * array.each {|item| block } -> array - * - * Calls <i>block</i> once for each element in <i>self</i>, passing that - * element as a parameter. - * + * ary.each {|item| block} -> ary + * ary.each -> Enumerator + * + * Calls the given block once for each element in +self+, passing that element + * as a parameter. Returns the array itself. + * + * If no block is given, an Enumerator is returned. + * * a = [ "a", "b", "c" ] * a.each {|x| print x, " -- " } - * + * * produces: - * + * * a -- b -- c -- */ VALUE -rb_ary_each(ary) - VALUE ary; +rb_ary_each(VALUE ary) { long i; - - for (i=0; i<RARRAY(ary)->len; i++) { - rb_yield(RARRAY(ary)->ptr[i]); + ary_verify(ary); + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + for (i=0; i<RARRAY_LEN(ary); i++) { + rb_yield(RARRAY_AREF(ary, i)); } return ary; } /* * call-seq: - * array.each_index {|index| block } -> array - * - * Same as <code>Array#each</code>, but passes the index of the element - * instead of the element itself. - * + * ary.each_index {|index| block} -> ary + * ary.each_index -> Enumerator + * + * Same as Array#each, but passes the +index+ of the element instead of the + * element itself. + * + * An Enumerator is returned if no block is given. + * * 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_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); - for (i=0; i<RARRAY(ary)->len; i++) { + for (i=0; i<RARRAY_LEN(ary); i++) { rb_yield(LONG2NUM(i)); } return ary; @@ -1176,29 +2169,32 @@ 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 - * order. - * + * ary.reverse_each {|item| block} -> ary + * ary.reverse_each -> Enumerator + * + * Same as Array#each, 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 = RARRAY(ary)->len; + long len; + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + len = RARRAY_LEN(ary); while (len--) { - rb_yield(RARRAY(ary)->ptr[len]); - if (RARRAY(ary)->len < len) { - len = RARRAY(ary)->len; + long nlen; + rb_yield(RARRAY_AREF(ary, len)); + nlen = RARRAY_LEN(ary); + if (nlen < len) { + len = nlen; } } return ary; @@ -1206,555 +2202,908 @@ 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 + * [].length #=> 0 */ 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 +true+ 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); + long len = RARRAY_LEN(ary); + VALUE dup = rb_ary_new2(len); + ary_memcpy(dup, 0, len, RARRAY_CONST_PTR_TRANSIENT(ary)); + ARY_SET_LEN(dup, len); - DUPSETUP(dup, ary); - MEMCPY(RARRAY(dup)->ptr, RARRAY(ary)->ptr, VALUE, RARRAY(ary)->len); - RARRAY(dup)->len = RARRAY(ary)->len; + ary_verify(ary); + ary_verify(dup); return dup; } +VALUE +rb_ary_resurrect(VALUE ary) +{ + return ary_make_partial(ary, rb_cArray, 0, RARRAY_LEN(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) +{ + 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; +} + +static void +ary_join_0(VALUE ary, VALUE sep, long max, VALUE result) { - return rb_ary_join(arg[0], arg[1]); + long i; + VALUE val; + + if (max > 0) rb_enc_copy(result, RARRAY_AREF(ary, 0)); + for (i=0; i<max; i++) { + val = RARRAY_AREF(ary, i); + if (i > 0 && !NIL_P(sep)) + rb_str_buf_append(result, sep); + rb_str_buf_append(result, val); + } } -VALUE -rb_ary_join(ary, sep) - VALUE ary, sep; +static void +ary_join_1(VALUE obj, VALUE ary, VALUE sep, long i, VALUE result, int *first) { - long len = 1, i; - int taint = Qfalse; - VALUE result, tmp; + VALUE val, tmp; - if (RARRAY(ary)->len == 0) return rb_str_new(0, 0); - if (OBJ_TAINTED(ary) || OBJ_TAINTED(sep)) taint = Qtrue; + for (; i<RARRAY_LEN(ary); i++) { + if (i > 0 && !NIL_P(sep)) + rb_str_buf_append(result, sep); - 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); - } - result = rb_str_buf_new(len); - for (i=0; i<RARRAY(ary)->len; i++) { - tmp = RARRAY(ary)->ptr[i]; - switch (TYPE(tmp)) { - case T_STRING: - break; - case T_ARRAY: - if (rb_inspecting_p(tmp)) { - tmp = rb_str_new2("[...]"); + val = RARRAY_AREF(ary, i); + if (RB_TYPE_P(val, T_STRING)) { + str_join: + rb_str_buf_append(result, val); + if (*first) { + rb_enc_copy(result, val); + *first = FALSE; + } + } + else if (RB_TYPE_P(val, T_ARRAY)) { + obj = val; + ary_join: + if (val == ary) { + rb_raise(rb_eArgError, "recursive array join"); } else { - VALUE args[2]; + VALUE args[4]; - args[0] = tmp; + *first = FALSE; + 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); } - if (i > 0 && !NIL_P(sep)) - rb_str_buf_append(result, sep); - rb_str_buf_append(result, tmp); - if (OBJ_TAINTED(tmp)) taint = Qtrue; + else { + tmp = rb_check_string_type(val); + if (!NIL_P(tmp)) { + val = tmp; + goto str_join; + } + tmp = rb_check_array_type(val); + if (!NIL_P(tmp)) { + obj = val; + val = tmp; + goto ary_join; + } + val = rb_obj_as_string(val); + goto str_join; + } + } +} + +VALUE +rb_ary_join(VALUE ary, VALUE sep) +{ + long len = 1, i; + VALUE val, tmp, result; + + if (RARRAY_LEN(ary) == 0) return rb_usascii_str_new(0, 0); + + if (!NIL_P(sep)) { + StringValue(sep); + len += RSTRING_LEN(sep) * (RARRAY_LEN(ary) - 1); + } + for (i=0; i<RARRAY_LEN(ary); i++) { + val = RARRAY_AREF(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()); + 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); } - if (taint) OBJ_TAINT(result); + result = rb_str_new(0, len); + rb_str_set_len(result, 0); + + ary_join_0(ary, sep, RARRAY_LEN(ary), result); + return result; } /* * call-seq: - * array.join(sep=$,) -> str - * + * ary.join(separator=$,) -> str + * * Returns a string created by converting each element of the array to - * a string, separated by <i>sep</i>. - * + * a string, separated by the given +separator+. + * If the +separator+ is +nil+, it uses current <code>$,</code>. + * If both the +separator+ and <code>$,</code> are +nil+, + * it uses an empty string. + * * [ "a", "b", "c" ].join #=> "abc" * [ "a", "b", "c" ].join("-") #=> "a-b-c" + * + * For nested arrays, join is applied recursively: + * + * [ "a", [1, 2, [:x, :y]], "b" ].join("-") #=> "a-1-2-x-y-b" */ 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; - + if (rb_check_arity(argc, 0, 1) == 0 || NIL_P(sep = argv[0])) { + sep = rb_output_fs; + if (!NIL_P(sep)) { + rb_warn("$, is set to non-nil value"); + } + } + return rb_ary_join(ary, sep); } +static VALUE +inspect_ary(VALUE ary, VALUE dummy, int recur) +{ + long i; + VALUE s, str; + + if (recur) return rb_usascii_str_new_cstr("[...]"); + str = rb_str_buf_new2("["); + for (i=0; i<RARRAY_LEN(ary); i++) { + s = rb_inspect(RARRAY_AREF(ary, i)); + 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, "]"); + return str; +} + /* * call-seq: - * array.to_s -> string - * - * Returns _self_<code>.join</code>. - * - * [ "a", "e", "i", "o" ].to_s #=> "aeio" + * ary.inspect -> string + * ary.to_s -> string * + * Creates a string representation of +self+, by calling #inspect + * on each element. + * + * [ "a", "b", "c" ].to_s #=> "[\"a\", \"b\", \"c\"]" */ -VALUE -rb_ary_to_s(ary) - VALUE ary; +static VALUE +rb_ary_inspect(VALUE ary) { - if (RARRAY(ary)->len == 0) return rb_str_new(0, 0); - - return rb_ary_join(ary, rb_output_fs); + if (RARRAY_LEN(ary) == 0) return rb_usascii_str_new2("[]"); + return rb_exec_recursive(inspect_ary, ary, 0); } -static ID inspect_key; +VALUE +rb_ary_to_s(VALUE ary) +{ + return rb_ary_inspect(ary); +} -struct inspect_arg { - VALUE (*func)(); - VALUE arg1, arg2; -}; +/* + * call-seq: + * ary.to_a -> ary + * + * Returns +self+. + * + * If called on a subclass of Array, converts the receiver to an Array object. + */ static VALUE -inspect_call(arg) - struct inspect_arg *arg; +rb_ary_to_a(VALUE ary) { - return (*arg->func)(arg->arg1, arg->arg2); + if (rb_obj_class(ary) != rb_cArray) { + VALUE dup = rb_ary_new2(RARRAY_LEN(ary)); + rb_ary_replace(dup, ary); + return dup; + } + return ary; } +/* + * call-seq: + * ary.to_h -> hash + * ary.to_h {|item| block } -> hash + * + * Returns the result of interpreting <i>ary</i> as an array of + * <tt>[key, value]</tt> pairs. + * + * [[:foo, :bar], [1, 2]].to_h + * # => {:foo => :bar, 1 => 2} + * + * If a block is given, the results of the block on each element of + * the array will be used as pairs. + * + * ["foo", "bar"].to_h {|s| [s.ord, s]} + * # => {102=>"foo", 98=>"bar"} + */ + static VALUE -get_inspect_tbl(create) - int create; +rb_ary_to_h(VALUE ary) { - 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); + long i; + VALUE hash = rb_hash_new_with_size(RARRAY_LEN(ary)); + int block_given = rb_block_given_p(); + + for (i=0; i<RARRAY_LEN(ary); i++) { + const VALUE e = rb_ary_elt(ary, i); + const VALUE elt = block_given ? rb_yield_force_blockarg(e) : e; + const VALUE key_value_pair = rb_check_array_type(elt); + if (NIL_P(key_value_pair)) { + rb_raise(rb_eTypeError, "wrong element type %"PRIsVALUE" at %ld (expected array)", + rb_obj_class(elt), i); } + if (RARRAY_LEN(key_value_pair) != 2) { + rb_raise(rb_eArgError, "wrong array length at %ld (expected 2, was %ld)", + i, RARRAY_LEN(key_value_pair)); + } + rb_hash_aset(hash, RARRAY_AREF(key_value_pair, 0), RARRAY_AREF(key_value_pair, 1)); } - 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; + return hash; } +/* + * call-seq: + * ary.to_ary -> ary + * + * Returns +self+. + */ + static VALUE -inspect_ensure(obj) - VALUE obj; +rb_ary_to_ary_m(VALUE ary) { - VALUE inspect_tbl; - - inspect_tbl = get_inspect_tbl(Qfalse); - if (!NIL_P(inspect_tbl)) { - rb_ary_pop(inspect_tbl); - } - return 0; + return ary; } -VALUE -rb_protect_inspect(func, obj, arg) - VALUE (*func)(ANYARGS); - VALUE obj, arg; +static void +ary_reverse(VALUE *p1, VALUE *p2) { - 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); + while (p1 < p2) { + VALUE tmp = *p1; + *p1++ = *p2; + *p2-- = tmp; } - 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; +rb_ary_reverse(VALUE ary) { - 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)); -} + VALUE *p2; + long len = RARRAY_LEN(ary); -static VALUE -inspect_ary(ary) - VALUE ary; -{ - int tainted = OBJ_TAINTED(ary); - long i; - VALUE s, str; - - 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; - if (i > 0) rb_str_buf_cat2(str, ", "); - rb_str_buf_append(str, s); + rb_ary_modify(ary); + if (len > 1) { + RARRAY_PTR_USE_TRANSIENT(ary, p1, { + p2 = p1 + len - 1; /* points last item */ + ary_reverse(p1, p2); + }); /* WB: no new reference */ } - rb_str_buf_cat2(str, "]"); - if (tainted) OBJ_TAINT(str); - return str; + return ary; } /* * call-seq: - * array.inspect -> string + * ary.reverse! -> ary * - * Create a printable version of <i>array</i>. + * Reverses +self+ in place. + * + * a = [ "a", "b", "c" ] + * a.reverse! #=> ["c", "b", "a"] + * a #=> ["c", "b", "a"] */ static VALUE -rb_ary_inspect(ary) - VALUE ary; +rb_ary_reverse_bang(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); + return rb_ary_reverse(ary); } /* * call-seq: - * array.to_a -> array - * - * Returns _self_. If called on a subclass of Array, converts - * the receiver to an Array object. + * 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_to_a(ary) - VALUE ary; +rb_ary_reverse_m(VALUE ary) { - if (rb_obj_class(ary) != rb_cArray) { - VALUE dup = rb_ary_new2(RARRAY(ary)->len); - rb_ary_replace(dup, ary); - return dup; + long len = RARRAY_LEN(ary); + VALUE dup = rb_ary_new2(len); + + if (len > 0) { + const VALUE *p1 = RARRAY_CONST_PTR_TRANSIENT(ary); + VALUE *p2 = (VALUE *)RARRAY_CONST_PTR_TRANSIENT(dup) + len - 1; + do *p2-- = *p1++; while (--len > 0); } - return ary; + ARY_SET_LEN(dup, RARRAY_LEN(ary)); + return dup; } -/* - * call-seq: - * array.to_ary -> array - * - * Returns _self_. - */ +static inline long +rotate_count(long cnt, long len) +{ + return (cnt < 0) ? (len - (~cnt % len) - 1) : (cnt % len); +} -static VALUE -rb_ary_to_ary_m(ary) - VALUE ary; +static void +ary_rotate_ptr(VALUE *ptr, long len, long cnt) { - return ary; + --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); } VALUE -rb_ary_reverse(ary) - VALUE ary; +rb_ary_rotate(VALUE ary, long cnt) { - 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 (cnt != 0) { + long len = RARRAY_LEN(ary); + if (len > 0 && (cnt = rotate_count(cnt, len)) > 0) { + RARRAY_PTR_USE_TRANSIENT(ary, ptr, ary_rotate_ptr(ptr, len, cnt)); + return ary; + } } - return ary; + return Qnil; } /* * call-seq: - * array.reverse! -> array - * - * Reverses _self_ in place. - * - * a = [ "a", "b", "c" ] - * a.reverse! #=> ["c", "b", "a"] - * a #=> ["c", "b", "a"] + * ary.rotate!(count=1) -> ary + * + * Rotates +self+ in place so that the element at +count+ comes first, and + * returns +self+. + * + * If +count+ is negative then it rotates in the opposite direction, starting + * from the end of the array where +-1+ is the last element. + * + * 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_reverse_bang(ary) - VALUE ary; +rb_ary_rotate_bang(int argc, VALUE *argv, VALUE ary) { - return rb_ary_reverse(ary); + long n = (rb_check_arity(argc, 0, 1) ? NUM2LONG(argv[0]) : 1); + rb_ary_rotate(ary, n); + return ary; } /* * call-seq: - * array.reverse -> an_array - * - * Returns a new array containing <i>self</i>'s elements in reverse order. - * - * [ "a", "b", "c" ].reverse #=> ["c", "b", "a"] - * [ 1 ].reverse #=> [1] + * ary.rotate(count=1) -> new_ary + * + * Returns a new array by rotating +self+ so that the element at +count+ is + * the first element of the new array. + * + * If +count+ is negative then it rotates in the opposite direction, starting + * from the end of +self+ where +-1+ is the last element. + * + * 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_reverse_m(ary) - VALUE ary; +rb_ary_rotate_m(int argc, VALUE *argv, VALUE ary) { - return rb_ary_reverse(rb_ary_dup(ary)); + VALUE rotated; + const VALUE *ptr; + long len; + long cnt = (rb_check_arity(argc, 0, 1) ? NUM2LONG(argv[0]) : 1); + + len = RARRAY_LEN(ary); + rotated = rb_ary_new2(len); + if (len > 0) { + cnt = rotate_count(cnt, len); + ptr = RARRAY_CONST_PTR_TRANSIENT(ary); + len -= cnt; + ary_memcpy(rotated, 0, len, ptr + cnt); + ary_memcpy(rotated, len, cnt, ptr); + } + ARY_SET_LEN(rotated, RARRAY_LEN(ary)); + return rotated; } struct ary_sort_data { VALUE ary; - VALUE *ptr; - long len; + struct cmp_opt_data cmp_opt; }; -static void -ary_sort_check(data) - struct ary_sort_data *data; +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; + VALUE args[2]; int n; - n = rb_cmpint(retval, *a, *b); - ary_sort_check(data); + args[0] = a; + args[1] = b; + retval = rb_yield_values2(2, args); + 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) && CMP_OPTIMIZABLE(data->cmp_opt, 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) && CMP_OPTIMIZABLE(data->cmp_opt, String)) { + return rb_str_cmp(a, b); + } + if (RB_FLOAT_TYPE_P(a) && CMP_OPTIMIZABLE(data->cmp_opt, Float)) { + return rb_float_cmp(a, b); } - retval = rb_funcall(a, id_cmp, 1, b); + retval = rb_funcallv(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; +/* + * call-seq: + * ary.sort! -> ary + * ary.sort! {|a, b| block} -> ary + * + * Sorts +self+ in place. + * + * Comparisons for the sort will be done using the <code><=></code> operator + * or using an optional code block. + * + * The block must implement a comparison between +a+ and +b+ and return + * an integer less than 0 when +b+ follows +a+, +0+ when +a+ and +b+ + * are equivalent, or an integer greater than 0 when +a+ follows +b+. + * + * The result is not guaranteed to be stable. When the comparison of two + * elements returns +0+, the order of the elements is unpredictable. + * + * ary = [ "d", "a", "e", "c", "b" ] + * ary.sort! #=> ["a", "b", "c", "d", "e"] + * ary.sort! {|a, b| b <=> a} #=> ["e", "d", "c", "b", "a"] + * + * See also Enumerable#sort_by. + */ + +VALUE +rb_ary_sort_bang(VALUE ary) { - struct ary_sort_data data; + rb_ary_modify(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; + long len = RARRAY_LEN(ary); + RBASIC_CLEAR_CLASS(tmp); + data.ary = tmp; + data.cmp_opt.opt_methods = 0; + data.cmp_opt.opt_inited = 0; + RARRAY_PTR_USE(tmp, ptr, { + ruby_qsort(ptr, len, sizeof(VALUE), + rb_block_given_p()?sort_1:sort_2, &data); + }); /* WB: no new reference */ + rb_ary_modify(ary); + if (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); + } + ary_memcpy(ary, 0, ARY_EMBED_LEN(tmp), ARY_EMBED_PTR(tmp)); + ARY_SET_LEN(ary, ARY_EMBED_LEN(tmp)); + } + else { + if (!ARY_EMBED_P(ary) && ARY_HEAP_PTR(ary) == ARY_HEAP_PTR(tmp)) { + FL_UNSET_SHARED(ary); + ARY_SET_CAPA(ary, RARRAY_LEN(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 { + ary_heap_free(ary); + } + ARY_SET_PTR(ary, ARY_HEAP_PTR(tmp)); + ARY_SET_HEAP_LEN(ary, len); + ARY_SET_CAPA(ary, ARY_HEAP_LEN(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_SET_CLASS_RAW(tmp, rb_cArray); /* rb_cArray must be marked */ + } + ary_verify(ary); + return ary; +} + +/* + * call-seq: + * 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 must implement a comparison between +a+ and +b+ and return + * an integer less than 0 when +b+ follows +a+, +0+ when +a+ and +b+ + * are equivalent, or an integer greater than 0 when +a+ follows +b+. + * + * The result is not guaranteed to be stable. When the comparison of two + * elements returns +0+, the order of the elements is unpredictable. + * + * ary = [ "d", "a", "e", "c", "b" ] + * ary.sort #=> ["a", "b", "c", "d", "e"] + * ary.sort {|a, b| b <=> a} #=> ["e", "d", "c", "b", "a"] + * + * To produce the reverse order, the following can also be used + * (and may be faster): + * + * ary.sort.reverse! #=> ["e", "d", "c", "b", "a"] + * + * See also Enumerable#sort_by. + */ - 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); +VALUE +rb_ary_sort(VALUE ary) +{ + ary = rb_ary_dup(ary); + rb_ary_sort_bang(ary); return ary; } +static VALUE rb_ary_bsearch_index(VALUE ary); + +/* + * call-seq: + * ary.bsearch {|x| block } -> elem + * + * By using binary search, finds a value from this array which meets + * the given condition in O(log n) where n is the size of the array. + * + * You can use this method in two modes: a find-minimum mode and + * a find-any mode. In either case, the elements of the array must be + * monotone (or sorted) with respect to the block. + * + * In find-minimum mode (this is a good choice for typical use cases), + * the block must always return true or false, and there must be an index i + * (0 <= i <= ary.size) so that: + * + * - the block returns false for any element whose index is less than + * i, and + * - the block returns true for any element whose index is greater + * than or equal to i. + * + * This method returns the i-th element. If i is equal to ary.size, + * it returns nil. + * + * ary = [0, 4, 7, 10, 12] + * ary.bsearch {|x| x >= 4 } #=> 4 + * ary.bsearch {|x| x >= 6 } #=> 7 + * ary.bsearch {|x| x >= -1 } #=> 0 + * ary.bsearch {|x| x >= 100 } #=> nil + * + * In find-any mode (this behaves like libc's bsearch(3)), the block + * must always return a number, and there must be two indices i and j + * (0 <= i <= j <= ary.size) so that: + * + * - the block returns a positive number for ary[k] if 0 <= k < i, + * - the block returns zero for ary[k] if i <= k < j, and + * - the block returns a negative number for ary[k] if + * j <= k < ary.size. + * + * Under this condition, this method returns any element whose index + * is within i...j. If i is equal to j (i.e., there is no element + * that satisfies the block), this method returns nil. + * + * ary = [0, 4, 7, 10, 12] + * # try to find v such that 4 <= v < 8 + * ary.bsearch {|x| 1 - x / 4 } #=> 4 or 7 + * # try to find v such that 8 <= v < 10 + * ary.bsearch {|x| 4 - x / 2 } #=> nil + * + * You must not mix the two modes at a time; the block must always + * return either true/false, or always return a number. It is + * undefined which value is actually picked up at each iteration. + */ + static VALUE -sort_unlock(ary) - VALUE ary; +rb_ary_bsearch(VALUE ary) { - FL_UNSET(ary, ARY_TMPLOCK); - return ary; + VALUE index_result = rb_ary_bsearch_index(ary); + + if (FIXNUM_P(index_result)) { + return rb_ary_entry(ary, FIX2LONG(index_result)); + } + return index_result; } /* * call-seq: - * array.sort! -> array - * array.sort! {| a,b | block } -> array - * - * 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"] + * ary.bsearch_index {|x| block } -> int or nil + * + * By using binary search, finds an index of a value from this array which + * meets the given condition in O(log n) where n is the size of the array. + * + * It supports two modes, depending on the nature of the block. They are + * exactly the same as in the case of the #bsearch method, with the only difference + * being that this method returns the index of the element instead of the + * element itself. For more details consult the documentation for #bsearch. */ -VALUE -rb_ary_sort_bang(ary) - VALUE ary; +static VALUE +rb_ary_bsearch_index(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); + long low = 0, high = RARRAY_LEN(ary), mid; + int smaller = 0, satisfied = 0; + VALUE v, val; + + RETURN_ENUMERATOR(ary, 0, 0); + while (low < high) { + mid = low + ((high - low) / 2); + val = rb_ary_entry(ary, mid); + v = rb_yield(val); + if (FIXNUM_P(v)) { + if (v == INT2FIX(0)) return INT2FIX(mid); + smaller = (SIGNED_VALUE)v < 0; /* Fixnum preserves its sign-bit */ + } + else if (v == Qtrue) { + satisfied = 1; + smaller = 1; + } + else if (v == Qfalse || v == Qnil) { + smaller = 0; + } + else if (rb_obj_is_kind_of(v, rb_cNumeric)) { + const VALUE zero = INT2FIX(0); + switch (rb_cmpint(rb_funcallv(v, id_cmp, 1, &zero), v, zero)) { + case 0: return INT2FIX(mid); + case 1: smaller = 1; break; + case -1: smaller = 0; + } + } + else { + rb_raise(rb_eTypeError, "wrong argument type %"PRIsVALUE + " (must be numeric, true, false or nil)", + rb_obj_class(v)); + } + if (smaller) { + high = mid; + } + else { + low = mid + 1; + } } - return ary; + if (!satisfied) return Qnil; + return INT2FIX(low); +} + + +static VALUE +sort_by_i(RB_BLOCK_CALL_FUNC_ARGLIST(i, dummy)) +{ + return rb_yield(i); } /* * 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 - * 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"] + * ary.sort_by! {|obj| block} -> ary + * ary.sort_by! -> Enumerator + * + * Sorts +self+ in place using a set of keys generated by mapping the + * values in +self+ through the given block. + * + * The result is not guaranteed to be stable. When two keys are equal, + * the order of the corresponding elements is unpredictable. + * + * If no block is given, an Enumerator is returned instead. + * + * See also Enumerable#sort_by. */ -VALUE -rb_ary_sort(ary) - VALUE ary; +static VALUE +rb_ary_sort_by_bang(VALUE ary) { - ary = rb_ary_dup(ary); - rb_ary_sort_bang(ary); + VALUE sorted; + + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + 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 - * new array containing the values returned by the block. - * See also <code>Enumerable#collect</code>. - * + * ary.collect {|item| block} -> new_ary + * ary.map {|item| block} -> new_ary + * ary.collect -> Enumerator + * ary.map -> Enumerator + * + * Invokes the given block once for each element of +self+. + * + * Creates a new array containing the values returned by the block. + * + * See also Enumerable#collect. + * + * 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"] + * a.collect {|x| x + "!"} #=> ["a!", "b!", "c!", "d!"] + * a.map.with_index {|x, i| x * i} #=> ["", "b", "cc", "ddd"] + * 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_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + collect = rb_ary_new2(RARRAY_LEN(ary)); + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_ary_push(collect, rb_yield(RARRAY_AREF(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! -> Enumerator + * ary.map! -> Enumerator + * + * Invokes the given block once for each element of +self+, replacing the + * element with the value returned by the block. + * + * See also Enumerable#collect. + * + * If no block is given, an Enumerator is returned instead. * - * Invokes the block once for each element of _self_, replacing the - * element with the value returned by _block_. - * See also <code>Enumerable#collect</code>. - * * a = [ "a", "b", "c", "d" ] - * a.collect! {|x| x + "!" } - * a #=> [ "a!", "b!", "c!", "d!" ] + * a.map! {|x| x + "!" } + * a #=> [ "a!", "b!", "c!", "d!" ] + * a.collect!.with_index {|x, i| x[0...i] } + * a #=> ["", "b", "c!", "d!"] */ static VALUE -rb_ary_collect_bang(ary) - VALUE ary; +rb_ary_collect_bang(VALUE ary) { long i; + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); 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_AREF(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, const VALUE *argv, VALUE (*func) (VALUE, long)) { VALUE result = rb_ary_new2(argc); long beg, len, i, j; @@ -1765,15 +3114,13 @@ rb_values_at(obj, olen, argc, argv, func) continue; } /* check if idx is Range */ - switch (rb_range_beg_len(argv[i], &beg, &len, olen, 0)) { - case Qfalse: - break; - case Qnil: - continue; - default: - for (j=0; j<len; j++) { - rb_ary_push(result, (*func)(obj, j+beg)); + if (rb_range_beg_len(argv[i], &beg, &len, olen, 1)) { + long end = olen < beg+len ? olen : beg+len; + for (j = beg; j < end; j++) { + rb_ary_push(result, (*func)(obj, j)); } + if (beg + len > j) + rb_ary_resize(result, RARRAY_LEN(result) + (beg + len) - j); continue; } rb_ary_push(result, (*func)(obj, NUM2LONG(argv[i]))); @@ -1781,118 +3128,297 @@ rb_values_at(obj, olen, argc, argv, func) return result; } -/* +static VALUE +append_values_at_single(VALUE result, VALUE ary, long olen, VALUE idx) +{ + long beg, len; + if (FIXNUM_P(idx)) { + beg = FIX2LONG(idx); + } + /* check if idx is Range */ + else if (rb_range_beg_len(idx, &beg, &len, olen, 1)) { + if (len > 0) { + const VALUE *const src = RARRAY_CONST_PTR_TRANSIENT(ary); + const long end = beg + len; + const long prevlen = RARRAY_LEN(result); + if (beg < olen) { + rb_ary_cat(result, src + beg, end > olen ? olen-beg : len); + } + if (end > olen) { + rb_ary_store(result, prevlen + len - 1, Qnil); + } + } + return result; + } + else { + beg = NUM2LONG(idx); + } + return rb_ary_push(result, rb_ary_entry(ary, beg)); +} + +/* * 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. + * + * See also Array#select. * - * Returns an array containing the elements in - * _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) - * a.values_at(-1, -3, -5, -7) - * a.values_at(1..3, 2...5) + * a.values_at(1, 3, 5) # => ["b", "d", "f"] + * a.values_at(1, 3, 5, 7) # => ["b", "d", "f", nil] + * a.values_at(-1, -2, -2, -7) # => ["f", "e", "e", nil] + * a.values_at(4..6, 3...6) # => ["e", "f", nil, "d", "e", "f"] */ 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); + long i, olen = RARRAY_LEN(ary); + VALUE result = rb_ary_new_capa(argc); + for (i = 0; i < argc; ++i) { + append_values_at_single(result, ary, olen, argv[i]); + } + RB_GC_GUARD(ary); + return result; } + /* * call-seq: - * array.select {|item| block } -> an_array - * - * Invokes the block passing in successive elements from <i>array</i>, - * returning an array containing those elements for which the block - * returns a true value (equivalent to <code>Enumerable#select</code>). - * - * a = %w{ a b c d e f } - * a.select {|v| v =~ /[aeiou]/} #=> ["a", "e"] + * ary.select {|item| block} -> new_ary + * ary.select -> Enumerator + * ary.filter {|item| block} -> new_ary + * ary.filter -> Enumerator + * + * Returns a new array containing all elements of +ary+ + * for which the given +block+ returns a true value. + * + * If no block is given, an Enumerator is returned instead. + * + * [1,2,3,4,5].select {|num| num.even? } #=> [2, 4] + * + * a = %w[ a b c d e f ] + * a.select {|v| v =~ /[aeiou]/ } #=> ["a", "e"] + * + * See also Enumerable#select. + * + * Array#filter is an alias for Array#select. */ static VALUE -rb_ary_select(ary) - VALUE ary; +rb_ary_select(VALUE ary) { VALUE result; long i; - result = rb_ary_new2(RARRAY(ary)->len); - for (i = 0; i < RARRAY(ary)->len; i++) { - if (RTEST(rb_yield(RARRAY(ary)->ptr[i]))) { + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + result = rb_ary_new2(RARRAY_LEN(ary)); + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) { rb_ary_push(result, rb_ary_elt(ary, i)); } } return result; } +struct select_bang_arg { + VALUE ary; + long len[2]; +}; + +static VALUE +select_bang_i(VALUE a) +{ + volatile struct select_bang_arg *arg = (void *)a; + VALUE ary = arg->ary; + long i1, i2; + + for (i1 = i2 = 0; i1 < RARRAY_LEN(ary); arg->len[0] = ++i1) { + VALUE v = RARRAY_AREF(ary, i1); + if (!RTEST(rb_yield(v))) continue; + if (i1 != i2) { + rb_ary_store(ary, i2, v); + } + arg->len[1] = ++i2; + } + return (i1 == i2) ? Qnil : ary; +} + +static VALUE +select_bang_ensure(VALUE a) +{ + volatile struct select_bang_arg *arg = (void *)a; + VALUE ary = arg->ary; + long len = RARRAY_LEN(ary); + long i1 = arg->len[0], i2 = arg->len[1]; + + if (i2 < len && i2 < i1) { + long tail = 0; + if (i1 < len) { + tail = len - i1; + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr + i2, ptr + i1, VALUE, tail); + }); + } + ARY_SET_LEN(ary, i2 + tail); + } + return ary; +} + +/* + * call-seq: + * ary.select! {|item| block } -> ary or nil + * ary.select! -> Enumerator + * ary.filter! {|item| block } -> ary or nil + * ary.filter! -> Enumerator + * + * Invokes the given block passing in successive elements from +self+, + * deleting elements for which the block returns a +false+ value. + * + * The array may not be changed instantly every time the block is called. + * + * If changes were made, it will return +self+, otherwise it returns +nil+. + * + * If no block is given, an Enumerator is returned instead. + * + * See also Array#keep_if. + * + * Array#filter! is an alias for Array#select!. + */ + +static VALUE +rb_ary_select_bang(VALUE ary) +{ + struct select_bang_arg args; + + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + rb_ary_modify(ary); + + args.ary = ary; + args.len[0] = args.len[1] = 0; + return rb_ensure(select_bang_i, (VALUE)&args, select_bang_ensure, (VALUE)&args); +} + +/* + * call-seq: + * ary.keep_if {|item| block} -> ary + * ary.keep_if -> Enumerator + * + * Deletes every element of +self+ for which the given block evaluates to + * +false+, and returns +self+. + * + * 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"] + * a #=> ["a", "e"] + * + * See also Array#select!. + */ + +static VALUE +rb_ary_keep_if(VALUE ary) +{ + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + rb_ary_select_bang(ary); + return ary; +} + +static void +ary_resize_smaller(VALUE ary, long len) +{ + rb_ary_modify(ary); + if (RARRAY_LEN(ary) > len) { + ARY_SET_LEN(ary, len); + if (len * 2 < ARY_CAPA(ary) && + ARY_CAPA(ary) > ARY_DEFAULT_SIZE) { + ary_resize_capa(ary, len * 2); + } + } +} + /* * 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 - * 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. - * + * ary.delete(obj) -> item or nil + * ary.delete(obj) {block} -> item or result of block + * + * Deletes all items from +self+ that are equal to +obj+. + * + * Returns the last deleted item, or +nil+ if no matching item is found. + * + * If the optional code block is given, the result of the block is returned if + * the item is not found. (To remove +nil+ elements and get an informative + * return value, use Array#compact!) + * * a = [ "a", "b", "b", "b", "c" ] * a.delete("b") #=> "b" * a #=> ["a", "c"] * a.delete("z") #=> nil - * a.delete("z") { "not found" } #=> "not found" + * a.delete("z") {"not found"} #=> "not found" */ 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_AREF(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); } return Qnil; } - 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; + ary_resize_smaller(ary, i2); + + ary_verify(ary); + return v; +} + +void +rb_ary_delete_same(VALUE ary, VALUE item) +{ + long i1, i2; + + for (i1 = i2 = 0; i1 < RARRAY_LEN(ary); i1++) { + VALUE e = RARRAY_AREF(ary, i1); + + if (e == item) { + continue; } + if (i1 != i2) { + rb_ary_store(ary, i2, e); + } + i2++; + } + if (RARRAY_LEN(ary) == i2) { + return; } - return item; + ary_resize_smaller(ary, i2); } 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; @@ -1902,52 +3428,48 @@ 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_AREF(ary, pos); + RARRAY_PTR_USE_TRANSIENT(ary, ptr, { + MEMMOVE(ptr+pos, ptr+pos+1, VALUE, len-pos-1); + }); + ARY_INCREASE_LEN(ary, -1); + ary_verify(ary); return del; } /* * call-seq: - * array.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 ) + * ary.delete_at(index) -> obj or nil + * + * Deletes the element at the specified +index+, returning that element, or + * +nil+ if the +index+ is out of range. + * + * See also Array#slice! + * + * a = ["ant", "bat", "cat", "dog"] * a.delete_at(2) #=> "cat" * a #=> ["ant", "bat", "dog"] * a.delete_at(99) #=> nil */ 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 - * - * 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 - * + * 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 up to +length+ + * elements) or by a +range+. + * + * Returns the deleted object (or objects), or +nil+ if the +index+ is out of + * range. + * * a = [ "a", "b", "c" ] * a.slice!(1) #=> "b" * a #=> ["a", "c"] @@ -1958,194 +3480,312 @@ 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; + long pos, len, orig_len; - if (rb_scan_args(argc, argv, "11", &arg1, &arg2) == 2) { - pos = NUM2LONG(arg1); - len = NUM2LONG(arg2); + rb_ary_modify_check(ary); + if (argc == 2) { + pos = NUM2LONG(argv[0]); + len = NUM2LONG(argv[1]); delete_pos_len: + if (len < 0) return Qnil; + orig_len = RARRAY_LEN(ary); if (pos < 0) { - pos = RARRAY(ary)->len + pos; + pos += orig_len; + if (pos < 0) return Qnil; + } + else if (orig_len < pos) return Qnil; + if (orig_len < pos + len) { + len = orig_len - pos; } - arg2 = rb_ary_subseq(ary, pos, len); - rb_ary_splice(ary, pos, len, Qnil); /* Qnil/rb_ary_new2(0) */ + if (len == 0) return rb_ary_new2(0); + arg2 = rb_ary_new4(len, RARRAY_CONST_PTR_TRANSIENT(ary)+pos); + RBASIC_SET_CLASS(arg2, rb_obj_class(ary)); + rb_ary_splice(ary, pos, len, 0, 0); return arg2; } - if (!FIXNUM_P(arg1) && rb_range_beg_len(arg1, &pos, &len, RARRAY(ary)->len, 1)) { - goto delete_pos_len; + rb_check_arity(argc, 1, 2); + arg1 = argv[0]; + + if (!FIXNUM_P(arg1)) { + switch (rb_range_beg_len(arg1, &pos, &len, RARRAY_LEN(ary), 0)) { + case Qtrue: + /* valid range */ + goto delete_pos_len; + case Qnil: + /* invalid range */ + return Qnil; + default: + /* not a range */ + break; + } } return rb_ary_delete_at(ary, NUM2LONG(arg1)); } -/* - * call-seq: - * array.reject! {|item| block } -> array or nil - * - * 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>. - */ +static VALUE +ary_reject(VALUE orig, VALUE result) +{ + long i; + + for (i = 0; i < RARRAY_LEN(orig); i++) { + VALUE v = RARRAY_AREF(orig, i); + + if (!RTEST(rb_yield(v))) { + rb_ary_push(result, v); + } + } + return result; +} static VALUE -rb_ary_reject_bang(ary) - VALUE ary; +reject_bang_i(VALUE a) { + volatile struct select_bang_arg *arg = (void *)a; + VALUE ary = arg->ary; long i1, i2; - rb_ary_modify(ary); - for (i1 = i2 = 0; i1 < RARRAY(ary)->len; i1++) { - VALUE v = RARRAY(ary)->ptr[i1]; + for (i1 = i2 = 0; i1 < RARRAY_LEN(ary); arg->len[0] = ++i1) { + VALUE v = RARRAY_AREF(ary, i1); if (RTEST(rb_yield(v))) continue; if (i1 != i2) { rb_ary_store(ary, i2, v); } - i2++; + arg->len[1] = ++i2; } - if (RARRAY(ary)->len == i2) return Qnil; - if (i2 < RARRAY(ary)->len) - RARRAY(ary)->len = i2; + return (i1 == i2) ? Qnil : ary; +} - return ary; +static VALUE +ary_reject_bang(VALUE ary) +{ + struct select_bang_arg args; + rb_ary_modify_check(ary); + args.ary = ary; + args.len[0] = args.len[1] = 0; + return rb_ensure(reject_bang_i, (VALUE)&args, select_bang_ensure, (VALUE)&args); } /* * call-seq: - * array.reject {|item| block } -> an_array - * - * Returns a new array containing the items in _self_ - * for which the block is not true. + * ary.reject! {|item| block} -> ary or nil + * ary.reject! -> Enumerator + * + * Deletes every element of +self+ for which the block evaluates to +true+, + * if no changes were made returns +nil+. + * + * The array may not be changed instantly every time the block is called. + * + * See also Enumerable#reject and Array#delete_if. + * + * If no block is given, an Enumerator is returned instead. */ static VALUE -rb_ary_reject(ary) - VALUE ary; +rb_ary_reject_bang(VALUE ary) { - ary = rb_ary_dup(ary); - rb_ary_reject_bang(ary); - return ary; + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + rb_ary_modify(ary); + return ary_reject_bang(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>. - * - * a = [ "a", "b", "c" ] - * a.delete_if {|x| x >= "b" } #=> ["a"] + * ary.reject {|item| block } -> new_ary + * ary.reject -> Enumerator + * + * Returns a new array containing the items in +self+ for which the given + * block is not +true+. The ordering of non-rejected elements is maintained. + * + * See also Array#delete_if + * + * If no block is given, an Enumerator is returned instead. */ static VALUE -rb_ary_delete_if(ary) - VALUE ary; +rb_ary_reject(VALUE ary) { - rb_ary_reject_bang(ary); + VALUE rejected_ary; + + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + rejected_ary = rb_ary_new(); + ary_reject(ary, rejected_ary); + return rejected_ary; +} + +/* + * call-seq: + * ary.delete_if {|item| block} -> ary + * ary.delete_if -> Enumerator + * + * Deletes every element of +self+ for which block evaluates to +true+. + * + * The array is changed instantly every time the block is called, not after + * the iteration is over. + * + * See also Array#reject! + * + * If no block is given, an Enumerator is returned instead. + * + * scores = [ 97, 42, 75 ] + * scores.delete_if {|score| score < 80 } #=> [97] + */ + +static VALUE +rb_ary_delete_if(VALUE ary) +{ + ary_verify(ary); + RETURN_SIZED_ENUMERATOR(ary, 0, 0, ary_enum_length); + ary_reject_bang(ary); return ary; } +static VALUE +take_i(RB_BLOCK_CALL_FUNC_ARGLIST(val, cbarg)) +{ + VALUE *args = (VALUE *)cbarg; + if (args[1] == 0) rb_iter_break(); + else args[1]--; + 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; + if (rb_check_block_call(obj, idEach, 0, 0, take_i, (VALUE)args) == Qundef) + rb_raise(rb_eTypeError, "wrong argument type %"PRIsVALUE" (must respond to :each)", + rb_obj_class(obj)); + return result; +} + + /* * call-seq: - * array.zip(arg, ...) -> an_array - * array.zip(arg, ...) {| arr | block } -> nil - * - * Converts any arguments to arrays, then merges elements of - * <i>self</i> 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 - * invoked for each output array, otherwise an array of arrays is - * returned. - * + * ary.zip(arg, ...) -> new_ary + * ary.zip(arg, ...) {|arr| block} -> nil + * + * Converts any arguments to arrays, then merges elements of +self+ with + * corresponding elements from each argument. + * + * This generates a sequence of <code>ary.size</code> _n_-element arrays, + * where _n_ is one more than the count of arguments. + * + * If the size of any argument is less than the size of the initial array, + * +nil+ 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]] + * [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; + long len = RARRAY_LEN(ary); + VALUE result = Qnil; 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); + int arity = rb_block_arity(); - 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)); + if (arity > 1) { + VALUE work, *tmp; + + tmp = ALLOCV_N(VALUE, work, argc+1); + + for (i=0; i<RARRAY_LEN(ary); i++) { + tmp[0] = RARRAY_AREF(ary, i); + for (j=0; j<argc; j++) { + tmp[j+1] = rb_ary_elt(argv[j], i); + } + rb_yield_values2(argc+1, tmp); + } + + if (work) ALLOCV_END(work); + } + else { + for (i=0; i<RARRAY_LEN(ary); i++) { + VALUE tmp = rb_ary_new2(argc+1); + + rb_ary_push(tmp, RARRAY_AREF(ary, i)); + for (j=0; j<argc; j++) { + rb_ary_push(tmp, rb_ary_elt(argv[j], i)); + } + rb_yield(tmp); } - rb_yield(tmp); } - return Qnil; } - len = RARRAY(ary)->len; - result = rb_ary_new2(len); - for (i=0; i<len; i++) { - VALUE tmp = rb_ary_new2(argc+1); + else { + result = rb_ary_new_capa(len); + + for (i=0; i<len; i++) { + VALUE tmp = rb_ary_new_capa(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(tmp, RARRAY_AREF(ary, i)); + for (j=0; j<argc; j++) { + rb_ary_push(tmp, rb_ary_elt(argv[j], i)); + } + rb_ary_push(result, tmp); } - 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 - * rows and columns. - * + * 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]] + * + * If the length of the subarrays don't match, an IndexError is raised. */ 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)); @@ -2156,75 +3796,111 @@ 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 + * ary.initialize_copy(other_ary) -> ary + * + * Replaces the contents of +self+ with the contents of +other_ary+, + * 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 shared_root = 0; + + if (ARY_OWNS_HEAP_P(copy)) { + ary_heap_free(copy); + } + else if (ARY_SHARED_P(copy)) { + shared_root = ARY_SHARED_ROOT(copy); + FL_UNSET_SHARED(copy); + } + FL_SET_EMBED(copy); + ary_memcpy(copy, 0, RARRAY_LEN(orig), RARRAY_CONST_PTR_TRANSIENT(orig)); + if (shared_root) { + rb_ary_decrement_share(shared_root); + } + ARY_SET_LEN(copy, RARRAY_LEN(orig)); + } + else { + VALUE shared_root = ary_make_shared(orig); + if (ARY_OWNS_HEAP_P(copy)) { + ary_heap_free(copy); + } + else { + rb_ary_unshare_safe(copy); + } + FL_UNSET_EMBED(copy); + ARY_SET_PTR(copy, ARY_HEAP_PTR(orig)); + ARY_SET_LEN(copy, ARY_HEAP_LEN(orig)); + rb_ary_set_shared(copy, shared_root); + } + ary_verify(copy); 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); + if (ARY_SHARED_P(ary)) { + if (!ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + FL_SET_EMBED(ary); + ARY_SET_EMBED_LEN(ary, 0); + } + } + else { + ARY_SET_LEN(ary, 0); + if (ARY_DEFAULT_SIZE * 2 < ARY_CAPA(ary)) { + ary_resize_capa(ary, ARY_DEFAULT_SIZE * 2); + } } + ary_verify(ary); 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 - * 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. - * + * 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 +obj+. + * + * A +start+ of +nil+ is equivalent to zero. + * + * A +length+ of +nil+ is equivalent to the length of the array. + * + * The last three forms fill the array with the value of the given block, + * which is passed the absolute index of each element to be filled. + * + * Negative values of +start+ count from the end of the array, where +-1+ is + * the last element. + * * a = [ "a", "b", "c", "d" ] * a.fill("x") #=> ["x", "x", "x", "x"] * a.fill("z", 2, 2) #=> ["x", "x", "z", "z"] @@ -2234,18 +3910,12 @@ 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; + VALUE item = Qundef, arg1, arg2; long beg = 0, end = 0, len = 0; - VALUE *p, *pend; - int block_p = Qfalse; if (rb_block_given_p()) { - block_p = Qtrue; rb_scan_args(argc, argv, "02", &arg1, &arg2); argc += 1; /* hackish */ } @@ -2255,20 +3925,20 @@ 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); @@ -2279,91 +3949,138 @@ 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; + ary_mem_clear(ary, RARRAY_LEN(ary), end - RARRAY_LEN(ary)); + ARY_SET_LEN(ary, end); } - if (block_p) { + if (item == Qundef) { VALUE v; long i; 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; + ARY_SET(ary, i, v); } } else { - p = RARRAY(ary)->ptr + beg; - pend = p + len; - while (p < pend) { - *p++ = item; - } + ary_memfill(ary, beg, len, item); } return ary; } -/* +/* * call-seq: - * array + other_array -> an_array + * ary + other_ary -> new_ary * - * Concatenation---Returns a new array built by concatenating the + * 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 ] + * a = [ "a", "b", "c" ] + * c = a + [ "d", "e", "f" ] + * c #=> [ "a", "b", "c", "d", "e", "f" ] + * a #=> [ "a", "b", "c" ] + * + * Note that + * x += y + * is the same as + * x = x + y + * This means that it produces a new array. As a consequence, + * repeated use of <code>+=</code> on arrays can be quite inefficient. + * + * See also Array#concat. */ VALUE -rb_ary_plus(x, y) - VALUE x, y; +rb_ary_plus(VALUE x, VALUE y) { VALUE z; - long len; + long len, xlen, ylen; y = to_ary(y); - len = RARRAY(x)->len + RARRAY(y)->len; + xlen = RARRAY_LEN(x); + ylen = RARRAY_LEN(y); + len = xlen + ylen; 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; + + ary_memcpy(z, 0, xlen, RARRAY_CONST_PTR_TRANSIENT(x)); + ary_memcpy(z, xlen, ylen, RARRAY_CONST_PTR_TRANSIENT(y)); + ARY_SET_LEN(z, len); return z; } -/* +static VALUE +ary_append(VALUE x, VALUE y) +{ + long n = RARRAY_LEN(y); + if (n > 0) { + rb_ary_splice(x, RARRAY_LEN(x), 0, RARRAY_CONST_PTR_TRANSIENT(y), n); + } + return x; +} + +/* * call-seq: - * array.concat(other_array) -> array + * ary.concat(other_ary1, other_ary2, ...) -> ary + * + * Appends the elements of <code>other_ary</code>s to +self+. + * + * [ "a", "b" ].concat( ["c", "d"]) #=> [ "a", "b", "c", "d" ] + * [ "a" ].concat( ["b"], ["c", "d"]) #=> [ "a", "b", "c", "d" ] + * [ "a" ].concat #=> [ "a" ] + * + * a = [ 1, 2, 3 ] + * a.concat( [ 4, 5 ]) + * a #=> [ 1, 2, 3, 4, 5 ] * - * Appends the elements in other_array to _self_. - * - * [ "a", "b" ].concat( ["c", "d"] ) #=> [ "a", "b", "c", "d" ] + * a = [ 1, 2 ] + * a.concat(a, a) #=> [1, 2, 1, 2, 1, 2] + * + * See also Array#+. */ - -VALUE -rb_ary_concat(x, y) - VALUE x, y; +static VALUE +rb_ary_concat_multi(int argc, VALUE *argv, VALUE ary) { - y = to_ary(y); - if (RARRAY(y)->len > 0) { - rb_ary_splice(x, RARRAY(x)->len, 0, y); + rb_ary_modify_check(ary); + + if (argc == 1) { + rb_ary_concat(ary, argv[0]); } - return x; + else if (argc > 1) { + int i; + VALUE args = rb_ary_tmp_new(argc); + for (i = 0; i < argc; i++) { + rb_ary_concat(args, argv[i]); + } + ary_append(ary, args); + } + + ary_verify(ary); + return ary; } +VALUE +rb_ary_concat(VALUE x, VALUE y) +{ + return ary_append(x, to_ary(y)); +} -/* +/* * 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_. + * Repetition --- With a String argument, equivalent to + * <code>ary.join(str)</code>. + * + * Otherwise, returns a new array built by concatenating the +int+ copies of + * +self+. * * * [ 1, 2, 3 ] * 3 #=> [ 1, 2, 3, 1, 2, 3, 1, 2, 3 ] @@ -2372,11 +4089,11 @@ 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; + const VALUE *ptr; + long t, len; tmp = rb_check_string_type(times); if (!NIL_P(tmp)) { @@ -2384,37 +4101,48 @@ 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); - } - OBJ_INFECT(ary2, ary); - + ARY_SET_LEN(ary2, len); + + ptr = RARRAY_CONST_PTR_TRANSIENT(ary); + t = RARRAY_LEN(ary); + if (0 < t) { + ary_memcpy(ary2, 0, t, ptr); + while (t <= len/2) { + ary_memcpy(ary2, t, t, RARRAY_CONST_PTR_TRANSIENT(ary2)); + t *= 2; + } + if (t < len) { + ary_memcpy(ary2, t, len-t, RARRAY_CONST_PTR_TRANSIENT(ary2)); + } + } + out: return ary2; } -/* +/* * call-seq: - * array.assoc(obj) -> an_array or nil + * ary.assoc(obj) -> element_ary or nil + * + * Searches through an array whose elements are also arrays comparing +obj+ + * with the first element of each contained array using <code>obj.==</code>. + * + * Returns the first contained array that matches (that is, the first + * associated array), or +nil+ if no match is found. * - * Searches through an array whose elements are also arrays - * comparing _obj_ with the first element of each contained array - * using obj.==. - * Returns the first contained array that matches (that - * is, the first associated array), - * or +nil+ if no match is found. - * See also <code>Array#rassoc</code>. + * See also Array#rassoc * * s1 = [ "colors", "red", "blue", "green" ] * s2 = [ "letters", "a", "b", "c" ] @@ -2425,17 +4153,15 @@ 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 = RARRAY(ary)->ptr[i]; - if (TYPE(v) == T_ARRAY && - 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_AREF(ary, i)); + if (!NIL_P(v) && RARRAY_LEN(v) > 0 && + rb_equal(RARRAY_AREF(v, 0), key)) return v; } return Qnil; @@ -2443,55 +4169,79 @@ rb_ary_assoc(ary, key) /* * call-seq: - * array.rassoc(key) -> an_array or nil - * - * Searches through the array whose elements are also arrays. Compares - * <em>key</em> with the second element of each contained array using - * <code>==</code>. Returns the first contained array that matches. See - * also <code>Array#assoc</code>. - * + * ary.rassoc(obj) -> element_ary or nil + * + * Searches through the array whose elements are also arrays. + * + * Compares +obj+ with the second element of each contained array using + * <code>obj.==</code>. + * + * Returns the first contained array that matches +obj+. + * + * See also Array#assoc. + * * 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]; - if (TYPE(v) == T_ARRAY && - RARRAY(v)->len > 1 && - rb_equal(RARRAY(v)->ptr[1], value)) + for (i = 0; i < RARRAY_LEN(ary); ++i) { + v = RARRAY_AREF(ary, i); + if (RB_TYPE_P(v, T_ARRAY) && + RARRAY_LEN(v) > 1 && + rb_equal(RARRAY_AREF(v, 1), value)) return v; } return Qnil; } static VALUE -recursive_equal(ary1, ary2) - VALUE ary1, ary2; +recursive_equal(VALUE ary1, VALUE ary2, int recur) { - long i; - - for (i=0; i<RARRAY(ary1)->len; i++) { - if (!rb_equal(rb_ary_elt(ary1, i), rb_ary_elt(ary2, i))) - return Qfalse; + long i, len1; + const VALUE *p1, *p2; + + if (recur) return Qtrue; /* Subtle! */ + + /* rb_equal() can evacuate ptrs */ + p1 = RARRAY_CONST_PTR(ary1); + p2 = RARRAY_CONST_PTR(ary2); + len1 = RARRAY_LEN(ary1); + + for (i = 0; i < len1; i++) { + if (*p1 != *p2) { + if (rb_equal(*p1, *p2)) { + len1 = RARRAY_LEN(ary1); + if (len1 != RARRAY_LEN(ary2)) + return Qfalse; + if (len1 < i) + return Qtrue; + p1 = RARRAY_CONST_PTR(ary1) + i; + p2 = RARRAY_CONST_PTR(ary2) + i; + } + else { + return Qfalse; + } + } + p1++; + p2++; } 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 - * Object.==) the corresponding element in the other array. + * Equality --- Two arrays are equal if they contain the same number of + * elements and if each element is equal to (according to Object#==) the + * corresponding element in +other_ary+. * * [ "a", "c" ] == [ "a", "c", 7 ] #=> false * [ "a", "c", 7 ] == [ "a", "c", 7 ] #=> true @@ -2500,28 +4250,27 @@ recursive_equal(ary1, ary2) */ 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) { - if (!rb_respond_to(ary2, rb_intern("to_ary"))) { + if (!RB_TYPE_P(ary2, T_ARRAY)) { + if (!rb_respond_to(ary2, idTo_ary)) { return Qfalse; } return rb_equal(ary2, ary1); } - if (RARRAY(ary1)->len != RARRAY(ary2)->len) return Qfalse; - if (rb_inspecting_p(ary1)) return Qfalse; - return rb_protect_inspect(recursive_equal, ary1, ary2); + if (RARRAY_LEN(ary1) != RARRAY_LEN(ary2)) return Qfalse; + if (RARRAY_CONST_PTR_TRANSIENT(ary1) == RARRAY_CONST_PTR_TRANSIENT(ary2)) return Qtrue; + return rb_exec_recursive_paired(recursive_equal, ary1, ary2, ary2); } static VALUE -recursive_eql(ary1, ary2) - VALUE ary1, ary2; +recursive_eql(VALUE ary1, VALUE ary2, int recur) { long i; - 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; } @@ -2530,99 +4279,106 @@ recursive_eql(ary1, ary2) /* * 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, - * or are both arrays with the same content. + * Returns +true+ if +self+ and +other+ are the same object, + * or are both arrays with the same content (according to Object#eql?). */ 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; - if (rb_inspecting_p(ary1)) return Qfalse; - return rb_protect_inspect(recursive_eql, ary1, ary2); -} - -static VALUE recursive_hash _((VALUE ary)); - -static VALUE -recursive_hash(ary) - VALUE ary; -{ - long i, h; - VALUE n; - - 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); - } - return LONG2FIX(h); + if (!RB_TYPE_P(ary2, T_ARRAY)) return Qfalse; + if (RARRAY_LEN(ary1) != RARRAY_LEN(ary2)) return Qfalse; + if (RARRAY_CONST_PTR_TRANSIENT(ary1) == RARRAY_CONST_PTR_TRANSIENT(ary2)) return Qtrue; + return rb_exec_recursive_paired(recursive_eql, ary1, ary2, ary2); } /* * call-seq: - * array.hash -> fixnum + * ary.hash -> integer + * + * Compute a hash-code for this array. + * + * Two arrays with the same content will have the same hash code (and will + * compare using #eql?). * - * 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>). + * See also Object#hash. */ static VALUE -rb_ary_hash(ary) - VALUE ary; +rb_ary_hash(VALUE ary) { - if (rb_inspecting_p(ary)) { - return LONG2FIX(0); + long i; + st_index_t h; + VALUE n; + + h = rb_hash_start(RARRAY_LEN(ary)); + h = rb_hash_uint(h, (st_index_t)rb_ary_hash); + for (i=0; i<RARRAY_LEN(ary); i++) { + n = rb_hash(RARRAY_AREF(ary, i)); + h = rb_hash_uint(h, NUM2LONG(n)); } - return rb_protect_inspect(recursive_hash, ary, 0); + h = rb_hash_end(h); + return ST2FIX(h); } /* * call-seq: - * array.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>), - * <code>false</code> otherwise. - * + * ary.include?(object) -> true or false + * + * Returns +true+ if the given +object+ is present in +self+ (that is, if any + * element <code>==</code> +object+), otherwise returns +false+. + * * 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)) { + VALUE e; + + for (i=0; i<RARRAY_LEN(ary); i++) { + e = RARRAY_AREF(ary, i); + if (rb_equal(e, item)) { return Qtrue; } } return Qfalse; } -VALUE -recursive_cmp(ary1, ary2) - VALUE ary1, ary2; +static VALUE +rb_ary_includes_by_eql(VALUE ary, VALUE item) +{ + long i; + VALUE e; + + for (i=0; i<RARRAY_LEN(ary); i++) { + e = RARRAY_AREF(ary, i); + if (rb_eql(item, e)) { + return Qtrue; + } + } + return Qfalse; +} + +static VALUE +recursive_cmp(VALUE ary1, VALUE ary2, int recur) { long i, len; - 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)); + VALUE e1 = rb_ary_elt(ary1, i), e2 = rb_ary_elt(ary2, i); + VALUE v = rb_funcallv(e1, id_cmp, 1, &e2); if (v != INT2FIX(0)) { return v; } @@ -2630,266 +4386,707 @@ recursive_cmp(ary1, ary2) 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 <code>+1</code>) if this + * array is less than, equal to, or greater than +other_ary+. + * + * Each object in each array is compared (using the <=> operator). + * + * Arrays are compared in an "element-wise" manner; the first element of +ary+ + * is compared with the first one of +other_ary+ using the <=> operator, then + * each of the second elements, etc... + * As soon as the result of any such comparison is non zero (i.e. the two + * corresponding elements are not equal), that result is returned for the + * whole array comparison. + * + * If all the elements are equal, then the result is based on a comparison of + * the array lengths. Thus, two arrays are "equal" according to Array#<=> 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. + * + * +nil+ is returned if the +other_ary+ is not an array or if the comparison + * of two elements returned +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 - * (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 - * comparison of the array lengths. Thus, two arrays are - * ``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 + * [ 1, 2 ] <=> [ 1, :two ] #=> nil * */ 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); - if (rb_inspecting_p(ary1)) return INT2FIX(0); - v = rb_protect_inspect(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++) { + VALUE elt = RARRAY_AREF(ary, i); + rb_hash_add_new_element(hash, elt, elt); } - 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(VALUE ary) +{ + long size = RARRAY_LEN(ary); + VALUE hash = rb_hash_new_with_size(size); + + RBASIC_CLEAR_CLASS(hash); + return hash; +} + +static VALUE +ary_make_hash(VALUE ary) +{ + VALUE hash = ary_tmp_hash_new(ary); + 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); + rb_hash_add_new_element(hash, k, v); } return hash; } -/* +static VALUE +ary_make_hash_by(VALUE ary) +{ + VALUE hash = ary_tmp_hash_new(ary); + return ary_add_hash_by(hash, ary); +} + +static inline void +ary_recycle_hash(VALUE hash) +{ + assert(RBASIC_CLASS(hash) == 0); + if (RHASH_ST_TABLE_P(hash)) { + st_table *tbl = RHASH_ST_TABLE(hash); + st_free_table(tbl); + RHASH_ST_CLEAR(hash); + } +} + +/* * 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 all + * occurrences of any item that also appear in +other_ary+. The order is + * preserved from the original array. * - * 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 - * library class Set.) + * It compares elements using their #hash and #eql? methods for efficiency. * * [ 1, 1, 2, 2, 3, 3, 4, 5 ] - [ 1, 2, 4 ] #=> [ 3, 3, 5 ] + * + * Note that while 1 and 2 were only present once in the array argument, and + * were present twice in the receiver array, all occurrences of each Integer are + * removed in the returned array. + * + * If you need set-like behavior, see the library class Set. + * + * See also Array#difference. */ static VALUE -rb_ary_diff(ary1, ary2) - VALUE ary1, ary2; +rb_ary_diff(VALUE ary1, VALUE ary2) { VALUE ary3; - volatile VALUE hash; + VALUE hash; long i; - hash = ary_make_hash(to_ary(ary2), 0); + ary2 = 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; + if (RARRAY_LEN(ary1) <= SMALL_ARRAY_LEN || RARRAY_LEN(ary2) <= SMALL_ARRAY_LEN) { + for (i=0; i<RARRAY_LEN(ary1); i++) { + VALUE elt = rb_ary_elt(ary1, i); + if (rb_ary_includes_by_eql(ary2, elt)) continue; + rb_ary_push(ary3, elt); + } + return ary3; + } + + hash = ary_make_hash(ary2); + for (i=0; i<RARRAY_LEN(ary1); i++) { + if (rb_hash_stlike_lookup(hash, RARRAY_AREF(ary1, i), NULL)) continue; rb_ary_push(ary3, rb_ary_elt(ary1, i)); } + ary_recycle_hash(hash); return ary3; } -/* +/* * call-seq: - * array & other_array + * ary.difference(other_ary1, other_ary2, ...) -> new_ary + * + * Array Difference + * + * Returns a new array that is a copy of the original array, removing all + * occurrences of any item that also appear in +other_ary+. The order is + * preserved from the original array. + * + * It compares elements using their #hash and #eql? methods for efficiency. + * + * [ 1, 1, 2, 2, 3, 3, 4, 5 ].difference([ 1, 2, 4 ]) #=> [ 3, 3, 5 ] * - * Set Intersection---Returns a new array - * containing elements common to the two arrays, with no duplicates. + * Note that while 1 and 2 were only present once in the array argument, and + * were present twice in the receiver array, all occurrences of each Integer are + * removed in the returned array. * - * [ 1, 1, 3, 5 ] & [ 1, 2, 3 ] #=> [ 1, 3 ] + * Multiple array arguments can be supplied and all occurrences of any element + * in those supplied arrays that match the receiver will be removed from the + * returned array. + * + * [ 1, 'c', :s, 'yep' ].difference([ 1 ], [ 'a', 'c' ]) #=> [ :s, "yep" ] + * + * If you need set-like behavior, see the library class Set. + * + * See also Array#-. + */ + +static VALUE +rb_ary_difference_multi(int argc, VALUE *argv, VALUE ary) +{ + VALUE ary_diff; + long i, length; + volatile VALUE t0; + bool *is_hash = ALLOCV_N(bool, t0, argc); + ary_diff = rb_ary_new(); + length = RARRAY_LEN(ary); + + for (i = 0; i < argc; i++) { + argv[i] = to_ary(argv[i]); + is_hash[i] = (length > SMALL_ARRAY_LEN && RARRAY_LEN(argv[i]) > SMALL_ARRAY_LEN); + if (is_hash[i]) argv[i] = ary_make_hash(argv[i]); + } + + for (i = 0; i < RARRAY_LEN(ary); i++) { + int j; + VALUE elt = rb_ary_elt(ary, i); + for (j = 0; j < argc; j++) { + if (is_hash[j]) { + if (rb_hash_stlike_lookup(argv[j], RARRAY_AREF(ary, i), NULL)) + break; + } + else { + if (rb_ary_includes_by_eql(argv[j], elt)) break; + } + } + if (j == argc) rb_ary_push(ary_diff, elt); + } + + ALLOCV_END(t0); + + return ary_diff; +} + + +/* + * call-seq: + * ary & other_ary -> new_ary + * + * Set Intersection --- Returns a new array containing unique elements common to the + * two arrays. The order is preserved from the original array. + * + * It compares elements using their #hash and #eql? methods for efficiency. + * + * [ 1, 1, 3, 5 ] & [ 3, 2, 1 ] #=> [ 1, 3 ] + * [ 'a', 'b', 'b', 'z' ] & [ 'a', 'b', 'c' ] #=> [ 'a', 'b' ] + * + * See also Array#uniq. */ 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_new(); + if (RARRAY_LEN(ary1) == 0 || RARRAY_LEN(ary2) == 0) 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)) { + if (RARRAY_LEN(ary1) <= SMALL_ARRAY_LEN && RARRAY_LEN(ary2) <= SMALL_ARRAY_LEN) { + for (i=0; i<RARRAY_LEN(ary1); i++) { + v = RARRAY_AREF(ary1, i); + if (!rb_ary_includes_by_eql(ary2, v)) continue; + if (rb_ary_includes_by_eql(ary3, v)) continue; rb_ary_push(ary3, v); } + return ary3; } + hash = ary_make_hash(ary2); + + for (i=0; i<RARRAY_LEN(ary1); i++) { + v = RARRAY_AREF(ary1, i); + vv = (st_data_t)v; + if (rb_hash_stlike_delete(hash, &vv, 0)) { + rb_ary_push(ary3, v); + } + } + ary_recycle_hash(hash); + return ary3; } -/* +/* * call-seq: - * array | other_array -> an_array + * ary.intersection(other_ary1, other_ary2, ...) -> new_ary + * + * Set Intersection --- Returns a new array containing unique elements common + * to +self+ and <code>other_ary</code>s. Order is preserved from the original + * array. + * + * It compares elements using their #hash and #eql? methods for efficiency. * - * Set Union---Returns a new array by joining this array with - * other_array, removing duplicates. + * [ 1, 1, 3, 5 ].intersection([ 3, 2, 1 ]) # => [ 1, 3 ] + * [ "a", "b", "z" ].intersection([ "a", "b", "c" ], [ "b" ]) # => [ "b" ] + * [ "a" ].intersection #=> [ "a" ] * - * [ "a", "b", "c" ] | [ "c", "d", "a" ] - * #=> [ "a", "b", "c", "d" ] + * See also Array#&. */ static VALUE -rb_ary_or(ary1, ary2) - VALUE ary1, ary2; +rb_ary_intersection_multi(int argc, VALUE *argv, VALUE ary) +{ + VALUE result = rb_ary_dup(ary); + int i; + + for (i = 0; i < argc; i++) { + result = rb_ary_and(result, argv[i]); + } + + return result; +} + +static int +ary_hash_orset(st_data_t *key, st_data_t *value, st_data_t arg, int existing) +{ + if (existing) return ST_STOP; + *key = *value = (VALUE)arg; + return ST_CONTINUE; +} + +static void +rb_ary_union(VALUE ary_union, VALUE ary) { - VALUE hash, ary3; - VALUE v, vv; long i; + for (i = 0; i < RARRAY_LEN(ary); i++) { + VALUE elt = rb_ary_elt(ary, i); + if (rb_ary_includes_by_eql(ary_union, elt)) continue; + rb_ary_push(ary_union, elt); + } +} + +static void +rb_ary_union_hash(VALUE hash, VALUE ary2) +{ + long i; + for (i = 0; i < RARRAY_LEN(ary2); i++) { + VALUE elt = RARRAY_AREF(ary2, i); + if (!rb_hash_stlike_update(hash, (st_data_t)elt, ary_hash_orset, (st_data_t)elt)) { + RB_OBJ_WRITTEN(hash, Qundef, elt); + } + } +} + +/* + * call-seq: + * ary | other_ary -> new_ary + * + * Set Union --- Returns a new array by joining +ary+ with +other_ary+, + * excluding any duplicates and preserving the order from the given arrays. + * + * It compares elements using their #hash and #eql? methods for efficiency. + * + * [ "a", "b", "c" ] | [ "c", "d", "a" ] #=> [ "a", "b", "c", "d" ] + * [ "c", "d", "a" ] | [ "a", "b", "c" ] #=> [ "c", "d", "a", "b" ] + * + * See also Array#union. + */ + +static VALUE +rb_ary_or(VALUE ary1, VALUE ary2) +{ + VALUE hash, ary3; ary2 = to_ary(ary2); - ary3 = rb_ary_new2(RARRAY(ary1)->len+RARRAY(ary2)->len); - hash = ary_make_hash(ary1, ary2); + if (RARRAY_LEN(ary1) + RARRAY_LEN(ary2) <= SMALL_ARRAY_LEN) { + ary3 = rb_ary_new(); + rb_ary_union(ary3, ary1); + rb_ary_union(ary3, ary2); + 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)) { - rb_ary_push(ary3, v); + hash = ary_make_hash(ary1); + rb_ary_union_hash(hash, ary2); + + ary3 = rb_hash_values(hash); + ary_recycle_hash(hash); + return ary3; +} + +/* + * call-seq: + * ary.union(other_ary1, other_ary2, ...) -> new_ary + * + * Set Union --- Returns a new array by joining <code>other_ary</code>s with +self+, + * excluding any duplicates and preserving the order from the given arrays. + * + * It compares elements using their #hash and #eql? methods for efficiency. + * + * [ "a", "b", "c" ].union( [ "c", "d", "a" ] ) #=> [ "a", "b", "c", "d" ] + * [ "a" ].union( ["e", "b"], ["a", "c", "b"] ) #=> [ "a", "e", "b", "c" ] + * [ "a" ].union #=> [ "a" ] + * + * See also Array#|. + */ + +static VALUE +rb_ary_union_multi(int argc, VALUE *argv, VALUE ary) +{ + int i; + long sum; + VALUE hash, ary_union; + + sum = RARRAY_LEN(ary); + for (i = 0; i < argc; i++) { + argv[i] = to_ary(argv[i]); + sum += RARRAY_LEN(argv[i]); + } + + if (sum <= SMALL_ARRAY_LEN) { + ary_union = rb_ary_new(); + + rb_ary_union(ary_union, ary); + for (i = 0; i < argc; i++) rb_ary_union(ary_union, argv[i]); + + return ary_union; + } + + hash = ary_make_hash(ary); + for (i = 0; i < argc; i++) rb_ary_union_hash(hash, argv[i]); + + ary_union = rb_hash_values(hash); + ary_recycle_hash(hash); + return ary_union; +} + +/* + * call-seq: + * ary.max -> obj + * ary.max {|a, b| block} -> obj + * ary.max(n) -> array + * ary.max(n) {|a, b| block} -> array + * + * Returns the object in _ary_ with the maximum value. The + * first form assumes all objects implement Comparable; + * the second uses the block to return <em>a <=> b</em>. + * + * ary = %w(albatross dog horse) + * ary.max #=> "horse" + * ary.max {|a, b| a.length <=> b.length} #=> "albatross" + * + * If the +n+ argument is given, maximum +n+ elements are returned + * as an array. + * + * ary = %w[albatross dog horse] + * ary.max(2) #=> ["horse", "dog"] + * ary.max(2) {|a, b| a.length <=> b.length } #=> ["albatross", "horse"] + */ +static VALUE +rb_ary_max(int argc, VALUE *argv, VALUE ary) +{ + struct cmp_opt_data cmp_opt = { 0, 0 }; + VALUE result = Qundef, v; + VALUE num; + long i; + + if (rb_check_arity(argc, 0, 1) && !NIL_P(num = argv[0])) + return rb_nmin_run(ary, num, 0, 1, 1); + + if (rb_block_given_p()) { + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_AREF(ary, i); + if (result == Qundef || rb_cmpint(rb_yield_values(2, v, result), v, result) > 0) { + result = 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)) { - rb_ary_push(ary3, v); + else { + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_AREF(ary, i); + if (result == Qundef || OPTIMIZED_CMP(v, result, cmp_opt) > 0) { + result = v; + } } } - return ary3; + if (result == Qundef) return Qnil; + return result; +} + +/* + * call-seq: + * ary.min -> obj + * ary.min {| a,b | block } -> obj + * ary.min(n) -> array + * ary.min(n) {| a,b | block } -> array + * + * Returns the object in _ary_ with the minimum value. The + * first form assumes all objects implement Comparable; + * the second uses the block to return <em>a <=> b</em>. + * + * ary = %w(albatross dog horse) + * ary.min #=> "albatross" + * ary.min {|a, b| a.length <=> b.length} #=> "dog" + * + * If the +n+ argument is given, minimum +n+ elements are returned + * as an array. + * + * ary = %w[albatross dog horse] + * ary.min(2) #=> ["albatross", "dog"] + * ary.min(2) {|a, b| a.length <=> b.length } #=> ["dog", "horse"] + */ +static VALUE +rb_ary_min(int argc, VALUE *argv, VALUE ary) +{ + struct cmp_opt_data cmp_opt = { 0, 0 }; + VALUE result = Qundef, v; + VALUE num; + long i; + + if (rb_check_arity(argc, 0, 1) && !NIL_P(num = argv[0])) + return rb_nmin_run(ary, num, 0, 0, 1); + + if (rb_block_given_p()) { + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_AREF(ary, i); + if (result == Qundef || rb_cmpint(rb_yield_values(2, v, result), v, result) < 0) { + result = v; + } + } + } + else { + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_AREF(ary, i); + if (result == Qundef || OPTIMIZED_CMP(v, result, cmp_opt) < 0) { + result = v; + } + } + } + if (result == Qundef) return Qnil; + return result; +} + +/* + * call-seq: + * ary.minmax -> [obj, obj] + * ary.minmax {| a,b | block } -> [obj, obj] + * + * Returns a two element array which contains the minimum and the + * maximum value in the array. + * + * Can be given an optional block to override the default comparison + * method <code>a <=> b</code>. + */ +static VALUE +rb_ary_minmax(VALUE ary) +{ + if (rb_block_given_p()) { + return rb_call_super(0, NULL); + } + return rb_assoc_new(rb_ary_min(0, 0, ary), rb_ary_max(0, 0, ary)); +} + +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_. - * Returns <code>nil</code> if no changes are made (that is, no - * duplicates are found). - * + * 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. + * + * It compares values using their #hash and #eql? methods for efficiency. + * + * +self+ is traversed in order, and the first occurrence is kept. + * + * Returns +nil+ 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; - long i, j; + VALUE hash; + long hash_size; - hash = ary_make_hash(ary, 0); - - if (RARRAY(ary)->len == RHASH(hash)->tbl->num_entries) { + rb_ary_modify_check(ary); + if (RARRAY_LEN(ary) <= 1) + return Qnil; + if (rb_block_given_p()) + hash = ary_make_hash_by(ary); + else + hash = ary_make_hash(ary); + + hash_size = RHASH_SIZE(hash); + if (RARRAY_LEN(ary) == hash_size) { return Qnil; } - 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); - } + rb_ary_modify_check(ary); + ARY_SET_LEN(ary, 0); + if (ARY_SHARED_P(ary) && !ARY_EMBED_P(ary)) { + rb_ary_unshare(ary); + FL_SET_EMBED(ary); } - RARRAY(ary)->len = j; + ary_resize_capa(ary, hash_size); + rb_hash_foreach(hash, push_value, ary); + 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. + * + * It compares values using their #hash and #eql? methods for efficiency. + * + * +self+ is traversed in order, and the first occurrence is kept. + * * 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; + + if (RARRAY_LEN(ary) <= 1) { + hash = 0; + uniq = rb_ary_dup(ary); + } + else if (rb_block_given_p()) { + hash = ary_make_hash_by(ary); + uniq = rb_hash_values(hash); + } + else { + hash = ary_make_hash(ary); + uniq = rb_hash_values(hash); + } + RBASIC_SET_CLASS(uniq, rb_obj_class(ary)); + if (hash) { + ary_recycle_hash(hash); + } + + return uniq; } -/* +/* * call-seq: - * array.compact! -> array or nil + * ary.compact! -> ary or nil + * + * Removes +nil+ elements from the array. * - * Removes +nil+ elements from array. - * Returns +nil+ if no changes were made. + * Returns +nil+ if no changes were made, otherwise returns the array. * * [ "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 = (VALUE *)RARRAY_CONST_PTR_TRANSIENT(ary); /* WB: no new reference */ + 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_CONST_PTR_TRANSIENT(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_resize_smaller(ary, n); 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); @@ -2898,149 +5095,1846 @@ rb_ary_compact(ary) /* * call-seq: - * array.nitems -> int - * - * Returns the number of non-<code>nil</code> elements in _self_. - * May be zero. - * - * [ 1, nil, 3, nil, 5 ].nitems #=> 3 + * ary.count -> int + * ary.count(obj) -> int + * ary.count {|item| block} -> int + * + * Returns the number of elements. + * + * If an argument is given, counts the number of elements which equal +obj+ + * using <code>==</code>. + * + * If a block is given, counts the number of elements for which the block + * returns a true value. + * + * ary = [1, 2, 4, 2] + * ary.count #=> 4 + * ary.count(2) #=> 2 + * ary.count {|x| x%2 == 0} #=> 3 + * */ static VALUE -rb_ary_nitems(ary) - VALUE ary; +rb_ary_count(int argc, VALUE *argv, VALUE ary) { - long n = 0; - VALUE *p, *pend; + long i, n = 0; + + if (rb_check_arity(argc, 0, 1) == 0) { + VALUE v; - p = RARRAY(ary)->ptr; - pend = p + RARRAY(ary)->len; + if (!rb_block_given_p()) + return LONG2NUM(RARRAY_LEN(ary)); - while (p < pend) { - if (!NIL_P(*p)) n++; - p++; + for (i = 0; i < RARRAY_LEN(ary); i++) { + v = RARRAY_AREF(ary, i); + if (RTEST(rb_yield(v))) n++; + } + } + else { + VALUE obj = argv[0]; + + if (rb_block_given_p()) { + rb_warn("given block not used"); + } + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (rb_equal(RARRAY_AREF(ary, i), obj)) n++; + } } + return LONG2NUM(n); } -static long -flatten(ary, idx, ary2, memo) - VALUE ary; - long idx; - VALUE ary2, memo; +static VALUE +flatten(VALUE ary, int level) { - VALUE id; - long i = idx; - long n, lim = idx + RARRAY(ary2)->len; - - id = rb_obj_id(ary2); - if (rb_ary_includes(memo, id)) { - rb_raise(rb_eArgError, "tried to flatten recursive array"); - } - rb_ary_push(memo, id); - rb_ary_splice(ary, idx, 1, ary2); - while (i < lim) { - VALUE tmp; - - tmp = rb_check_array_type(rb_ary_elt(ary, i)); - if (!NIL_P(tmp)) { - n = flatten(ary, i, tmp, memo); - i += n; lim += n; + long i; + VALUE stack, result, tmp, elt, vmemo; + st_table *memo; + st_data_t id; + + for (i = 0; i < RARRAY_LEN(ary); i++) { + elt = RARRAY_AREF(ary, i); + tmp = rb_check_array_type(elt); + if (!NIL_P(tmp)) { + break; + } + } + if (i == RARRAY_LEN(ary)) { + return ary; + } else if (tmp == ary) { + rb_raise(rb_eArgError, "tried to flatten recursive array"); + } + + result = ary_new(0, RARRAY_LEN(ary)); + ary_memcpy(result, 0, i, RARRAY_CONST_PTR_TRANSIENT(ary)); + ARY_SET_LEN(result, i); + + stack = ary_new(0, ARY_DEFAULT_SIZE); + rb_ary_push(stack, ary); + rb_ary_push(stack, LONG2NUM(i + 1)); + + vmemo = rb_hash_new(); + RBASIC_CLEAR_CLASS(vmemo); + memo = st_init_numtable(); + rb_hash_st_table_set(vmemo, memo); + st_insert(memo, (st_data_t)ary, (st_data_t)Qtrue); + st_insert(memo, (st_data_t)tmp, (st_data_t)Qtrue); + + ary = tmp; + i = 0; + + while (1) { + while (i < RARRAY_LEN(ary)) { + elt = RARRAY_AREF(ary, i++); + if (level >= 0 && RARRAY_LEN(stack) / 2 >= level) { + rb_ary_push(result, elt); + continue; + } + tmp = rb_check_array_type(elt); + if (RBASIC(result)->klass) { + RB_GC_GUARD(vmemo); + st_clear(memo); + rb_raise(rb_eRuntimeError, "flatten reentered"); + } + if (NIL_P(tmp)) { + rb_ary_push(result, elt); + } + else { + id = (st_data_t)tmp; + if (st_is_member(memo, id)) { + st_clear(memo); + rb_raise(rb_eArgError, "tried to flatten recursive array"); + } + st_insert(memo, id, (st_data_t)Qtrue); + rb_ary_push(stack, ary); + rb_ary_push(stack, LONG2NUM(i)); + ary = tmp; + i = 0; + } } - i++; + if (RARRAY_LEN(stack) == 0) { + break; + } + id = (st_data_t)ary; + st_delete(memo, &id, 0); + tmp = rb_ary_pop(stack); + i = NUM2LONG(tmp); + ary = rb_ary_pop(stack); } - rb_ary_pop(memo); - return lim - idx - 1; /* returns number of increased items */ + st_clear(memo); + + RBASIC_SET_CLASS(result, rb_obj_class(ary)); + return result; } /* * call-seq: - * array.flatten! -> array or nil - * - * Flattens _self_ in place. - * Returns <code>nil</code> if no modifications were made (i.e., - * <i>array</i> contains no subarrays.) - * + * ary.flatten! -> ary or nil + * ary.flatten!(level) -> ary or nil + * + * Flattens +self+ in place. + * + * Returns +nil+ if no modifications were made (i.e., the array contains no + * subarrays.) + * + * The optional +level+ argument determines the level of recursion to flatten. + * * a = [ 1, 2, [3, [4, 5] ] ] * a.flatten! #=> [1, 2, 3, 4, 5] * a.flatten! #=> nil * a #=> [1, 2, 3, 4, 5] + * a = [ 1, 2, [3, [4, 5] ] ] + * a.flatten!(1) #=> [1, 2, 3, [4, 5]] */ static VALUE -rb_ary_flatten_bang(ary) - VALUE ary; +rb_ary_flatten_bang(int argc, VALUE *argv, VALUE ary) { - long i = 0; - int mod = 0; - VALUE memo = Qnil; + int mod = 0, level = -1; + VALUE result, lv; - while (i<RARRAY(ary)->len) { - VALUE ary2 = RARRAY(ary)->ptr[i]; - VALUE tmp; + lv = (rb_check_arity(argc, 0, 1) ? argv[0] : Qnil); + rb_ary_modify_check(ary); + if (!NIL_P(lv)) level = NUM2INT(lv); + if (level == 0) return Qnil; - tmp = rb_check_array_type(ary2); - if (!NIL_P(tmp)) { - if (NIL_P(memo)) { - memo = rb_ary_new(); - } - i += flatten(ary, i, tmp, memo); - mod = 1; - } - i++; + result = flatten(ary, level); + if (result == ary) { + return Qnil; } - if (mod == 0) 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 - * - * 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. - * + * ary.flatten -> new_ary + * ary.flatten(level) -> new_ary + * + * Returns a new array that is a one-dimensional flattening of +self+ + * (recursively). + * + * That is, for every element that is an array, extract its elements into + * the new array. + * + * The optional +level+ 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] - * a.flatten #=> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10 + * a.flatten #=> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] + * a = [ 1, 2, [3, [4, 5] ] ] + * a.flatten(1) #=> [1, 2, 3, [4, 5]] */ static VALUE -rb_ary_flatten(ary) - VALUE ary; +rb_ary_flatten(int argc, VALUE *argv, VALUE ary) +{ + int level = -1; + VALUE result; + + if (rb_check_arity(argc, 0, 1) && !NIL_P(argv[0])) { + level = NUM2INT(argv[0]); + if (level == 0) return ary_make_shared_copy(ary); + } + + result = flatten(ary, level); + if (result == ary) { + result = ary_make_shared_copy(ary); + } + + return result; +} + +#define OPTHASH_GIVEN_P(opts) \ + (argc > 0 && !NIL_P((opts) = rb_check_hash_type(argv[argc-1])) && (--argc, 1)) +static ID id_random; + +#define RAND_UPTO(max) (long)rb_random_ulong_limited((randgen), (max)-1) + +/* + * call-seq: + * ary.shuffle! -> ary + * ary.shuffle!(random: rng) -> ary + * + * Shuffles elements in +self+ in place. + * + * a = [ 1, 2, 3 ] #=> [1, 2, 3] + * a.shuffle! #=> [2, 3, 1] + * a #=> [2, 3, 1] + * + * The optional +rng+ argument will be used as the random number generator. + * + * a.shuffle!(random: Random.new(1)) #=> [1, 3, 2] + */ + +static VALUE +rb_ary_shuffle_bang(int argc, VALUE *argv, VALUE ary) +{ + VALUE opts, randgen = rb_cRandom; + long i, len; + + if (OPTHASH_GIVEN_P(opts)) { + VALUE rnd; + ID keyword_ids[1]; + + keyword_ids[0] = id_random; + rb_get_kwargs(opts, keyword_ids, 0, 1, &rnd); + if (rnd != Qundef) { + randgen = rnd; + } + } + rb_check_arity(argc, 0, 0); + rb_ary_modify(ary); + i = len = RARRAY_LEN(ary); + RARRAY_PTR_USE(ary, ptr, { + while (i) { + long j = RAND_UPTO(i); + VALUE tmp; + if (len != RARRAY_LEN(ary) || ptr != RARRAY_CONST_PTR_TRANSIENT(ary)) { + rb_raise(rb_eRuntimeError, "modified during shuffle"); + } + tmp = ptr[--i]; + ptr[i] = ptr[j]; + ptr[j] = tmp; + } + }); /* WB: no new reference */ + return ary; +} + + +/* + * call-seq: + * ary.shuffle -> new_ary + * ary.shuffle(random: rng) -> new_ary + * + * Returns a new array with elements of +self+ shuffled. + * + * a = [ 1, 2, 3 ] #=> [1, 2, 3] + * a.shuffle #=> [2, 3, 1] + * a #=> [1, 2, 3] + * + * The optional +rng+ argument will be used as the random number generator. + * + * a.shuffle(random: Random.new(1)) #=> [1, 3, 2] + */ + +static VALUE +rb_ary_shuffle(int argc, VALUE *argv, VALUE ary) { ary = rb_ary_dup(ary); - rb_ary_flatten_bang(ary); + rb_ary_shuffle_bang(argc, argv, ary); + return ary; +} + + +/* + * call-seq: + * 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 +nil+ and the second form + * returns an empty array. + * + * a = [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10 ] + * a.sample #=> 7 + * a.sample(4) #=> [6, 4, 2, 5] + * + * The optional +rng+ argument will be used as the random number generator. + * + * a.sample(random: Random.new(1)) #=> 6 + * a.sample(4, random: Random.new(1)) #=> [6, 10, 9, 2] + */ + + +static VALUE +rb_ary_sample(int argc, VALUE *argv, VALUE ary) +{ + VALUE nv, result; + VALUE opts, randgen = rb_cRandom; + long n, len, i, j, k, idx[10]; + long rnds[numberof(idx)]; + long memo_threshold; + + if (OPTHASH_GIVEN_P(opts)) { + VALUE rnd; + ID keyword_ids[1]; + + keyword_ids[0] = id_random; + rb_get_kwargs(opts, keyword_ids, 0, 1, &rnd); + if (rnd != Qundef) { + randgen = rnd; + } + } + len = RARRAY_LEN(ary); + if (rb_check_arity(argc, 0, 1) == 0) { + if (len < 2) + i = 0; + else + i = RAND_UPTO(len); + + return rb_ary_elt(ary, i); + } + nv = argv[0]; + 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] = RAND_UPTO(len - i); + } + } + k = len; + len = RARRAY_LEN(ary); + if (len < k && n <= numberof(idx)) { + for (i = 0; i < n; ++i) { + if (rnds[i] >= len) return rb_ary_new_capa(0); + } + } + if (n > len) n = len; + switch (n) { + case 0: + return rb_ary_new_capa(0); + case 1: + i = rnds[0]; + return rb_ary_new_from_values(1, &RARRAY_AREF(ary, i)); + case 2: + i = rnds[0]; + j = rnds[1]; + if (j >= i) j++; + return rb_ary_new_from_args(2, RARRAY_AREF(ary, i), RARRAY_AREF(ary, j)); + case 3: + i = rnds[0]; + j = rnds[1]; + k = rnds[2]; + { + long l = j, g = i; + if (j >= i) l = i, g = ++j; + if (k >= l && (++k >= g)) ++k; + } + return rb_ary_new_from_args(3, RARRAY_AREF(ary, i), RARRAY_AREF(ary, j), RARRAY_AREF(ary, k)); + } + memo_threshold = + len < 2560 ? len / 128 : + len < 5120 ? len / 64 : + len < 10240 ? len / 32 : + len / 16; + if (n <= numberof(idx)) { + long sorted[numberof(idx)]; + sorted[0] = idx[0] = rnds[0]; + for (i=1; i<n; i++) { + k = rnds[i]; + 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_new_capa(n); + RARRAY_PTR_USE_TRANSIENT(result, ptr_result, { + for (i=0; i<n; i++) { + ptr_result[i] = RARRAY_AREF(ary, idx[i]); + } + }); + } + else if (n <= memo_threshold / 2) { + long max_idx = 0; +#undef RUBY_UNTYPED_DATA_WARNING +#define RUBY_UNTYPED_DATA_WARNING 0 + VALUE vmemo = Data_Wrap_Struct(0, 0, st_free_table, 0); + st_table *memo = st_init_numtable_with_size(n); + DATA_PTR(vmemo) = memo; + result = rb_ary_new_capa(n); + RARRAY_PTR_USE(result, ptr_result, { + for (i=0; i<n; i++) { + long r = RAND_UPTO(len-i) + i; + ptr_result[i] = r; + if (r > max_idx) max_idx = r; + } + len = RARRAY_LEN(ary); + if (len <= max_idx) n = 0; + else if (n > len) n = len; + RARRAY_PTR_USE_TRANSIENT(ary, ptr_ary, { + for (i=0; i<n; i++) { + long j2 = j = ptr_result[i]; + long i2 = i; + st_data_t value; + if (st_lookup(memo, (st_data_t)i, &value)) i2 = (long)value; + if (st_lookup(memo, (st_data_t)j, &value)) j2 = (long)value; + st_insert(memo, (st_data_t)j, (st_data_t)i2); + ptr_result[i] = ptr_ary[j2]; + } + }); + }); + DATA_PTR(vmemo) = 0; + st_free_table(memo); + } + else { + result = rb_ary_dup(ary); + RBASIC_CLEAR_CLASS(result); + RB_GC_GUARD(ary); + RARRAY_PTR_USE(result, ptr_result, { + 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_SET_CLASS_RAW(result, rb_cArray); + } + ARY_SET_LEN(result, n); + + return result; +} + +static VALUE +rb_ary_cycle_size(VALUE self, VALUE args, VALUE eobj) +{ + long mul; + VALUE n = Qnil; + if (args && (RARRAY_LEN(args) > 0)) { + n = RARRAY_AREF(args, 0); + } + if (RARRAY_LEN(self) == 0) return INT2FIX(0); + if (n == Qnil) return DBL2NUM(HUGE_VAL); + mul = NUM2LONG(n); + if (mul <= 0) return INT2FIX(0); + n = LONG2FIX(mul); + return rb_fix_mul_fix(rb_ary_length(self), n); +} + +/* + * call-seq: + * ary.cycle(n=nil) {|obj| block} -> nil + * ary.cycle(n=nil) -> Enumerator + * + * Calls the given block for each element +n+ times or forever if +nil+ is + * given. + * + * Does nothing if a non-positive number is given or the array is empty. + * + * 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(int argc, VALUE *argv, VALUE ary) +{ + long n, i; + + rb_check_arity(argc, 0, 1); + + RETURN_SIZED_ENUMERATOR(ary, argc, argv, rb_ary_cycle_size); + if (argc == 0 || NIL_P(argv[0])) { + n = -1; + } + else { + n = NUM2LONG(argv[0]); + if (n <= 0) return Qnil; + } + + while (RARRAY_LEN(ary) > 0 && (n < 0 || 0 < n--)) { + for (i=0; i<RARRAY_LEN(ary); i++) { + rb_yield(RARRAY_AREF(ary, i)); + } + } + return Qnil; +} + +#define tmpary(n) rb_ary_tmp_new(n) +#define tmpary_discard(a) (ary_discard(a), RBASIC_SET_CLASS_RAW(a, rb_cArray)) + +/* + * Build a ruby array of the corresponding values and yield it to the + * associated block. + * Return the class of +values+ for reentry check. + */ +static int +yield_indexed_values(const VALUE values, const long r, const long *const p) +{ + const VALUE result = rb_ary_new2(r); + long i; + + for (i = 0; i < r; i++) ARY_SET(result, i, RARRAY_AREF(values, p[i])); + ARY_SET_LEN(result, r); + rb_yield(result); + return !RBASIC(values)->klass; +} + +/* + * Compute permutations of +r+ elements of the set <code>[0..n-1]</code>. + * + * When we have a complete permutation of array indices, copy the values + * at those indices 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 + * used: an array of booleans: whether a given index is already used + * values: the Ruby array that holds the actual values to permute + */ +static void +permute0(const long n, const long r, long *const p, char *const used, const VALUE values) +{ + long i = 0, index = 0; + + for (;;) { + const char *const unused = memchr(&used[i], 0, n-i); + if (!unused) { + if (!index) break; + i = p[--index]; /* pop index */ + used[i++] = 0; /* index unused */ + } + else { + i = unused - used; + p[index] = i; + used[i] = 1; /* mark index used */ + ++index; + if (index < r-1) { /* if not done yet */ + p[index] = i = 0; + continue; + } + for (i = 0; i < n; ++i) { + if (used[i]) continue; + p[index] = i; + if (!yield_indexed_values(values, r, p)) { + rb_raise(rb_eRuntimeError, "permute reentered"); + } + } + i = p[--index]; /* pop index */ + used[i] = 0; /* index unused */ + p[index] = ++i; + } + } +} + +/* + * Returns the product of from, from-1, ..., from - how_many + 1. + * http://en.wikipedia.org/wiki/Pochhammer_symbol + */ +static VALUE +descending_factorial(long from, long how_many) +{ + VALUE cnt; + if (how_many > 0) { + cnt = LONG2FIX(from); + while (--how_many > 0) { + long v = --from; + cnt = rb_int_mul(cnt, LONG2FIX(v)); + } + } + else { + cnt = LONG2FIX(how_many == 0); + } + return cnt; +} + +static VALUE +binomial_coefficient(long comb, long size) +{ + VALUE r; + long i; + if (comb > size-comb) { + comb = size-comb; + } + if (comb < 0) { + return LONG2FIX(0); + } + else if (comb == 0) { + return LONG2FIX(1); + } + r = LONG2FIX(size); + for (i = 1; i < comb; ++i) { + r = rb_int_mul(r, LONG2FIX(size - i)); + r = rb_int_idiv(r, LONG2FIX(i + 1)); + } + return r; +} + +static VALUE +rb_ary_permutation_size(VALUE ary, VALUE args, VALUE eobj) +{ + long n = RARRAY_LEN(ary); + long k = (args && (RARRAY_LEN(args) > 0)) ? NUM2LONG(RARRAY_AREF(args, 0)) : n; + + return descending_factorial(n, k); +} + +/* + * call-seq: + * ary.permutation {|p| block} -> ary + * ary.permutation -> Enumerator + * ary.permutation(n) {|p| block} -> ary + * ary.permutation(n) -> Enumerator + * + * When invoked with a block, yield all permutations of length +n+ of the + * elements of the array, then return the array itself. + * + * If +n+ is not specified, yield all permutations of all elements. + * + * The implementation makes no guarantees about the order in which the + * permutations are yielded. + * + * If no block is given, an Enumerator is returned instead. + * + * Examples: + * + * a = [1, 2, 3] + * a.permutation.to_a #=> [[1,2,3],[1,3,2],[2,1,3],[2,3,1],[3,1,2],[3,2,1]] + * a.permutation(1).to_a #=> [[1],[2],[3]] + * a.permutation(2).to_a #=> [[1,2],[1,3],[2,1],[2,3],[3,1],[3,2]] + * a.permutation(3).to_a #=> [[1,2,3],[1,3,2],[2,1,3],[2,3,1],[3,1,2],[3,2,1]] + * a.permutation(0).to_a #=> [[]] # one permutation of length 0 + * a.permutation(4).to_a #=> [] # no permutations of length 4 + */ + +static VALUE +rb_ary_permutation(int argc, VALUE *argv, VALUE ary) +{ + long r, n, i; + + n = RARRAY_LEN(ary); /* Array length */ + RETURN_SIZED_ENUMERATOR(ary, argc, argv, rb_ary_permutation_size); /* Return enumerator if no block */ + r = n; + if (rb_check_arity(argc, 0, 1) && !NIL_P(argv[0])) + r = NUM2LONG(argv[0]); /* Permutation size from argument */ + + 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_LEN(ary); i++) { + rb_yield(rb_ary_new3(1, RARRAY_AREF(ary, i))); + } + } + else { /* this is the general case */ + volatile VALUE t0; + long *p = ALLOCV_N(long, t0, r+roomof(n, sizeof(long))); + char *used = (char*)(p + r); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC_CLEAR_CLASS(ary0); + + MEMZERO(used, char, n); /* initialize array */ + + permute0(n, r, p, used, ary0); /* compute and yield permutations */ + ALLOCV_END(t0); + RBASIC_SET_CLASS_RAW(ary0, rb_cArray); + } + return ary; +} + +static void +combinate0(const long len, const long n, long *const stack, const VALUE values) +{ + long lev = 0; + + MEMZERO(stack+1, long, n); + stack[0] = -1; + for (;;) { + for (lev++; lev < n; lev++) { + stack[lev+1] = stack[lev]+1; + } + if (!yield_indexed_values(values, n, stack+1)) { + rb_raise(rb_eRuntimeError, "combination reentered"); + } + do { + if (lev == 0) return; + stack[lev--]++; + } while (stack[lev+1]+n == len+lev+1); + } +} + +static VALUE +rb_ary_combination_size(VALUE ary, VALUE args, VALUE eobj) +{ + long n = RARRAY_LEN(ary); + long k = NUM2LONG(RARRAY_AREF(args, 0)); + + return binomial_coefficient(k, n); +} + +/* + * call-seq: + * ary.combination(n) {|c| block} -> ary + * ary.combination(n) -> Enumerator + * + * When invoked with a block, yields all combinations of length +n+ of elements + * from the array and then returns the array itself. + * + * The implementation makes no guarantees about the order in which the + * combinations are yielded. + * + * If no block is given, an Enumerator is returned instead. + * + * Examples: + * + * a = [1, 2, 3, 4] + * a.combination(1).to_a #=> [[1],[2],[3],[4]] + * a.combination(2).to_a #=> [[1,2],[1,3],[1,4],[2,3],[2,4],[3,4]] + * a.combination(3).to_a #=> [[1,2,3],[1,2,4],[1,3,4],[2,3,4]] + * 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(VALUE ary, VALUE num) +{ + long i, n, len; + + n = NUM2LONG(num); + RETURN_SIZED_ENUMERATOR(ary, 1, &num, rb_ary_combination_size); + len = RARRAY_LEN(ary); + if (n < 0 || len < n) { + /* yield nothing */ + } + else if (n == 0) { + rb_yield(rb_ary_new2(0)); + } + else if (n == 1) { + for (i = 0; i < RARRAY_LEN(ary); i++) { + rb_yield(rb_ary_new3(1, RARRAY_AREF(ary, i))); + } + } + else { + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + volatile VALUE t0; + long *stack = ALLOCV_N(long, t0, n+1); + + RBASIC_CLEAR_CLASS(ary0); + combinate0(len, n, stack, ary0); + ALLOCV_END(t0); + RBASIC_SET_CLASS_RAW(ary0, rb_cArray); + } + return ary; +} + +/* + * Compute repeated permutations of +r+ elements of the set + * <code>[0..n-1]</code>. + * + * When we have a complete repeated permutation of array indices, copy the + * values at those indices 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 + * values: the Ruby array that holds the actual values to permute + */ +static void +rpermute0(const long n, const long r, long *const p, const VALUE values) +{ + long i = 0, index = 0; + + p[index] = i; + for (;;) { + if (++index < r-1) { + p[index] = i = 0; + continue; + } + for (i = 0; i < n; ++i) { + p[index] = i; + if (!yield_indexed_values(values, r, p)) { + rb_raise(rb_eRuntimeError, "repeated permute reentered"); + } + } + do { + if (index <= 0) return; + } while ((i = ++p[--index]) >= n); + } +} + +static VALUE +rb_ary_repeated_permutation_size(VALUE ary, VALUE args, VALUE eobj) +{ + long n = RARRAY_LEN(ary); + long k = NUM2LONG(RARRAY_AREF(args, 0)); + + if (k < 0) { + return LONG2FIX(0); + } + if (n <= 0) { + return LONG2FIX(!k); + } + return rb_int_positive_pow(n, (unsigned long)k); +} + +/* + * call-seq: + * ary.repeated_permutation(n) {|p| block} -> ary + * ary.repeated_permutation(n) -> Enumerator + * + * When invoked with a block, yield all repeated permutations of length +n+ of + * the elements of the array, 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_SIZED_ENUMERATOR(ary, 1, &num, rb_ary_repeated_permutation_size); /* 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_AREF(ary, i))); + } + } + else { /* this is the general case */ + volatile VALUE t0; + long *p = ALLOCV_N(long, t0, r); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC_CLEAR_CLASS(ary0); + + rpermute0(n, r, p, ary0); /* compute and yield repeated permutations */ + ALLOCV_END(t0); + RBASIC_SET_CLASS_RAW(ary0, rb_cArray); + } + return ary; +} + +static void +rcombinate0(const long n, const long r, long *const p, const long rest, const VALUE values) +{ + long i = 0, index = 0; + + p[index] = i; + for (;;) { + if (++index < r-1) { + p[index] = i; + continue; + } + for (; i < n; ++i) { + p[index] = i; + if (!yield_indexed_values(values, r, p)) { + rb_raise(rb_eRuntimeError, "repeated combination reentered"); + } + } + do { + if (index <= 0) return; + } while ((i = ++p[--index]) >= n); + } +} + +static VALUE +rb_ary_repeated_combination_size(VALUE ary, VALUE args, VALUE eobj) +{ + long n = RARRAY_LEN(ary); + long k = NUM2LONG(RARRAY_AREF(args, 0)); + if (k == 0) { + return LONG2FIX(1); + } + return binomial_coefficient(k, n + k - 1); +} + +/* + * call-seq: + * ary.repeated_combination(n) {|c| block} -> ary + * ary.repeated_combination(n) -> Enumerator + * + * When invoked with a block, yields all repeated combinations of length +n+ of + * elements from the array and then returns the array 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_SIZED_ENUMERATOR(ary, 1, &num, rb_ary_repeated_combination_size); /* 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 < RARRAY_LEN(ary); i++) { + rb_yield(rb_ary_new3(1, RARRAY_AREF(ary, i))); + } + } + else if (len == 0) { + /* yield nothing */ + } + else { + volatile VALUE t0; + long *p = ALLOCV_N(long, t0, n); + VALUE ary0 = ary_make_shared_copy(ary); /* private defensive copy of ary */ + RBASIC_CLEAR_CLASS(ary0); + + rcombinate0(len, n, p, n, ary0); /* compute and yield repeated combinations */ + ALLOCV_END(t0); + RBASIC_SET_CLASS_RAW(ary0, 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 +self+ and + * the argument arrays. + * + * If given a block, #product 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([]) #=> [] + */ + +static VALUE +rb_ary_product(int argc, VALUE *argv, VALUE ary) +{ + int n = argc+1; /* How many arrays we're operating on */ + volatile VALUE t0 = tmpary(n); + volatile VALUE t1 = Qundef; + VALUE *arrays = RARRAY_PTR(t0); /* The arrays we're computing the product of */ + int *counters = ALLOCV_N(int, t1, n); /* 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; + + RBASIC_CLEAR_CLASS(t0); + + /* 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; + + /* 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]); + } + } + 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]); + if (k == 0) { + result = rb_ary_new2(0); + goto done; + } + if (MUL_OVERFLOW_LONG_P(resultlen, k)) + rb_raise(rb_eRangeError, "too big to product"); + resultlen *= k; + } + result = rb_ary_new2(resultlen); + } + for (;;) { + int m; + /* fill in one subarray */ + VALUE subarray = rb_ary_new2(n); + for (j = 0; j < n; j++) { + rb_ary_push(subarray, rb_ary_entry(arrays[j], counters[j])); + } + + /* put it on the result array */ + 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 + * and increment the one before it. + */ + m = n-1; + counters[m]++; + while (counters[m] == RARRAY_LEN(arrays[m])) { + counters[m] = 0; + /* If the first counter overflows, we are done */ + if (--m < 0) goto done; + counters[m]++; + } + } +done: + tmpary_discard(t0); + ALLOCV_END(t1); + + return NIL_P(result) ? ary : result; +} + +/* + * call-seq: + * ary.take(n) -> new_ary + * + * Returns first +n+ elements from the array. + * + * If a negative number is given, raises an ArgumentError. + * + * See also Array#drop + * + * a = [1, 2, 3, 4, 5, 0] + * a.take(3) #=> [1, 2, 3] + * + */ + +static VALUE +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 {|obj| block} -> new_ary + * ary.take_while -> 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. + * + * See also Array#drop_while + * + * a = [1, 2, 3, 4, 5, 0] + * a.take_while {|i| i < 3} #=> [1, 2] + * + */ + +static VALUE +rb_ary_take_while(VALUE ary) +{ + long i; + + RETURN_ENUMERATOR(ary, 0, 0); + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (!RTEST(rb_yield(RARRAY_AREF(ary, i)))) break; + } + return rb_ary_take(ary, LONG2FIX(i)); +} + +/* + * call-seq: + * ary.drop(n) -> new_ary + * + * Drops first +n+ elements from +ary+ and returns the rest of the elements in + * an array. + * + * If a negative number is given, raises an ArgumentError. + * + * See also Array#take + * + * a = [1, 2, 3, 4, 5, 0] + * a.drop(3) #=> [4, 5, 0] + * + */ + +static VALUE +rb_ary_drop(VALUE ary, VALUE n) +{ + VALUE result; + long pos = NUM2LONG(n); + if (pos < 0) { + rb_raise(rb_eArgError, "attempt to drop negative size"); + } + + result = rb_ary_subseq(ary, pos, RARRAY_LEN(ary)); + if (result == Qnil) result = rb_ary_new(); + return result; +} + +/* + * call-seq: + * ary.drop_while {|obj| block} -> new_ary + * ary.drop_while -> Enumerator + * + * Drops elements up to, but not including, the first element for 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. + * + * See also Array#take_while + * + * a = [1, 2, 3, 4, 5, 0] + * a.drop_while {|i| i < 3 } #=> [3, 4, 5, 0] + * + */ + +static VALUE +rb_ary_drop_while(VALUE ary) +{ + long i; + + RETURN_ENUMERATOR(ary, 0, 0); + for (i = 0; i < RARRAY_LEN(ary); i++) { + if (!RTEST(rb_yield(RARRAY_AREF(ary, i)))) break; + } + return rb_ary_drop(ary, LONG2FIX(i)); +} + +/* + * call-seq: + * ary.any? [{|obj| block} ] -> true or false + * ary.any?(pattern) -> true or false + * + * See also Enumerable#any? + */ + +static VALUE +rb_ary_any_p(int argc, VALUE *argv, VALUE ary) +{ + long i, len = RARRAY_LEN(ary); + + rb_check_arity(argc, 0, 1); + if (!len) return Qfalse; + if (argc) { + if (rb_block_given_p()) { + rb_warn("given block not used"); + } + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_funcall(argv[0], idEqq, 1, RARRAY_AREF(ary, i)))) return Qtrue; + } + } + else if (!rb_block_given_p()) { + for (i = 0; i < len; ++i) { + if (RTEST(RARRAY_AREF(ary, i))) return Qtrue; + } + } + else { + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) return Qtrue; + } + } + return Qfalse; +} + +/* + * call-seq: + * ary.all? [{|obj| block} ] -> true or false + * ary.all?(pattern) -> true or false + * + * See also Enumerable#all? + */ + +static VALUE +rb_ary_all_p(int argc, VALUE *argv, VALUE ary) +{ + long i, len = RARRAY_LEN(ary); + + rb_check_arity(argc, 0, 1); + if (!len) return Qtrue; + if (argc) { + if (rb_block_given_p()) { + rb_warn("given block not used"); + } + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (!RTEST(rb_funcall(argv[0], idEqq, 1, RARRAY_AREF(ary, i)))) return Qfalse; + } + } + else if (!rb_block_given_p()) { + for (i = 0; i < len; ++i) { + if (!RTEST(RARRAY_AREF(ary, i))) return Qfalse; + } + } + else { + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (!RTEST(rb_yield(RARRAY_AREF(ary, i)))) return Qfalse; + } + } + return Qtrue; +} + +/* + * call-seq: + * ary.none? [{|obj| block} ] -> true or false + * ary.none?(pattern) -> true or false + * + * See also Enumerable#none? + */ + +static VALUE +rb_ary_none_p(int argc, VALUE *argv, VALUE ary) +{ + long i, len = RARRAY_LEN(ary); + + rb_check_arity(argc, 0, 1); + if (!len) return Qtrue; + if (argc) { + if (rb_block_given_p()) { + rb_warn("given block not used"); + } + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_funcall(argv[0], idEqq, 1, RARRAY_AREF(ary, i)))) return Qfalse; + } + } + else if (!rb_block_given_p()) { + for (i = 0; i < len; ++i) { + if (RTEST(RARRAY_AREF(ary, i))) return Qfalse; + } + } + else { + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) return Qfalse; + } + } + return Qtrue; +} + +/* + * call-seq: + * ary.one? [{|obj| block} ] -> true or false + * ary.one?(pattern) -> true or false + * + * See also Enumerable#one? + */ + +static VALUE +rb_ary_one_p(int argc, VALUE *argv, VALUE ary) +{ + long i, len = RARRAY_LEN(ary); + VALUE result = Qfalse; + + rb_check_arity(argc, 0, 1); + if (!len) return Qfalse; + if (argc) { + if (rb_block_given_p()) { + rb_warn("given block not used"); + } + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_funcall(argv[0], idEqq, 1, RARRAY_AREF(ary, i)))) { + if (result) return Qfalse; + result = Qtrue; + } + } + } + else if (!rb_block_given_p()) { + for (i = 0; i < len; ++i) { + if (RTEST(RARRAY_AREF(ary, i))) { + if (result) return Qfalse; + result = Qtrue; + } + } + } + else { + for (i = 0; i < RARRAY_LEN(ary); ++i) { + if (RTEST(rb_yield(RARRAY_AREF(ary, i)))) { + if (result) return Qfalse; + result = Qtrue; + } + } + } + return result; +} + +/* + * call-seq: + * ary.dig(idx, ...) -> object + * + * Extracts the nested value specified by the sequence of <i>idx</i> + * objects by calling +dig+ at each step, returning +nil+ if any + * intermediate step is +nil+. + * + * a = [[1, [2, 3]]] + * + * a.dig(0, 1, 1) #=> 3 + * a.dig(1, 2, 3) #=> nil + * a.dig(0, 0, 0) #=> TypeError: Integer does not have #dig method + * [42, {foo: :bar}].dig(1, :foo) #=> :bar + */ + +static VALUE +rb_ary_dig(int argc, VALUE *argv, VALUE self) +{ + rb_check_arity(argc, 1, UNLIMITED_ARGUMENTS); + self = rb_ary_at(self, *argv); + if (!--argc) return self; + ++argv; + return rb_obj_dig(argc, argv, self, Qnil); +} + +static inline VALUE +finish_exact_sum(long n, VALUE r, VALUE v, int z) +{ + if (n != 0) + v = rb_fix_plus(LONG2FIX(n), v); + if (r != Qundef) { + /* r can be an Integer when mathn is loaded */ + if (FIXNUM_P(r)) + v = rb_fix_plus(r, v); + else if (RB_TYPE_P(r, T_BIGNUM)) + v = rb_big_plus(r, v); + else + v = rb_rational_plus(r, v); + } + else if (!n && z) { + v = rb_fix_plus(LONG2FIX(0), v); + } + return v; +} + +/* + * call-seq: + * ary.sum(init=0) -> number + * ary.sum(init=0) {|e| expr } -> number + * + * Returns the sum of elements. + * For example, [e1, e2, e3].sum returns init + e1 + e2 + e3. + * + * If a block is given, the block is applied to each element + * before addition. + * + * If <i>ary</i> is empty, it returns <i>init</i>. + * + * [].sum #=> 0 + * [].sum(0.0) #=> 0.0 + * [1, 2, 3].sum #=> 6 + * [3, 5.5].sum #=> 8.5 + * [2.5, 3.0].sum(0.0) {|e| e * e } #=> 15.25 + * [Object.new].sum #=> TypeError + * + * The (arithmetic) mean value of an array can be obtained as follows. + * + * mean = ary.sum(0.0) / ary.length + * + * This method can be used for non-numeric objects by + * explicit <i>init</i> argument. + * + * ["a", "b", "c"].sum("") #=> "abc" + * [[1], [[2]], [3]].sum([]) #=> [1, [2], 3] + * + * However, Array#join and Array#flatten is faster than Array#sum for + * array of strings and array of arrays. + * + * ["a", "b", "c"].join #=> "abc" + * [[1], [[2]], [3]].flatten(1) #=> [1, [2], 3] + * + * + * Array#sum method may not respect method redefinition of "+" methods + * such as Integer#+. + * + */ + +static VALUE +rb_ary_sum(int argc, VALUE *argv, VALUE ary) +{ + VALUE e, v, r; + long i, n; + int block_given; + + v = (rb_check_arity(argc, 0, 1) ? argv[0] : LONG2FIX(0)); + + block_given = rb_block_given_p(); + + if (RARRAY_LEN(ary) == 0) + return v; + + n = 0; + r = Qundef; + for (i = 0; i < RARRAY_LEN(ary); i++) { + e = RARRAY_AREF(ary, i); + if (block_given) + e = rb_yield(e); + if (FIXNUM_P(e)) { + n += FIX2LONG(e); /* should not overflow long type */ + if (!FIXABLE(n)) { + v = rb_big_plus(LONG2NUM(n), v); + n = 0; + } + } + else if (RB_TYPE_P(e, T_BIGNUM)) + v = rb_big_plus(e, v); + else if (RB_TYPE_P(e, T_RATIONAL)) { + if (r == Qundef) + r = e; + else + r = rb_rational_plus(r, e); + } + else + goto not_exact; + } + v = finish_exact_sum(n, r, v, argc!=0); + return v; + + not_exact: + v = finish_exact_sum(n, r, v, i!=0); + + if (RB_FLOAT_TYPE_P(e)) { + /* + * Kahan-Babuska balancing compensated summation algorithm + * See http://link.springer.com/article/10.1007/s00607-005-0139-x + */ + double f, c; + double x, t; + + f = NUM2DBL(v); + c = 0.0; + goto has_float_value; + for (; i < RARRAY_LEN(ary); i++) { + e = RARRAY_AREF(ary, i); + if (block_given) + e = rb_yield(e); + if (RB_FLOAT_TYPE_P(e)) + has_float_value: + x = RFLOAT_VALUE(e); + else if (FIXNUM_P(e)) + x = FIX2LONG(e); + else if (RB_TYPE_P(e, T_BIGNUM)) + x = rb_big2dbl(e); + else if (RB_TYPE_P(e, T_RATIONAL)) + x = rb_num2dbl(e); + else + goto not_float; + + if (isnan(f)) continue; + if (isnan(x)) { + f = x; + continue; + } + if (isinf(x)) { + if (isinf(f) && signbit(x) != signbit(f)) + f = NAN; + else + f = x; + continue; + } + if (isinf(f)) continue; + + t = f + x; + if (fabs(f) >= fabs(x)) + c += ((f - t) + x); + else + c += ((x - t) + f); + f = t; + } + f += c; + return DBL2NUM(f); + + not_float: + v = DBL2NUM(f); + } + + goto has_some_value; + for (; i < RARRAY_LEN(ary); i++) { + e = RARRAY_AREF(ary, i); + if (block_given) + e = rb_yield(e); + has_some_value: + v = rb_funcall(v, idPLUS, 1, e); + } + return v; +} + +static VALUE +rb_ary_deconstruct(VALUE ary) +{ + return ary; +} -/* 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. + * + * == Creating Arrays + * + * A new array can be created by using the literal constructor + * <code>[]</code>. Arrays can contain different types of objects. For + * example, the array below contains an Integer, a String and a Float: + * + * ary = [1, "two", 3.0] #=> [1, "two", 3.0] + * + * An array can also be created by explicitly calling Array.new with zero, one + * (the initial size of the Array) or two arguments (the initial size and a + * default object). + * + * ary = Array.new #=> [] + * Array.new(3) #=> [nil, nil, nil] + * Array.new(3, true) #=> [true, true, true] + * + * Note that the second argument populates the array with references to the + * same object. Therefore, it is only recommended in cases when you need to + * instantiate arrays with natively immutable objects such as Symbols, + * numbers, true or false. + * + * To create an array with separate objects a block can be passed instead. + * This method is safe to use with mutable objects such as hashes, strings or + * other arrays: + * + * Array.new(4) {Hash.new} #=> [{}, {}, {}, {}] + * Array.new(4) {|i| i.to_s } #=> ["0", "1", "2", "3"] + * + * This is also a quick way to build up multi-dimensional arrays: + * + * empty_table = Array.new(3) {Array.new(3)} + * #=> [[nil, nil, nil], [nil, nil, nil], [nil, nil, nil]] + * + * An array can also be created by using the Array() method, provided by + * Kernel, which tries to call #to_ary, then #to_a on its argument. + * + * Array({:a => "a", :b => "b"}) #=> [[:a, "a"], [:b, "b"]] + * + * == Example Usage + * + * In addition to the methods it mixes in through the Enumerable module, the + * Array class has proprietary methods for accessing, searching and otherwise + * manipulating arrays. + * + * Some of the more common ones are illustrated below. + * + * == Accessing Elements + * + * Elements in an array can be retrieved using the Array#[] method. It can + * take a single integer argument (a numeric index), a pair of arguments + * (start and length) or a range. Negative indices start counting from the end, + * with -1 being the last element. + * + * arr = [1, 2, 3, 4, 5, 6] + * arr[2] #=> 3 + * arr[100] #=> nil + * arr[-3] #=> 4 + * arr[2, 3] #=> [3, 4, 5] + * arr[1..4] #=> [2, 3, 4, 5] + * arr[1..-3] #=> [2, 3, 4] + * + * Another way to access a particular array element is by using the #at method + * + * arr.at(0) #=> 1 + * + * The #slice method works in an identical manner to Array#[]. + * + * To raise an error for indices outside of the array bounds or else to + * provide a default value when that happens, you can use #fetch. + * + * arr = ['a', 'b', 'c', 'd', 'e', 'f'] + * arr.fetch(100) #=> IndexError: index 100 outside of array bounds: -6...6 + * arr.fetch(100, "oops") #=> "oops" + * + * The special methods #first and #last will return the first and last + * elements of an array, respectively. + * + * arr.first #=> 1 + * arr.last #=> 6 + * + * To return the first +n+ elements of an array, use #take + * + * arr.take(3) #=> [1, 2, 3] + * + * #drop does the opposite of #take, by returning the elements after +n+ + * elements have been dropped: + * + * arr.drop(3) #=> [4, 5, 6] + * + * == Obtaining Information about an Array + * + * Arrays keep track of their own length at all times. To query an array + * about the number of elements it contains, use #length, #count or #size. + * + * browsers = ['Chrome', 'Firefox', 'Safari', 'Opera', 'IE'] + * browsers.length #=> 5 + * browsers.count #=> 5 + * + * To check whether an array contains any elements at all + * + * browsers.empty? #=> false + * + * To check whether a particular item is included in the array + * + * browsers.include?('Konqueror') #=> false + * + * == Adding Items to Arrays + * + * Items can be added to the end of an array by using either #push or #<< + * + * arr = [1, 2, 3, 4] + * arr.push(5) #=> [1, 2, 3, 4, 5] + * arr << 6 #=> [1, 2, 3, 4, 5, 6] + * + * #unshift will add a new item to the beginning of an array. + * + * arr.unshift(0) #=> [0, 1, 2, 3, 4, 5, 6] + * + * With #insert you can add a new element to an array at any position. + * + * arr.insert(3, 'apple') #=> [0, 1, 2, 'apple', 3, 4, 5, 6] + * + * Using the #insert method, you can also insert multiple values at once: + * + * arr.insert(3, 'orange', 'pear', 'grapefruit') + * #=> [0, 1, 2, "orange", "pear", "grapefruit", "apple", 3, 4, 5, 6] + * + * == Removing Items from an Array + * + * The method #pop removes the last element in an array and returns it: + * + * arr = [1, 2, 3, 4, 5, 6] + * arr.pop #=> 6 + * arr #=> [1, 2, 3, 4, 5] + * + * To retrieve and at the same time remove the first item, use #shift: + * + * arr.shift #=> 1 + * arr #=> [2, 3, 4, 5] + * + * To delete an element at a particular index: + * + * arr.delete_at(2) #=> 4 + * arr #=> [2, 3, 5] + * + * To delete a particular element anywhere in an array, use #delete: + * + * arr = [1, 2, 2, 3] + * arr.delete(2) #=> 2 + * arr #=> [1,3] + * + * A useful method if you need to remove +nil+ values from an array is + * #compact: + * + * arr = ['foo', 0, nil, 'bar', 7, 'baz', nil] + * arr.compact #=> ['foo', 0, 'bar', 7, 'baz'] + * arr #=> ['foo', 0, nil, 'bar', 7, 'baz', nil] + * arr.compact! #=> ['foo', 0, 'bar', 7, 'baz'] + * arr #=> ['foo', 0, 'bar', 7, 'baz'] + * + * Another common need is to remove duplicate elements from an array. + * + * It has the non-destructive #uniq, and destructive method #uniq! + * + * arr = [2, 5, 6, 556, 6, 6, 8, 9, 0, 123, 556] + * arr.uniq #=> [2, 5, 6, 556, 8, 9, 0, 123] + * + * == Iterating over Arrays + * + * Like all classes that include the Enumerable module, Array has an each + * method, which defines what elements should be iterated over and how. In + * case of Array's #each, all elements in the Array instance are yielded to + * the supplied block in sequence. + * + * Note that this operation leaves the array unchanged. + * + * arr = [1, 2, 3, 4, 5] + * arr.each {|a| print a -= 10, " "} + * # prints: -9 -8 -7 -6 -5 + * #=> [1, 2, 3, 4, 5] + * + * Another sometimes useful iterator is #reverse_each which will iterate over + * the elements in the array in reverse order. + * + * words = %w[first second third fourth fifth sixth] + * str = "" + * words.reverse_each {|word| str += "#{word} "} + * p str #=> "sixth fifth fourth third second first " + * + * The #map method can be used to create a new array based on the original + * array, but with the values modified by the supplied block: + * + * arr.map {|a| 2*a} #=> [2, 4, 6, 8, 10] + * arr #=> [1, 2, 3, 4, 5] + * arr.map! {|a| a**2} #=> [1, 4, 9, 16, 25] + * arr #=> [1, 4, 9, 16, 25] + * + * == Selecting Items from an Array + * + * Elements can be selected from an array according to criteria defined in a + * block. The selection can happen in a destructive or a non-destructive + * manner. While the destructive operations will modify the array they were + * called on, the non-destructive methods usually return a new array with the + * selected elements, but leave the original array unchanged. + * + * === Non-destructive Selection + * + * arr = [1, 2, 3, 4, 5, 6] + * arr.select {|a| a > 3} #=> [4, 5, 6] + * arr.reject {|a| a < 3} #=> [3, 4, 5, 6] + * arr.drop_while {|a| a < 4} #=> [4, 5, 6] + * arr #=> [1, 2, 3, 4, 5, 6] + * + * === Destructive Selection + * + * #select! and #reject! are the corresponding destructive methods to #select + * and #reject + * + * Similar to #select vs. #reject, #delete_if and #keep_if have the exact + * opposite result when supplied with the same block: + * + * arr.delete_if {|a| a < 4} #=> [4, 5, 6] + * arr #=> [4, 5, 6] + * + * arr = [1, 2, 3, 4, 5, 6] + * arr.keep_if {|a| a < 4} #=> [1, 2, 3] + * arr #=> [1, 2, 3] + * */ 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_alloc_func(rb_cArray, empty_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_h", rb_ary_to_h, 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); rb_define_method(rb_cArray, "==", rb_ary_equal, 1); rb_define_method(rb_cArray, "eql?", rb_ary_eql, 1); @@ -3052,12 +6946,17 @@ Init_Array() rb_define_method(rb_cArray, "fetch", rb_ary_fetch, -1); rb_define_method(rb_cArray, "first", rb_ary_first, -1); rb_define_method(rb_cArray, "last", rb_ary_last, -1); - rb_define_method(rb_cArray, "concat", rb_ary_concat, 1); + rb_define_method(rb_cArray, "concat", rb_ary_concat_multi, -1); + rb_define_method(rb_cArray, "union", rb_ary_union_multi, -1); + rb_define_method(rb_cArray, "difference", rb_ary_difference_multi, -1); + rb_define_method(rb_cArray, "intersection", rb_ary_intersection_multi, -1); rb_define_method(rb_cArray, "<<", rb_ary_push, 1); rb_define_method(rb_cArray, "push", rb_ary_push_m, -1); - rb_define_method(rb_cArray, "pop", rb_ary_pop, 0); - rb_define_method(rb_cArray, "shift", rb_ary_shift, 0); + rb_define_alias(rb_cArray, "append", "push"); + rb_define_method(rb_cArray, "pop", rb_ary_pop_m, -1); + rb_define_method(rb_cArray, "shift", rb_ary_shift_m, -1); rb_define_method(rb_cArray, "unshift", rb_ary_unshift_m, -1); + rb_define_alias(rb_cArray, "prepend", "unshift"); rb_define_method(rb_cArray, "insert", rb_ary_insert, -1); rb_define_method(rb_cArray, "each", rb_ary_each, 0); rb_define_method(rb_cArray, "each_index", rb_ary_each_index, 0); @@ -3065,20 +6964,26 @@ Init_Array() rb_define_method(rb_cArray, "length", rb_ary_length, 0); rb_define_alias(rb_cArray, "size", "length"); rb_define_method(rb_cArray, "empty?", rb_ary_empty_p, 0); - 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, "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, "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, "filter", rb_ary_select, 0); + rb_define_method(rb_cArray, "filter!", 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); @@ -3106,14 +7011,41 @@ Init_Array() rb_define_method(rb_cArray, "&", rb_ary_and, 1); rb_define_method(rb_cArray, "|", rb_ary_or, 1); + rb_define_method(rb_cArray, "max", rb_ary_max, -1); + rb_define_method(rb_cArray, "min", rb_ary_min, -1); + rb_define_method(rb_cArray, "minmax", rb_ary_minmax, 0); + rb_define_method(rb_cArray, "uniq", rb_ary_uniq, 0); rb_define_method(rb_cArray, "uniq!", rb_ary_uniq_bang, 0); rb_define_method(rb_cArray, "compact", rb_ary_compact, 0); rb_define_method(rb_cArray, "compact!", rb_ary_compact_bang, 0); - rb_define_method(rb_cArray, "flatten", rb_ary_flatten, 0); - rb_define_method(rb_cArray, "flatten!", rb_ary_flatten_bang, 0); - rb_define_method(rb_cArray, "nitems", rb_ary_nitems, 0); - - id_cmp = rb_intern("<=>"); - inspect_key = rb_intern("__inspect_key__"); + 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, "count", rb_ary_count, -1); + 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); + rb_define_method(rb_cArray, "take_while", rb_ary_take_while, 0); + rb_define_method(rb_cArray, "drop", rb_ary_drop, 1); + rb_define_method(rb_cArray, "drop_while", rb_ary_drop_while, 0); + rb_define_method(rb_cArray, "bsearch", rb_ary_bsearch, 0); + rb_define_method(rb_cArray, "bsearch_index", rb_ary_bsearch_index, 0); + rb_define_method(rb_cArray, "any?", rb_ary_any_p, -1); + rb_define_method(rb_cArray, "all?", rb_ary_all_p, -1); + rb_define_method(rb_cArray, "none?", rb_ary_none_p, -1); + rb_define_method(rb_cArray, "one?", rb_ary_one_p, -1); + rb_define_method(rb_cArray, "dig", rb_ary_dig, -1); + rb_define_method(rb_cArray, "sum", rb_ary_sum, -1); + + rb_define_method(rb_cArray, "deconstruct", rb_ary_deconstruct, 0); + + id_random = rb_intern("random"); } |